Warfare, Violence and Slavery in Prehistory: Proceedings of a Prehistoric Society conference at Sheffield University 9781841718163, 9781407328065

19 papers presented at the Proceedings of a Prehistoric Society conference at Sheffield University in February 2001.

180 6 20MB

English Pages [239] Year 2005

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Front Cover
Title Page
Copyright
Table of Contents
The ancient origins of warfare and violence
Warfare, violence and slavery in later prehistory: An introduction
Aggression and nonhuman primates
Sociobiology, cultural anthropology and the causes of warfare
The physical evidence of warfare - subtle stigmata?
The head burials from Ofnet cave: An example of warlike conflict in the Mesolithic
Assessing rank and warfare-strategy in prehistoric hunter-gatherer society: A study of representational warrior figures in rock-art from the Spanish Levant, southeastern Spain
The emergence of warfare in the Early Bronze Age: The Nitra group in Slovakia and Moravia, 2200-1800 BC
Warfare, redistribution and society in western Iberia
Warfare, violence and the construction of masculinity in the Iron Age rock art of Valcamonica, northern Italy
The dead of Tormarton - Middle Bronze Age combat victims?
Giving up weapons
Ritual bondage, violence, slavery and sacrifice in later European prehistory
Fragmentation, mutilation and dismemberment: An interpretation of human remains on Iron Age sites
The origins of warfare: Later prehistory in southeastern Iberia
Weaponry, statues and petroglyphs: The ideology of war in Atlantic Iron Age Iberia
A palaeodemographic investigation of warfare in prehistory
War in prehistoric society: Modern views of ancient violence
Ambushed by a grotesque: Archaeology, slavery and the third paradigm
Recommend Papers

Warfare, Violence and Slavery in Prehistory: Proceedings of a Prehistoric Society conference at Sheffield University
 9781841718163, 9781407328065

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

BAR S1374 2005 PARKER PEARSON & THORPE (Eds)

Warfare, Violence and Slavery in Prehistory Edited by

Mike Parker Pearson I. J. N. Thorpe

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY

B A R red cover template.indd 1

BAR International Series 1374 2005

16/12/2010 10:33:11

Warfare, Violence and Slavery in Prehistory Proceedings of a Prehistoric Society conference at Sheffield University Edited by

Mike Parker Pearson I. J. N. Thorpe

BAR International Series 1374 2005

ISBN 9781841718163 paperback ISBN 9781407328065 e-format DOI https://doi.org/10.30861/9781841718163 A catalogue record for this book is available from the British Library

BAR

PUBLISHING

Contents Chapter 1 The ancient origins of warfare and violence .........................................................................................1 I J N Thorpe Chapter 2 Warfare, violence and slavery in later prehistory: an introduction .....................................................19 Mike Parker Pearson Chapter 3 Aggression and nonhuman primates ...................................................................................................35 Pia Nystrom Chapter 4 Sociobiology, cultural anthropology and the causes of warfare..........................................................41 Robert Layton Chapter 5 The physical evidence of warfare - subtle stigmata?...........................................................................49 Christopher J Knüsel Chapter 6 The head burials from Ofnet cave: an example of warlike conflict in the Mesolithic.........................67 Jörg Orschiedt Chapter 7 Assessing rank and warfare-strategy in prehistoric hunter-gatherer society: a study of representational warrior figures in rock-art from the Spanish Levant, southeastern Spain .................75 George Nash Chapter 8 The emergence of warfare in the Early Bronze Age: the Nitra group in Slovakia and Moravia, 2200-1800 BC.....................................................................................................................................87 Andreas Hårde Chapter 9 Warfare, redistribution and society in western Iberia .......................................................................107 Eduardo Sánchez-Moreno Chapter 10 Warfare, violence and the construction of masculinity in the Iron Age rock art of Valcamonica, northern Italy.....................................................................................................................................127 Lynne Bevan Chapter 11 The dead of Tormarton - Middle Bronze Age combat victims? .......................................................139 Richard Osgood Chapter 12 Giving up weapons............................................................................................................................145 David Fontijn Chapter 13 Ritual bondage, violence, slavery and sacrifice in later European prehistory ...................................155 Miranda Aldhouse Green

i

Chapter 14 Fragmentation, mutilation and dismemberment: an interpretation of human remains on Iron Age sites ....................................................................................................................................165 Rebecca Craig, Christopher J Knüsel and Gillian Carr Chapter 15 The origins of warfare: later prehistory in southeastern Iberia..........................................................181 Gonzalo Aranda Jiménez and Margarita Sánchez Romero Chapter 16 Weaponry, statues and petroglyphs: the ideology of war in Atlantic Iron Age Iberia.......................195 José Freire Chapter 17 A palaeodemographic investigation of warfare in prehistory............................................................201 Neil A Bishop and Christopher J Knüsel Chapter 18 War in prehistoric society: modern views of ancient violence..........................................................217 John Carman and Patricia Carman Chapter 19 Ambushed by a grotesque: archaeology, slavery and the third paradigm .........................................225 Tim Taylor

ii

The ancient origins of warfare and violence I.J.N. Thorpe Department of Archaeology, King Alfred’s College, Winchester

Recent DNA analysis suggests that some 500,000 years ago our Homo erectus ancestors were reduced in number to perhaps only a thousand individuals. David Woodruff, senior biologist on the team, believes that genocide was the most likely cause: ‘In our success came a need to remove the competitors. It’s my personal favourite’ (Brookes 1999: 35). Why is violence the first explanation to come to mind when considering our distant ancestors?

1997). From an American perspective, with the relatively recent and dramatic change in view of the Maya from peace-loving astronomer-priests (Thompson 1954) to warring rulers of city-states continually at war (Culbert 1988; Webster 2000) Keeley’s critique makes sense. But for a European audience, which has always seen the Bronze Age and Iron Age as times of warriors, his overall case is far less persuasive.

Warfare, and more generally conflict, is clearly a hot topic in both archaeology and anthropology, as the number of recent overviews and volumes of case studies amply demonstrates (e.g. Keegan 1993; Redmond 1994; Reyna and Downs 1994; Keeley 1996; Martin and Frayer 1997; Carman and Harding 1999; Raaflaub and Rosenstein 1999; Osgood et al. 2000; Lambert 2002; Gilchrist 2003). Not only is it a subject of great importance in its own right, but it also has a bearing on several other major issues, such as the attempt by evolutionary psychology to co-opt archaeological evidence, the history and biases of archaeology as a discipline and the nature of archaeological evidence.

Certainly, Keeley’s starting point in his claims of systematic attempts to downplay the role of warfare in prehistory, the ‘interpretative “pacification”’ of the causewayed enclosures of Britain, is far less clear-cut than he implies (1996: 18; Thorpe in press). Keeley criticises Whittle (Whittle 1985: 219-20) for discussing the boundary earthworks of enclosures in terms of the symbolism of exclusion. He then contrasts Whittle with ‘the archaeologists who have conducted extensive excavations of some of these enclosures’, who have uncovered traces of palisades and at several camps ‘thousands of flint arrowheads, concentrated along the palisade and especially at the gates’ (Keeley 1996: 18). Ironically, Whittle has actually carried out a major excavation at the most famous of all British causewayed enclosures – Windmill Hill (Whittle et al. 1999). The ditches at Windmill Hill contain not only deliberately dumped mounds of feasting debris and items which arrived there through long-distance exchange, but also burials. In the case of Windmill Hill, a wide range of activities was carried out at this major site, but these did not include either settlement or defence. There are also sites in eastern England which have been intensively investigated but present no evidence for defence or attack and little of possible settlement – Briar Hill (Bamford 1985), Etton (Pryor 1998) and Haddenham (Evans 1988; Hodder 1992), together with a number of examples in Sussex (Drewett 1994).

Origins are always interesting subjects and the origin of war is no exception, having been considered recently from the varied perspectives of archaeology (e.g. Keeley 1996), military history (e.g. Keegan 1993; O’Connell 1995), cultural evolution (e.g. Dawson 2001), biological anthropology (e.g. Van der Dennen 1995; Wrangham 1999) and social anthropology (e.g. Otterbein 1997; Kelly 2000). Of these fields, the most interest in early war and conflict has been shown by anthropologists of varying kinds, as it is generally considered to be central to that elusive quality, ‘human nature’. Archaeology has been largely an onlooker in this argument (with Ferguson 1997a being a significant exception) that has been almost entirely fought out using evidence from contemporary societies, which have conveniently been assumed to correspond in some way to ancient pre-agricultural societies.

Moreover, Saville (2002) notes that, in terms of numbers of arrowheads, only Carn Brea (Mercer 1981), Crickley Hill (Dixon 1988) and Hembury (Liddell 1930; 1931; 1932; 1935; Todd 1984) stand out, with other enclosures having a consistently minor presence of the type. Even these sites, however, possess several hundred rather than thousands of arrowheads. Indeed, Crickley Hill is the only British causewayed enclosure with a concentration of arrowheads at the palisade and entrance area. Keeley’s characterization of these sites may therefore be described as a case of interpretative ‘warrification’ rather than pacification.

Recent interest in the ancient evidence for warfare has been stimulated by Keeley’s polemical War Before Civilization: the myth of the peaceful savage (1996). As the title states, he seeks to demolish the ‘myth’ which he believes has been peddled by archaeologists and anthropologists who have attempted to pacify the past. Keeley’s broadside has prompted widely different reactions, with responses varying from whole-hearted approval, with Kristiansen (1999: 188) describing it as ‘inspirational’ and Le Blanc (1999: 3) as ‘an excellent critique of archeologists’ prior attitudes towards warfare’, to the highly critical (e.g. Ferguson 1997b; Otterbein

Vandkilde (2003) notes that some significant publications on warfare in European prehistory had appeared just before Keeley’s book was published (e.g. Randsborg 1

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY slaughtered by their own kind, were the victims of leopard predation.

1995; Treherne 1995). She attributes this revival in the popularity of ‘the warrior tale’ (2003: 136) to the ethnic wars in the Balkans. From the narrowly archaeological perspective one could also point to the impact of individual discoveries such as the Linearbandkeramik site of Talheim in Germany (Wahl and KĘnig 1987), dating to c. 5000 BC, where a mass grave contained 34 men, women and children, killed by axe and adze blows to the head.

This is not simply a case of newer scientific techniques correcting the errors produced by earlier methods. For, as Cartmill notes (1993: 20), the original connection made by Dart (1925) between the baboon skull and the Taung australopithecine infant was not one forced by the evidence at all. The instant interpretation, that the find of a baboon skull with a hole must represent the remains of a kill and subsequent consumption of the brain by Australopithecines, appears to owe more to preconceptions than anything else. Dart, who served as a member of the medical corps in England during World War I, might well have been influenced by memories of the injuries he witnessed in service.

The relative lack of attention given to conflict and war in the literature on early prehistory also needs to be set in historical perspective. Certainly some of the wariness concerning war in the Palaeolithic comes from a sense of necessary correction to the excesses of past interpretations, such as Dart’s view of the Australopithecines of southern Africa. Based on his studies of australopithecine skeletal material, Dart argued (1953: 209) that ancestral humans were:

A more accurate characterization of the history of the study of war within archaeology may therefore be one of swings between the warrior tale and the peasant tale, as Vandkilde argues (2003). Even if one were to follow Keeley (1996) in believing that the importance of warfare has been underestimated in social anthropology (despite Ferguson 1997b; Otterbein 1997), as others had argued earlier (e.g. Carneiro 1994), a crucial difficulty with Keeley’s argument remains. Although he rightly points to the high incidence of ethnographically recorded conflict and of more formal inter-group warfare - making a good case that this is not ‘ritual’ but real war, with a high level of casualties - this does not in itself tell us anything significant about ancient warfare. For there are ethnographic examples of relatively peaceful societies (as Keeley accepts – 1996: 27-32) so we cannot assume the existence of warfare, as Keeley himself appears to do for the Neolithic. Here we are in the realm of interpretation, unaided by participant accounts or direct outside observation.

‘… confirmed killers: carnivorous creatures that seized living quarries by violence, battered them to death, tore apart their broken bodies, dismembered them limb from limb, slaking their ravenous thirst with the hot blood of victims and greedily devouring livid writhing flesh.’ Dart went on to conclude that the course of human history had been set unalterably by the bloodthirsty Australopithecines (Dart and Craig 1959: 201): ‘The loathsome cruelty of mankind to man is the inescapable by-product of his blood lust; this differentiative human characteristic is explicable only in terms of man’s carnivorous and cannibalistic origin.’ ‘The blood-bespattered, slaughter-gutted archives of human history from the earliest Egyptian and Sumerian records to the most recent atrocities of the Second World War accord with early universal cannibalism, with animal and human sacrificial practices or their substitutes in formalized religions and with the world-wide scalping, head-hunting, bodymutilating and necrophiliac practices of mankind in proclaiming this common bloodlust differentiator, this predaceous habit, this mark of Cain that separates man dietetically from his anthropoidal relatives and allies him rather with the deadliest of the Carnivora.’

Before considering the major schools of interpretation of the origins of conflict I should provide a definition of warfare, for this has been a point of discussion in the past – it is defined here rather simply as organized aggression between autonomous political units, and thus of wider significance than individual acts of violence. This is certainly a far less complex definition than many others (e.g. the seven point scheme proposed in Kelly 2000), but it is one which may be usable when considering early prehistory. One might distinguish warfare proper from feuding, with feuds being conflicts between small and specific elements of autonomous political units, normally households (Otterbein 2000). I see no great value in drawing too definite a distinction, however, since feuds frequently escalate into war. As Koch puts it (1974: 523): ‘linguistic distinctions between raids, feuds, and war tend to obscure rather than elucidate the problem of explaining why people resort to violent methods of confrontation in pursuit of their interests’. The majority of conflicts recorded in the ethnographic record occur between closely related groups, with the warring parties frequently acting as exchange or marriage partners before

The influence of Dart’s views grew significantly after the appearance of the dramatist Robert Ardrey’s popularizing books of the 1960s (e.g. Ardrey 1961; 1967). For Ardrey (1961: 316) ‘man is a predator whose natural instinct is to kill with a weapon’. However, a later reanalysis of the skeletal material by Brain (1981) demonstrated quite clearly that the Australopithecines, far from being

2

I.J.N. THORPE: THE ANCIENT ORIGINS OF WARFARE AND VIOLENCE warfare is that which sees humans as shaped by a longpast ancestral environment, the environment of evolutionary adaptation (EEA). Thus Leda Cosmides and John Tooby propose (in their on-line ‘Primer of Evolutionary Psychology’ – available at http://www.psych.ucsb.edu/research/cep/primer.html) as their Principal 5 that ‘Our modern skulls house a stone age mind’. The EEA is generally identified as the Palaeolithic and Mesolithic (e.g. Cosmides et al. 1992; Pinker 1998: 42), with the development of agriculture making a crucial break. Unfortunately, within this model the assumption soon developed that human evolution took place on the savannah of Africa, so that specific adaptations needed to be traced back to their value in this context. Tooby and Cosmides (1990: 386-7) have rejected the savannah interpretation, which they believe to be a misunderstanding of their views, instead arguing that the EEA:

and after. As Clastres (1994: 162) notes, this runs counter to the notion first developed by Thomas Hobbes in Leviathan (1651) that war prevents exchange. It is certainly the case that some past societies have distinguished between feuding and warfare. Thus Redmond focuses on long-distance raiding in her important study of South American warfare (1994), as the Jívaro make a distinction between local feuding and true warfare, even though the former includes the assassination of individuals (Harner 1972). Jívaro true warfare consists of head-hunting raids carried out against distant groups (Descola 1993) – a far more prestigious activity, as it involves expeditions into lands ‘filled ... with evil spirits’ (Cotlow 1953: 144). So we are perhaps dealing here with high (long-distance) and low (local) status warfare. The role of head-hunting appears to have undergone a dramatic transformation in the late nineteenth century (Steel 1999). The original motive lay in the Jívaro belief that power could be extracted from the spirit trapped in the victim’s shrunken head, but a trade developed at that time of heads for guns. We seem to have here an example of cultural reasons for warring being overtaken by economic ones. A similar division between types of fighting is made by the Yukpa of Venezuela (Halbmayer 2001): external war involves ritual preparation, attacks on inhabitants of other valleys and the glorification of the warriors; vendettas have no rituals, are carried out against known individuals and are kept secret.

‘… is a statistical composite of the adaptation relevant properties of the ancestral environments encountered by members of ancestral populations, weighted by their frequency and their fitness consequences.’ As yet there has been no attempt to begin compiling this ‘statistical composite’, which suggests that it was more of a rhetorical device to keep the theory alive than an attempt to provide a better theoretical basis of understanding human behaviour. Whatever the problems with the concept of an environment of evolutionary adaptation, the three main competing theories of warfare situated within evolutionary psychology – territorial, reproductive and status competition – should all be susceptible to analysis from the archaeological evidence of early prehistory.

Definitions of the opposing theorists have also long been a source of debate, with lines of thought which can be traced back to Hobbes’ Leviathan (1651) and his ‘State of Warre’, and Jean-Jacques Rousseau’s The Social Contract (1762) and his vision of a golden age of peace before the state. Of course, neither Hobbes nor Rousseau would recognize many of the arguments stated by opponents to be following their lead. An equally clear dichotomy is maintained by Otterbein (1997; 1999) in his discussion of the study of warfare within anthropology, dividing authors into ‘doves’ or ‘hawks’ with obvious reference to the similar popular divide in contemporary American politics. The fundamental danger for the student of ancient warfare is that such a divide will take over the archaeological debate, thereby forcing out consideration of variability in the evidence.

The territorial model as it exists currently was developed by E.O. Wilson, who argued from the sociobiological strand of evolutionary psychology that ethnocentricity was a product of natural selection (1978: 119): ‘Our brains do appear to be programmed to the following extent: we are inclined to partition other people into friends and aliens, in the same sense that birds are inclined to learn territorial songs and to navigate by the polar constellations. We tend to fear deeply the actions of strangers and to solve conflicts by aggression. These learning rules are most likely to have evolved during the past hundreds or thousands of years of human evolution and, thus, to have conferred a biological advantage on those who conformed to them with the greatest fidelity.’

Some theories of war Much of the current public interest in conflict and warfare in Britain and North America stems from the dominance of neo-Darwinian thought in society at large. From the sociobiology of the 1960s to the evolutionary psychology of the 1990s there has been a considerable growth in the popular acceptance of a significant role for biology in human culture, especially in terms of the ‘nature vs. nurture’ debate, now taking a slightly less confrontational turn (e.g. Ridley 2003). The strand of evolutionary psychology most clearly relevant to the study of early

The difficulties with this supposed rule are two-fold. It would also be adaptive to create good relations with neighbouring groups, perhaps especially so when 3

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY may be a significant over-simplification. Research is now also revealing significant variations in the prevalence of aggressive behaviour between groups among gorilla populations, with western gorillas having far lower levels of violence than mountain gorillas (Bradley et al. 2004).

ancestral hominids were the hunted rather than the hunters, as with the Australopithecines. Second, Cashdan’s wide-ranging survey (2001) found no correlation between ethnocentrism and xenophobia, undermining Wilson’s claims for universality. More specifically, Wrangham (1999) has argued for continuity of a territorial instinct from the common ancestor of common chimpanzees and humans. He argues that a territorial instinct exists in modern common chimpanzees, with the patrolling by young male chimpanzee of territorial borders leading to conflicts of extermination with neighbouring groups, improving the victors’ access to resources. He argues (Wrangham and Peterson 1996: 63) that:

Considering Wrangham’s main ethnographic study, the comparison between common chimpanzee behaviour and the Yanomamö is particularly implausible. The Yanomamö do not patrol the bounds of a territory, most of the fighting is done by older men, and those they kill are often relatives in closely related villages with whom they have close trade relations during times of peace (Chagnon 1996a; Ferguson 2001).

‘modern chimpanzees are … surprisingly excellent models of our direct ancestors. It suggests that chimpanzee-like violence preceded and paved the way for human war, making modern humans the dazed survivors of a continuous, 5-million-year habit of lethal aggression.’

A more general difficulty with the territorial theory is proposed by Harrison (1993: 14-21). He suggests that, in Melanesia at least, the view that groups make war is the wrong way round, and that it is actually the process of war which enables groups to form. This could relate to the suggestion that humans have created a layer of social organization - the regional community - which is not known among chimpanzees (Layton and Barton 2001). This regional community frequently takes a territorial form in areas where resources are dense and predictable or, among horticulturalists, producing boundary defences. (However, this interpretation runs counter to Ember and Ember’s [1992] ethnographic survey, which concluded that resource unpredictability is a predictor of warfare.) As Layton and Barton note, this would mean that warfare could not be an ancestral trait. Indeed, returning to the EEA theory, warfare would appear at the very end of this time, long after the appearance of modern humans.

These chimpanzee behaviours are then compared with the territorial nature of modern American gang culture, with the link provided by comparisons with the Yanomamö of the Amazon as an example of primitive culture. There are several difficulties with this theory of a universal territorial instinct. The model is in essence derived from observations of chimpanzee behaviour in which it is assumed that chimpanzee behaviour is best interpreted in human terms; to complete the circular argument chimpanzee behaviour is assumed to be a guide to human societies. Too little consideration has been given to the question of whether a close relationship between common chimpanzee and ancestral human behaviour can reasonably be assumed.

The reproductive theory of warfare is based on analogies with common chimpanzee behaviour in which malecentred competition, over access to females, results in violence (e.g. Wrangham and Peterson 1996). However, several detailed studies of warring human societies suggest that violence is no guarantee of reproductive success (e.g. Moore 1990 on Cheyenne war chiefs, and Knauft 1987 and Kelly 2000: 20-35 on the !Kung, Mbuti, Central Eskimo, Semai, Hadza and Gebusi). The most famous claim for a link between warfare and reproductive success is Chagnon on the Yanomamö (1988), but his single set of figures cannot demonstrate that killers are more successful in reproducing over the lifetime (see Thorpe 2003 for further discussion). Moreover, the Yanomamö provide evidence against the general theory. Chagnon notes (1988: 985) that in a Yanomamö revenge attack the aggressors ‘always hope to dispatch the original killer’, while other ethnographers have recorded the deaths of a number of those who are multiple killers while they were still fertile (Albert 1990). If this is a more general pattern then killers could indeed be less successful over the lifetime. It is also worth noting that ethnographers working among other Yanomamö groups (Good and Chanoff 1991) have noted far lower levels of

Alternatively, it can be argued that choosing common chimpanzees over pygmy chimpanzees or bonobos (de Waal 1989) is a result of pre-determining the appropriate comparison. Bonobos show little sign of violence (infanticide, cannibalism or inter-group conflict), do not hunt much and are not dominated by males. Attempts to downplay the significance of bonobos suggest that they are atypical (e.g. Wrangham and Peterson 1996), the differences to common chimpanzees being explained by ecological factors, with bonobos occupying large, undisturbed, rainforests. However, as de Waal (2001) notes, if ecology is the key variable then one should expect bonobos to have a more gorilla-like social organization of a leading male and a following group of females. In addition, it is important to note that a recent survey of common chimpanzee field studies revealed ‘significant cultural variation … far more extensive than … previously documented for any animal species except humans’ (Whiten et al. 1999: 682). Thus the standardized picture of common chimpanzee violence

4

I.J.N. THORPE: THE ANCIENT ORIGINS OF WARFARE AND VIOLENCE innate. Headhunting raids required special magic, which placed the fighters in a trance-like state of dissociation and relieved them of accountability for their actions; it was supposed to make them capable of killing even their own wives and children. That is to say, the ability to kill had to be imparted by magic and ritual, and deliberately removed at the end of raids. But for so long as the magic was in effect, the capacity to kill was quite indiscriminate and turned the fighters into a dangerous menace to all other people, including their own families.’

violence, so we should not assume a uniform pattern applies. A related but rather more sophisticated approach sees warfare as the inevitable outcome of violent competition between young males striving for status and prestige (Maschner and Reedy-Maschner 1998), channelling their natural aggression outside the group. Again the Yanomamö are used as an example, but provide a poor one. As Ferguson notes (2001: 109), the vast majority of killers were over 40, and possibly none were below 25. Although some archaeological case studies quoted (Maschner and Reedy-Maschner 1998; Reedy-Maschner and Maschner 1999) do support the theory (e.g. Lambert 1997 on southern California), others such as Jebel Sahaba in the Sudan (Wendorf 1968) are far less clear-cut: here only slightly more males than females show traces of violence (see below). Other archaeological case studies could be used to argue a quite different case. For the Arikara of the Great Plains of North America (Holliman 2000), for example, of 19 identifiable cases of traumatic injury among males where the age at death could be established, only one was below 30, and the two individuals with projectile points embedded in bone were both over 35.

As a strategy for increasing reproductive success this is implausible, and suggests that any biological urging to kill had to be heavily reinforced by cultural methods. In the Americas, even the famously aggressive Yanomamö use drugs to work themselves up into a suitable state to fight chest-pounding duels against members of neighbouring villages (Chagnon 1990). These duels sometimes result in fatalities, which then precipitate a cycle of village raiding. The Avatip, among other New Guinea societies, socialize strongly for warfare (Harrison 1993: 23-6), as do many groups in North and South America (Carneiro 1994; Clastres 1994: 172-3). The Sambia of New Guinea practice a rather severe form of such training (Herdt 1987): warriors are removed from their families in boyhood to be raised in an all-male setting. Although details differ between societies, the groups concerned clearly believed that warriors had to be created by human intervention. It is this kind of evidence Malinowski had in mind when he argued (1941: 23) ‘human beings fight not because they are biologically impelled but because they are culturally induced’.

Although status and prestige are clearly major factors in the creation of warriors (Clastres 1994: 169-200), just as with the reproductive theory, however, dubbing violence the business of men (e.g. Daly and Wilson 1994; Gilbert 1994; Van der Dennen 1995) is an unacceptable generalization (Goldstein 2001). It ignores the considerable ethnographic and historical evidence, from Africa (Alpern 1998), Asia (Rolle 1989; Guliaev 2003) and the Americas (Koehler 1997; Hollimon 2001), of female warriors and even female war chiefs. To comment that ‘women’s participation in warfare is rare’ (Maschner and Reedy-Maschner 1998: 23) simply sidesteps the issue of variation, while Van der Dennen’s assumption that ‘it is universally the males who are the warriors’ (1995: 593) is fundamentally in error.

The other fundamental problem for the sociobiology/evolutionary anthropology school comes from the archaeological record. As many critics have pointed out (e.g. Knauft 1991; Foley 1996) few archaeologists accept the assumption of an unchanging environment of evolutionary adaptation over millions of years until the advent of agriculture. Given this, there is no reason to believe that the pattern of conflict and warfare will not have varied both geographically and chronologically. This possibility is examined below for the Palaeolithic and Mesolithic. In particular there is no room in this model for the dramatic changes believed by many (e.g. Klein and Edgar 2002) to characterize the Upper Palaeolithic ‘revolution’.

A more general problem for all three theories considered under the sociobiology/evolutionary anthropology umbrella is copious evidence of the need to produce an altered mental state before taking part in warfare. Specific preparations that have been recorded ethnographically include fasting, possession by spirits, dances, special costumes, vows, rehearsals and drugtaking (Kennedy 1971; Ehrenreich 1997). The dangers of the altered mental state thus created are noted by Harrison (1993: 27) for the Avatip of New Guinea. Avatip men were head-hunters in the recent past and used headhunting as a form of status competition, but they do not believe themselves to be born violent:

The other main camp within anthropology is materialism. Materialists start from the standpoint that warfare has to be understood in terms of material advantages which may accrue from successful warring: one would only risk one’s life in combat when there was a real need for land or, more immediately, food (e.g. Ferguson 1990). Ferguson (1990) argues that motivations as stated by the participants hide the real motives of achieving basic

‘Although [Avatip] men attach a positive value to aggression, they do not conceive of human nature as inherently violent, or assume that homicidal aggression is in some sense natural or 5

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY authority without fighting, or was there actually a decline in conflict through time?

material goals. Even in the classic ethnographic area of New Guinean warfare, however, recent analyses suggest that there is no simple relationship between land shortage and warfare, with some of the most warlike societies having fairly low population densities (Knauft 1990). In the case of the Waorani of Ecuador (Robarchek and Robarchek 1998), with very high levels of violence recorded, there are no apparent shortages of resources, but a historical context of war against Inca and later Spanish expansion during which violence can be said to have become institutionalized and in which all deaths are interpreted as the result of the evil actions of others, whether these be war or witchcraft.

Finally, an approach based on historical contingency would reject any of the unifying theories presented above in favour of the examination of the particular circumstances of each conflict, and indeed, of each example of lack of conflict (e.g. Keeley 1996; Otterbein 1997; Robarchek and Robarchek 1998; Guilaine and Zammit 2001). In other models, certainly, a lack of warfare is difficult to explain. In a general sense, the greater the degree of variability observed, both within a single society and between different societies, the more difficult it is to fit all warfare into any overarching structure, an issue which clearly troubles Dawson (1999: 97-8). The approach taken here will therefore be similar to the regionalized and comparative anthropology of war advocated by Brandt (2002: 19-31).

For Ferguson (1995) the Yanomamö fight for material resources, specifically the steel axes with which more land could be cleared. However, ethnographers of the Yanomamö do not recognize this picture (e.g. Peters 1998: 216), while the notion of Yanomamö groups controlling trade routes seems particularly far-fetched (Chagnon 1996b). In general, there seems to be no good reason except a belief that warfare must be a rational undertaking to ignore the reasons given by groups themselves as to why they fight (Thorpe 2003). Of course, this is not to deny that in some cases the reasons given are materialistic, e.g. over salmon runs on the Pacific coast of North America. The point is that material considerations are not convincing as a total or sole explanation.

Archaeology and the origins of conflict and warfare Turning to the Palaeolithic and Mesolithic, there are three main areas of possible evidence for conflict – the existence of weapons, depictions of warfare and skeletal remains demonstrating traumatic injuries. Weapons have often been presented as the most straightforward category, but here we run up against the fraught issue of symbolism. Although past generations of archaeologists have blithely dubbed various types of prehistoric artefacts as weapons – battle-axes, daggers etc. - this provides no guarantee that they were actually used as such. On the other hand, axes (usually seen as everyday tools of forest clearance if small, or prestige objects if large) were certainly used as weapons. Chapman’s (1999) coinage of the term tool-weapons seems very appropriate here. The presence of possible weapons in itself says little until we reach the Bronze Age and the production of weapons proper such as swords.

Perhaps best located here as another rationalizing theory, although clothed in evolutionary terms, is Gat’s (1999) first strike theory. He argues that technology allowed the development in humans of a capacity to strike first if a cost/benefit analysis suggests this is sensible. Not only is it difficult to see this in the light of natural selection operating on the individual, neither does it take account of long-distance raids such as those engaged in by the Jívaro of the Amazon (Descola 1993) and many other groups. Nor does it account for the tendency of headmen, such as those of the Yanomamö (Chagnon 1997: 198), to be cajoled into leading raids by virtue of their position, lest their authority be undermined.

Levantine Spanish rock art (Beltrán 1982) is often presented as the clearest body of evidence for conflict in the Mesolithic (e.g. O’Connell 1995; Nash 2000). It is now widely argued, however, that the commonly suggested date of the Mesolithic is mistaken and that Levantine art is actually Neolithic (Fairén 2004). Upper Palaeolithic representations of possible violence are considered below.

A recent reappearance in warfare studies is cultural evolution (Dawson 1996; 1999; 2001), which harks back to 1960s notions of a simple division of prehistoric societies into bands, tribes, chiefdoms and states. For Dawson a key development was agriculture, although he does propose that in prehistory the pattern of warfare might have been cyclical rather than linear. Certainly no clear line of development can be traced in Britain, where the best evidence for warfare occurs in the Early Neolithic with the attacks on defended enclosures and the deaths by arrowshot of their defenders (Mercer 1999). The Early Bronze Age, which traditionally sees the emergence of warrior aristocrats ruling over chiefdoms, fails to provide examples of defended settlements and only a handful of cases of skeletal trauma (Thorpe in press). So did warrior aristocracies maintain their

For the Lower and Middle Palaeolithic, skeletal evidence is our only source. In the Lower Palaeolithic individual cases of healed injuries have been proposed, such as Swanscombe in Britain (Brothwell 1964) with three apparent healed depressions on the skull, Ehringsdorf in Germany (Keith 1931: 319) with a major wound penetrating the forehead and Fontéchevade in France with a hole in the skullcap (Vallois 1961). Outside Europe there is Broken Hill, Zambia, with a partially healed perforation that could be from a weapon (Keith 1928: 6

I.J.N. THORPE: THE ANCIENT ORIGINS OF WARFARE AND VIOLENCE Overall the Lower Palaeolithic evidence is strongly suggestive of conflict during this period, along with cannibalism. However, many of the older claims are unsatisfactory and re-examination taking into account the possibility of damage occurring to skulls in a cave environment is clearly required (Brothwell 1999), and if the cannibalism at sites such as Gran Dolina is gastronomic in nature, then there is no reason to assume that it results from the consumption of the enemy.

418) or more likely a carnivore bite (Tappen 1987). The earliest consistent skeletal evidence of conflict comes perhaps from Sima de los Huesos, Atapuerca, Spain, the enigmatic cave containing at least 27 human skeletons dating to c. 350,000 BC (Carbonell et al. 2003). Several skulls have signs of healed impact fractures, with Cranium 5 possessing 13. Whether the excavators interpret these all as evidence of conflict is not yet clear, as only preliminary information is available at present, although this is thought likely in the case of an adolescent with a deep but partially healed blow (Arsuaga et al. 2003).

Turning to the Neanderthal skeletal record, the same caveats broadly apply. Brothwell (1999) concludes from the frequency of traumatic injury that there might have been Neanderthal conflict, and argues that these cannot simply be accidents. Examples of traumatic injuries include the fractured vertebrae from Kebara (Palestine), the healed rib from La Chapelle-aux-Saints (France), a scar on the skull, and a broken arm and collarbone from Krapina (Croatia), a broken and healed leg from La Ferrassie (France), a broken arm from the Neander Thal site, and a damaged leg at Tabun (Palestine) (Klein 1999: 475). At Shanidar in Iraq (Trinkaus 1983: 401-23) Burial 1 had numerous signs of trauma, especially a heavy blow to the left side of the face and a withered right arm; Burial 3 had a partially healed rib injury; Burial 4 had a healed rib fracture; and Burial 5 had a large scar on the head.

Similar examples of healed depressed fractures have been claimed for five of the 11 skulls, four of them female, from Ngandong in Java (the dating of which is intensely debated – Swisher et al. 1995), along with the removal of part of the skull (Coon 1962: 391). The recently rediscovered Sambungmacan 3 cranium from Indonesia also has a possible healed depression (Márquez et al. 2001). At Choukoutien in China four of the 14 Homo erectus specimens, both females and males, have been reported as having depressed fractures of the cranium (Weidenrich 1943; Courville 1950). However, as Roper (1969) notes, Weidenrich states (1943: 190) that the human bones had been washed into their present location together with boulders and pebbles, so the injuries could be accidental. Unfortunately the material has now disappeared, so re-examination is impossible. Atapuerca also provides the oldest possible case of cannibalism: the Gran Dolina cave has produced the remains of six individuals (Fernández-Jalvo et al. 1999) identified as the victims of cannibalism on the following grounds: the use of analogous butchering techniques in humans and animals; similar breakage patterns to extract the marrow; and identical patterns of discard for human and animal bone. However, the Gran Dolina case is argued to represent gastronomic cannibalism – the consumption of members of the group after their death – rather than conflict. Certainly at Gran Dolina the vertical spread of human bones suggests occasional consumption over a long period rather than a massacre. The Homo erectus cranium from Bodo in Ethiopia has ancient cutmarks on it that seem to be evidence of defleshing, so the Gran Dolina case is not entirely alone (White 1986).

There thus appears to be a good case supporting Brothwell for a high level of conflict, if not warfare, among the Neanderthals. But other analysts have concluded that Neanderthals suffered a higher level of trauma than other hominids - comparable in intensity and location to modern rodeo riders - because of their practice of short-range hunting with spears (Berger and Trinkaus 1995). This could also be the cause of the spear-like injury which perforated the lung of Shanidar 3 (Trinkaus 1983: 414-15). Again there have been dubious claims of violence, such as the Grotta Guattari (Monte Circeo) skull from Italy, with supposed lethal wounds (Blanc 1961) that have been reidentified as hyena damage (Stiner 1991; White and Toth 1991). The most recent claim of Neanderthal violence comes from reanalysis of the young adult, possibly male, from St. Césaire, France (Zollikofer et al. 2002). A healed skull fracture is argued to be the result of an attack.

Even earlier cut-marks have been found on the Sterkfontein early hominid from South Africa (Pickering et al. 2000), which may be c. two million years old, relating to the removal of the mandible from the cranium. Removal of parts of the skull from Arago in France may relate to consumption of the brain (de Lumley and de Lumley 1974). Other claims of cannibalism, such as have been made for the finds from Choukoutien, have been reconsidered and found unconvincing. At Choukoutien, the fragmentary nature of the human skeletons and the missing parts of skulls are probably the result of destruction by hyenas (Binford and Ho 1985).

Cannibalism has also been suggested quite frequently for Neanderthals, not always on a sound basis. Cut-marks on the skull of the Engis child from Belgium are now believed to be marks made by earlier researchers preparing and measuring it (White and Toth 1989). Hundreds of fragmentary bones of Neanderthals come from the Krapina rock shelter, some of them cut-marked, have been considered in terms of cannibalism (Trinkaus 1985). The specific pattern of cut-marks has, however, been argued to be a better fit with an interpretation of defleshing as part of mortuary ritual (Russell 1987). The best case seems to be Moula-Guercy in France (Defleur et

7

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY al. 1999), but this is believed to represent gastronomic cannibalism, so not related to war.

do suggest a move towards the use of projectiles as weapons (Crosby 2002).

It is not, however, so clear that a strong distinction can be made between the Neanderthal pattern of traumatic injuries and that of early anatomically modern humans, if we distinguish them from the people of the Upper Palaeolithic, after c. 40,000 BC. Similar evidence for a skull injury from Klasies river, South Africa, dates to 90,000 years ago (Deacon and Deacon 1999: 103). From Palestine the Skhnjl IX adult (McCown and Keith 1939: 373) has a spear-like injury in the left hip and the child from the cave has a fractured skull (McCown and Keith 1939: 309); the Qafzeh 11 adolescent (Tillier 1995) has a scar at the front of the skull. These injuries could certainly be explained in the same terms as those found on the Neanderthal skeletons.

Klein argues that, in comparison to the Middle Palaeolithic, ‘skeletal evidence for deliberate injury is … rare’ (1999: 555). He notes healed fractures from male skulls at Chancelade in France (itself doubted by Sollas [1927] after his examination of the skull, preferring an interpretation of an accidental fall) and Dolní VČstonice in Moravia and the discredited cases of skull injuries from Cro-Magnon in France and Choukoutien in China. (A number of old accounts of skull trauma collected by Courville [1950] may be similarly suspect.) This is something of an understatement, however, as examination of the isolated human remains (i.e. not from formal burials) from Dolní VČstonice Site II has produced two partial skulls with healed depressions (Trinkaus et al. 2000) in addition to two complete skeletons with similar evidence of trauma (Vlþek 1995). This does, however, seem to make Dolní VČstonice rather an exception, pointing to the pattern of variability seen much more clearly in the Mesolithic (see below).

Considering the Upper Palaeolithic, and thus only anatomically modern humans from now on, the first crucial points to note are the significantly larger number of skeletal remains surviving for analysis and the development of burial outside caves, with a consequently less complex variety of possible posthumous fates for the skeletal material.

The most dramatic revelations concerning conflict at the end of the Palaeolithic come from Jebel Sahaba in the Sudan (Wendorf 1968), a cemetery overlooking the Nile with a radiocarbon date from one of the burials of c. 13,500 BC (Wendorf and Schild in press). Of 59 burials, 24 had chert projectile points either embedded in the bones or found within the grave fill. The 110 chert points excavated were ‘almost all in positions which indicate they had penetrated the body either as points or barbs on projectiles or spears’ (Wendorf 1968: 959). Slightly more males than females show traces of violence, and several children were also killed by projectiles, not on the face of it an argument for war being the business of men. Environmental pressure is suggested by Wendorf (1968: 993; Wendorf and Schild in press) as the cause of the violence, but this could be a special purpose burial place for those suffering a ‘bad death’ (Thorpe 2003). Certainly the extremely high level of violent death suggests that Jebel Sahaba is an exceptional site which stands out against a generally low level of traumatic injury for the Upper Palaeolithic.

There are also additional lines of evidence that may be pursued for the Upper Palaeolithic: the development of weaponry and possible depictions of violence in cave paintings and engravings. The earliest bows probably appeared in the Late Upper Palaeolithic, after c. 20,000 BC, but definitive evidence in the form of arrows or bows is missing. (Projectile injuries are considered below.) There are some seven depictions of anthropomorphs apparently pierced by projectiles (perhaps spears) recorded in European cave art (Guilaine and Zammit 2001: 85-9): whether they depict actual killings, wishedfor killings or magical killings cannot easily be determined. Indeed, Lewis-Williams (2002: 277-81) recently argued that the spears represent the pricking sensation felt by a shaman when entering a trance. In terms of skeletal evidence, there are two final Upper Palaeolithic bodies with flint points lodged in them, both from Italy (Bachechi et al. 1997). One, from San Teodoro cave in Sicily, is that of a woman with a flint point in her pelvis. The other is a child with a flint point in its backbone, found in the Grotta dei Fanciulli, Grimaldi, on the Italian mainland (Henry-Gambier 2001). Whether the fatal points were spear-tips or arrowheads is unclear, but the excavators in both cases thought they were probably arrows. In the less definitive case of Wadi Kubbaniya, Egypt (Wendorf and Schild 1986), two bladelets found between the ribs and backbone might have been attached to a projectile and the cause of death. The poorly dated cave site of Montfort Saint-Lizier, where a quartzite blade was embedded in a human vertebra (Bégouën et al. 1922), may also belong here, although it is equally likely to date to the Early Mesolithic. Although they are few in number, these cases

Considering the Mesolithic, the volume of the evidence requires a regional approach to be taken. A considerable body of literature exists on conflict and war among the gatherer-hunter groups of North America, many of whom fought against the expanding settler states from the sixteenth century onwards. Some of the earliest skeletons to be recovered from North America (dating to c. 10,0008000 BC) have produced clear evidence of violent lives and death. Kennewick Man from Washington State (Chatters 2001) had a dent in the forehead, a healing injury on the temple, several fractured ribs, a chipped shoulder socket and a stone projectile point embedded in the pelvis. The teenage male from Grimes Point in Nevada had two stab wounds in the ribs; Spirit Cave Man from Nevada had a skull fracture, two ruptured discs in 8

I.J.N. THORPE: THE ANCIENT ORIGINS OF WARFARE AND VIOLENCE evidence of high rates of skeletal trauma among maritime societies. Burials from a cemetery of the Chinchorro culture of Chile, dating to c. 2000 BC, demonstrate high levels of male skull fractures (probably from thrown cobblestones), amounting to one in three males, together with some female examples, and almost no trace of violence directed at sub-adults (Standen and Arriaza 2000). Traumatic injuries resulting from conflict, however, have not been reported from other early skeletons in South America (Lessa and Guidon 2002), so the historical background is rather different.

his spine and hand injuries; ‘Stick Man’ from Washington State had a minor dent on his skull, as did Marmes III Man from Washington State; Horn Shelter Man from Texas had foot and collarbone injuries (ibid.: 208). Altogether six of the 12 known early males (although the Grimes Point skeleton may be that of a female) from North America produced evidence of conflict. Although evidence from a scatter of contexts is easy to over-interpret, the level of violence certainly suggests conflict. From a slightly later date (c. 6000 BC) the Windover Pond cemetery in Florida (Doran 2002) contained a man with a shattered eye orbit and a ‘parry’ fracture of the arm (Dickel et al. 1989), another man with a sharpened antler point embedded in the pelvis, whose hand had been removed, and five individuals (both adults and children) with depressed fractures of the skull. A level of traumatic injury is present in eastern North America from the oldest skeletal finds, but a significant increase in the evidence for conflict occurs in skeletal remains dating to Middle to Late Archaic period, the time of the development of shell middens (Milner 1999). Projectile injuries become common, with graves cut into midden deposits sometimes containing more than one body killed by arrowshot. Archaic period (6000-500 BC) skeletal remains from western Tennessee, mostly of men, show lethal projectile point injuries, cut-marks and missing bones (possible forearm-trophy taking; Smith 1997). By contrast, on the Great Plains of Canada and the United States, there is almost no evidence for conflict before AD 950 and the formation of villages (Owsley 1994).

In Australia both rock art and skeletal remains have been used to consider the prevalence of conflict. Taçon and Chippindale (1994) have argued that the rock art of Arnhem Land in the Northern Territory demonstrates the emergence of conflict as a theme around 8000 BC, with depictions of two opposed armed figures. Then about 4000 BC large groups of warriors (up to 110 altogether) are painted throughout western Arnhem Land showing two groups in conflict, with spears flying in volleys, speared figures and warriors waving hooked sticks and stone axes. Taçon and Chippindale conclude that the change in the depiction of conflict reflects the appearance of a more complex social organization and wars over resources. In terms of skeletal material, there are examples of traumatic injury from the late Pleistocene, including the Burke’s Bridge skull with a spear injury to the skull (Webb 1995: 74-5), and the Roonka and Willandra Lakes sites with skeletons exhibiting ‘parry’ fractures and depressed fractures of the skull (Pretty and Kricun 1989; Webb 1995: 54-6). Later skeletal finds from South Australia, of uncertain date, show a large number of ‘parry’ fractures and common cranial trauma in the form of circular depressions, some very deep (Webb 1995: 188-216). In general females have a higher proportion of skull injuries than males, although there are clear regional differences. In the tropics the percentage of males with traumatic skull injuries is lowest, and the proportion of these which are injuries to the back of the head is highest – perhaps indicating that these result from domestic disputes and that the level of conflict between groups was very low. Again here, as in the Americas, the pattern is of a fairly consistent low level of violence with specific areas showing much higher levels of conflict and probably warfare.

In British Columbia, Canada, there is a high level of trauma in the period 3500-1500 BC, with six of 57 skeletons showing signs of violence, mostly in the form of healed injuries, but also one projectile wound and possible examples of beheading (Cybulski 1994). Studies of skeletal material from the northern Channel Islands of California (Lambert 1997), especially from shell middens, show that small healed fractures of the skull are present in all time periods, even before evidence for lethal injuries caused by projectile points appear around 4000 BC. So perhaps a level of social tension existed throughout (either within or between communities) which slipped into actual warfare at times. Lower levels of traumatic injury (but still high when compared to other collections), mostly cranial, and some lethal projectile injuries are also reported from central California (Jurmain 2001). In the far North, analysis of the skeletal material from the Saunaktuk site in the Canadian Arctic suggests a mass killing of at least 35 Inuit by Amerindians around AD 1400 (Melbye and Fairgreave 1994), matching oral traditions of a massacre of children, women and the elderly by raiders. There are clearly areas where raiding and warfare are distinct possibilities, but across the area as a whole there are also significant differences.

The absence of direct evidence of conflict is especially striking for the Natufian of Palestine (Roper 1975; Peterson 2002: 137), where there is a large skeletal sample (over 400 inhumations, most from large cemeteries), yet very low levels of identifiable violence. Only a handful of traumatic injuries have been reported: a depressed fractures of the skull at Nahel Oren (Ferembach 1959) and a mandible with new bone growth and a possible healed leg fracture at Hayonim cave (Peterson 2002: 83). The depression in the skull of a probable male from El Wad appears to have been

Less study has been carried out on the gatherer-hunters of South America, but once again there is convincing 9

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY identified to be both the result of traumatic injury (BelferCohen 1995) and a product of artificial deformation of the head (Belfer-Cohen et al. 1991). Population pressure has often been suggested as a factor in the Natufian development of an agricultural economy (see Thorpe 1996: 12-13 for a discussion) and yet there is no trace in the archaeological record that this led to conflict.

400 skeletons, and roughly ten individuals with fractures. These two sites thus provide the vast bulk of cases from just one-fifth of the total burial population that has been examined, suggesting that dramatic cultural variability existed in the level of conflict within this small part of Europe. Even at Vlasac the phases before and after Period II lack evidence of violence (ibid.).

Turning to Europe, and beginning with the Mediterranean, remarkably little trace of violence can be detected from the Mesolithic skeletal evidence (Cordier 1990; Vencl 1991; Grünberg 2000). For the central and eastern Mediterranean the relatively small sample size (Grünberg 2000) is certainly a limiting factor, with just a single example of violent death from Franchthi cave, Greece (Cullen 1995). It may be significant that relatively high levels of healed cranial trauma do appear at Neolithic sites in Greece, such as Alepotrypa cave (Papathanasiou et al. 2000).

To the east, projectile injuries apparently bringing about the death of the victims (Cordier 1990; Vencl 1991; 1999; Alekšin 1994; Lillie 2001) are reported from Volos´ke (two cases, one also with a fractured skull) and Vasylivka I in the Ukraine; also in the Ukraine the Vasylivka III cemetery produced two burials with arrow injuries, another with a bone point wedged between vertebrae and several with apparently crushed skulls. Lillie (2001: 56) suggests that these ‘Ukrainian cemeteries functioned as territorial markers, establishing a group’s rights to regional hunting grounds’. These hunting grounds could, therefore, also have become the focus of conflicts over such claimed rights.

The limitation of a small sample size cannot be said to be a factor in the western Mediterranean round to the Atlantic coast, where at least 400 burials have been found in Portuguese shell middens, at sites such as Moita do Sebastião (Lubell et al. 1989; Cunha et al. 2003). A projectile wound in the foot at Moita do Sebastião and a broken arm and a skull wound at Cabeço da Arruda are the only clear traces of violence here. There appears to be little evidence of traumatic injury from elsewhere in Iberia (Cabel and Garralda 1995). Parry fractures have been reported in some numbers from Portuguese sites (Grünberg 2000: 190) – these fractures of the forearm are usually interpreted as the result of an attempt to fend off a blow directed at the head or upper body. However, recent examinations of prehistoric Californian skeletal material (see above) show there to be no link at the level of the individual site between the frequencies of head injuries thought to result from attacks and of parry fractures (Larsen 1997: 112). There are, in fact, a large number of accidents that can result in ‘parry’ fractures (Lovell 1997). Along the Mediterranean and up into the Atlantic, therefore, there appears to be a quite consistent pattern of low levels of identifiable violence.

In central Europe we have the oldest definite evidence for arrows and thus bows, at Stellmoor in Germany, c. 8500 BC (Guilaine and Zammit 2001: 97-103). Skeletal material from Germany also points to the existence of conflicts occurring on a much larger scale. At Ofnet cave in Bavaria two pits held the skulls and vertebrae of 38 individuals, all stained with red ochre, dating to around 6500 BC (Frayer 1997; Orschiedt 1998; 1999). Most of the dead were children; two-thirds of the adults were females. Finds of deer teeth and shells, presumed to be decoration, were associated only with adult females and children. Half the individuals were wounded before death by blunt mace-like weapons, with males and females and children (even infants) all injured, but males having the most wounds (Frayer 1997). There were cutmarks on the vertebrae of one in three of all individuals, relating to the removal of the head. The scale of the apparent massacre suggests an attempt to wipe out a whole community, followed by the ceremonial burial of ‘trophy skulls’ (Keeley 1996: 102). Certainly, there are many accounts in the ethnographic record which demonstrate that skulls taken in warfare could be very carefully curated (e.g. Sterpin 1993). It should be noted, however, that Orschiedt (1998; 1999) has lower injury counts than Frayer, although this may well relate to a smaller sample being available to him. Orschiedt suggests that not all the skulls were deposited at the same time, indeed possibly over a long time, although the skulls he does accept as definitely injured are found together. The chronological question can only be resolved by further radiocarbon dating.

Moving away from the Mediterranean, projectile injuries that were probably the cause of death are known from Schela Cladovei and Vlasac Period II in the Iron Gates area on the Danube (Radovanovic 1996; Chapman 1999). At Schela Cladovei the level of violence appears to be very high among the 56 skeletons excavated, with six cases of projectile injuries (four male, one female, one unsexed, mostly from bone points) along with half a dozen examples of cranial injuries (generally not healed before death), so that some one-third of all adults from the site suffered fatal traumatic injuries. Within the region’s Mesolithic cemeteries as a whole, however, Schela Cladovei and Vlasac appear to be exceptional (Chapman 1999), with Lepenski Vir, for example, providing no cases of trauma from 85 skeletons – the latest overall figures are eight projectile injuries out of

In addition to the Ofnet site, the nearby cave site of Hohlenstein-Stadel (Orschiedt 1998; 1999) produced the skulls of a man, a woman and a young child (18 months to two years old) in a pit, resting on a layer of stones. Fish teeth found around the female skull may represent a 10

I.J.N. THORPE: THE ANCIENT ORIGINS OF WARFARE AND VIOLENCE necklace. All three skulls had wounds from a blunt instrument and cut-marks demonstrated that these individuals had also been decapitated. The diagnosis of hydrocephaly (Orschiedt 1998) for the infant has suggested an interpretation of the killing of socially unacceptable members of the group (Gronenbourn 1999). Despite these apparently intense episodes of violence, we should also note that none of the conditions of sedentism, territoriality and status competition attributed to the Mesolithic elsewhere in Europe have been claimed for the German situation (Jochim 1998).

form of fractures and impressions, but a very low level of fractures to other parts of the body, again suggesting the lack of a linkage between parry fractures and other injuries. Most of the individuals with injuries were male, with Fischer (2002: 372) concluding that ‘nearly half of the male skulls exhibit lesions that seem to have been caused by blows with an axe or some similar weapon’, but with some examples of female head wounds (see Thorpe 2003 for specific examples for both sexes). This pattern may be the result of fighting duels with heavy wooden clubs.

At the Téviec and Hoèdic shell midden sites on the coast of Brittany (Péquart et al. 1937; Péquart and Péquart 1954), men, women and children were interred in stonelined cists. Twenty-three burials were found at Téviec and 14 at Hoèdic. Several skeletons had healed fractures of the forearm and collarbone and at Téviec a young adult male burial had two flint points embedded in his spine (Péquart 1931), and a healed fracture of the jaw, while another young adult (possibly male) had traces of facial injuries and a partially healed hole in the skull (Newell et al. 1979: 132-7).

At Dyrholmen in Jutland a large number of bones from at least nine individuals were discovered (Degerbøhl 1942). There are traces of cut-marks and fractures of long bones and mandibles apparently to reach the marrow, with cutmarks on the skull of a ten-year-old child suggesting scalping (Anger and Dieck 1978: 166-7). Claims of cannibalism may be exaggerated but the deliberate treatment of human bones as if they came from animals seems to be the pattern here. Cannibalism is also sometimes used in South American societies as a way of disrespecting the enemy, eating their flesh ‘like animal meat’ (Conklin 1995).

Finally reaching southern Scandinavia, the overwhelming impression is of a significantly higher level of conflict visible in the archaeological record than in the areas considered before. In Sweden several sites have produced evidence of fatal weapon injuries (Strassburg 2000: 162-5). Bone points were found in the chests of burials at Bäckaskog (an adult female) and Stora Bjärs (an adult male) (Albrethsen and Brinch Petersen 1976). The Stora Bjärs man also had a partially healed head injury and a fresh wound to the jaw (Lindqvist and Possnert 1999). At the Skateholm I ‘cemetery’ in southern Sweden an arrowhead was lodged in the pelvic bone of an adult male who had also suffered blows to the legs and arms (Larsson 1989), and a bone point found with another male (Vencl 1991) may be the cause of death; three further individuals had healed skull injuries and another a broken and healed jaw (Persson and Persson 1984; 1988). The Skateholm II ‘cemetery’ contained the burial of a woman with the tip of an arrowhead in the chest cavity (Persson and Persson 1988) and one individual with a healed cranial depression (Persson and Persson 1984). At Tågerup (Karsten and Knarrström 2003: 204) an arrowhead with a broken tip was found at the hip of a ten-year-old child.

The overall conclusion of this review should be clear: the variation revealed between regions, areas and sites is significant in assessing the value of general theories, and a generalized approach masks crucial cultural variability. Despite the limitations of sample size in many areas, and the lack of collections of data beyond the anecdotal, there seem to be changes through space and time and specific areas and periods (and even single episodes) of greatly elevated levels of violence. It should be clear therefore that the Palaeolithic and Mesolithic evidence cannot be fitted neatly into any one of the over-arching explanations considered at the beginning of this paper. The biological theories (unless they admit a considerable degree of cultural influence) imply a constant level of violence not supported by the archaeological evidence, which demonstrates significant variations in evidence for conflict from virtually none to apparent massacres. The materialist theory would be far more convincing if there were a significant increase in conflict with the adoption of agriculture. We need, therefore, to turn to historical factors and considerations of the specific societies concerned. Mesolithic conflict need not have been over economic resources (as Haas [2001: 338] notes), but a strong degree of internal territoriality would certainly be consistent with other indications (Layton and Barton 2001), and seems to fit well with the evidence of elevated levels of conflict in coastal areas where areas of shellbeds were an important resource and shell middens in some way related to attempts to control such resources.

In Denmark, at the Vedbæk ‘cemetery’ on Zealand one of the individuals (an adult, probably male [Bennicke 1985: 102] although Meiklejohn et al. 2000 argue for a female attribution) in a grave containing three bodies had a bone point through the throat (Albrethsen and Brinch Petersen 1976). The apparently simultaneous burial of the man, woman and child has led to the suggestion that all three died suddenly and violently (ibid.).

However, many other causes of wars among gathererhunters have been noted in the ethnographic record. Warfare in early prehistory might well often have arisen

Bennicke’s examination of trauma (1985: 98-101) shows that there were a high number of cranial injuries in the 11

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Belfer-Cohen, A., Schepartz, L.A. and Arensburg, B. 1991. New biological data for the Natufian populations in Israel. In O. Bar-Yosef and F.R. Valla (eds) The Natufian Culture in the Levant. Ann Arbor: International Monographs in Prehistory 1. 411-24. Beltrán, A. 1982. Rock Art of the Spanish Levant. Cambridge: Cambridge University Press. Bennicke, P. 1985. Palaeopathology of Danish Skeletons. København: Akademisk Forlag. Berger, T.D. and Trinkaus, E. 1995. Patterns of trauma among the Neanderthals. Journal of Archaeological Science 22: 841-52. Binford, L.R. and Ho, C.K. 1985. Taphonomy at a distance: Zhoukoudien, ‘the cave home of Beijing Man’? Current Anthropology 26: 413-42. Blanc, M.H. 1961. Some evidences for the ideologies of early man. In S.L. Washburn (ed.) Social Life of Early Man. New York: Viking Fund Publications in Anthropology 31. 119-36. Bradley, B.J., Doran-Sheehy, D.M., Lukas, D., Boesch, C. and Vigilant, L. 2004. Dispersed male networks in western gorillas. Current Biology 14: 510-3. Brain, C.K. 1981. The Hunters or the Hunted? An Introduction to African Cave Taphonomy. Chicago: Chicago University Press. Brandt, E. 2002. On War and Anthropology: a history of debates concerning the New Guinea Highlands and the Balkans. Amsterdam: Rozenburg. Brookes, M. 1999. Apocalypse Then. New Scientist 14/8/99: 32-5. Brothwell, D.R. 1964. Further comments on the right parietal from Swanscombe: anomalies and endocranial features. In C.D. Ovey (ed.) The Swanscombe Skull: a survey of research on a Pleistocene site. London: Royal Anthropological Institute, Occasional Paper 20. 173-4. Brothwell, D.R. 1999. Biosocial and bioarchaeological aspects of conflict and warfare. In J. Carman and A. Harding (eds) Ancient Warfare: archaeological perspectives. Stroud: Sutton. 25-38. Cabel, P.A. and Garralda, M.D. 1995. Les sépultures epipaléolithiques de la Cueva de los Canes (Asturies, Espagne). In M. Otte (ed.) Nature et Culture. Liège: Etudes et Recherches Archéologiques de l’Université de Liège 68. 871-97. Carbonell, E., Mosquera, M., Ollé, A., Rodríguez, X.P., Sala, R., Vergès, J.M., Arsuaga, J.L. and Bermúdez de Castro, J.M. Les premiers comportements funéraires auraient-iles pris place à Atapuerca, il y a 350 000 ans? L’Anthropologie 107: 1-14. Carman, J. and Harding, A. (eds) 1999. Ancient Warfare: archaeological perspectives. Stroud: Sutton. Carneiro, R.L. 1994. War and peace: alternating realities in human history. In S.P. Reyna and R.E. Downs (eds) Studying War: anthropological perspectives. Langhorne: Gordon and Breach. 3-27. Cartmill, M. 1993. A View to a Death in the Morning: hunting and nature through history. London: Harvard University Press.

from matters of personal honour – such as slights, insults, marriages going wrong, or theft. In a small community, everyone is related to each other. An attack on one group member may well therefore be treated as an attack on the whole family. A personal feud may quickly involve the whole community. From there it is a small step to war. Acknowledgements I wish to thank those who have discussed the topic or who have supplied me with information during the writing of this paper, especially Dr Jörg Orschiedt, Neanderthal Museum, Germany, Dr Chris Meiklejohn, University of Winnipeg, Professor Keith Otterbein, State University of New York, Sheila Coulsen, University of Oslo, and Per Karsten, Swedish National Heritage Board. I also heard with profit an important paper by Kath McSweeney on evidence for violence in the Iron Gates area at the European Association of Archaeologists meeting in Lisbon, September 2000. Finally, I should like to thank Mike Parker Pearson for waiting patiently for this paper to be written. Bibliography Albert, B. 1990. On Yanomami warfare: rejoinder. Current Anthropology 31: 558-63. Albrethsen, S.E. and Brinch Petersen, E. 1976. Excavation of a Mesolithic cemetery at Vedbæk, Denmark. Acta Archaeologica 47: 1-28. Alekšin, V.A. 1994. Mesolithische Gräberfelder der Ukraine (chronologische, kulturelle und soziologische Aspekte der Interpretation). Zeitschrift fur Archaologie 28: 163-89. Alpern, S.B. 1998. Amazons of Black Sparta: the women warriors of Dahomey. New York: New York University Press. Anger, S. and Dieck, A. 1978. Skalpieren in Europa seit dem Neolithikum bis um 1767 nach Chr. Bonner Hefte zur Vorgeschichte 17: 153-240. Ardrey, R. 1961. African Genesis. London: Collins. Ardrey, R. 1967. The Territorial Imperative. London: Collins. Arsuaga, J.L., Carbonell, E. and Bermúdez de Castro, J.M. 2003. The First Europeans: treasures from the hills of Atapuerca. New York: American Museum of Natural History. Bachechi, L., Fabri, P-F. and Mallegni, F. 1997. An arrow-caused lesion in a Late Upper Palaeolithic human pelvis. Current Anthropology 38: 134-40. Bamford, H.N. 1985. Briar Hill. Northampton: Northamptonshire Development Corporation Archaeological Monograph 3. Bégouën, H., Cugulieres, [?] and Miquel, H. 1922. Vertèbre humaine traversée par une lame en quartzite. Revue Anthropologique 32: 230-2. Belfer-Cohen, A. 1995. Rethinking social stratification in the Natufian culture: the evidence from burials. In S. Campbell and A. Green (eds) The Archaeology of Death in the Near East. Oxford: Oxbow. 9-16. 12

I.J.N. THORPE: THE ANCIENT ORIGINS OF WARFARE AND VIOLENCE Coast. In C.S. Larsen and G.R. Milner (eds) In the Wake of Contact: biological responses to conquest. New York: Wiley-Liss. 75-85. Daly, M. and Wilson, M. 1994. Evolutionary psychology of male violence. In J. Archer (ed.) Male Violence. London: Routledge. 253-88. Dart, R. 1925. Australopithecus africanus: the man-ape of South Africa. Nature 115: 195-9. Dart, R. 1953. The predatory transition from ape to man. International Anthropological and Linguistic Review 1: 201-13. Dart, R. and Craig, D. 1959. Adventures with the Missing Link. New York: Harper. Dawson, D. 1996. The origins of war: biological and anthropological theories. History and Theory 35: 128. Dawson, D. 1999. Evolutionary theory and group selection: the question of warfare. History and Theory 38: 79-100. Dawson, D. 2001. The First Armies. London: Cassell. de Lumley, H. and de Lumley, M.A. 1974. PreNeanderthal human remains from Arago cave in southwestern France. Yearbook of Physical Anthropology 17: 162-8. de Waal, F.B.M. 1989. Peacemaking Among Primates. Cambridge MA: Harvard University Press. de Waal, F.B.M. 2001. Apes from Venus. In F.B.M. de Waal (ed.) Tree of Origin. Cambridge MA: Harvard University Press. 41-68. Deacon, H.J. and Deacon, J. 1999. Human Beginnings in South Africa. Walnut Creek: Altamira Press. Defleur, A., White, T., Valensi, P., Slimak, L. and Crégut-Bonnoure, E. 1999. Neanderthal cannibalism at Moula-Guercy, Ardèche, France. Science 286: 12831. Degerbøhl, M. 1942. Et knoglemateriale fra DyrholmenBopladsen, en ældre Stenalder-køkkenmødding. Med særlight henblik paa uroksens køns-dimorphisme og paa kannibalismen i Danmark. In T. Mathiassen, M. Degerbøhl and J. Troels-Smith (eds) Dyrholmen: en Stenalderboplads paa Djursland. København: Det Kongelige Danske Videnskabernes Selskab, Arkæologisk-Kunsthistoriske Skrifter 1. 77-135. Descola, P. 1993. Les affinités sélectives: alliance, guerre et predations dans l’ensemble jivaro. L’Homme 126128: 171-90. Dickel, D.N., Aker, C.G., Barton, B.K. and Doran, G.H. 1989. An orbital floor and ulna fracture from the early Archaic of Florida. Journal of Palaeopathology 2: 165-70. Dixon, P. 1988. The Neolithic settlements on Crickley Hill. In C. Burgess, P. Topping, C. Mordant and M. Maddison (eds) Enclosures and Defences in the Neolithic of Western Europe. Oxford: BAR International Series 403. 75-87. Doran, G.H. (ed) 2002. Windover: multidisciplinary investigations of an Early Archaic Florida cemetery. Gainesville: University Press of Florida. Drewett, P. 1994. Dr V. Seton Williams’ excavations at Combe Hill, 1962, and the role of Neolithic

Cashdan, E. 2001. Ethnocentrism and xenophobia: a cross-cultural study. Current Anthropology 42: 760-5. Chagnon, N. 1988. Life histories, blood revenge, and warfare in a tribal population. Science 239: 985-92. Chagnon, N. 1990. Reproductive and somatic conflicts of interest in the genesis of violence and warfare among tribesmen. In J. Haas (ed.) The Anthropology of War. Cambridge: Cambridge University Press. 77-104. Chagnon, N. 1996a. Chronic problems in understanding tribal violence and warfare. In G.R. Bock and J.A. Goode (eds) Genetics of Criminal and Antisocial Behavior. New York: John Wiley. 202-36. Chagnon, N. 1996b. Review of Yanomami Warfare: a political history. American Anthropologist 98: 670-2. Chagnon, N. 1997. Yanomamö. (5th edition). Fort Worth: Harcourt Brace. Chapman, J. 1999. The origins of warfare in the prehistory of central and eastern Europe. In J. Carman and A. Harding (eds) Ancient Warfare: archaeological perspectives. Stroud: Sutton. 101-42. Chatters, J.C. 2001. Ancient Encounters: Kennewick man and the first Americans. New York: Simon and Schuster. Clastres, P. 1994. Archeology of Violence. New York: Semiotext(e). Conklin, B.A. 1995. ‘Thus are our Bodies, thus was our Custom’: mortuary cannibalism in an Amazonian society. American Ethnologist 22: 75-101. Coon, C.S. 1962. The Origin of Races. New York: Alfred A. Knopf. Cordier, G. 1990. Blessures préhistoriques animales et humaines avec armes ou projectiles conservés. Bulletin de la Société Préhistorique Française 87: 462-81. Cosmides, L., Tooby, J. and Barkow, J. 1992. Introduction: evolutionary psychology and conceptual integration. In J. Barkow, L. Cosmides and J. Tooby (eds) The Adapted Mind. Oxford: Oxford University Press. 3-15. Cotlow, L. 1953. Amazon Head-Hunters. New York: Henry Holt. Courville, C.B. 1950. Cranial injuries in prehistoric man. Yearbook of Physical Anthropology 6: 185-205. Crosby, A.W. 2002. Throwing Fire: projectile technology through history. Cambridge: Cambridge University Press. Culbert, T.P. 1988. The collapse of Classic Maya civilization. In N. Yoffee and G.L. Cowgill (eds)) The Collapse of Ancient States and Civilizations. Tucson: University of Arizona Press. 69-101. Cullen, T. 1995. Mesolithic mortuary ritual at Franchthi cave, Greece. Antiquity 69: 270-89. Cunha, E., Cardoso, F. and Umbelino, C. 2003. Inferences about Mesolithic life style on the basis of anthropological data: the case of the Portuguese shell middens. In L. Larsson, H. Kindgren, K. Knutsson, D. Loeffler and A. Åkerlund (eds) Mesolithic on the Move. Oxford: Oxbow. 184-8. Cybulski, J.S. 1994. Culture change, demographic history, and health and disease on the Northwest 13

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Goldstein, S. 2001. War and Gender. Cambridge: Cambridge University Press. Good, K. and Chanoff, D. 1991. Into the Heart: one man’s pursuit of love and knowledge among the Yanomama. New York: Simon and Schuster. Gronenbourn, D. 1999. A variation on a basic theme: the transition to farming in southern central Europe. Journal of World Prehistory 13: 123-210. Grünberg, J.M. 2000. Mesolithische Bestattungen in Europa. Rahden: Verlag Marie Leidorf. Guilaine, J. and Zammit, J. 2001. Le Sentier de la Guerre. Paris: Editions du Seuil. Guliaev, V. 2003. Amazons in the Scythia: new finds at the middle Don, southern Russia. World Archaeology 35: 112-25. Haas, J. 2001. Warfare and the evolution of culture. In G.M. Feinman and T.D. Price (eds) Archaeology at the Millennium. New York: Kluwer/Plenum. 329-50. Halbmayer, E. 2001. Socio-cosmological contexts and forms of violence: war, vendetta, duels and suicide among the Yukpa of north-western Venezuela. In B.E. Schmidt and I.W. Schröder (eds) Anthropology of Violence and Conflict. London: Routledge. 49-75. Harner, M.J. 1972. The Jívaro: people of the sacred waterfalls. Garden City NY: Natural History Press. Harrison, S. 1993. The Mask of War. Manchester: Manchester University Press. Henry-Gambier, D. 2001. Le Sépulture des Enfants de Grimaldi. Paris: Comité des Travaux Scientifiques et Historiques, Documents Préhistorique 14. Herdt, G. 1987. The Sambia: ritual and gender in New Guinea. New York: McGraw-Hill. Hodder, I. 1992. The Haddenham causewayed enclosure – a hermeneutic circle. In I. Hodder (ed.) Theory and Practice in Archaeology. London: Routledge. 213-40. Holliman, S.E. 2000. Sex, health and gender roles among the Arikara of the northern Plains. In A.E. Rautman (ed.) Reading the Body. Philadelphia: University of Pennsylvania Press. 25-37. Hollimon, S. 2001. Warfare and gender in the northern Plains: osteological evidence of trauma reconsidered. In B. Arnold and N. Wicker (eds) Gender and the Archaeology of Death. Walnut Creek: Altamira Press. 179-93. Jochim, M.A. 1998 A Hunter-Gatherer Landscape: southwest Germany in the late Paleolithic and Mesolithic. New York: Plenum. Jurmain, R. 2001. Palaeoepidemiological patterns of trauma in a prehistoric population from central California. American Journal of Physical Anthropology 115: 13-23. Karsten, P. and Knarrström, B. 2003. The Tågerup Excavations. Lund: National Heritage Board. Keegan, J. 1993. A History of Warfare. London: Hutchinson. Keeley, L.H. 1996. War Before Civilization: the myth of the peaceful savage. Oxford & New York: Oxford University Press. Keith, A. 1928. The Antiquity of Man. (2nd edition). Philadelphia: Lippincott.

causewayed enclosures in Sussex. Sussex Archaeological Collections 132: 7-24. Ehrenreich, B. 1997. Blood Rites. New York: Metropolitan Books. Ember, C.R. and Ember, M. 1992. Resource unpredictability, mistrust, and war: a cross-cultural study. Journal of Conflict Resolution 36: 242-62. Evans, C. 1988. Excavations at Haddenham, Cambridgeshire: a ‘planned’ enclosure and its regional affinities. In C. Burgess, P. Topping, C. Mordant and M. Maddison (eds) Enclosures and Defences in the Neolithic of Western Europe. Oxford: BAR International Series 403. 47-73. Fairén, S. 2004. Rock-art and the transition to farming: the Neolithic landscape of the central Mediterranean coast of Spain. Oxford Journal of Archaeology 23: 119. Ferembach, D. 1959. Note sur un crane brachycephále et deux mandibles du Mésolithique d’Israel. Israel Exploration Journal 9: 65-73. Ferguson, R.B. 1990. Explaining war. In J. Haas (ed.) The Anthropology of War. Cambridge: Cambridge University Press. 26-55. Ferguson, R.B. 1995. Yanomami Warfare: a political history. Santa Fe NM: School of American Research Press. Ferguson, R.B. 1997a. Violence and war in prehistory. In D.L. Martin and D.W. Frayer (eds) Troubled Times. New York: Gordon and Breach. 321-55. Ferguson, R.B. 1997b. Review of War Before Civilization: the myth of the peaceful savage. American Anthropologist 99: 424-5. Ferguson, R.B. 2001. Materialist, cultural and biological theories on why Yanomami make war. Anthropological Theory 1: 99-116. Fernández-Jalvo, Y., Díez, J.C., Cáceres, I. and Rosell, J. 1999. Human cannibalism in the early Pleistocene of Europe. Journal of Human Evolution 37: 591-622. Fischer, A. 2002. Food for feasting? An evaluation of explanations of the neolithisation of Denmark and southern Sweden. In A. Fischer and K. Kristiansen (eds) The Neolithisation of Denmark: 150 years of debate. Sheffield: J.R. Collis Publications. 343-93. Foley, R.A. 1996. An evolutionary and chronological framework for human social behaviour. Proceedings of the British Academy 88: 95-117. Frayer, D.W. 1997. Ofnet: evidence of a Mesolithic massacre. In D.L. Martin and D.W. Frayer (eds) Troubled Times. New York: Gordon and Breach. 181216. Gat, A. 1999. The pattern of fighting in simple smallscale, prestate societies. Journal of Anthropological Research 55: 563-83. Gilbert, P. 1994. Male violence: towards an integration. In J. Archer (ed.) Male Violence. London: Routledge. 352-89. Gilchrist, R. (ed) 2003. Special issue ‘The social commemoration of warfare’. World Archaeology 35 (1).

14

I.J.N. THORPE: THE ANCIENT ORIGINS OF WARFARE AND VIOLENCE Keith, A. 1931. New Discoveries Relating to the Antiquity of Man. London: Williams and Norgate. Kelly, R.C. 2000. Warless Societies and the Origin of War. Ann Arbor: University of Michigan Press. Kennedy, J. 1971. Ritual and intergroup murder: comments on war, primitive and modern. In M.N. Walsh (ed.) War and the Human Race. Amsterdam: Elsevier. 40-61. Klein, R.G. 1999. The Human Career. (2nd edition). Chicago: University of Chicago Press. Klein, R.G. and Edgar, B. 2002. The Dawn of Human Culture. Chichester: Wiley. Knauft, B.M. 1987. Reconsidering violence in simple human societies: homicide among the Gebusi of New Guinea. Current Anthropology 28: 457-500. Knauft, B.M. 1990. Melanesian warfare: a theoretical history. Oceania 60: 250-311. Knauft, B.M. 1991. Violence and sociality in human evolution. Current Anthropology 32: 391-428. Koch, K-F. 1974. The Anthropology of Warfare. Reading MA: Addison-Wesley. Koehler, L. 1997. Earth mothers, warriors, horticulturalists, artists, and chiefs: women among the Mississippian and Mississippian-Oneota peoples, A.D. 1000 to 1750. In C. Claasen and R.A. Joyce (eds) Women in Prehistory: North America and Mesoamerica. Philadelphia: University of Pennsylvania Press. 211-26. Kristiansen, K. 1999. The emergence of warrior aristocracies in later European prehistory and their long-term history. In J. Carman and A. Harding (eds) Ancient Warfare: archaeological perspectives. Stroud: Sutton. 175-89. Lambert, P.M. 1997. Patterns of violence in prehistoric hunter-gatherer societies of coastal southern California. In D.L. Martin and D.W. Frayer (eds) Troubled Times. New York: Gordon and Breach. 77110. Lambert, P.M. 2002. The archaeology of war: a North American perspective. Journal of Archaeological Research 10: 207-41. Larsen, C.S. 1997. Bioarchaeology. Cambridge: Cambridge University Press. Larsson, L. 1989. Late Mesolithic settlements and cemeteries at Skateholm, southern Sweden. In C. Bonsall (ed.) The Mesolithic in Europe. Edinburgh: John Donald. 367-78. Layton, R. and Barton, R. 2001. Warfare and human social evolution. In K.J. Fewster and M. Zvelebil (eds) Ethnoarchaeology and HunterGatherers: pictures at an exhibition. Oxford: BAR Publishing, BAR International Series 955. 53-63. Le Blanc, S.A. 1999. Prehistoric Warfare in the American Southwest. Salt Lake City: University of Utah Press. Lessa, A. and Guidon, N. 2002. Osteobiographic analysis of skeleton 1, sítio Toca dos Coqueiros, Serra da Caprivara national park, Brazil, 11060 BP: first results. American Journal of Physical Anthropology 118: 99-110.

Lewis-Williams, D. 2002. The Mind in the Cave: consciousness and the origins of art. London: Thames and Hudson. Liddell, D.M. 1930. Report on the excavations at Hembury Fort, Devon, 1930. Proceedings of the Devon Archaeological Exploration Society 1: 40-63. Liddell, D.M. 1931. Report on the excavations at Hembury Fort, Devon, 1931. Proceedings of the Devon Archaeological Exploration Society 1: 90-119. Liddell, D.M. 1932. Report on the excavations at Hembury Fort, Devon. Proceedings of the Devon Archaeological Exploration Society 1: 162-90. Liddell, D.M. 1935. Report on the excavations at Hembury Fort, Devon. Proceedings of the Devon Archaeological Exploration Society 2: 135-75. Lillie, M.C. 2001. Mesolithic cultures of Ukraine. In K.J. Fewster and M. Zvelebil (eds) Ethnoarchaeology and Hunter-Gatherers: pictures at an exhibition. Oxford: BAR Publishing, BAR International Series 955. 53-63. Lindqvist, C. and Possnert, G. 1999. The first seal hunter families on Gotland. Current Swedish Archaeology 7: 65-87. Lovell, N.C. 1997. Trauma analysis in palaeo-pathology. Yearbook of Physical Anthropology 40: 139-70. Lubell, D., Jackes, M. and Meiklejohn, C. 1989. Archaeology and human biology of the MesolithicNeolithic transition in southern Portugal: a preliminary report. In C. Bonsall (ed.) The Mesolithic in Europe. Edinburgh: John Donald. 632-40. McCown, T.D. and Keith, A. 1939. The Stone Age Men of Mount Carmel: the fossil human remains from the Lavalloiso-Mousterian. Oxford: Clarendon. Malinowski, B. 1941. War – past, present and future, In J. D. Clarkson and T.C. Cochran (eds) War as a Social Institution. New York: Columbia University. 21-31. Márquez, S., Mowbray, K., Sawyer, G.J., Jacob, T. and Silvers, A. 2001. New fossil hominid calvaria from Indonesia – Sambungmacan 3. The Anatomical Record 262: 344-68. Martin, D.L. and Frayer, D.W. (eds) 1997. Troubled Times. Langhorne: Gordon and Breach. Maschner, H.D.G. and Reedy-Maschner, K.L. 1998. Raid, retreat, defend (repeat): the archaeology and ethnohistory of warfare on the north Pacific rim. Journal of Anthropological Archaeology 17: 19-51. Meiklejohn, C., Brinch Petersen, E. and Alexandersen, V. 2000. The anthropology and archaeology of Mesolithic gender in the western Baltic. In M. Donald and L. Hurcombe (eds) Gender and Material Culture in Archaeological Perspective. London: Macmillan. 222-37. Melbye, J. and Fairgreave, S.I. 1994. A massacre and possible cannibalism in the Canadian Arctic: new evidence from the Saunaktuk site (NgTn-1). Arctic Anthropology 31: 57-77. Mercer, R.J. 1981. Excavations at Carn Brea, Illogan, Cornwall, 1970-73. Cornish Archaeology 20: 1-204. Mercer, R.J. 1999. The origins of warfare in the British Isles. In J. Carman and A. Harding (eds) Ancient

15

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Persson, O. and Persson, E. 1988. Anthropological report concerning the interred Mesolithic populations from Skateholm, southern Sweden. Excavation seasons 1983-1984. In L. Larsson (ed.) The Skateholm Project I. Stockholm: Almqvist and Wiksell International. 89105. Peters, J.F. 1998. Life Among the Yanomami. Orchard Park NY: Broadview Press. Peterson, J. 2002. Sexual Revolutions: gender and labor at the dawn of agriculture. Walnut Creek: Altamira. Pickering, T.R., White, T.D. and Toth, N. 2000. Brief communication: cut-marks on a Plio-Pleistocene hominid from Sterkfontein, South Africa. American Journal of Physical Anthropology 111: 579-84. Pinker, S. 1998. How the Mind Works. London: Penguin. Pretty, G.L. and Kricun, M.E. 1989. Prehistoric health status of the Roonka population. World Archaeology 21: 198-224. Pryor, F. 1998. Etton: excavations at a Neolithic causewayed enclosure near Maxey, Cambridgeshire. London: English Heritage Archaeological Report 18. Raaflaub, K. and Rosenstein, N. (eds) 1999. War and Society in the Ancient and Medieval Worlds. Washington D.C.: Center for Hellenic Studies. Radovanovic, I. 1996. The Iron Gates Mesolithic. Ann Arbor: International Monographs in Prehistory 11. Randsborg, K. 1995. Hjortspring: warfare and sacrifice in early Europe. Århus: Århus University Press. Redmond, E.M. 1994. Tribal and Chiefly Warfare in South America. Ann Arbor: University of Michigan, Memoirs of the Museum of Anthropology 28. Reedy-Maschner, K.L. and Maschner, H.D.G. 1999. Marauding middlemen: western expansion and violent conflict in the Subarctic. Ethnohistory 46: 703-43. Reyna, S.P. and Downs, R.E. (eds) 1994. Studying War: anthropological perspectives. Langhorne: Gordon and Breach. Ridley, E.M. 2003. Nature Via Nurture: genes, experience and what makes us human. London: Fourth Estate. Robarchek, C. and Robarchek, C. 1998. Waorani: the contexts of violence and war. Fort Worth: Harcourt Brace. Rolle, R. 1989. The World of the Scythians. London: Batsford. Roper, M.K. 1969. A survey of the evidence for intrahuman killing in the Pleistocene. Current Anthropology 10: 427-59. Roper, M.K. 1975. Evidence of warfare in the Near East from 10,000-4,300 B.C. In M.A. Nettleship, R. Dalegivens and A. Nettleship (eds) War: its causes and correlates. The Hague: Mouton. 299-343. Russell, M.D. 1987. Bone breakage in the Krapina hominid collection. American Journal of Physical Anthropology 72: 373-80. Saville, A. 2002. Lithic artefacts from Neolithic causewayed enclosures: character and meaning. In G. Varndell and P. Topping (eds) Enclosures in Neolithic Europe. Oxford: Oxbow. 91-105.

Warfare: archaeological perspectives. Stroud: Sutton. 143-56. Milner, G.R. 1999. Warfare in prehistoric and early historic eastern North America. Journal of Archaeological Research 7: 105-51. Moore, J.H. 1990. The reproductive success of Cheyenne war chiefs: a contrary case to Chagnon’s Yanomamö. Current Anthropology 31: 322-30. Nash, G. 2000. The power and the violence of the archers. Archaeological Review from Cambridge 17: 81-98. Newell, R.R., Constandse-Westermann, T.S. and Meiklejohn, C. 1979. The skeletal remains of Mesolithic man in western Europe: an evaluative catalogue. Journal of Human Evolution 8: 1-228. O’Connell, R.L. 1995. Ride of the Second Horseman. Oxford: Oxford University Press. Orschiedt, J. 1998. Ergebnisse einer neuen Untersuchung der spätmesolithischen Kopfbestattungen aus Süddeutschland. In N. Conard (ed.) Aktuelle Forschungen zum Mesolithikum. Tübingen: Urgeschichtliche Materialhelfte 12. 147-60. Orschiedt, J. 1999. Manipulationen an menschlichen Skelettresten. Bingen: Urgeschichtliche Materialhelfte 13. Osgood, R.H. and Monks, S. with Toms, J. 2000. Bronze Age Warfare. Stroud: Sutton. Otterbein, K.F. 1997. The origins of war. Critical Review 11: 251-77. Otterbein, K.F. 1999. A history of research on warfare in anthropology. American Anthropologist 101: 794805. Otterbein, K.F. 2000. Five feuds: an analysis of homicides in eastern Kentucky in the late nineteenth century. American Anthropologist 102: 231-43. Owsley, D.W. 1994. Warfare in Coalescent Tradition populations of the northern Plains. In D.W. Owsley and R.L. Jantz (eds) Skeletal Biology in the Great Plains: migration, warfare, health and subsistence. Washington D.C.: Smithsonian Institution Press. 33344. Papathanasiou, A., Larsen, C.S. and Norr, L. 2000. Bioarchaeological inferences from a Neolithic ossuary from Alepotrypa cave, Diros, Greece. International Journal of Osteoarchaeology 10: 210-28. Péquart, M. and Péquart, S-J. 1954. Hoèdic: deuxieme station-nécropole du mésolithique côtier armoricain. Anvers: de Sikkel. Péquart, M., Péquart, S-J., Boule, M. and Vallois, H-V. 1937. Téviec, station-nécropole mésolithique du Morbihan. Paris: Archives de l’Institut de Paléontologie Humaine 18. Péquart, S-J. 1931. Sur une vertèbre humaine mésolithique percée d’une fleche. Association Française pour l’Avancement des Sciences 55: 321-4. Persson, O. and Persson, E. 1984. Anthropological Report on the Mesolithic Graves from Skateholm, Southern Sweden, I. Excavation Seasons 1980-1982. Lund: Papers of the Archaeological Institution (N.S.) 5. 16

I.J.N. THORPE: THE ANCIENT ORIGINS OF WARFARE AND VIOLENCE Smith, M.O. 1997. Osteological indications of warfare in the Archaic period of the western Tennessee Valley. In D.L. Martin and D.W. Frayer (eds) Troubled Times. New York: Gordon and Breach. 241-65. Sollas, W.J. 1927. The Chancelade skull. Journal of the Royal Anthropological Institute 30: 89-122. Standen, V.G. and Arriaza, B.T. 2000. Trauma in the Preceramic coastal populations of northern Chile: violence or occupational hazards? American Journal of Physical Anthropology 112: 239-49. Steel, D. 1999. Trade goods and Jívaro warfare: the Shuar 1850-1957, and the Achuar, 1940-1978. Ethnohistory 46: 745-76. Sterpin, A. 1993. La chasse aux scalps chez les Nivacle du Gran Chaco. Journal de la Société des Americanistes 79: 33-66. Stiner, M.C. 1991. The faunal remains from Grotta Guattari: a taphonomic perspective. Current Anthropology 32: 103-17. Strassburg, J. 2000. Shamanic Shadows. Stockholm: Stockholm Studies in Archaeology 20. Swisher, C.C., Rink, W.J., Anton, S.C., Schwarcz, H.P., Curtis, G.H., Suprijo, A. and Widiasmoro. 1995. Latest Homo erectus of Java: potential contemporaneity with Homo sapiens in southeast Asia. Science 274: 1870-4. Taçon, P. and Chippindale, C. 1994. Australia’s ancient warriors: changing depictions of fighting in the rock art of Arnhem Land, N.T. Cambridge Archaeological Journal 4: 211-48. Tappen, N.C. 1987. Circum-mortem damage to some ancient African hominid crania: a taphonomic and evolutionary essay. African Archaeological Review 5: 39-47. Thompson, J.E.S. 1954. The Rise and Fall of Maya Civilization. Norman OK: University of Oklahoma Press. Thorpe, I.J.N. 1996. The Origins of Agriculture in Europe. London: Routledge. Thorpe, I.J.N. 2003. Anthropology, archaeology and the origin of warfare. World Archaeology 35: 145-65. Thorpe I.J.N. In press. Fighting and feuding in Neolithic and Bronze Age Britain and Ireland. In H. Vandkilde (ed.) War and Society. Århus: Århus University Press. Tillier, A-M. 1995. Paléoanthropologie et pratiques funéraires au Levant méditerranéen durant le Paléolithique Moyen: le cas des sujets non-adultes. Paléorient 21: 63-76. Todd, M. 1984. Excavations at Hembury (Devon), 198083: a summary report. Antiquaries Journal 64: 25168. Tooby, J. and Cosmides, L. 1990. The past explains the present: emotional adaptations and the structure of ancestral environments. Ethology and Sociobiology 11: 375-424. Treherne, P. 1995. The warrior’s beauty: the masculine body and self-identity in Bronze-Age Europe. Journal of European Archaeology 3: 105-44. Trinkaus, E. 1983. The Shanidar Neandertals. London: Academic Press.

Trinkaus, E. 1985. Cannibalism and burial at Krapina. Journal of Human Evolution 14: 203-16. Trinkaus, E., Svoboda, S., West, D.L., Sládek, V., Hillson, S.W., Drozdová, E. and Fišáková, M. 2000. Human remains from the Moravian Gravettian: morphology and taphonomy of isolated elements from the Dolní VČstonice II site. Journal of Archaeological Science 27: 1115-32. Vallois, H.V. 1961. The evidence of skeletons. In S.L. Washburn (ed.) Social Life of Early Man. New York: Viking Fund Publications in Anthropology 31. 21435. Van der Dennen, J. 1995. The Origin of War: the evolution of a male-coalitional reproductive strategy. Groningen: Origin Press. Vandkilde, H. 2003. Commemorative tales: archaeological responses to modern myth, politics and war. World Archaeology 35: 126-44. Vencl, S. 1991. Interprétation des blessures causées par les armes au Mésolithique. L'Anthropologie 95: 21928. Vencl, S. 1999. Stone Age warfare. In J. Carman and A. Harding (eds) Ancient Warfare: archaeological perspectives. Stroud: Sutton. 57-72. Vlþek, E. 1995. Genetische und paläoethnographische Aspekte bei dur Beurteilung der Mammutjägerpopulation von Dolní VČstonice in Südmähren. In H. Ullrich (ed.) Man and Environment in the Palaeolithic. Liège: Études et Recherches Archéologiques de l’Université de Liège 62: 209-21. Wahl, J. and König, H.G. 1987. Anthropologischtraumatologische Untersuchung der menschlichen Skelettreste aus dem Bandkeramischen Massengrab bei Talheim, Kreis Heilbronn. Fundberichte aus Baden-Württemberg 12: 65-193. Webb, S. 1995. Palaeopathology of Aboriginal Australians. Cambridge: Cambridge University Press. Webster, D. 2000. The not so peaceful civilization: a review of Maya war. Journal of World Prehistory 14: 65-119. Weidenrich, F. 1943. The skull of Sinanthropus pekinensis: a comparative study on a primitive hominid skull. Palaeontologia Sinica (N.S.) 10D: 1485. Wendorf, F. 1968. Site 117: a Nubian Final Palaeolithic graveyard near Jebel Sahaba, Sudan. In F. Wendorf (ed.) The Prehistory of Nubia, vol. 2. Dallas: Southern Methodist University Press. 954-1040. Wendorf, F. and Schild, R. 1986. The Wadi Kubbaniya Skeleton: a Late Paleolithic burial from southern Egypt. Dallas: Southern Methodist University Press. Wendorf, F. and Schild, R. In press. Late Palaeolithic warfare in Nubia: the evidence and causes. Adumatu 10. White, T.D. 1986. Cut-marks on the Bodo cranium: a case of prehistoric defleshing. American Journal of Physical Anthropology 69: 503-9. White, T.D. and Toth, N. 1989. Engis: preparation damage, not ancient cut-marks. American Journal of Physical Anthropology 78: 361-8. 17

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Wilson, M.L. and Wrangham, R.W. 2003. Intergroup relations in chimpanzees. Annual Review of Anthropology 32. 363-92. Wrangham, R.W. 1999. The evolution of coalitionary killing. Yearbook of Physical Anthropology 42: 1-30. Wrangham, R.W. and Peterson, D. 1996. Demonic Males: apes and the origins of human violence. London: Bloomsbury. Zollikofer, C.P.E., Ponce de Leon, M.S., Vandermeersch, B. and Lévêque, F. 2002. Evidence for interpersonal violence in the St. Césaire Neanderthal. Proceedings of the National Academy of Sciences 99: 6444-8.

White, T.D. and Toth, N. 1991. The question of ritual cannibalism at Grotta Guattari. Current Anthropology 32: 118-24. Whiten, A., Goodall, J., McGrew, W.C., Nishida, T., Reynolds, V., Sugiyama, Y., Tutin, C.E., Wrangham, R.W. and Boesch, C. 1999. Cultures in chimpanzees. Nature 399: 682-5. Whittle, A.W.R. 1985. Neolithic Europe: a survey. Cambridge: Cambridge University Press. Whittle, A.W.R. et al. 1999. The Harmony of Symbols: the Windmill Hill causewayed enclosure, Wiltshire. Oxford: Oxbow. Wilson, E.O. 1978. On Human Nature. Cambridge MA: Harvard University Press.

18

Warfare, violence and slavery in later prehistory: an introduction Mike Parker Pearson Department of Archaeology, University of Sheffield

Of the hundreds, if not thousands, of military conflicts in the ‘post-war’ period (since the Second World War) the vast majority have been fought in the poor nations of the ‘South’ with weapons bought from the richer ‘North’. At any time, somewhere on the planet, someone is at war.

In 1991 one of the most extraordinary archaeological discoveries was made within a glacier in the Upper Ötztal area of the Alps. Here was the completely preserved body of a prehistoric man together with his clothing and equipment. With the kind of evidence that only survives in extraordinary circumstances, archaeologists relished this opportunity to gain insights into Alpine Copper Age life and society. By 1993 Konrad Spindler had published a book on the ‘Iceman’, identifying him as an itinerant shepherd who, in the early autumn, had fled some form of disaster - probably a violent fight or attack at his village which had left him with broken ribs. Fleeing into the mountains alone, ‘Ötzi’ succumbed to the cold during a blizzard or a sudden fog and froze to death (Spindler 1993: 163-5, 248-51). Almost a decade after the discovery, archaeologists discovered that he had probably died in early summer, on the basis of pollen in his gut. Not all archaeologists were convinced about the evidence for violence; the unhealed breaks in his ribs could easily have occurred at any time after death (Spindler 1993: 180; Chamberlain and Parker Pearson 2001: 130-1).

War and peace: the politics of post-war archaeology One could be forgiven for thinking that, in a world of continuous warfare and violence, that such matters would have made a deep impact on archaeologists studying the prehistoric past. In one sense the answer is that indeed they have, expressed in archaeologists’ strong aversion to accepting anything but the most rigorous evidence for these distasteful practices. Most members of the public would probably wonder what all the fuss was about mention prehistory and what comes to mind for many is probably some version of Thomas Hobbes’ eighteenthcentury dictum of prehistoric life having been ‘nasty, brutish and short’. Some writers have clung to a vision of noble savages living in harmony with nature, followed by the modern world’s fall from grace, whereas others have preferred a model of evolutionary progress, from the endemic violence of grunting cavemen to the technological precision of the surgical strike.

The suggestion that this individual suffered a violent assault during his final days was finally confirmed in 2001 by the discovery of an arrow wound in his shoulder (Gostner and Egarter Vigl 2002). In 2003 DNA analysis identified the blood of four other people on his belongings - on his flint dagger blade, on his cloak, on an arrowhead and on a broken arrow shaft (Ives 2003). The new interpretation was that he had died as the result of a border skirmish in which he had stabbed someone with his dagger but had himself suffered an injury to the hand as well as to the ribs before being shot in the back. That he was not alone during the fight was inferred from the fact that the arrow shaft had been snapped, presumably in an attempt to pull it out of the wound.

In Paul Ricoeur’s terms (1988), post-war archaeologists have generally written the past as ‘other’ rather than the past as ‘same’. Breaking with the militarist backgrounds of earlier archaeologists such as Pitt-Rivers and Wheeler, these more recent generations have presented images of prehistoric communities which, consciously or not, have tended to emphasize their non-violent aspects. As Roger Mercer reminded me at the conference that spawned this volume, even in recent decades feelings among archaeologists were so strong that a session on warfare at a TAG conference in Durham in the 1980s attracted demonstrators on the door of the auditorium, protesting at an unnecessary glorification of war. In response to such unease, one can only question whether the study of a particularly repellent subject necessarily turns us into aficionados of evil - does the student of slavery develop a yearning for its return? Is the historian of Hitler necessarily a Nazi? Richard Morris has recently addressed this issue as follows:

The continuing history of Ötzi’s post-mortem is a good example not only of the problems of identifying evidence of violence and warfare from archaeological evidence but also of the difficulty of persuading the archaeological community to accept such conclusions - I myself was one of those who were very sceptical of Spindler’s initial conclusions about Ötzi’s evidence for pre-mortem violence. Like cannibalism, human-on-human violence has been something of an archaeological taboo. As a Swedish archaeologist explained to me recently, he would not even attempt to apply for research grants to study prehistoric warfare. It would be anathema to Swedish academic morality: Sweden is a country that avoids war at all costs - although it has done very well supplying armaments to others. Of course, it would be disingenuous to pretend that the conundrum of pacifism versus the supply of weapons is limited to Sweden alone.

‘Why do we study such remains? Indeed, should we study them? We should ask, for there are some who have qualms. I have heard it said that an interest in military remains reflects a devotion (implicitly male) to militarism, while others are nervous lest nostalgia for wartime structures should fan Eurosceptic insularity. I have never heard the contrary argument, that the 19

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY consciousness of archaeologists, leading to a new interest which developed around 1995. Her thesis of a circumstance-driven agenda is very plausible but, on closer examination, it seems that, for a good number of archaeologists at least, the interest has been evidencedriven. Firstly, her case for a lack of warfare studies prior to 1995 - including in her own field of Scandinavian archaeology - can be seriously questioned. Secondly, I would argue that it is the discoveries of dramatic evidence such as the massacre sites at Wassenaar (Netherlands; Louwe Koojimans 1993) and Talheim (Germany; Wahl and König 1987)2 which forced prehistorians to reconsider their pacifist assumptions.

demolition of wartime fabric would cause jingoism to abate, perhaps because its silliness when put that way round becomes all too evident.’ (Morris 2003: 2). Roger Mercer is among those archaeologists ruminating on their colleagues’ avoidance of matters martial and he has suggested two explanations, one being the painful scar left on the twentieth century by warfare and the other being the dominance of ‘Marxist and post-Marxist models’ in archaeology, which sought to find conflict and struggle within prehistoric societies rather than between them (Mercer 1999: 143). Oddly enough, although this latter point sounds initially convincing, the Marxists are among the few who have attempted to discuss warfare as an important element of social formations and their transformations. If we revisit classic works such as Friedman and Rowlands’ The Evolution of Social Systems (1978) or Spriggs’ Marxist Perspectives in Archaeology (1984) together with other papers of the period (Gillman 1976; 1981; Frankenstein and Rowlands 1978), warfare is certainly in there along with class struggle, structural contradictions and prestige goods. If we want to find a culprit lurking in the theoretical woodshed then it is probably the mature post-processual archaeology of the later 1980s and 1990s, particularly as practised by a younger generation reacting against their seniors’ transplanting of Classical writings on Celtic chiefs and warfare onto the archaeological remains of the British and European Iron Age as well as earlier periods.

According to Vandkilde, twentieth-century Scandinavian prehistorians have written happily about warriors but rarely have they considered war to be a driving force in society other than as a kind of theatrical (and bloodless) activity (2003: 130-2). Her evidence is drawn explicitly from studies of the Neolithic and Bronze Age but, had she addressed Iron Age studies, her argument would have been less persuasive. The investigation of the weapon deposits at Illerup and other Scandinavian sites left the excavators and other Iron Age archaeologists in no doubt that these were votive offerings of war booty (Ilkjaer and Lønstrup 1982), supporting a tradition of warfare interpretation that went back to the earliest excavations of these offering sites in the nineteenth century (Englehardt 1866). Studies of fighting methods and militia formations (Ørsnes 1968) and of the impact of war on society and landscape (Hedeager 1990; Parker Pearson 1984; 1989) were similarly being carried out prior to the 1990s. Although not interpreted as warfare victims, the Iron Age bog bodies of Denmark were also understood by pre- and post-World War II archaeologists as being subjected to extraordinary violence whether as victims of human sacrifice, executions or other murderous acts (Glob 1969; Munksgaard 1984; van der Sanden 1996).

We now live in a period of backlash. Lawrence Keeley (1996) has written of the ‘pacification of the past’ and has argued that evidence of conflict and defence in European prehistory has been misinterpreted as ‘ritual’: defended enclosures, for example, are reinterpreted as ceremonial gathering places, and Bronze Age swords and spears are described as ornamental items to be paraded rather than weapons to be used in anger. In an eloquent dissection of the attitudes behind ‘de-militarized’ interpretations of the British and European Iron Age, Simon James (forthcoming) has recently criticized what some have called the ‘Habitat© view of the past’ in which people spent peaceful lives ‘crocheting their own yoghurt’ (James 1997).1

The battle offerings of destroyed weaponry in the lakes and bogs of Denmark and northern Germany are archaeological evidence for war that is hard to refute. Prehistorians may argue - as many did in the 1980s and 1990s - that swords, spears and other weaponry were just symbols of power designed principally for show, but the large weapon deposits could not be explained in similarly pacific terms. From the 1990s and before a growing corpus of evidence was being amassed of unambiguous signs of violence on Neolithic and Bronze Age skeletons, from mass burials such as Talheim and Wassenaar, together with finds from Middle Bronze Age sites such as Tormarton (Parker Pearson 1993: 118; Osgood this volume) and Velim in the Czech Republic (Hrala et al. 1992; Harding 1999: 158-9). The journal World Archaeology published a special issue on Weapons and

The recent revival of archaeological interest in warfare has been attributed by Helle Vandkilde (2003) to the return of conflict to European soil in the 1990s. In her view, archaeologists draw their analogies in very direct and literal ways from the political, economic and military circumstances around them. Despite an unbroken sequence of wars, atrocities and conflicts in different parts of the world from the close of the Second World War, Vandkilde considers that only the Balkan War was real enough and close enough to impact on the creative

2

At Wassenaar, three of the male skeletons show evidence of blows while a fourth has a flint arrowhead in his ribs. At Talheim, the positioning of bodies and the evidence of blows to the back of the head suggest that men and women were lined up and executed (Paul Pettitt pers. comm.).

1 This phrase was apparently coined by Alexei Sayle and has since been modified to ‘yoghurt-weaving’ as a (self) derogatory term for practitioners of the ‘good life’.

20

MIKE PARKER PEARSON: WARFARE, VIOLENCE AND SLAVERY IN LATER PREHISTORY: AN INTRODUCTION in which societies flip back and forth from one state to the other, peace and war are specific social outcomes which may well differ from the state of uncertainty and insecurity in which probably most of the planet has often lived. This ‘third way’ is characterized by casual but sustained violence which operates at many different levels, from feuding and vendetta to larger scale skirmishing and fighting between clans and kin groups. Honour and shame are structuring principles in such contexts, in which men (and sometimes women) compete for honour and fight to expunge shame. Fighting and warfare may also be deeply embedded in exchange relationships; as social anthropologists have known for a long time, we often fight those families into which we marry. Even so, reworkings of social conceptions of exchange, such as those by Bourdieu (1977) and Godelier (1999), sometimes pay scant attention to the agonistic dimensions of such relationships.

warfare in 1987 whilst Classical archaeologists (e.g. Snodgrass 1964; Hanson 1989) and social anthropologists (e.g. Ferguson 1984; Haas 1990b) were already pointing the way prior to the 1990s. That said, most prehistorians were unwilling to accept the significance of war in the past until forced to do so by this new prehistoric evidence. The archaeology of war has come of age only in the last five years, in a flurry of books and television programmes. Our decision to hold a conference on warfare, violence and slavery at Sheffield in 2000 was prompted entirely by undergraduate students who saw this as an exciting topic of wide appeal which was largely denied to them in their taught modules and courses. Warfare has almost become an area of specialism in archaeology, formed out of the nexus of osteology, settlement studies and military history. This potential ghetto-ization is probably the very worst direction in which to go because the archaeology of violence is surely not to be relegated to a ‘special interest’ group in the same way as standing buildings or mummies. It has to be rescued from the wargamers and reinstated within the heart of a social archaeology. Warfare, violence and slavery are social institutions involving human agency within specific social contexts.

The insecurity of life in societies without war or peace runs through many aspects of everyday behaviour. Protection may be organized by controlling and directing the violence of the young men of the kin group, who intimidate others by openly carrying weapons. These informal militias may also extort food, clothing, sexual favours and other forms of ‘tribute’ from those being protected. Dogs are important not just for animal husbandry but also for canine alarm calls during the night. A small palisade or picket fence, or similar ‘lowlevel’ defences will be enough to deter the unwelcome visitor. Living in an isolated hamlet may also be a means of hiding away, off the beaten track, a method of avoiding attack alternative to congregating together in a large but highly visible settlement. Keeping weapons to hand is a necessary exercise day and night. In such paranoid circumstances few sleep easily in their beds: even though actual attacks may be few and far between it is the fear of the violence that is all pervasive and omnipresent. Readers will recognize that these features of social insecurity are by no means restricted solely to societies outside the state, appearing also in places and at times when state control is feeble (see Layton this volume).

Throwing babies out with bathwater is an activity with a long history in archaeology. The problem with some of the current rethinking is that it follows the swing of the pendulum. All of a sudden, prehistoric evidence for warfare and violence is found to be everywhere, as previous ‘politically correct’ authorized versions of the past are unmasked, their ritual veneer stripped away to reveal the pragmatic and ugly face of what really went on. Archaeologists suddenly lose their critical faculties, to pronounce on the pervasive grisliness of ancient ancestors - to the delight of television producers looking for sensationalist thrills - as proponents of a new creed which demands a similar orthodoxy amongst its converts. To move from a hyper-sceptical position of denial to one which recognizes those moments of humanity’s inhumanity is not simply a matter of switching interpretations. For example, even given the evidence of violence at Hambledon Hill and other Neolithic enclosures in western Europe (Mercer 1999; Bradley 1998: 197-200), we would be rash to reject the notion that such places had ritual and ceremonial significance simply because some have evidence of being attacked. This would be tantamount to inferring that the burning of a church or synagogue with its worshippers inside reveals its ‘true’ function as a refuge or defensive building.

Contextualizing insecurity Societies which have not been incorporated into state organizations, or only weakly so, may lack overarching state control but we should avoid considering them lawless. Stealing, fighting, killing and other misdemeanours all result in penalties. Anthropologists have occasionally talked of warfare in these small-scale societies as being more highly ritualized (see Carman and Carman this volume) but this sometimes confuses the dispensing of justice with the conduct of war. A fight between two lineages or clans may actually revolve around the need to redress the wrongs perpetrated by one individual, and the battle is over once that person has been injured or killed.

Beyond war and peace As Simon James has perceptively noted, both peace and war are engineered social and political situations that can require considerable resources and force to implement (forthcoming). Rather than forming a dialectical tension

21

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY One of the most graphic descriptions of life in a rural feuding society is Edith Durham’s High Albania (1909), written after a lifetime spent amongst gun-toting hill farmers, many of whose lives were cut short in blood feuds lasting five or more generations. A fearlessly independent Victorian lady, Durham left Britain on account of her ill health and in Albania soon established her reputation for courage and integrity. Marriage exchanges were often the cause of feuds since brides in arranged marriages would often run away from their new husbands to live with the man they loved. To wipe away the shame and ‘clean’ his honour, the husband and his male kin felt obliged to kill the other man, as a result of which his death would be avenged by his kin, thereby setting in motion a cycle that could last for generations. One man told Durham that his family had been involved in a blood feud for five generations, with men of each generation killing men of the other family and themselves being killed. ‘The chief man in this feud ... lamented his position bitterly. Five generations were too much. The quarrel had had nothing whatever to do with him, but he was liable to be shot for it after all these years.’ (Durham 1909: 166-7). As far as the men were concerned, ‘blood feuds ... were almost all the fault of women. Women were very wicked.’ (ibid.: 184)

x

downplay individual emotions and wants; finally, they reinforce their morality of peacefulness by having an ethnocentric attitude towards outsiders (humans and spirits) who are epitomized as violent, dangerous and evil (McCauley 1990: 13-15).

These societies may live peacefully but historically they have not been able entirely to live in peace, occasionally having to flee from raiders. The foregrounding of violence in order to maintain peace is an interesting strategy because it suggests that an element of fear (of outsiders in this case) may be deployed to restrict and avoid violence. Such aspects are also found in nation states in which ‘the king’s peace’ is maintained by state executions and cruel punishments for transgressors whilst mass violence is directed through warfare against rival states and other ‘bogeyman’ outsiders. One of the implications of living in a peaceful society is that perhaps we need outsiders to hate and fear, that in a sense good cannot be considered to exist without evil. The Semai and Buid can show us how the smallscale societies of world prehistory might potentially have been able to live peacefully but they cannot be used to claim that they must have done so. What they do help us to grasp is that violence and aggression are socially, culturally and contextually situated rather than being an inescapable condition of human nature. Of course, this goes very much against the grain of earlier thinking about the innateness of human aggression, as something instinctive and fundamental, set out by Freud (1964) and later by writers such as Lorenz (1967) and Ardrey (1967). Brian Ferguson has offered a useful critique of the innatist position (1984: 8-14) and pointed towards the complexity of explaining human aggression in terms of psychological, materialist, social and motivational factors.

Feuding and vendetta are chronicled from many parts of the world (see, for example, Black-Michaud 1975) and, for the European reader, are most notorious in the context of the mafia of Sicily and southern Italy as well as in Corsica and Sardinia. This cycle of deadly violence is characteristic of societies with strong gender hierarchies in which relationships between men are defined prominently in terms of honour and shame. We can expect the so-called ‘warrior societies’ of the European Bronze and Iron Ages to have followed such principles, given the stress on male bodily presentation and its likely implications for strong codes of honour (Treherne 1995). It could be argued that the Trojan War was the outcome of one such feud.

The two inter-linked theoretical positions which have had the greatest impact on anthropological studies of the causes and conditions of warfare are materialist or culture ecological approaches, and the modelling of social complexity (see Thorpe this volume). Not surprisingly, both have been prominent elements of the New Archaeology whose processual approaches dominated archaeological interpretations in the 1970s and 1980s. These cultural ecological approaches explain the causes of warfare in functionalist terms as being due to processes beyond individual agency or events: over-population and scarcity of agricultural land (Vayda 1976), shortage of protein (Harris 1974), and population regulation by creation of a ‘male supremacist complex’ (Harris 1984). Social and political evolutionary models of complexity from egalitarian to ranked and stratified societies and states (Fried 1967) have also been employed to generalize on the evolution of warfare. Ferguson writes: ‘As one moves up the scale of political evolution, military activity dramatically increases in scale and organizational complexity... the social institution of war becomes progressively more autonomous. It is not only less

Causes and conditions of war Interpersonal violence is institutionalized in many societies around the world. Japanese sumo, Yanomamö club fighting, European soccer violence and drunks fighting outside British pubs are just some of these institutions which may be state-sanctioned or not. Ethnographers have recorded a very few societies which shun violence, such as the Semai of Malaysia and the Buid of the Philippines (Gibson 1990; McCauley 1990). Violence amongst these two egalitarian groups is almost unknown and there appear to be several reasons for this: x they have an ‘anti-violent’ morality which rewards the generous and gentle whilst stigmatizing those who boast, quarrel or show stinginess, anger or violence; x they are small-scale, face-to-face and open-ended groups with little or no competition for status or other such forms of dominating each other, and they 22

MIKE PARKER PEARSON: WARFARE, VIOLENCE AND SLAVERY IN LATER PREHISTORY: AN INTRODUCTION 5.

constrained by social structure and ecology, it becomes less embedded in social life in general...’ (Ferguson 1984: 59).

6. We can also mention sociobiological approaches to explaining warfare. Probably the best known of these is Napoleon Chagnon’s investigation of Yanomamö warfare (1990; see Layton this volume), in which he claims that conflict and warfare are motivated primarily by reproductive issues although shortage of resources are also significant: ‘[I]n a word, individuals fight for mates when resources are abundant, and fight for the means to acquire mates when resources become scarce.’ (Chagnon 1990: 89). McCauley (1990) has noted that sociobiological explanations have become prominent just as biological innatist ‘killer instinct’ theories have faded. At the heart of sociobiological accounts are assumptions that behaviour which maximises inclusive fitness is biologically selected (cf Nystrom this volume on aggression in primates). Through aggression and warfare, individual males can increase their reproductive success by capturing women and producing offspring (see Thorpe this volume).

At the time of writing, during the American and British occupation of Iraq, it does seem that Ferguson’s strategic objectives may be almost custom-designed to fit the explicit and inferred objectives of the Iraq invasion. Perhaps Ferguson’s materialist perspective really does describe the United States’ contemporary political and military strategies but does it necessarily apply at every other time and place? For example, do these objectives adequately explain the attack on America on 11th September 2001 and the associated jihad? Although it is undeniably difficult to separate religious from material motivations, it would require a breathtaking leap of inference to reduce the reasons for fundamentalist Moslem terrorism to the kinds of extreme materialist explanation that were once advocated by Marvin Harris (1974). We should question whether ‘strategic objectives’ as described by Ferguson are genuinely ‘material’ and ‘infrastructural’ as opposed to ‘superstructural’ and whether such a ‘vulgar materialist’ conception of society (Friedman 1979; Giddens 1984) is still defensible. Robarchek asks, for example, whether going to war to acquire more wives or more food resources is ‘material’ or actually depends on cultural values in which multiple wives are symbolic of virility and success, or in which certain foodstuffs have status or symbolic significance (1990: 69).

What is lacking or denied in functionalist or sociobiological approaches are concepts of intentionality, decision-making and human agency in any terms other than material or biological imperatives. Robarchek illustrates this problem with a case study from the Semai in which he describes how one man decided to benefit materially from the opening-up of a road and the concomitant influx of Western goods and technology into the region (1990: 74). This course of action was in opposition to Semai values - other members of the community perceived the potential for violence arising from the rift that they saw developing as a result of his future activities - and they consciously rejected it.

For historians of warfare and its causes, attempts at materialist reductionism no doubt appear to diverge from the facts. One has only to read some of the many books on the causes of the Second World War to realize the complexity and intermeshing of interests, events, processes and intentions, to understand that single-issue generalizing models are simply unrealistic. It is telling that anthropologists talk of warfare - a process or activity - whereas historians tend to talk of wars - specific historically situated events.

Materialist explanations of warfare are often so broad as to be virtually all-encompassing. Ferguson considers that we can generalize about motivation in that ‘wars occur when those who make the decision to fight estimate that it is in their material interests to do so’ (1990: 30). Material (or infrastructural) factors explain why war occurs (competition for essential material resources) and how that warfare is practised (i.e. the technology of war). Structural factors such as kinship affect how people are grouped to fight and determine why war starts just when and where it does. Finally, at the top of the formation, superstructural factors (values, rules and perceptions) have a very limited effect in Ferguson’s view, perhaps only influencing whether groups on the verge of war actually proceed to action or not. He lists six strategic objectives which arise from material interests: 1. 2. 3. 4.

to use external conflict as a means of enhancing the decision-makers’ position within their own society; to forestall attacks by others (Ferguson 1990: 30).

Archaeological evidence for later prehistoric warfare, violence and slavery Many parts of the world are littered with the remains of recent conflicts or marked by the monuments that commemorate those wars. Battlefields, aircraft crash sites, coastal defences and batteries, airfields, PoW camps, military bases and dockyards, frontier fortifications, castles, town defences, siege works, war memorials, war cemeteries and other military artefacts provide an extensive record of the last 2000 years, from the site of the German massacre of Varus’ legions in AD 9 to the abandoned Cold War missile bases and their peace camps. When we turn to the prehistoric evidence,

to increase access to fixed resources; to capture moveable valuables; to impose an exploitative relationship on another independent group; to conquer and incorporate another group; 23

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY The question of detecting prehistoric evidence for slavery is perhaps even more difficult. In the Late Iron Age of Britain there are metal slave chains (see Aldhouse Green this volume) which it would be hard to argue were used for any purpose other than restraining slaves or captives. Yet these are some of the very few artefacts that we can identify as evidence for prehistoric slavery. The archaeology of slavery has thrived within the Classical and historical periods, with studies of slave quarters, diets and lifestyles (see Taylor this volume), whereas there is next to nothing to say on prehistoric slavery and whether it was even practiced in periods for which we have no written records. For example, were slave raiding, slave trading and the keeping of slaves features of early empires only, and simply not known in less hierarchical pre-state societies until they became targets for such operations?

there are difficult and fundamentally different issues of survival and interpretation. Analysts of prehistoric warfare have generally recognized a series of categories into which the evidence may be arranged: x

x x

x x x x

hillforts and other defended sites (including settlements located for protection; e.g. Aranda Jiménez and Sánchez Romero this volume); burning and destruction of houses and stores; osteological evidence of trauma caused by wounding (e.g. Hårde this volume; Knüsel this volume; Craig et al. this volume; Osgood this volume); demographic patterns consonant with high mortality of young adult males (Bishop and Knüsel this volume); mass graves; martial equipment and weaponry (e.g. Fontijn this volume; Aranda Jiménez and Sánchez Romero this volume); iconographic representations of martial activity and conflict (e.g. Bevan this volume; Nash this volume).

Approaches to warfare in prehistoric societies There have been two main strands to archaeological studies of warfare and violence in prehistory. On the one hand, certain studies have attempted to provide empirical documentation of their existence (e.g. Harding 1999; Osgood et al. 2000). On the other, writers have attempted to develop explanatory frameworks of why wars happen and how innovations in warfare come about. In both cases the knotty problem of interpreting the ambiguities of archaeological evidence has to be faced. There is no standard ‘cook book’ formula of middlerange theory to provide a cut-and-dried clinical diagnosis of warfare, violence or slavery from archaeological remains: each case has to be built up as a closely argued and richly textured contextual regional study which draws on multiple strands of evidence which, when analysed in isolation, are often inconclusive and ambiguous. In many cases the evidence must be argued on a balance of probabilities.

Not only is the evidence often limited but it also invites considerable ambiguities in interpretation. Are flint arrowheads found outside a gateway the vestiges of a siege or are they votive deposits? Do a hillfort’s ramparts signal protection or are they a status symbol which never saw a single attack? Were Bronze Age swords and spears, like the ineffectively thin bronze shields and body armour, more significant as items of male dress and authority than for fighting? Were injuries to bones sustained pre- or post-mortem? Were buildings burnt down in funeral rituals rather than in sackings? Do artistic representations reflect reality in any direct way? When dealing with pre-Bronze Age societies how do we tell if and when bows and arrows, axes and other subsistence tools were used as weapons against fellow humans?

As mentioned above, Lawrence Keeley (1996) has described war as a near-constant, ubiquitous and deadly feature of prehistoric life. From the evidence of Neolithic mass graves such as Talheim, he has argued that total war was conducted with very limited means - artefacts which were otherwise used in domestic, farming or hunting activities. Keeley is sceptical of the division between ritualized warfare in traditional societies and total war as exercised by modern industrial complexes, and it is certainly the case that modern wars between world powers have their ritualized and rule-bound dimensions (white flags, the Geneva convention, the Falklands exclusion zone etc.).

There are further difficulties. Evidence for violence is relatively easy to detect in cases such as the Alpine ‘Iceman’ when a projectile or other weapon is embedded in the body or skeleton, or when the skeleton displays fractures caused by maces or clubs. Yet how do we move from inferring that the violence took place at the more coordinated level of warfare - purposeful, organized and sustained violence between groups who may be politically independent - rather than merely during a quarrel, feud, revolt or skirmish? Were the pre-state societies throughout much of prehistory actually capable, in any case, of the mass mobilization necessary to initiate and sustain warfare?

Yet should archaeologists discard that division between total war and ‘ritualized’ war (see Carman and Carman this volume)? Does a large body of ambiguous evidence, peppered with a few strong examples where the implication of violence is hard to deny, really constitute definitive evidence for violence on the scale of sustained

24

MIKE PARKER PEARSON: WARFARE, VIOLENCE AND SLAVERY IN LATER PREHISTORY: AN INTRODUCTION evidence for massacres, sackings and killing in the 1260s and total abandonment of the region by 1300. Haas characterizes warfare as the last resort of people faced with the stark collision of environmental stress, increasing population densities and uncertainty over food supplies.

inter-community warfare? It is not even certain whether the evidence even supports a case for continuous lowlevel violence in conditions of social insecurity (as outlined above) as opposed to all-out war. From Keeley’s perspective of near-constant endemic warfare, the temptation is to treat war in prehistory as a process beyond individual human agency and historically contingent events.

Kristian Kristiansen (1999) has proposed a similarly materialist model of warfare for northern European later prehistory but which is not driven simply by environmental determinants. In his view, warfare derives from competition for access to exchange networks of new producers and clients within an evolutionary process which shifts from extensive to more intensive farming. Following Earle’s work on chiefdoms (1997), Kristiansen posits that chiefdom networks were involved in raiding to increase their productive base (booty, cattle, slaves and women) which might have lead to over-exploitation, endemic warfare and political fragmentation. He also presents evidence for a short-term oscillation between stable-finance societies (settled agrarian societies investing in labour, land and hoards) and wealth-finance societies (expansionist warrior societies investing in moveable wealth and tombs; see Sanchez-Moreno this volume).

In his work on the Neolithic and Early Bronze Age of eastern Europe, John Chapman (1999) has addressed the issue of the evolution of specialized weaponry from multi-purpose tools. He develops a threefold evolutionary typology in which tool-weapons (points, bows and arrows, clubs etc.) were progressively superseded by weapon-tools (such as maceheads, spears, daggers and battle-axes) and, ultimately, weapons such as swords and shields. This is a useful beginning to the study of warfare as a growing specialization in social and bodily practices. Chapman’s critique of Keeley’s ‘rivers of blood’ vision of prehistory is devastating: not only does Keeley twist and squeeze the anthropological and prehistoric evidence but he hugely overstates the case, citing just 12 sites dating to a period of 30,000 years across a whole continent as evidence of wholesale and near-continuous warfare (Chapman 1999: 102). In contrast, Chapman shows that the large numbers of skeletons with no evidence of violent trauma from the Bavarian and Danish Mesolithic, from the central European Linearbandkeramik and the French Neolithic provide strong negative evidence, suggesting that these might have been relatively peaceful times and places. With Keeley’s overstatement brought down to earth, Chapman then develops models of where and when warfare was most likely - in frontier settings, in areas with differential and limited resources, and at times when it accompanied the emergence of warrior lifestyles in the Chalcolithic and later (ibid.: 140; cf Hårde this volume).

Klavs Randsborg (1995) has developed a rather less mechanistic approach to understanding changes in warfare during the first millennium BC within a broad social context, innovations in weaponry and tactics being the result of social organizational changes. Working from the evidence of the fourth-century BC votive deposit of a ship and weaponry at Hjortspring, Randsborg argues that the large numbers of spears and shields constitute evidence of phalanx fighting, a battle strategy better known from the hoplite armies of Iron Age Greece (see Carman and Carman this volume). As in the Greek and Aegean context, he identifies this innovation as linked to the emergence of a more egalitarian concept of citizenry, hastening the decline of aristocratic norms and lifestyles of the Late Bronze Age, in which fighting tactics revolved around individualized combat between champions. Randsborg’s analysis is particularly interesting in its argument that warfare and its tactics are linked to geographically widespread conventions and processes of change.

The American Southwest has produced dramatic evidence for prehistoric violence and cannibalism, and Jonathan Haas (1990a; 1999) has developed a detailed case study of one of these contexts of almost allconsuming warfare in the Anasazi culture of the northern Southwest between AD 1240 and AD 1300. He argues that climatic worsening, over-population and increased competition over scarce food resources led to warfare, which is inferred from the movement of large settlements into inaccessible and defensive locations on clifftops and from the use of cliff dwellings in large, protected rock shelters. He also points to the circumstantial evidence of the low frequency of young adult males in the cemeteries as well as to sites such as Sand Canyon Pueblo (a walled village largely burned down and with dead bodies lying on house floors) and Largo-Gallina (with the massacred remains of 11 individuals on a house floor). The war was relatively short-lived, starting in the 1240s with the beginnings of defended and isolated villages, followed by

In a short but influential paper Paul Treherne (1995) also develops this relationship between warfare and society. In a convincing argument for the rise of a warrior cult in the European Bronze Age, he interprets the warrior’s body as a heraldic device decorated with the values that proclaimed his honour. Similar emphases on the links with social and gender relations have also been explored by other scholars working on the European Bronze Age (Fontijn 2002 and this volume; cf Bevan this volume; Vandkilde 2003). The power of such approaches is that they place the development of tactics and technology of prehistoric 25

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY likely to have been clubs.3 Flint arrowheads, stone axeheads and a few wooden bows survive from this period but there is no physical evidence for any form of club other than the wooden hafts of stone axes and a range of antler and stone maceheads. Maceheads generally derive from the Middle-Late Neolithic and Early Bronze Age and it seems, therefore, that these injuries were inflicted by wooden clubs that have not survived. About half of the injuries that Schulting identified had healed, indicating that the victims had recovered from the attack, whereas the other half of the club wounds were presumably fatal.

warfare inside its social context rather than presenting it as a separate domain of human agency and achievement which is somehow outside and largely independent of the social. As an example of the latter, Hill and Wileman (2002: 178-80) have modelled a possible cycle of warfare from earliest times to the present day which runs on a continuous dynamic between aggressive and defensive postures - as new methods of attack develop so there arises a countering of these by innovations in defences which spark new attacking ploys and technology, and so on. From the development of early weapons for hunting and fighting comes the invention of personal armour and small-scale defences, which prompt the development of the first armies, which result in larger-scale defences which are in turn followed by a developing technology of siege warfare and so on. The model has some merit for examining technological developments in more recent industrialized times, particularly in the twentieth century, but it may only be of limited applicability on the long and broad canvas of prehistory.

Most of the skulls are those of men but there are also sufficient numbers of damaged female skulls to indicate that women were also regularly victims. Individuals buried in chambered tombs and long barrows represent a tiny proportion of society, less than 1% of the population, and so are not necessarily representative of the larger population, most of whose remains are absent. Were these individuals of a special social status or were there specific reasons why they were buried together? It is just possible that these communal tombs and barrows were not simply monuments for the ancestors but were also memorials to particular events such as wars, famines or plagues. Were some or all of them to have been war memorials containing the fallen, then this proportion of 5% of skulls showing injuries may well not be representative of causes of death throughout the Earlier Neolithic. Yet it is likely that these head injuries relate to a certain level of violence among the population at large, either through institutionalized combat or punishment, clandestine attacks and muggings, or more organized fighting and warfare.

Rethinking warfare, violence and slavery in European later prehistory By the end of the Upper Palaeolithic, burials from Nubia dating to 12,000-10,000 BC portray unequivocal evidence of trauma resulting from inter-personal violence (Wendorf 1968; Guilaine and Zammit 2005: 67-72; see Thorpe this volume for Palaeolithic and Mesolithic evidence). At around the same date in the Near East, the first phase of stone walls were built around the PrePottery Neolithic A village of Jericho, although these appear to have been primarily flood defences (Bar-Yosef 1986). Yet there is evidence for violence at this time. The headless body of an adult male, lying on its front on the floor of a burned-down PPNA house at Jerf el Ahmar in Syria is a likely indication that head-hunting was an element of the attack, possibly even the motive for it (Stordeur et al. 1997). An interest in heads, taken generally from dead ancestors rather than from fresh kills, seems to have been a strong feature of PPNA society as well as of the preceding Natufian period (12,800-10,000 BP) in the Near East (Kuijt 1996). Yet, with over 400 Natufian skeletons so far recovered, there is very little evidence for violence. This is all the more unusual given that some of the main theories for the adoption of agriculture in this part of the world posit widespread competition between households in feasting and for food resources generally (Bender 1978; Hodder 1990) or severe environmental stress (Binford 1968), both of which may be expected to have encouraged violence.

Cranial injuries from clubbing show up much more clearly than arrow wounds since the latter only occasionally cause damage to bone. There are a few instances of arrowheads lying within the body cavity or in such close association as to strongly suggest that these were injuries similar to the Iceman’s and just occasionally, as in the case of Neolithic burials at Ascottunder-Wychwood and Tulloch of Assery B (Mercer 1999: 149) or the Early Bronze Age archer from the ditch at Stonehenge (Evans 1983), the flint arrowheads or their broken-off tips are embedded in the bones themselves. In other cases, arrow impacts may leave only the smallest of peri-mortem injuries detectable only by expert analysis.4 Arrowheads can also be removed from wounds unless the victim is summarily buried, further stacking up the

Neolithic clubbing Rick Schulting and Mick Wysocki (2002) have recently estimated that 4% or 5% of skulls from British Earlier Neolithic contexts have evidence of healed or perimortem injuries caused by blunt instruments which are

3 The estimate of number of skulls with injuries is 7.4% but they think that some of the examples may be mis-diagnosed or the result of accidents (Rick Schulting pers. comm.) 4 An example is the adult human femur excavated in 2004 at Durrington Walls henge enclosure, Wiltshire, with two projectile impact injuries but no signs of any embedded flint tips (identified by Chris Knüsel).

26

MIKE PARKER PEARSON: WARFARE, VIOLENCE AND SLAVERY IN LATER PREHISTORY: AN INTRODUCTION The origins of slavery?

probability that many such injuries are likely to go undetected.5

The emergence of social inequality in the Neolithic and certainly by the third millennium BC is likely to have created the conditions in which humans had opportunities to enslave one other. Warfare and raiding are classic means of acquiring slaves and the numerous cases for organized war in the Neolithic (Mercer 1999; Vencl 1999; Chapman 1999) make it likely that the conditions for obtaining captives were already well in place by the fifth and fourth millennia BC. Debt bondage can also be a feature of non-egalitarian societies and so slaves might have been acquired by non-military means from these times onwards. Within Europe, the earliest evidence of slavery could be in the form of human accompaniments to high-status burials, although these may be the burials of captives, retainers or even relatives. One of these is the central primary burial within the large mound of Duggleby Howe in Yorkshire, dating to about 3000 BC (Mortimer 1905; Kinnes et al. 1983). The adult male in this grave was accompanied by an adult skull and a child, both of which appear to have been incorporated into the upper fill of the burial. Another example is the aristocratic burial at Leubingen, dating to 1952-1932 BC, which contains an adult male equipped with gold and bronze artefacts. Laid across his waist, as if just another of his grave goods, was the body of an adolescent or child (Coles and Harding 1979: 40-1; Harding 2000: 17, 978).6

One such summary burial is that of a man with an arrowhead in his chest cavity, lying face down in a ditch and covered by a thin layer of chalk rubble, apparently buried where he fell. The ditch forms part of the Neolithic defensive enclosure at Hambledon Hill where the excavators also found a second individual with similar evidence of having been shot. Roger Mercer has interpreted the box rampart, trivallate defences and burning of a long stretch of rampart at Hambledon as further compelling evidence of organized warfare (Mercer 1999) and, together with a group of four other enclosures in southwestern England with evidence of attack, the evidence from Hambledon most likely indicates a region-wide period of warfare during the second half of the fourth millennium BC. Mercer has been at pains to point out that these Neolithic causewayed enclosures also had a ceremonial role: what we may be seeing is the adoption of certain ceremonial sites as suitable places to fortify and defend against attack at particular times. It is interesting that causewayed enclosures elsewhere in Britain have not yielded similar evidence for warfare and the sites in the southwest probably relate to localized wars in this part of Britain. A resurgence of archaeological interest in the appearance of Corded Ware-using communities as immigrants into TRB-using communities in Scandinavia in the third millennium (Kristiansen 1989) coincided with a number of new finds from that period which provide evidence suggestive of warfare. Although Neolithic mass graves are known from central Europe in the much earlier LBK period, notably at Talheim (Germany) and AsparnSchletz (Austria; Stingeder et al. 1999; Romey 2003) around 5000 BC, a new find of a mass grave in a thirdmillennium BC settlement at Grødbygård on Bornholm (Nielsen 1996: fig. 21) may mark a similar atrocity. From the period around 2700 BC there are now half a dozen palisade enclosures - some with Corded Ware pottery and others with TRB ceramics - in the western Baltic that seem not to have had ceremonial or even domestic functions (Svensson 2002). The large quantities of projectile points on one of these defended sites at Gamleborg on Bornholm also point to armed struggle (F.O. Nielsen, pers. comm.). It may well be that this is one of the very few instances so far detected in the prehistoric record of immigrant groups waging war with indigenous communities, although the evidence can be interpreted in terms other than those of warfare and violence. Recent studies of strontium isotope signatures on casualties from Aspern have also claimed to identify ‘outsiders’ amongst the corpses, but this also raises possibilities of alternative explanations.

There is often a mistaken notion that slaves can only be kept so long as there are suitable physical restraints to prevent them from running away, yet slaves throughout the world, today as well as in more distant times, have been coerced by other means. These include magical spells which will cause them to die if they escape as well as more visceral threats of violence and death should they be captured. Colonial administrators were sometimes astonished to be told by slaves that they would not desert their masters. Edmund Leach recounts an administrator’s story from colonial Burma: ‘I asked him [the slave] why he did not run away from his Kachin masters, and he replied that, in the first place, he had no wish to do so, as he was very well treated; and in the second his own people who had sold him had told him that sooner or later they would buy him back.’ (Pritchard 1914 cited in Leach 1954: 303). The demise of the bow and arrow as a weapon: the rise of aristocratic violence Wooden Neolithic longbows from waterlogged deposits are not dissimilar to the medieval bows that brought victory to the English on the battlefields of Crecy and Agincourt. As Mercer has argued (1999), bows became

5

The numbers of Neolithic, Chalcolithic and Beaker-period cases of arrow injuries are, however, beginning to stack up. Guilaine and Zammit list 44 French sites with confirmed cases (2005: appendix 1) and the number of British and French examples is now estimated at around a hundred (Martin Smith, pers. comm.).

6 It could be argued that the adolescent is a relative of the deceased, placed lovingly in his lap, but the posture of of the adolescent is hardly dignified, stretched out across the man’s midriff, and is more suggestive of an offering than a companion.

27

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY efficient human-killing weapons in the Early Neolithic, with drawing-weights not so very different from medieval examples. Arrowheads also underwent evolutionary modifications in the Early Neolithic from multiple to single, heavy flint armatures whilst Late Neolithic and Early Bronze Age arrowheads in various parts of Europe were even more effective for killing people (see Aranda Jiménez and Sánchez Romero this volume). So it is something of a puzzle that this efficient means of killing at a distance declined in importance in central and western Europe after the Copper Age and Early Bronze Age, with the exception of communities on Mediterranean islands such as Sardinia (Osgood et al. 2000: 98)

of a warrior cult across Europe, embellished and embodied by arrays of bronze parade armour that was too thin for use in combat. The presentation of the warrior became focused increasingly on the male body, in the way that it was plucked or shaved with tweezers and razors, and adorned with precious metals and symbols of aggression such as boars’ tusks and ornate parade armour (Treherne 1995). This bodily aesthetic no doubt included comportment as much as physical appearance, training in use of rapier or dagger and a degree of freedom from agricultural labour. The rise of this pan-European warrior cult was the context through which the aristocratic societies of the Late Bronze Age and Early Iron Age came into being.

To the modern eye, this seems to be one of the strangest technological shifts in prehistoric warfare. The Mediterranean empires of the first millennia BC and AD fought eastern armies which made very effective use of the bow whilst Genghis Khan’s mounted archers were triumphant across Asia in the early thirteenth century. Although there are copper and bronze arrowheads from Anatolia, Greece and various parts of the Mediterranean and Europe as far west as Holland, Spain and France (Snodgrass 1964; Mercer 1970; Fontijn 2002: 223), there is only a handful from Britain7 and they are relatively rare in the rest of western Europe, in striking contrast to the thousands of swords and spearheads. Roger Mercer has argued convincingly that these metal arrowheads were used for hunting as opposed to the ‘doubtful use of the bow in the kind of in-fighting that would seem typical of the Bronze Age in Europe’ (1970: 203).

As David Fontijn points out (2002: 229; this volume), Bronze Age warriors were members of an imagined community, living at sometimes considerable distances from each other and only coming together at particular moments. He and other Dutch archaeologists (Fontijn 2002: 224-9; Fokkens 1999; 2001; Louwe Koojimans 1998; Roymans 1996) have explained Dutch Bronze Age martiality and violence as particularly linked to cattleraiding, resulting in small-scale, low-casualty conflicts with little need for defences and defended settlements. Fontijn further interprets Bronze Age sword-fighting as ritualized and guided by specific codes of courage and honour, possibly similar to those portrayed in the Iliad. In this respect, the multiple killings at Middle Bronze Age Velim (Hrala et al. 1992; Hrala et al. 2000) and Tormarton (Osgood this volume) hint at rather different situations of larger-scale atrocities on a par with Neolithic and Early Bronze Age mass graves.8

What caused the European decline of the bow and arrow as a prehistoric killing weapon? From late medieval and early modern times the importance of killing at a distance has dominated military technology to the extent that today’s killers can be hundreds of miles from their victims. Why engage in hand-to-hand combat when one can kill without having to get too close? One simple answer may be that the development of leather body armour and shields reduced the effectiveness of archery, in line with the aggressive-defensive cycle postulated by Hill and Wileman (2002). Yet stone and metal arrowheads can puncture armour of leather and even bronze whilst certain parts of the body - the face, armpits and legs - will always have been vulnerable. If we follow the arguments of Treherne and Randsborg, we should be thinking about the social context of violence and warfare within which technological innovation and use were possible.

The biblical story of David’s killing of Goliath by unexpected means is a well-known example of the subverting of these codes of martial conduct and it raises the issue of why ‘cheating’ - using bows and arrows and other projectiles - was not so widespread as to invalidate face-to-face combat.9 It is all very well to dismiss archery as a cowardly and ‘unsporting’ means by which ignoble farmers can bring down great champions in setpiece or guerrilla-style combat but it can certainly achieve the required end. One answer to this conundrum may lie in the emerging dominant mode of combat, away from mass encounters of lightly armed bands carrying out raids and ambushes, to honour-based confrontations between single individuals or small groups settling grievances and disputes in the manner of feuds and vendettas as well as extracting surplus and other forms of tribute from kin and outsiders. The Middle Bronze Age

What we see from the Middle Bronze Age onwards is the development of weaponry for hand-to-hand combat, as daggers lengthened into rapiers and moulded leather shields came into use. These were items marking the rise

8 Other British and Irish examples of Bronze Age violence, indicated by weapon cuts into bone, are known from Dorchester-on-Thames (see Osgood this volume), Sonna Demesne (Co. Westmeath) and Drumman More Lake (Co. Armagh; Sikora and Buckley 2003). 9 Projectile weaponry such as slings may not survive but must have been in use, perhaps throughout prehistory. Caches of beach pebbles, identified as sling stones, are known from British Iron Age hillforts such as Danebury and Maiden Castle. Otherwise they are difficult to recognize during excavations unless found en masse.

7 Currently there are 25 copper alloy arrowheads from Britain, most of them from East Anglia (Crawford and Wheeler 1921: 138; O’Connor 1980: 102-3; Portable Antiquities Scheme 2005).

28

MIKE PARKER PEARSON: WARFARE, VIOLENCE AND SLAVERY IN LATER PREHISTORY: AN INTRODUCTION Democratization and the State: aristocratic fighting

sees the beginnings of military specialism of warrior elites from which developed the armies and warrior hosts of Late Bronze Age and Iron Age states and tribal confederacies.

the decline of

Bronze swords continued to be used well into the Early Iron Age, as late as the eighth and early seventh centuries BC. Gordon Childe proposed many years ago (1946) that the widespread availability of iron ores freed Europe from the tyranny of bronze-controlling aristocracies, leading to a democratization of European society. Although his brand of technological determinism now seems out-dated, there were concomitant social changes across Europe. Randsborg (1999) has highlighted the importance of the Europe-wide inter-relationship between hoplite-style warfare and the decline of aristocratic and autocratic social structures. Even before about 700 BC there were the beginnings of this shift to lance and shield fighting. As Randsborg points out (ibid.: 202), the new style of fighting not only involved the entire ‘citizenry’, rich and poor, but also allowed for more decisive outcomes and for large-scale military expansion. Many writers have called attention to the strong associations of military force with early states (Goody 1971; Cohen 1984; Ferguson 1990; Brothwell 1999: 33-4) and both state and non-state societies in Late Iron Age Europe were able to wage large-scale and bloody wars.

The rapier - a useless weapon? The weapon that principally embodied this change in martial practices in the Middle Bronze Age was the rapier. At first glance this narrow bronze blade appears to be a peculiarly ineffectual weapon. Unlike Late Bronze Age swords which were cast in one, the rapier handle was merely riveted in place so that it would have been liable to break if used in a slashing motion. The effect of a flesh wound from a rapier would not generally have been enough to bring an enemy down. Archaeologists occasionally declare themselves baffled as to why the rapier, with such limitations, was to remain the weapon of choice for several centuries during the European Middle Bronze Age (Harding 1999: 166). Yet when considered as the successor to the copper, flint or bronze dagger, it is evident that the rapier is a development in a tradition of using stabbing weapons. Not only did it increase the reach of a Bronze Age bladesman but it also inflicted deeper wounds than the stubby daggers of the Early Bronze Age. Fortunately there is evidence of how rapiers were used, from depictions of duels on Minoan and Mycenaean seals, rings and gems. Held above the head, the rapier was used not to rain down blows on the enemy (as some scholars have concluded; Harding 1999: 166-7) but to stab the opponent in the face or throat. The rapier seems to have been adopted in tandem with the shield: it appears that duelling was conducted by jabbing at one’s opponent over the top of his shield, at the same time as attempting to unsight him by making him raise his shield so that he would be unable to see where the next stab was coming from. The groin would be particularly vulnerable at this point. Unlike flesh wounds to the muscle tissue on limbs and torso, stabbings in the face, throat and groin would have had considerable impact on the victim’s ability to continue fighting.

Conclusion Many of those attending the conference in Sheffield at which these papers were given, agreed with Chris Knüsel’s comment during the discussion that the growing realization of the depth and scope of human inhumanity in prehistory was something of a paradigm shift in archaeological perception. As someone who had hitherto subscribed largely to the ‘pacifist’ camp, I was myself particularly impressed by the results presented by many of the speakers. The papers in this volume present a variety of theoretical orientations and there has been no attempt to force the authors into adopting a common perspective. However, my own preference is for an understanding of warfare, violence and slavery within the total social fabric rather than as somewhat independent variables, either as functionalist processes or as technologically driven innovations. It seems that we no longer need to debate whether warfare, violence and slavery ever really happened in the prehistoric past and can now begin to ask more sophisticated questions about why and when and who.

The appearance of the slashing and thrusting sword in the Late Bronze Age, with its single-cast grip and blade, marks not a return to martial common sense but a technological improvement that gave the warrior a wider range of moves. Even so, despite the addition of greater weight to the blade by making it leaf-shaped, these small swords would not have been particularly effective in cleaving leather body armour. The key feature of these swords was also, therefore, their efficacy as stabbing and thrusting weapons. Even with the emergence of Iron Age hoplite phalanxes using spears and shields rather than swords, inflicting injury in the unprotected areas of face, throat and groin seems to have been the aim of combat.

Most archaeologists have moved ground in the last ten years, retreating from the standpoint which plays down the presence of warfare, violence and slavery in prehistoric societies and also, at the opposite extreme, from that which sees their effects everywhere. The latter perspective places too many demands on the archaeological evidence that simply cannot be met since archaeological remains which could add to our knowledge of this subject are often so poorly preserved – 29

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Binford, L.R. 1968. Post-Pleistocene adaptations. In L.R. Binford and S.R. Binford (eds) New Perspectives in Archeology. Chicago: Aldine. 313-41. Black-Michaud, J. 1975. Cohesive Force: feud in the Mediterranean and the Middle East. Oxford: Blackwell. Bourdieu, P. 1977. Outline of a Theory of Practice. Cambridge: Cambridge University Press. Bradley, R. 1998. Interpreting enclosures. In M. Edmonds and C. Richards (eds) Understanding the Neolithic of North-western Europe. Glasgow: Cruithne. 188-203. Brothwell, D.R. 1999. Biosocial and bioarchaeological aspects of conflict and warfare. In J. Carman and A. Harding (eds) Ancient Warfare: archaeological perspectives. Stroud: Sutton. 25-38. Chagnon, N.A. 1990. Reproductive and somatic conflicts of interest in the genesis of violence and warfare among tribesmen. In J. Haas (ed.) The Anthropology of War. New York & Cambridge: Cambridge University Press. 77-104. Chamberlain, A.T. and Parker Pearson, M. 2001. Earthly Remains: the history and science of preserved bodies. London: British Museum. Chapman, J. 1999. The origins of warfare in the prehistory of central and eastern Europe. In J. Carman and A. Harding (eds) Ancient Warfare: archaeological perspectives. Stroud: Sutton. 101-42. Childe, V.G. 1946. The social implications of the three ‘ages’ in archaeological classification. The Modern Quarterly (NS) 1: 18-33. Cohen, R. 1984. Warfare and state formation: wars make states and states make wars. In R.B. Ferguson (ed.) Warfare, Culture and Enviroment. New York: Academic Press. 329-58. Coles, J. and Harding, A.F. 1979. The Bronze Age in Europe. London: Methuen. Crawford, O.G.S. and Wheeler, R.E.M. 1921. The Llynfawr and other hoards of the Bronze Age. Archaeologia 71: 133-40. Durham, E. 1909. High Albania. London: Edward Arnold. Earle, T. 1997. How Chiefs Come to Power: the political economy in prehistory. Stanford: Stanford University Press. Englehardt, C. 1866. Denmark in the Early Iron Age; illustrated by recent discoveries in the peat mosses of Slesvig (Thorsbjerg and Nydam). London: Williams and Norgate. Evans, J.G. 1983. Stonehenge - the environment in the Late Neolithic and Early Bronze Age and a BeakerAge burial. Wiltshire Archaeological Magazine 78: 730. Ferguson, R.B. 1984. Introduction: studying war. In R.B. Ferguson (ed.) Warfare, Culture and Enviroment. New York: Academic Press. 1-81. Ferguson, R.B. 1990. Explaining war. In J. Haas (ed.) The Anthropology of War. New York & Cambridge: Cambridge University Press. 26-55.

even when human bones survive, the marks of violence can be difficult to interpret. Devotees of the murderous and bloody may still yearn for every other skeleton to show evidence of trauma, or for most settlements to have been burnt to the ground and so forth. Yet archaeologists only occasionally make such gory discoveries, to the disappointment of a media and public who lap up every over-sensationalized morsel. On the other hand, the pofaced denial of violence, warfare and even slavery in prehistory can no longer be sustained with any credibility. Prehistory was not the endless bloodbath envisaged by some but was still an insecure and dangerous place in which killings, violence, war and atrocities were committed on a regular and near-global basis. The twentieth century could be described as the century of atrocities (and the twenty-first may be going that way as well) and when we survey the material remains of warfare and violence in that century, on its industrial scale, it becomes apparent how insignificant are the traces from prehistoric times. Is this solely a factor of population sizes with so many more people alive to be slaughtered in the last hundred years? A Neolithic mass grave will never contain the thousands of corpses that are the hallmark of twentieth-century atrocities. We will never be able to answer questions such as whether we are more or less murderous than our prehistoric forebears: to pretend that we can do so not only imputes a dubious moral decline or ascendancy to our present state of ‘humanity’ but also leads to specious notions of what constitutes that humanity. Acknowledgements Although my contribution to the conference was no more than a few words of introduction and discussion, this paper arose out of some of the issues raised over those two days. Subsequently, it developed as a paper presented at a seminar in the Institute of Prehistoric Archaeology in Copenhagen and I would like to thank Klavs Randsborg for his invitation, hospitality and comments. My thanks for helping with the running of the conference go to Ash Lenton, representing the Sheffield University Archaeology Society, who organized the event with considerable efficiency and enthusiasm. Not all of the papers given at the Sheffield conference have been published in this volume since some (Schulting, Mercer, Louwe Koojimans, Fokkens) have been or will be published elsewhere. Bibliography Ardrey, R. 1967. The Territorial Imperative. London: Collins. Bar-Yosef, O. 1986. The walls of Jericho: an alternative explanation. Current Anthropology 27: 157-62. Bender, B. 1978. Gatherer-hunter to farmer: a social perspective. World Archaeology 10: 205-22.

30

MIKE PARKER PEARSON: WARFARE, VIOLENCE AND SLAVERY IN LATER PREHISTORY: AN INTRODUCTION Haas, J. (ed.) 1990b. The Anthropology of War. New York & Cambridge: Cambridge University Press. Haas, J. 1999. The origins of war and ethnic violence. In J. Carman and A. Harding (eds) Ancient Warfare: archaeological perspectives. Stroud: Sutton. 11-24. Hanson, V.D. 1989. The Western Way of War: infantry battle in Classical Greece. Oxford: Oxford University Press. Harding, A.F. 1999. Warfare: a defining characteristic of Bronze Age Europe? In J. Carman and A. Harding (eds) Ancient Warfare: archaeological perspectives. Stroud: Sutton. 157-73. Harding, A.F. 2000. European Societies in the Bronze Age. Cambridge: Cambridge University Press. Harris, M. 1974. Cows, Pigs, Wars and Witches: the riddles of culture. New York: Vintage Books. Harris, M. 1984. A cultural materialist theory of band and village warfare: the Yanamamo test. In R.B. Ferguson (ed.) Warfare, Culture and Environment. New York: Academic Press. 111-40. Hedeager, L. 1990. Iron-Age Societies: from tribe to state in northern Europe, 500 BC to AD 700. Oxford: Blackwell. Hill, P. and Wileman, J. 2002. Landscapes of War: the archaeology of aggression and defence. Stroud: Tempus. Hodder, I. 1990. The Domestication of Europe. Oxford: Blackwell. Hrala, J., Sedlácek, Z. and Vávra, M. 1992. Velim: a hilltop site of the Middle Bronze Age in Bohemia. Report on the excavations 1984-90. Památky Archeologické 83: 288-308. Hrala, J., Šumberová, R. and Vávra, M. (eds) 2000. Velim: a Bronze Age fortified site in Bohemia. Prague: Institute of Archaeology, Academy of Sciences of the Czech Republic. Ilkjaer, J. and Lønstrup, J. 1982. Interpretation of the great votive deposits of Iron Age weapons. Journal of Danish Archaeology 1: 95-103. Ives, S. 2003. Was ancient Alpine ‘iceman’ killed in battle? National Geographic News. http://news.nationalgeographic.com/news/2003/10/10 30_031030_icemandeath.html James, S.T. 1997. Iron Age warfare. British Archaeology 22: 10. James, S.T. forthcoming. A bloodless past: the pacification of Early Iron Age Britain. In C. Haselgrove and R.E. Pope (eds) The Earlier Iron Age in Britain and the Near Continent. Oxford: Oxbow. Keeley, L.H. 1996. War before Civilization: the myth of the peaceful savage. Oxford & New York: Oxford University Press. Kinnes, I., Schadla-Hall, T., Chadwick, P. and Dean, P. 1983. Duggleby Howe reconsidered. Archaeological Journal 140: 83-108. Kristiansen, K. 1989. Prehistoric migrations: the case of the Single Grave and Corded Ware cultures. Journal of Danish Archaeology 8: 211-25. Kristiansen, K. 1999. The emergence of warrior aristocracies in later European prehistory and their

Fokkens, H. 1999. Cattle and martiality: changing relations between man and landscape in the Late Neolithic and the Bronze Age. In C. Fabech and J. Ringtved (eds) Settlement and Landscape: proceedings of a conference in Århus, Denmark, May 4-7 1998. Århus: Jutland Archaeological Society. 3543. Fokkens, H. 2001. Cows, wars and wisdom: martiality and the constitution of man. Paper presented at conference ‘Warfare, violence and slavery in prehistory and protohistory’, University of Sheffield, 2-3 Feb 2001. Fontijn, D.R. 2002. Sacrificial Landscapes: cultural biographies of persons, objects, and ‘natural’ places in the Bronze Age of the southern Netherlands, 2300600 BC.. Leiden: Faculty of Archaeology, Analecta Praehistorica Leidensia 33/34. Frankenstein, S. and Rowlands, M.J. 1978. The internal structure and regional context of Early Iron Age society in south-western Germany. Bulletin of the Institute of Archaeology, London 15: 73-112. Freud, S. 1964. Why war? In L. Bramson and G. Goethals (eds) War: studies from psychology, sociology, anthropology. New York: Basic Books. 7180. Fried, M. 1967. The Evolution of Political Society: an essay in political anthropology. New York: Random House. Friedman, J. 1979. System, Structure and Contradiction: the evolution of ‘Asiatic’ social formations. Copenhagen: National Museum of Denmark. Friedman, J. and Rowlands, M.J. 1978. The Evolution of Social Systems. London: Duckworth. Gibson, T. 1990. Raiding, trading and tribal autonomy in insular southeast Asia. In J. Haas (ed.) The Anthropology of War. New York & Cambridge: Cambridge University Press. 125-45. Giddens, A. 1984. The Constitution of Society: outline of a theory of structuration. Cambridge: Polity Press. Gillman, A. 1976. Bronze Age dynamics in southeast Spain. Dialectical Anthropology 1: 307-19. Gillman, A. 1981. The development of social stratification in Bronze Age Europe. Current Anthropology 22: 1-23. Glob, P.V. 1969. The Bog People. London: Faber and Faber. Godelier, M. 1999. The Enigma of the Gift. Cambridge: Polity. Goody, J. 1971. Technology, Tradition and the State in Africa. Oxford: Oxford University Press. Gostner, P. and Egarter Vigl, E. 2002. Report of radiological-forensic findings on the Iceman. Journal of Archaeological Science 29: 323-6. Guilaine, J. and Zammit, J. 2005. The Origins of War: violence in prehistory. Oxford: Blackwell. Haas, J. 1990a. Warfare and the evolution of tribal polities in the prehistoric Southwest. In J. Haas (ed.) The Anthropology of War. New York & Cambridge: Cambridge University Press. 171-89.

31

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY long-term history. In J. Carman and A. Harding (eds) Ancient Warfare: archaeological perspectives. Stroud: Sutton. 175-89. Kuijt, I. 1996. Negotiating equality through ritual: a consideration of Late Natufian and Prepottery Neolithic A period mortuary practices. Journal of Anthropological Archaeology 15: 313-36. Leach, E.R. 1954. Political Systems of Highland Burma. Cambridge: Cambridge University Press. Lorenz, K. 1967. On Aggression. London: Methuen. Louwe Kooijmans, L.P. 1993. An Early/Middle Bronze Age multiple burial at Wassenaar, the Netherlands. Analecta Praehistorica Leidensia 26: 1-20. Louwe Kooijmans, L.P. 1998. Bronzezeitlichen Bauern in und um die niederländische Delta-Niederung. In B. Hänsel (ed.) Mensch und Umwelt in der Bronzezeit Europas. Kiel. 327-39. McCauley, C. 1990. Conference overview. In J. Haas (ed.) The Anthropology of War. New York & Cambridge: Cambridge University Press. 1-25. Mercer, R.J. 1970. Metal arrow-heads in the European Bronze and Early Iron Ages. Proceedings of the Prehistoric Society 36: 171-213. Mercer, R.J. 1999. The origins of warfare in the British Isles. In J. Carman and A. Harding (eds) Ancient Warfare: archaeological perspectives. Stroud: Sutton. 143-56. Morris, R. 2003. The archaeology of conflict: introduction. Conservation Bulletin 44: 2-3. Mortimer, J.R. 1905. Forty Years’ Researches in British and Anglo-Saxon Burial Mounds of East Yorkshire. London: A. Brown and Sons. Munksgaard, E. 1984. Bog bodies – a brief survey of interpretations. Journal of Danish Archaeology 3: 120-3. Nielsen, F.O. 1996. Forhistoriske Interesser. Rønne: Bornholms Amt. O’Connor, B. 1980. Cross-Channel Relations in the Later Bronze Age. Oxford: British Archaeological Reports (International Series) 91. Ørsnes, M. 1968. Der Moorfund von Ejsbøl bei Hadersleben und die Deutungsprobleme der grossen nordgermainschen Waffenopferfunde. In H. Jankuhn (ed.) Vorgeschichtliche Heiligtümer und Opferplätze in Mittel- und Nordeuropa. Göttingen: Vandenhoeck and Ruprecht. 172-87. Osgood, R.H. and Monks, S. with Toms, J. 2000. Bronze Age Warfare. Stroud: Sutton. Parker Pearson, M. 1984. Economic and ideological change: cyclical growth in the pre-state societies of Jutland. In D. Miller and C. Tilley (eds) Ideology and Power in Prehistory. Cambridge: Cambridge University Press. 69-92. Parker Pearson, M. 1989. Beyond the pale: barbarian social dynamics in western Europe. In J. Barrett, A.P. Fitzpatrick and L. MacInnes (eds) Barbarians and Romans in North-west Europe from the later Republic to late Antiquity. Oxford: BAR International Series 471. 198-226.

Parker Pearson, M. 1993. The powerful dead: archaeological relationships between the living and the dead. Cambridge Archaeological Journal 3: 20329. Portable Antiquities Scheme 2005. Stanningfield arrowhead. http://www.findsdatabase.org.uk/hms/pas_obj.php?ty pe=finds&id=0013EA1612501442. Randsborg, K. 1995. Hjortspring: warfare and sacrifice in early Europe. Århus: Århus University Press. Ricoeur, P. 1988. Time and Narrative, Volume 3. Chicago: Chicago University Press. Robarchek, C. 1990. Motivations and material causes: on the explanation of conflict and war. In J. Haas (ed.) The Anthropology of War. New York & Cambridge: Cambridge University Press. 56-76. Romey, K.M. 2003. Neolithic nightmare. Archaeology 56 (5): 13. Roymans, N. 1996. The sword or the plough: regional dynamics in the Romanisation of Belgic Gaul and the Rhineland area. In N. Roymans (ed.) From the Sword to the Plough: three studies on the earliest Romanisation of northern Gaul. Amsterdam: Amsterdam Archaeological Studies 1. 9-126. Schulting, R.J. and Wysocki, M. 2002. Cranial trauma in the British earlier Neolithic. Past 41: 4-6. Sikora, M. and Buckley, L. 2003. Casting new light on old excavations. Archaeology Ireland 17 (1): 16-19. Snodgrass, A. 1964. Early Greek Armour and Weapons from the Bronze Age to 600 BC. Edinburgh: Edinburgh University Press. Spindler, K. 1993. The Man in the Ice. London: Weidenfeld and Nicolson. Spriggs, M. (ed.) 1984. Marxist Perspectives in Archaeology. Cambridge: Cambridge University Press. Stordeur, D., Helmer, D. and Willcox, G. 1997. Jerf el Ahmar: un nouveau site de l'horizon PPNA sur le moyen Euphrate Syrien. Bulletin de la Société Préhistorique Française 94: 282-5. Svensson, M. 2002. Palisade enclosures - the second generation of enclosed sites in the Neolithic of northern Europe. In A.M. Gibson (ed.) Behind Wooden Walls: Neolithic palisaded enclosures in Oxford: BAR Publishing, BAR Europe. International Series 1013. 28-58. Teschler-Nicola, M., Gerold, F., Bujatti-Narbeshuber, M., Prohaska, T., Latkoczy, C., Stingeder, G. and Watkins, M. 1999. Evidence of genocide 7000 BP – Neolithic paradigm and geoclimatic reality. Collegium Antropologium 23: 437-50. Treherne, P. 1995. The warrior's beauty: the masculine body and self-identity in Bronze-Age Europe. Journal of European Archaeology 3: 105-44. van der Sanden, W. 1996. Through Nature to Eternity: the bog bodies of northwest Europe. Amsterdam: Batavian Lion International. H. 2003. Commemorative tales: Vandkilde, archaeological responses to modern myth, politics, and war. World Archaeology 35: 126-44. 32

MIKE PARKER PEARSON: WARFARE, VIOLENCE AND SLAVERY IN LATER PREHISTORY: AN INTRODUCTION Vayda, A.P. 1976. War in Ecological Perspective: persistence, change and adaptive processes in three Oceanian societies. New York: Plenum. Vencl, S. 1999. Stone Age warfare. In J. Carman and A. Harding (eds) Ancient Warfare. Stroud: Sutton. 57-72.

Wahl, J. and König, H.G. 1987. Anthropologischtraumatologische Unterschung der menschlichen Skelettreste aus dem bandkeramischen Massengrab bei Talheim, Kreis Heilbronn. Fundberichte aus Baden Württemberg 12: 65-193. Wendorf, F. (ed.) 1968. The Prehistory of Nubia. Dallas: Southern Methodist University Press.

33

Aggression and nonhuman primates Pia Nystrom Department of Archaeology, University of Sheffield

Zuckerman’s observations on the hamadryas baboons at London Zoo which, owing to their unnaturally crowded conditions, displayed high levels of aggression (Zuckerman 1932). However, reports from pioneering researchers who studied monkeys and apes in their natural habitat all stressed the lack of aggressive behaviours: fights were extremely rare and those that did occur were of very short duration and usually did not inflict wounds. It thus appeared that in their natural habitats nonhuman primates were peaceful and did not harm anyone, especially not their own (e.g. Russell and Russell 1968). This behaviour was perceived as a sharp contrast to that of humans and their hominid ancestors who were depicted as more than capable of assaulting and killing their own kind (Ardrey 1961; Washburn and Lancaster 1968; but see Russell and Russell 1968). It was therefore a shock when Jane Goodall reported in the early 1970s that some of the Gombe chimpanzees had attacked and killed members of another group (Goodall et al. 1979). Furthermore, reports of infant killings began to appear in scientific journals (e.g. Goodall 1977; Hrdy 1977; Mohnot 1971).

Traditionally we humans have placed ourselves apart from other animals and even other primates. We list many traits that define us as human and as distinct from other primate species, e.g. language, tool use, higher mental ability and culture. Yet we are primates, just as much as, for example, chimpanzees and baboons. As we learn more about the behaviour of our fellow primates, it is becoming increasingly clear that we may share a lot more with them than we anticipated in the past. There is, however, one sphere where we may exceed all other primates. Humans are an emotional species, capable of extreme compassion and extreme violence, not just towards conspecifics but also towards other species. The quest to understand and to seek the roots of our own behaviour has a long history, and the expression and function of aggressive behaviours have received special attention (e.g. de Wit and Hartup 1974; Karli 1991; Lorenz 1967; Silverberg and Gray 1991). Despite our close ties with other primates, it has been questioned if, owing to our empathetic and cognitive mind, nonhuman primates are good models for describing or explaining our own behaviours, and this is especially true when dealing with aggression and violent behaviour.

This newfound knowledge about bloodthirsty monkeys and apes led to a surge in behavioural observations on a range of different wild primates, with a special focus on intra-group and inter-group relationships. There was a need to explain why primates displayed such aggressive behaviours. Many wished to find extrinsic reasons for these aberrant behaviours, such as overcrowding, influence by humans, etc., and this became known as the social pathology theory (Bartlett et al. 1993; Curtin and Dolhinow 1979).

In this paper I shall give a brief overview of what we know about how primates deal with group living - the conflicts they may encounter, when tension may most likely be displayed and how they attempt to circumvent conflict. It is essential to stress that it is not possible to give a single view on aggression in primates because nonhuman primates are so diverse. Even though I shall attempt to give a holistic view, it is impossible to give equal weight to all species. More attention will be paid to more well studied species (e.g. baboons and chimpanzees) and those that have displayed some unusual behavioural pattern (e.g. langurs, the Asian leaf-eating monkeys).

Sociobiological theory, which was formulated from a basis of studies of insect and bird behaviour, states that the ultimate goal of each individual is to maximize his or her own reproductive success (Wilson 1975). According to this theory, competition for limited resources is the ultimate cause for aggressive interactions, both in case of intragroup and intergroup conflicts. Adult males are more likely to compete with each other for access to receptive females while females are more likely to compete for access to food sources to ensure the survival of their offspring. Since aggressive behaviours are used to establish and hold on to dominance or rank, at least in case of most males of a species, aggression would ultimately be beneficial.

Background Aggression has traditionally been considered an antisocial behaviour and yet, over the past few decades, much attention has been placed on the possible adaptive consequences of aggressive behaviours. It has been suggested that the threat of aggression is a tactic used in social interactions to gain a specific goal; such a threat may never be realised, however, because of the potential risks to the actor. Thus, aggression may be seen as having a pro-social effect rather than being antisocial (Aureli and de Waal 2000; de Waal 1996).

Over the past decade there has been a switch in research approaches away from looking for ultimate goals (which often are impossible to prove) to more proximate causes of behaviour. Less focus is placed on aggression and competition per se. Instead, conflict resolution and

Prior to the 1950s little was actually known about the behaviour of nonhuman primates, especially in their natural habitats. The most popular account was Lord 35

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY highly variable. Many factors can impinge on the behavioural pattern of a primate group:

affiliation have received more attention. Rather than focus on the individual, scholars give more attention to the complex social networks present in primate social groups and how individuals fit in to the greater whole of the social context. Frans de Waal proposes a ‘relational model’ in which aggression is a product of social decision making, with aggressive behaviour being just one of several options open to an animal when dealing with conflicts (de Waal 1996). Importantly, if an aggressive interaction has taken place, the participants tend to seek some form of reconciliation to reduce tension and prevent future conflicts. Over the past decade a great abundance of studies has examined how different primate species resolve conflicts (see Aureli and de Waal 2000).

1.

2.

3.

What is aggression and what may trigger its expression? To find a definition of aggression that is universal - in other words, one that can be used across species boundaries - has proven difficult. Silverberg and Gray (1991) suggest we should define aggression as ‘an act with the intent to harm’. The problem is that intent cannot be inferred in most animal species. To overcome this problem, people who study animals tend to use the behavioural term ‘agonism’. It encapsulates a continuum of behavioural expressions ranging from highly aggressive to non-aggressive or submissive behaviours. Each behaviour is recorded as a descriptive actionreaction sequence without any inference of intentionality. It is possible to rank aggressive or agonistic behaviours from low to high intensity. For example, the stare threat or flashing the white eye lids are relatively mild threats. Such a threat functions as a warning, indicating that the actor is not very pleased and may do something more drastic if the recipient does not alter its behaviour. An increase in aggressive intensity can be seen when a male yawns and shows off his large canine teeth. A further increase in aggressive intensity can be seen when an individual lunges at or chases another. When there is real trouble, there may be actual physical fighting.

4.

The social structure directly influences the type and frequency of social behaviour. The larger the social group, and the more adult males and females present, the more complex the social life. Dispersal and philopatry affect the social networks. Related individuals who remain in their natal group usually form the social core and interact at a higher frequency than do unrelated individuals. The female ovulation schedule is of great importance. In some species, females ovulate and are receptive once a year; in other species, females ovulate on approximately a monthly basis. Male-male competition escalates when females are receptive. It is not usual for females to copulate outside of the ovulatory phase. Humans and bonobos appear to be the only exceptions. The level of sexual dimorphism between males and females influences dominance relationships, and thus how much and what kind of social interaction will take place, especially aggressive interactions.

In most primate species, group members are organized into some sort of dominance hierarchy. Some rank high, while some rank low relative to each other. In general, immature individuals rank below adults and males more often than not rank above females. Where females are about the same size as males, dominance may be more equal or even reversed. There are very few species that can be considered egalitarian (e.g. bonobos, woolly spider monkey). The position of each individual within the dominance hierarchy is well known to and understood by each member of the social group. An individual’s position within the social dominance hierarchy is seldom stable over time. This is most obvious in the case of males where loss or gain of rank is often dictated by the individual’s age and health. Females form dominance hierarchies that are more stable over time and, in some species, a female infant, as it matures, gain the same rank as its mother.

Submissive behaviours may be shown in response to aggressive signals or when a dominant individual approaches. These include active avoidance of an aggressor, or giving a fear grimace or presenting the rear end to the opponent. Submissive behaviours are often used as appeasement signals.

Intragroup relationships Male-male interactions

No member of a social group is immune from aggression. While some individuals may be more likely targets than others, there is great variability according to species as well as to context (both social and environmental). Aggressive interactions are not common, however, and little time is actually devoted to such activities. Most primate species spend most of their day searching for food and eating (50-60%). Another 30-40% of the day may be taken up with resting, leaving some 5-15% for social activities. The daily activity pattern is, of course,

In general, males have an uneasy relationship with each other, especially if they are unrelated. A species’ social structure and mating pattern have a direct influence on the type of interactions that take place. In one-male groups (comprising a single adult male and several adult females and their offspring; e.g. gorillas), males are, in general, intolerant of each other. As male offspring mature, they are ejected from the group by the male leader. This intolerance of other males is most prevalent

36

PIA NYSTROM: AGGRESSION AND NONHUMAN PRIMATES usually well rewarded. Even the stoutest of friends may, however, fall out over a female.

when there are females around. In many species, extra males happily form bachelor groups where many affiliative interactions take place. The male leader (the silverback in case of gorillas) maintains constant vigilance against usurper males. In species where females mate once a year, this is a very turbulent time since much competition and fighting takes place as bachelor males seek to mate with the females and possibly to take over the leadership of the group. Injury is not uncommon during these conflicts, and no age or sex group is completely safe.

Female-female interactions Where females remain in their natal group, the social core is formed by different matrilines. Females form their own dominance hierarchies which tend to be stable over time. In species where females leave their natal group, females are less likely to form strong bonds. Migrating females have to establish their dominance rank as they enter a new group. This is mainly done through agonistic interactions with other females and a migrating female may solicit support from the new group’s males.

In multi-male groups, males have to deal with each other or at least accept each others’ presence. There are several mechanisms for doing so: 1.

2.

3.

There is often strong competition between matrilines for limited resources such as feeding or resting sites. Aggression between females is no less frequent than between males but inflicts less severe physical injury. It is, however, not unusual to see for example female baboons with scarred tails, a favourite place for females to bite each other. To protect themselves against harassment, females may seek friendships with adult males who, for a price, will protect them.

By establishing affiliative bonds. This is almost unheard of amongst male primates. The exceptions are species where related males form the core of the social group. Male chimpanzees are a prime example, where friendly physical contact, such as grooming, occurs between males. By forming a dominance hierarchy. Dominance hierarchies are usually based on dyadic interactions but males may form coalitions or alliances with other males to improve their position. However, the effect tends to be temporary. By avoidance. Males may disperse within the social group and keep at a distance from each other.

Females spend most of their social time affiliating, grooming their offspring, female friends and relatives, or male friends. Females have strong networks of kin and friends who will come to their aid if they are in trouble. In species where females are unrelated, their attention is more on their own offspring and the male leader. Female gorillas compete with unrelated females for access to either good food sources or the silverback male, and more often than not it is the male who has to settle these disputes. Bonobo females also leave their natal groups but in this species unrelated females form very strong affiliative bonds. Bonobos are unique in that most conflicts are solved by use of sex and all group members participate.

When a stranger male enters a new social group, he has to negotiate his position vis-à-vis the other individuals already in the group. When males first meet they tend to ‘size each other up’. All species have a set of behaviours which will give an initial evaluation of a male’s ability. This may entail showing big canine teeth, or swinging around in trees shaking branches, amongst other actions. Baboon males, for example, have a set of ritualized greeting behaviours that are used to confirm and reaffirm the social relationship between each pair. A greeting may entail just passing by each other rapidly while vocalizing. Or it may entail actual contact, where one or both participants inspects and touches the other’s rear end or genitals. These greetings take place on a daily basis and serve as a buffer against aggressive interactions. After a conflict, the males go through these ritualized greetings to reaffirm their dominance relationships.

Male-female interactions In groups where mating takes place once a year, males are often peripheral both physically and socially, especially if there is only a single adult male present. In such groups, females interact more with each other and their offspring. In species where females are receptive on a monthly basis, and there are more males around, males and females interact more frequently. Since females tend to be either pregnant or lactating for most of their adult life, very little time is actually spent in the ‘attractive’ receptive phase. Males use a variety of strategies to gain access to females. Where sexual dimorphism is great, males may use their size dominance to coerce a female to mate with them. In chimpanzee society, for example, a receptive female who does not immediately seize the opportunity to consort with an interested male may be attacked and beaten until she consents to a consortship. Likewise orangutan males,

Chimpanzee males also form dominance hierarchies. The establishment of these can be very noisy and violent affairs, with much posturing and fighting. After a conflict, however, chimpanzee males show a wide range of affiliative behaviours toward each other, such as touching and grooming. Chimpanzee males often form alliances with each other, either as part of the dominance hierarchy or during hunting. Friends and allies are

37

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY infants grow older, male friends (not necessarily fathers) often act as baby-sitters and may even play with and reassure them.

usually young males, have been observed to use force to mate with females (Galdikas 1985; Schürmann 1981). In baboons, where size dimorphism is also large, females may seek protection against an aggressive male from other males or kin. In general, when there are many males present females can express choice in who they interact or mate with. Special relationships may exist between specific males and females, so-called friendships, which may last for years (Smuts 1985). In such a relationship a male’s rank is not that important, as females appear to use other criteria to select male friends. Forming an affiliative relationship with a female can be an alternative strategy for a lower ranking male seeking to reproduce. The female benefits as a male friend protects her and her offspring against harassment from other group members. In some species, females appear to prefer to mate with their male friends, in others it is more random. If a male fails in his duty as a protector, the female will terminate the friendship. There is always another male waiting to take on the role.

A curious behaviour that has been recorded only for baboons and some other Old World monkeys is when a male carries or holds an infant while approaching a rival male. Why males do this has been given almost as many explanations as there are observations. It is very likely part of the quest for dominance, since the actors are often younger males in the process of advancing their rank. Most often the rival male will leave but sometimes a fight may ensue and occasionally the infant may get hurt or killed. The first reports of infanticide in the Asian leaf-eating langur monkeys some 30 years ago were believed to be an irregularity, restricted to a single population or species (Mohnot 1971). In the intervening years, however, an increasing number of cases of what appears to be infant killings have been reported, and from a multitude of different species and in different ecological and social settings (van Schaik 2000). Infants are killed mostly by immigrant or extra-group males who are seeking either high rank or access to receptive females, or both. Whether these infant killings are adaptive or an epiphenomenon is still hotly debated.

In species where sexual dimorphism is small or reversed, females have an easier time dealing with males, and may even boss the males around. Where females form the social core of the group, strong female alliances can be used effectively against harassing males. Females use counter strategies to appease potentially aggressive males such as, for example, spending a lot of time grooming group males, especially the male leader or friend. Bonobo females use sex as a means to appease males (de Waal 2000; White 1996).

To explain the behaviour of langurs, Hrdy (1977) evoked the sexual selection hypothesis. If a male kills an unweaned baby its mother will resume cycling and be available for mating sooner. The male would thus be able to ensure or improve his own reproductive success (see Hausfater and Hrdy 1984; van Schaik and Janson 2000). However, not everyone agrees that infanticide is an adaptive strategy to improve a male’s reproductive success. Rather it may be regarded as an abnormal behaviour, driven by external factors such as overpopulation, or temporary low resource availability (Dolhinow 1977). Alternatively, infant killing may be an epiphenomenon of overall male aggressiveness (Bartlett et al. 1993) with infants not being targets for killing but ‘in the way’ during male-male conflicts. This may be a likely explanation in multi-male social groups where infants often get in the middle of ‘squabbling’ between adult males, especially during intergroup encounters where both males and females may be involved in defending their territories.

Interactions with infants The birth of an infant elicits much attention from all group members, including adult males. In many primate species, newborn infants are distinctive by having a different coloured pelage (coat or fur) compared with juveniles and adults. The largest difference is seen in species where infant killings has been reported. For example, in some langurs, adults have a dark grey pelage while infants are bright orange. In general, it is the mother that provides parental care. Only in the South American marmosets and tamarins do males habitually participate in the rearing by carrying the infants. Baboon males may occasionally carry the offspring of their female friends during travel. Allomothering, the care-taking of an infant by unrelated females, is common in langurs and may be a counterstrategy to infanticide.

When considering these different hypotheses, however, it has to be taken into account that infant killings are not common. Maybe the potential for infanticide is higher but females have evolved successful counter-strategies. Intuitively, the killing of a conspecific may not appear as an adaptive strategy, especially as many conditions have to be met, e.g. females whose infants have been killed may not show a reduced interbirth interval, nor is there a guarantee that she will mate with the killer. Furthermore, it is not always obvious that the adult male and the target

Mothers are often reluctant and wary of allowing adult males close to their infants. The atmosphere is often tense when an adult male approaches, even if it is a friend. Males come in a variety of forms, however: some are gentle and cuddle or groom the infant while others seem more wary or even antagonistic. Owing to the underpinning tension, males often avoid infants but, as

38

PIA NYSTROM: AGGRESSION AND NONHUMAN PRIMATES infant are unrelated. Infant killing among primates remains a behavioural enigma.

adult males (ibid.). At Gombe females rarely participate on border raids.

Intergroup relationships

We do not yet understand the why and the how of chimpanzee border politics yet their behaviours show such purpose and strategic planning that they falls into a different category outside the usual primate territorial behaviour (Boesch and Boesch-Achermann 2000). Manson and Wrangham (1991) have suggested that the underpinning principles for intergroup aggression may be the same in chimpanzees and humans, e.g. male-male competition for access to females and preservation of kin groups.

Most primate species are not directly territorial but they do pay attention to who enters onto their turf. Group ranges usually show some overlap but each group has an exclusive core area that is actively defended. Neighbouring groups tend to avoid each other. Even when several groups utilize the same fruit trees, for example, the visits tend to be staggered, although group size often dictates who has priority of access. When group encounters do occur there is usually much excitement. One group may chase the other away, or they may take a good look at each other then walk off in different directions. In most primate species, group encounters more commonly entail aggressive posturing rather than actual physical combat. However, in the early 1970s Jane Goodall presented evidence that chimpanzees at Gombe in Tanzania actually conducted what appeared to be organized border raids and killed individuals from a neighbouring community (Goodall et al. 1979). Both males and females were targets during these raids. Sometimes infants of the attacked females were killed and some even cannibalized. The observations from Gombe are not unique: similar reports come from other sites in East Africa (e.g. Nishida et al. 1985).

Conclusion Primates are highly social animals and expert at manipulating their social environment. There are constant negotiations taking place between individuals to keep the peace, or at least to keep the cost to a minimum. Non-human primates do not usually display rampant, uncontrolled aggression. Conflicts do, of course, take place both within and between social groups but built into the social fabric are the means to control social tension and conflicts. There are ways to confirm and reaffirm individual positions in the group without having to resort to aggressive acts. Even if these unwritten social contracts are broken, aggression is only one of several options available to resolve the conflict.

These reports have resulted in much speculation. Ecological pressures and resource competition may be the basis for the attacks. Jane Goodall has proposed that the killers may have wanted to recruit young females to their own group – as they were never killed - by removing their mothers and male protectors. The behaviours displayed are curious because, according to Goodall, the attackers behaved as if they were hunting and the attacked individuals were treated as prey (Goodall 1986). Furthermore, the group that was attacked had once been part of the same community as the attackers. This begs the question when did the victims pass from the ‘us’ to ‘them’ status, and from being a ‘conspecific’ to ‘prey’?

The use of nonhuman primates as models to understand our own behaviours may function well if we consider everyday competition for limited resources. If we want to focus on what may be more disturbing aspects of human behaviour, notably violence and warfare, these models may hold less power. It is possible that a certain cognitive level has to be reached before planned violent aggression can be expressed. The more complex mental ability seen in humans may have provided a unique way of dealing with the world. Intentionality and the ability to project a false reality may be abilities limited to humans although there is some evidence that chimpanzees and bonobos may have at least some rudimentary level of these abilities (Savage-Rumbaugh and Lewin 1994). The behaviours displayed by chimpanzees during intercommunity conflicts therefore command more attention and contemplation. Even though border raids on neighbouring communities have been reported from all sites where chimpanzees are studied, only at the East African sites have these aggressive behaviours escalated to the point that individuals were killed. It is the highly planned and ritualistic behavioural pattern that makes these events so chillingly different.

Conflicts between neighbouring chimpanzee communities have also been reported from the Taï forest in West Africa, but here border attacks have never resulted in severe wounding or killings (Boesch and BoeschAchermann 2000). The Boesches have suggested that the behaviours displayed are directly tied to visibility. In Gombe long-distance visibility is good and groups can keep track of the strength of their neighbours (i.e. the number of adult males present), while in the Taï forest visibility is poor and it is not possible to know the strength of the neighbouring communities. Thus at Gombe the attacking group may have realized the asymmetry of power and therefore did not fear reprisals. In contrast, the Taï forest chimpanzees can never be sure about winning or losing; to improve the odds, adult females participate in attacks almost as frequently as

Bibliography Ardre, R. 1961. African Genesis. New York: Dell Publishing Co. Inc.

39

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Manson, J.H. and Wrangham, R.W. 1991. Intergroup aggression in chimpanzees and humans. Current Anthropology 32: 369-90. Mohnot, S.M. 1971. Some aspects of social changes and infant-killing in the hanuman langur, Presbytis entellus (Primates: Cercopithecidae) in western India. Mammalia 35: 175-98. Nishida, T., Hiraiwa-Hasegawa, M., Hasegawa, T. and Takahata, Y. 1985. Group extinction and female transfer in wild chimpanzees in the Mahale National Park, Tanzania. Zeitschrift für Tierpsychologie 67: 284-301. Russell. C. and Russell, W.M.S. 1968. Violence, Monkeys and Man. London: Macmillan. Savage-Rumbaugh, S. and Lewin, R. 1994. Kanzi: the ape at the brink of human mind. New York: John Wiley and Sons Inc. Schürmann, C.L. 1981. The courtship and mating behavior of wild orangutans in Sumatra. In A.B. Chiarelli and R.S. Corruccini (eds) Primate Behavior and Sociobiology. Berlin: Springer-Verlag. 130-5. Silverberg, J. and Gray, J.P. 1991. Violence and peacefulness as behavioural potentials of primates. In J. Silverberg and J.P. Gray (eds) Aggression and Peacefulness in Humans and Other Primates. Oxford: Oxford University Press. 1-36. Smuts, B.B. 1985. Sex and Friendship in Baboons. New York: Aldine de Gruyter. van Schaik, C.P. 2000. Infanticide by male primates: the sexual selection hypothesis revisited. In C.P. van Schaik and C.H. Janson (eds) Infanticide by Males and its Implications. Cambridge: Cambridge University Press. 27-60. van Schaik, C.P. and Janson, C.H. (eds) Infanticide by Males and its Implications. Cambridge: Cambridge University Press. Washburn, S.L. and Lancaster, C.S. 1968. The evolution of hunting. In S.L. Washburn and P.C. Jay (eds) Perspectives on Human Evolution, Vol. 1. New York: Holt, Rinehart and Winston. 213-29. Wilson, E.O. 1975. Sociobiology: the new synthesis. Cambridge MA: Harvard University Press. White, F.J. 1996. Comparative socio-ecology of Pan paniscus. In W.C. McGrew, L.F. Marchant and T. Nishida (eds) Great Ape Societies. Cambridge: Cambridge University Press. 29-41. Zuckerman, S. 1932. The Social Life Of Monkeys And Apes. London: Kegan, Paul, Trench, Trubner and Co.

Aureli, F. and de Waal, F.B.M. 2000. Why natural conflict resolution? In F. Aureli and F.B.M. de Waal (eds) Natural Conflict Resolution. London: University of California Press. 3-10. Bartlett, T.Q., Sussman, R.W. and Cheverud, J.M. 1993. Infant killing in primates: a review of observed cases with specific reference to the sexual selection hypothesis. American Anthropology 95: 958-90. Boesch, C. and Boesch-Achermann, H. 2000. The Chimpanzees of the Taï Forest: behavioural ecology and evolution. Oxford: Oxford University Press. Curtin, R.A. and Dolhinow, P.J. 1979. Infanticide among langurs – a solution to overcrowding? Science Today 13: 35-41. de Waal, F.B.M. 1996. Good Natured: the origins of right and wrong in humans and other animals. Cambridge MA: Harvard University Press. de Waal, F.B.M. 2000. The first kiss. Foundations of conflict resolution research in animals. In F. Aureli and F.B.M. de Waal (eds) Natural Conflict Resolution. London: University of California Press. 15-33. de Wit, J. and Hartup, W.W. 1974. Determinants and Origins of Aggressive Behaviours. The Hague: Mouton. Dolhinow, P.J. 1977. Normal monkeys? American Scientist 65: 266. Galdikas, B.M.F. 1985. Subadult male orangutan sociality and reproductive behavior at Tanjung Putting. International Journal of Primatology 9: 1-35. Goodall, J. 1977. Infant killing and cannibalism in freeliving chimpanzees. Folia Primatologica 28: 259-82. Goodall, J. 1986. The Chimpanzees of Gombe: patterns of behavior. London: Harvard University Press. Goodall, J., Bandora, A., Bergmann, E., Busse, C., Matama, H., Mpongo, E., Pierce, A. and Riss, D. 1979. Inter-community interactions in the chimpanzee population of the Gombe National Park. In D.A. Hamburg and E.R. McCown (eds) The Great Apes. Menlo Park CA: Benjamin/Cummings. 13-53. Hausfater, G. and Hrdy, S.B. 1984. Infanticide: comparative and evolutionary perspectives. New York: Aldine de Gruyter. Hrdy, S.B. 1977. Infanticide as an evolutionary strategy. American Scientist 65: 40-9. Karli, P. 1991. Animal and Human Aggression. Oxford: Oxford University Press. Lorenz, K. 1967. On Aggression. London: Methuen.

40

Sociobiology, cultural anthropology and the causes of warfare Robert Layton Department of Anthropology, University of Durham

Sillitoe defines war as ‘a relationship of mutual hostility between two groups where both try by armed force to secure some gain at the other’s expense’ (Sillitoe 1978: 252; cf Ember and Ember 1997: 3). The frequency of warfare among human populations has led some to argue that warfare is the product of an inherent human disposition, a genetically determined drive to aggression. Since Goodall (1986) and Nishida (1979; Nishida et al. 1985) reported cases of chimpanzees extending their territories by attack on adjacent groups, a direct connection has been alleged between male chimpanzee aggression and human warfare. Chagnon’s ethnography of the Yanomamö (1968), originally subtitled The Fierce People, has often been cited as evidence of an intrinsic link between warfare and natural selection.

his theory, Neel and Chagnon greatly exacerbated and probably started an epidemic of measles that killed ‘hundreds, perhaps thousands’ of Yanomami, by injecting them with a virulent vaccine (Edmonston B). The vaccine was counter-indicated by medical experts for use on isolated populations with no prior exposure to measles. ‘This nightmarish story’, Turner and Sponsel wrote, [is] ‘a real anthropological heart of darkness beyond the imagining of even a Joseph Conrad (though not, perhaps, a Josef Mengele)’. Chagnon’s work has been directed toward portraying the Yanomamö as exactly the kind of originary human society envisaged by Neel, displaying a Hobbsian state of savagery. Chagnon had ‘recooked’ his data to fit sociobiological predictions and deliberately fomented conflicts between Yanomami as, for example, during Asch’s films The Feast and The Ax Fight, where artificial villages were used for sets. Not only was Chagnon’s work exploited by Venezuelan politicians and gold miners to justify massacres of Yanomami and expropriation of their land, Chagnon himself joined forces with corrupt politicians to gain control of Yanomami lands for illegal gold mining and continued anthropological access.

The Yanomamö debate The shocking claims of malpractice levelled by Tierney against Chagnon and his supervisor Neel have recently renewed debate on the ‘naturalness’ of warfare in simple human societies. Tierney, a journalist who had worked in the South American rain forest, interviewed anthropologists, missionaries and others who were familiar with Chagnon’s work among the Yanomamö. News of Tierney’s book Darkness in El Dorado circulated, before its publication, through an email which Terry Turner and Leslie Sponsel sent to Louise Lamphere, President of the American Anthropological Association in late August 2000. The email was rapidly disseminated across the Internet.

Within a couple of months, the Provost of the University of Michigan (under whose auspices the research had been conducted) issued a rebuttal. Turner and Sponsel had been interviewed by Tierney in 1995 and had therefore long been aware of Tierney’s forthcoming book. Roche’s experiments on radioiodine uptake in the thyroid among Yanomami relied on a commonly used technique to investigate the effects of altitude. The Edmonston B vaccine, moreover, was not responsible for the 1968 measles epidemic, but instead greatly reduced its impact. Tierney’s claim that Chagnon was directly or indirectly responsible for endemic warfare among the Yanomamö was absurd, since warfare has existed in South America for 3,500 years. Yanomami describe themselves as fierce and valiant; Chagnon merely borrowed their own term as a subtitle for his book. Gold miners are often violent toward indigenous people, so Chagnon could not be blamed for the violent behaviour of 40,000 gold miners in Yanomami territory.

Turner and Sponsel reported that they had seen galley proofs of Tierney’s book, which would reveal the whole Yanomamö research project conducted by Chagnon and his supervisor James Neel was an outgrowth of the U.S. Atomic Energy Commission’s secret programme of experiments on human subjects. Neel was in charge of a research project to study the effects of the atomic bombs dropped on Hiroshima and Nagasaki on survivors, of which the Yanomamö project was an offshoot. Neel believed in the existence of a gene for ‘leadership’ or ‘innate ability’. In small-scale societies, male carriers of this gene would gain access to a disproportionate share of the available females, thus reproducing their own gene more frequently than less ‘innately able’ males.

Some of the Michigan University rebuttals are valid, others questionable. The vehemence of Turner and Sponsel’s email is an expression of a current debate in the U.S. between sociobiology and cultural anthropology. Among the most outspoken of sociobiologists are Tooby and Cosmides, who accuse cultural anthropologists of treating the human mind as an empty vessel waiting to be filled with learned cultural conventions. Cosmides and

A colleague of Neel’s, Marcel Roche, conducted experiments on Yanomami that involved injecting them with radioactive iodine and analysing blood samples for genetic data. Roche’s project apparently led Neel to choose the Yanomamö as a test case for his theory. Neel wanted to discover the impact that an epidemic would have on this process of natural selection. In 1968, to test 41

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Tooby argue evolution has, on the contrary, given the human mind a complex structure with many specific skills that enable the individual to decide, for example, who to mate with, when to co-operate with others, and how much parental care to expend on each offspring (Cosmides et al. 1992: 73). These skills developed by natural selection during the long period our species lived by hunting and gathering and are virtually universal among members of our species. Variation in human behaviour can be explained as the emergence of local adaptations predicated on the mind’s inherent skills rather than random cultural variation. As Fischer comments,

Sciences accepts that Edmonston B was more reactive than later vaccines, and that some Yanomami reacted with fever after inoculation, but argues the vaccine was licensed and has been successfully used on other populations similar to the Yanomamö (statement by President Bruce Alberts regarding Darkness in El Dorado, on the National Academies web site). Fischer insinuates that Neel might have obtained the somewhat outdated Edmonston B vaccine cheaply. ‘The donation of drugs near expiry by pharmaceutical companies to gain a tax write-off is not necessarily a bad thing, although it can lead to abuses’ (Fischer 2001a: 11). Fischer cites a letter Neel wrote to a missionary in April 1968 discussing using a vaccine that is ‘a little old’. He recommends doubling the standard dose and ignoring the label’s recommendation not to use gamma globulin. Neel asked the missionary for feedback on how the dosage worked, ‘feedback’ comments Fischer ‘that would be useful for the pharmaceutical company and for Neel in obtaining more donations’. The American Anthropological Association Task Force concluded Neel used the Edmonston B vaccine either because the manufacturers did not make the (newer) alternative available to him, or because he believed the more reactive vaccine would be more effective. Neel planned the vaccination campaign before the epidemic began, but he did observe the course of the epidemic, leaving room for a conflict of interest between the 1968 expedition’s humanitarian and scientific aims (American Anthropological Association 2002: 1.25-7).

‘What seems to infuriate cultural anthropologists about sociobiologists is their insistence on extrapolating from quite interesting statistics of animal mating and patterns of investment in care of offspring, and the various predictive models that can be made of these patterns, to the Vietnam War or the decisions of the Supreme Court’ (Fischer 2001a: 13). Tooby quickly jumped to Chagnon’s defence (http://slate.msn.com/HeyWait/00-10-24/HeyWait.asp), and pointed out many inaccuracies in Tierney’s citations (see also Ruby 2000). Tooby also pointed out that Turner and Sponsel were long-time adversaries of Chagnon (http://www.psych.ucsb.edu/research/cep/eldorado/witchc raft.html). The dispute between sociobiology and cultural anthropology is partly an issue of the preferred level of analysis. Cosmides and Tooby’s primary targets are Durkheim and Geertz. Geertz is interested in culturally specific ‘webs of significance’. His research method is dedicated to resolving the problems of interpretation posed by trying to understand an exotic culture’s values, figures of speech and assumptions (e.g. Geertz 1973a; 1973b). If one wanted to understand the cultural significance of the headdresses worn in Highland New Guinea during warfare and competitive feasting, for example, one would need to gain entry to culturally specific worlds of meaning. There is a more fundamental issue identified by Durkheim (1938 [1901]), that of the emergent properties of interaction. Evolutionary theorists debate whether the primary motor of evolution is the gene, or the ecological system that exerts selective pressures on genetic variations in a population. Kauffman (1993) and Conway Morris (1998) argue the environment to which organisms adapt is transformed by the emergent properties of interaction. At least some of the cognitive skills cited by Tooby and Cosmides (language, formation of coalitions) are only adaptive in a social environment, i.e. an environment characterized by the emergent properties of social interaction.

The University of Michigan’s exoneration of Chagnon, on the grounds that gold miners are frequently violent toward indigenous people, is outrageous. In 1988, Chagnon published a paper in Science in which he attempted to demonstrate that Yanomami men who have killed an enemy enjoy greater reproductive success. The American Anthropological Association’s Task Force, which investigated Tierney’s claims, noted that Chagnon’s Science paper coincided with a disastrous moment in the Yanomami struggle for land rights, when the Brazilian president authorized the division of Yanomami land into reserves in order to bring them under control. The Brazilian anthropologist Carneiro da Cunha pointed out in 1989 that Chagnon’s paper had been widely reported in the popular press, both in Brazil and the U.S. It is highly likely his arguments were damaging to the Yanomami, and Chagnon has not done enough to counter these negative images, despite toning down his language in later editions of his ethnography (American Anthropological Association 2002: 1.32-4). Tierney is more cautious than Turner and Sponsel in constructing a link between Neel and Chagnon’s work and Nazi eugenics. Tierney merely states Neel and Chagnon were employed by the Atomic Energy Commission to collect Yanomami blood samples to compare with those of Japanese atom bomb survivors (Tierney 2000: 43). Even this claim is refuted by the National Academy of Sciences, who state that the Atomic

It is now acknowledged that Neel and Chagnon did not cause the 1968 epidemic of measles, although questions remain as to whether the vaccine used caused a severe reaction that resembled infection. The U.S. Academy of 42

ROBERT LAYTON: SOCIOBIOLOGY, CULTURAL ANTHROPOLOGY AND THE CAUSES OF WARFARE Bomb Casualty Commission was an independent study, financially supported (albeit sometimes unwillingly) by the AEC.

involved (American Anthropological Association 2002: 1.87-8). The report quotes a Yanomami spokesperson, José Seripino, who told a member of the Task Force, ‘In those days we didn’t have our own motors and he came with all that material – his research materials. The Yanomami needed these things – we were getting them from peasants. So one community has them and another not. Then other communities will get “fighting mad” (Spanish bravo)’ (American Anthropological Association 2002: 2.97-8, parenthesis in original). Davi Kopenawa reported that his relatives returned from the area where Chagnon was working and claimed Chagnon was saying ‘Whoever is the most courageous will earn more pans. If you kill ten more people I will pay more’ (ibid.: 98). However Kopenawa was only ten years old at the time, and I think this assertion should be treated cautiously. Tierney’s claim that Chagnon and Asch’s films were unreliable has been rebutted on the grounds that they conformed to standards of ethnographic film making of the time. The village in which The Feast was filmed was an abandoned village, which Chagnon persuaded the community to reoccupy by offering them steel tools (Biella 2000).

Tierney does not claim that Chagnon started warfare between Yanomamö communities, merely that his behaviour in the field exacerbated it. As an editorial in Nature commented, Chagnon’s critics and defenders had both responded too hastily, prompting ‘a series of missteps that have clouded a fair assessment of the book’s charges’ (Nature 2000: 755). The Yanomami attribute the deaths of previously healthy adults to sorcery, and revenge on the supposed sorcerers motivated many attacks during the period between 1932 and 1956 (before Chagnon started fieldwork). According to Chagnon’s own doctoral thesis, a third of all war fatalities over a fifty-year period in Namowei villages occurred during the thirteen months of his fieldwork (Tierney 2000: 29). Chagnon provoked warfare by distributing machetes and other metal goods to win the favour of Yanomami from whom he needed to collect blood samples and genealogies. The desire for steel implements drew Yanomami from other villages toward Chagnon, allowing disease to spread. Chagnon had a hand in, and even encouraged, some of the raids he reported (ibid.: 878, 107-110, 166). Chagnon’s 1988 Science paper claimed that men who killed had twice as many wives as nonkillers, and three times as many children. Wrangham and Peterson, in their 1991 book Demonic Males relied on Chagnon’s argument to draw a parallel between humans and chimpanzees (Tierney 2000: 14, 158). Time magazine summarized Chagnon’s theory as one that showed how Yanomamö social structure paralleled that of many primates, where males compete for females, blood relatives form coalitions and groups undergo fission when they reach a certain size (ibid.: 22).

Chagnon’s association with corrupt politicians to gain continued access to the Yanomamö is unquestioned (it was not challenged even in the University of Michigan’s response to Tierney). This extraordinary incident arose in the early 1990s, when Chagnon was encountering increasing resistance to his return to the field. He joined forces with Charles Brewer, a naturalist and businessman, and the Venezuelan president’s mistress Cecilia Matos, to establish FUNDAFACI, an organization that sought to turn Yanomamö territory into a private reserve under their control. The American Anthropological Association describes Tierney’s account of this episode as ‘one of the most serious and better supported allegations of the book’ (American Anthropological Association 2002: 1.41). Brewer had been a member of Neel’s 1968 expedition, but later made two attempts at illegal mining in Yanomami territory. Neel’s belief in the genetic correlates of leadership is also supported. The American Anthropological Association reports, ‘Neel was definitely interested in the possibility that the differential reproductive rate of headmen constituted a positive selective pressure in their populations (Neel 1980)’ (American Anthropological Association 2002: 1.25). Fischer, citing the same source, puts it more strongly: Neel wanted to find a ‘gene’ or ‘index of innate ability’ for ‘headmanship’ that had a connection with reproductive success (Fischer 2001a: 13).

Tierney points to a number of weaknesses in Chagnon’s data. Chagnon includes unmarried men in his data on reproductive success. Chagnon could not confirm the number of reported killings; some were supposed killings achieved by sorcery, while on other occasions the killer and his helpers claimed the same killing. Very few women are actually abducted. Even Chagnon’s low figure of 17% is higher than that recorded elsewhere among the Yanomamö, and some of these were probably willing elopements (Tierney 2000: 159-164). Bisaasi-teri was the only village Chagnon studied with a surplus of females (ibid.: 68, citing Chagnon’s unpublished Ph.D. thesis). Yet, after Chagnon’s arrival, Bisaasi-teri launched one of the most intense wars in Yanomamö history against another village that had a sharp deficit of women.

Finally, one of Tierney’s most important points is that the Yanomami were not representative of the original human condition when Chagnon studied them. Far from ‘uncontaminated’ by contact with the outside world, they had interacted with outsiders since the eighteenth century, as victims of slave raiders, enemies of settlers and subjects of missionary endeavours. Fischer agrees that one of the most disconcerting aspects of writing about the

The American Anthropological Association Task Force concluded that while Chagnon helped a Yanomami man negotiate a peace-making alliance at risk to himself, he also transported a raiding party in his canoe. Tierney does not point out that the party could not locate its intended victims, but Chagnon should not have got 43

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Yanomamö is the way that their long history of contact with slavers, rubber tappers and others has been ignored. Fischer commends Ferguson’s (1995) argument that warfare may be as much a response to other communities’ positions in trade networks as it is competition for women (Fischer 2001a: 10). He concludes that Tierney ‘does what journalists often do: put together with great detail a vivid picture, much of which the principles in the story find wrong and misconceived, but which sometimes puts a new set of questions on the agenda’ (Fischer 2001b: 18).

gatherers function to maintain the regional network of social relationships upon which rights of mutual access depend. We concluded that permeable territorial boundaries are most adaptive in a sparse, patchy and unpredictable environment (cf Davies and Houston 1984; Dyson-Hudson and Smith 1978). Modern humans are thought to have evolved in a semi-arid environment and we hypothesized that the genetic capability for flexible movement between bands evolved after the separation of the human and chimpanzee lines of evolution. Chimpanzee territorial behaviour cannot therefore be interpreted as the ancestral human pattern.

War in small-scale societies Mutual access is least adaptive in environments where resources are dense and patchy, but seasonally predictable in distribution. The best-documented exception to inter-access between neighbouring band territories is that of the Northwest Coast of North America, where resources are densely distributed, and predictable in their seasonal abundance. Northwest Coast territories were held, and their boundaries defended, by lineages. Trespassers were killed (Boas 1966: 35); land could be alienated and slaves taken during warfare (Garfield and Wingert 1966: 14, 29).

During the 1960s, writers such as Ardrey (1967) and Lorenz (1966 [1963]) popularized the idea that warfare was linked to ‘instinctive’ defence of territories, and therefore part of human nature. Dart’s alleged evidence for cannibalism among Australopithecines (Dart 1925; 1959) appeared to confirm the antiquity of such behaviour. Close parallels were drawn between human territoriality and that of other species. Subsequent research into territorial behaviour among animals revealed that territoriality was much more flexible than Lorenz and others had supposed. Australopithecines were shown to have been victims of animal predators and not other members of their own species (Brain 1981).

If we want to understand when and why human warfare began, we need to look at the emergent properties of social and ecological systems. Layton and Barton hypothesized human warfare first occurred when huntergatherers moved into environments with dense and predictable resources. In the areas of the world best studied archaeologically, this will have occurred in postglacial times, during the Mesolithic. The cultural invention of farming, creating defended fields of dense crops, will have exacerbated the trend. Warfare is an emergent phenomenon precipitated by specific natural and social conditions. Ember and Ember (1997) found that hunter-gatherer societies were not particularly peaceful but had a lower frequency of war than nonforagers. They concluded that fear of natural disaster is a frequent precipitate of war. The victors in those societies that fight at least once every two years almost always take land or other resources from the defeated. Foragers, they argue, ‘have less war than food producers because they have significantly less threat of natural disasters and less socialization for mistrust’ (ibid.: 11).

More recent observations of inter-group violence and the defence of territories among chimpanzees have again raised the prospect that warfare may be a phylogenetic trait. However, in a paper written with my colleague Robert Barton (Layton and Barton 2001) we argue that a comparison of human and chimpanzee territoriality reveals that hunter-gatherers have developed flexible forms of territorial behaviour which generally circumvent the conditions that apparently lead to inter-group violence among chimpanzees. Chimpanzees live in social groups comparable in size to human hunter-gatherer bands (20100 individuals), but chimpanzee groups are autonomous, whereas hunter-gatherers in low latitudes can move freely between bands within a larger regional community (comprising about 10-15 bands; often c. 500 people, but may be up to 1500). This phenomenon is what Gamble (1998) called ‘the release from proximity’, the emergence of social networks which depend on uniquely human genetic skills yet transform the environment into which the individual human is born (cf Geertz 1973c).

The Northwest Coast of North America has been inhabited since 9000 BC (Maschner 1997). During the long period between 9000 and 3500 BC, groups were small and mobile. The first evidence for conflict on the Northwest Coast occurs by 3000 BC, coinciding with the earliest shell middens, and is seen primarily in non-lethal skeletal injuries. Maschner cautions that violent conflict might have occurred earlier, without generating archaeological evidence. From AD 200-500, however, the onset of warfare is evident in the construction of defensive sites, the amalgamation of what might have been single lineage communities into large villages and

E.A. Smith (1988) argues that when all hunter-gatherer bands in a region suffer uncertainty as to current resource distribution, and the risk of local resource failure is unsynchronized, mutual access to each other's territories is an adaptive strategy. If one band's territory experiences better rainfall than its neighbours, the band will benefit from allowing other bands to share its windfall, provided those bands in turn allow their former hosts to camp with them when the unpredictable sequence of rainfall favours them. The patterns of interband visiting and gift exchange characteristic of hunter-

44

ROBERT LAYTON: SOCIOBIOLOGY, CULTURAL ANTHROPOLOGY AND THE CAUSES OF WARFARE breakdown of social order rarely if ever results in total anarchy or lack of social interaction. When the existing social order does break down, ethnicity and religion are two key dimensions on which to reconstruct trust, but on a smaller scale, between people who interact and claim exclusive rights to resources. Ethnicity and religion are frequently closely related, as in the Balkans or southern Sudan.

population decline. The bow and arrow were introduced to the region at that time. Maschner does not attempt to identify a single cause for war on the Northwest Coast. In historic times, ‘raids were planned and wars were fought over status, prestige, revenge, women, raw materials, and occasionally, food and territory’ (Maschner 1997: 270). In general, Maschner discounts pressure on subsistence resources as a cause, but identifies slave raiding, revenge killing, the theft of women and disputes over rights to candlefish spawning areas as well-attested motives. ‘The wars that did result in changes in territory, at least in every recorded case, were the result of expansionist activities by the most populous and strongest group in a region, and the group that had the greatest amount of subsistence resources in their own territory’ (ibid.: 292). Those with least territory had neither the wealth nor the numbers to undertake a successful attack.

There are a number of dramatic cases where the existing social order broke down because it became economically unsustainable. Duffield (1994) argues that the increasing frequency of large-scale complex disasters since the 1980s is a consequence of emerging global interdependence. Disasters have social, not natural causes (ibid.: 50-1). Expansion of mechanized agriculture in the north of the Sudan prompted the State to appropriate nomads’ grazing land. Peasant producers benefited but nomads, including Baggara Arabs, suffered. By the 1980s, the process of asset transfer had acquired a violent, sectarian and inter-ethnic character. Its spread rekindled civil war between Arab and African Sudanese. The Baggara (Muslim Arabs), for example, turned on their southern neighbours, the Dinka (Pagan/Christian Nilotes), asserting they were not ‘fit’ to own such wealth in cattle. This made the Baggara ‘ready tools in the government’s strategy of arming irregular militias’ (ibid.: 54), a strategy that both satisfied the economic needs of the government’s local clients, and the State’s wider aims of opening the south to Arab exploitation. There are striking parallels with Milosevic’s strategy in former Yugoslavia.

Sillitoe (1978) examined the role of Big Men in warfare in New Guinea. He argued that ‘in this area immediate objectives, such as redress, revenge, economic gain or religious necessity, hide the political aims behind many wars’ (ibid.: 253). Big Men are not simply strong men who can push others around; while admired as skilful organizers and talkers, they must respect the Melanesian principle that all men are equal and free to control their own affairs. Their scope depends on the flexibility of local social organization, and the extent to which it allows disaffected individuals to join influential leaders in other villages. Big Men fear the encroachment of rivals and try to sustain or extend their own influence. Insecure Big Men are more likely to foment discord (ibid.: 265, table 4). Sillitoe distinguishes between minor wars that punish a breakdown in reciprocity between groups who regularly trade and exchange marriage partners, and deep-rooted wars that persist between groups that lack such interrelationships but seek constantly to revenge past killings by the enemy. ‘There is a good chance that a war of redress will end in the rout of the defeated and the pillage of their settlement’ (ibid.: 263), but routing the enemy is more popular in major wars between settlements not normally linked by exchange.

Duffield argues that the key to understanding the survival of the State is to recognize that disasters have winners as well as losers. The winners implement social strategies that advance their own interests. The states that emerged from the break-up of Yugoslavia have a strong ethnic and religious basis. In Somalia and northern Albania, on the other hand, the State has broken down and has been replaced by a clan structure. Turton (1997) argues ethnicity does not emerge from nowhere, but requires specific historical conditions to flourish. James (2002), indeed, notes that the Baggara ethnic group has a flexible membership, supporting Barth’s original work on ethnic groups and boundaries (Barth 1967; 1969). Ethnic conflict in the Balkans after 1989 was not, according to Turton, the result of a clock which Tito had stopped ticking, starting again. Ethnic rivalry had continued in a muted form. Competition for the political and economic resources held by the Serb centre was carried out through the idioms of ethnicity and nationalism because Tito, the president of Yugoslavia, had created an ethnically-based federal structure without genuine power-sharing. After Tito’s death communist elites in Serbia and Croatia embraced ethnic nationalism to ensure their survival. ‘Pan-Yugoslav consciousness existed and there is evidence it was growing in the 1980s

Sillitoe notes that different types of military engagement tend to be found in different New Guinea environments. Battles and frequent raids occur in regions of mixed forest and grassland, whereas infrequent raids occur in regions of swamp and dense rainforest. Swamp and dense rainforest support a lower density of population, so there are both fewer occasions for people to meet and less scope for Big Men to construct inter-community networks. Sillitoe rejects, however, a simple correlation between population pressure and frequent war (Sillitoe 1978: 269; see also Sillitoe 1977). War and ethnic conflict In complex societies, recent ethnic conflicts have been associated with instability in the nation state. The 45

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY but it proved too weak’ to overcome ethnic conflict (Gallagher 1997: 48).

Disappearing World film We are all neighbours which charts the breakdown of trust in an ethnically mixed village in Bosnia as the war approaches. Croats originally joined forces with Bosnians to counter the Serbian advance but, once it seemed likely that Serbs would successfully take control of eastern Bosnia, the Croats turned against Bosnian Muslims and attempted to form a Croat state in western Bosnia. At the start of the film, Croat and Muslim neighbours affirm that they have always been friends and that the disintegration of the community is unthinkable. As Serb shelling approaches, the Muslims find themselves increasingly isolated and are forced to flee as Croat soldiers train their artillery on Muslim houses.

Some writers on the Balkans have explained the spread of ethnic conflict in terms of a divide separating rural from urban populations. Over the past fifty years, ethnic identity had become irrelevant for many city dwellers, particularly in Bosnia, but it remained strong in the countryside. Nationalist leaders had to draw their support from rural communities. In the countryside Serbs, Croats and Muslims were far more insular: conflicts during World War II were more clearly remembered, and villages had not experienced the modernization that occurred in cities. Wartime weapons were retained, especially by Serbs. Intermarriage and even contact with other communities were much less frequent and villagers were ‘ready to take up arms against cities as mythical places of affluence and sin’ (Gallagher 1997: 66). Gallagher concludes that it is only a slight exaggeration to describe recent nationalist conflicts in Eastern Europe as conflicts between urban (tolerant) and rural (nationalist) communities. State-controlled TV is often the only source of information about the outside world reaching rural communities. Milosevic exploited rural ethnic chauvinism by building up a well-armed police force and rewarding irregular militias. Troebst’s study, asking why ethnic conflict did not occur in Macedonia during the mid-1990s, is rather out of date, but still makes some useful points. Serb residents did not present a threat in Macedonia because only a small proportion of the Serbian minority lives in peasant villages. Most live in towns (Troebst 1997: 84). Greater ethnic tension exists between the large Albanian minority and ethnic Macedonians. ‘The overwhelming majority of Albanians in Macedonia form a highly traditional community isolated from almost all other sections of society’ (ibid.: 89).

The breakdown of mutual trust can therefore also be interpreted as the consequence of moving from a nonzero sum game to a zero-sum game. If everyone’s wellbeing can be enhanced by co-operation, they have a strong motive to work together (a non-zero sum game). If resources are perceived to be fixed, then conflict will break out in the scramble to secure the largest portion for oneself. Turton argues that the revival of ethnic identity requires a very selective reading of history but it must have enough connection with the past to be plausible and meaningful. The question is, what gives historic reference points contemporary salience (Turton 1997: 11)? Turton follows Glazier and Moynihan (1979), arguing that where the state holds finite resources (a zerosum game) the best strategy is to make claims as a group that is small enough for its members to make significant gains. Ethnicity satisfies this need better than class and can be particularly effective because it combines an interest with an affective, ‘primordial’ tie. Duffield argues, ‘The more direct or coercive the form of transfer’ of resources from losers to winners, ‘the more likely it is that winners have mobilized ethnic, national or religious sectarianism as justification for their extra-legal activity’ (Duffield 1994: 52).

In northern Albania, the State has been replaced by feuding lineages. Schwandner-Sievers (1999) reports the revival of feuding in Albania after the collapse of communism. She argues that the State’s weak hold over local communities leads it covertly to rely upon traditional leaders. The collapse of Communism created a ‘law vacuum’, causing a reversion to longer-established social procedures that had tacitly been allowed to continue.

The trick performed by ethnic or nationalist extremists is to convince members of a multi-ethnic community that they can dispense with each other’s help in future and instead fight for the largest share of limited resources by claiming an inalienable, primordial entitlement. After the Yugoslav Communist Party was disbanded in January 1990, nationalism came to the fore. Serbs recalled the wartime atrocities of the Ustashe, pointing to the fact that Tudjman, the Croatian leader, had revived the chequerboard Croatian flag last flown by the Ustashe. Croatians countered by recalling massacres and forced relocations of Croats perpetuated by the wartime Serb Chetniks, and the killing of tens of thousands of anti-Communist refugees turned back at the Austrian border by the British Army (Denich 1994: 379; Tanner 1997: 160). As Denich writes, ‘Conflicts over various issues in shifting localities were symbolically manipulated to polarize public opinion along the lines of resurgent ethnic identities’. Nationalists seized upon ‘the random constellations that

There are several ways of analyzing the breakdown of mutual trust. In his study of the evolution of cooperation, Axelrod (1990) showed that, for reciprocity to persist, people must not only have experience of each other as trustworthy partners in previous social exchanges, but must also anticipate that they will remain dependent on one another indefinitely. If partners in mutual aid know they are participating for the last time, and will not depend on each other for further cooperation, they will have no incentive to give their time and resources to help others and will withdraw into selfishness (ibid.: 10-13). This is well documented in the

46

ROBERT LAYTON: SOCIOBIOLOGY, CULTURAL ANTHROPOLOGY AND THE CAUSES OF WARFARE create opportunities for those who lurk off-stage with alternate scripts’ (Denich 1994: 369).

social relationships constructed and sustained, or repudiated by the communities within which they operate.

Urban people who grew up in post-Second World War Yugoslavia to whom I spoke in 1997-8 emphasized they had always thought of themselves as Yugoslavs rather than Serbs or Croatians. As Yugoslavian unity broke down, however, so many found it increasingly expedient, not only to secure a national identity, but to increase that nation’s share of the limited area of land within former Yugoslavia’s borders. Denich quotes a Croatian Serb: ‘So long as Yugoslavia’s federal structure was emphasized, we didn’t raise questions about national [Serb or Croatian] consciousness and national institutions. We considered Yugoslavia to be our state…. But now that there are fewer and fewer Yugoslavs and more and more Croats, Slovenians, Serbs, Albanians and so on, we realized that we Serbs in Croatia need to return to our own national identity’ (Denich 1994: 377). The same experience is described by other authors. Jansen quotes the novelist Dubravka Ugrešiü: ‘suddenly everything had to change: address books, the language and our names, our identity…. Everything changed with astonishing speed into old garbage’. People who had not discarded their Yugoslav identities became known as ‘Yugozombies’ (Jansen 2000: 95-6). Vulliamy quotes a Bosnian who told him:

Bibliography American Anthropological Association. 2002. The El Dorado Task Force Report. http://www.aaanet.org/edtt/final/preface.htm Ardrey, R. 1967. The Territorial Imperative: a personal enquiry into the animal origins of property and nations. London: Collins. Axelrod, R. 1990. The Evolution of Co-operation. Harmondsworth: Penguin (first published 1984, New York: Basic Books). Barth, F. 1967. On the study of social change. American Anthropologist 69: 661-9. Barth, F. (ed.) 1969. Ethnic Groups and Boundaries: the social organization of culture difference. London: Allen and Unwin. Biella, P. 2000. Visual anthropology in the plague year: Tierney and the Yanomamö films of Asch and Chagnon. Anthropology News December 2000: 5-6. Boas, F. 1966. Kwakiutl Ethnography (ed. H. Codere). Chicago: Chicago University Press. Brain, C.K. 1981. The Hunters or the Hunted? An introduction to African cave taphonomy. Chicago: Chicago University Press. Chagnon, N. 1968. Yanomamö: the fierce people. New York: Holt, Rinehart and Winston. Chagnon, N. 1988. Life histories, blood revenge and warfare in a tribal population. Science 239: 985-92. Conway-Morris, S. 1998. The Crucible of Creation: the Burgess shale and the rise of animals. Oxford: Oxford University Press. Cosmides, L., Tooby, J. and Barkow, J. 1992. Introduction: evolutionary psychology and conceptual integration. In J.H. Barkow, L. Cosmides and J. Tooby (eds) The Adapted Mind: evolutionary psychology and the generation of culture. New York & Oxford: Oxford University Press. 4-136 Dart, R. 1925. Australopithecus africanus: the man-ape of South Africa. Nature 115: 195-9. Dart, R. 1959. Adventures with the Missing Link. London: Hamilton. Davies, N.B. and Houston, A.I. 1984. Territory economics. In J.R. Krebs and N.B. Davies (eds) Behavioural Ecology: an evolutionary approach. Oxford: Blackwell. 148-69. Denich, B. 1994. Dismembering Yugoslavia: nationalist ideologies and the symbolic revival of genocide. American Ethnologist 21: 367-90. Duffield, M. 1994. The political economy of internal war. In J. Macrae and A. Zwi (eds) War and Hunger. London: Zed Press. 50-69. Dunbar, R. 1993. Co-evolution of neocortical size, group size and language in humans. Behavioural and Brain Sciences Evolution 16: 681-735.

‘I never thought of myself as a Muslim. I don’t know how to pray. I never went to a mosque. I’m European like you. I don’t want the Arab world to help us, I want Europe to help us. But now I have to think of myself as a Muslim, not in a religious way, but as a member of a people. Now we are faced with obliteration, I have to understand what it is about me and my people they wish to obliterate’ (Vulliamy 1994: 65; quoted by Gallagher 1997: 63). Conclusion The arguments for a direct link between chimpanzee inter-group aggression and human warfare, and for the existence of a gene for ‘leadership’, are simplistic. Humans differ from chimpanzees in their ability to construct social relationships on a wider scale, with individuals beyond the local band. There is surely a genetic component in our ability to keep track of the state of play in reciprocal relationships (cf Dunbar 1993), but the environment to which such a condition is adaptive is socially constructed. Human warfare occurs within the social environment; it arises when the web of social relationships that enable the ‘release from proximity’ is compromised. Competition for resources is important but mates and territory are only two among several types of resource. The manipulative activities of leaders play a part in fomenting war, whether they are local Big Men in small-scale, uncentralized societies, or the leaders of nation states. Leaders, however, can only manipulate

47

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Durkheim, E. 1938. The Rules of Sociological Method (trans. S.A. Solovay and J.H. Mueller). London: Macmillan. Dyson-Hudson, R. and Smith, E.A. 1978. Human territoriality: an ecological assessment. American Anthropologist 80: 21-41. Ember, C.R. and Ember, N. 1997. Violence in the ethnographic record: results of cross-cultural research on war and aggression. In D.L. Martin and D.W. Frayer (eds) Troubled Times: violence and warfare in the past. Amsterdam: Gordon and Breach. 1-20. Ferguson, B. 1995. Yanomami Warfare. Santa Fe: School for American Research. Fischer, M. 2001a. In the science zone: the Yamomami and the fight for representation. Anthropology Today 17 (4): 9-14. Fischer, M. 2001b. In the science zone, part 3. Anthropology Today 17 (10): 16-19. Gallagher, T. 1997. My neighbour, my enemy: the manipulation of ethnic identity and the origins and conduct of war in Yugoslavia. In D. Turton (ed.) War and Ethnicity: global connections and local violence. San Marino: University of Rochester Press. 47-75. Gamble, C. 1998. Palaeolithic society and the release from proximity: a network approach to intimate relations. World Archaeology 29: 426-49. Garfield, V. and Wingert, P. 1966. The Tsimshian Indians and their Arts. Seattle: University of Washington Press. Geertz, C. 1973a. Thick description: towards an interpretive theory of culture. In C. Geertz, The Interpretation of Cultures. London: Hutchinson. 3-30. Geertz, C. 1973b. Deep play: notes on the Balinese cock fight. In C. Geertz, The Interpretation of Cultures. London: Hutchinson. 412-53. Geertz, C. 1973c. The growth of culture and the evolution of mind. In C. Geertz, The Interpretation of Cultures. London: Hutchinson. 55-83. Glazier, N. and Moynihan, D.P. 1979. Why ethnicity? In D.R. Colburn and G.E. Pozzatta (eds) America and the New Ethnicity. Port Washington NY: National University Publications and Kennikat Press. 29-42 Goodall, J. 1986. The Chimpanzees of Gombe: principles of behaviour. Cambridge MA: Harvard and Bellknap. James, W. 2002. The anthropological family: from ancestors to affines. Presidential address distributed with Anthropology Today 18 (6). Jansen, S. 2000. Victims, rebels, underdogs: discursive practices on resistance in Serbian protest. Critique of Anthropology 20: 393-419. Kauffman, S. 1993. The Origins of Order: selforganisation and selection in evolution. Oxford: Oxford University Press. Layton, R. and Barton, R. 2001. Warfare and human social evolution. In K.J. Fewster and M. Zvelebil (eds) Ethnoarchaeology and Hunter-Gatherers: pictures at an exhibition. Oxford: BAR Publishing, BAR International Series 955.

Lorenz, K. 1966 [1963]. On Aggression. (trans. M. Latzke). London: Methuen. Maschner, H. 1997. The evolution of Northwest Coast warfare. In D.L. Martin and D.W. Frayer (eds) Troubled Times: violence and warfare in the past. Amsterdam: Gordon and Breach. 267-302. Nature 2000. Editorial dated 14.12.2000. Nature 408: 755. Neel, J. 1980. On being headman. Perspectives in Biology and Medicine 23: 277-93. Nishida, T. 1979. The social structure of chimpanzees in the Mahale Mountains. In D.A. Hamburg and E.R. McCown (eds) The Great Apes. Menlo Park CA: Benjamin/Cummings. 73-122. Nishida, T., Haraiwa-Hasegawa, M. and Takahata, Y. 1985. Group extinction and female transfer in wild chimpanzees in the Mahale National Park, Tanzania. Zeitschrift für Tierpsychologie 67: 284-301. Ruby, J. 2000. Tierney’s claims about Tim Asch. Anthropology News December 2000: 7. Schwandner-Sievers, S. 1999. Humiliation and reconciliation in northern Albania: the logics of feuding in symbolic and diachronic perspectives. In G. Elwert, S. Feuchtwang and D. Neubert (eds) Dynamics of Violence: processes in escalation and de-escalation of violent group conflicts. Berlin: Duncker and Humblot. 133-52. Sillitoe, P. 1977. Land shortage and war in New Guinea. Ethnology 16: 71-81. Sillitoe, P. 1978. Big Men and war in New Guinea. Man (N.S.) 13: 252-71. Smith, E.A. 1988. Risk and uncertainty in the ‘original affluent society’: evolutionary ecology of resourcesharing and land tenure. In T. Ingold, J. Woodburn and R. Riches (eds) Hunters and Gatherers: history, evolution and social change. Oxford: Berg. 222-51. Tanner, M. 1997. Croatia: a nation forged in war. New Haven: Yale University Press. Tierney, P. 2000. Darkness in El Dorado. New York: Norton. Troebst, S. 1997. An ethnic war that did not take place: Macedonia, its minorities and its neighbours in the 1990s. In D. Turton (ed.) War and Ethnicity: global connections and local violence. San Marino: University of Rochester Press. 77-103. Turton, D. 1997. Introduction: war and ethnicity. In D. Turton (ed.) War and Ethnicity: global connections and local violence. San Marino: University of Rochester Press. 1-45. Vulliamy, E. 1994. Seasons in Hell: understanding Bosnia. London: Simon and Schuster. Wrangham, R. and Peterson, D. 1996. Demonic Males: apes and the origins of human violence. London: Bloomsbury.

48

The physical evidence of warfare - subtle stigmata? Christopher J. Knüsel Department of Archaeological Sciences, University of Bradford

1993) rather than period-specific or broad diachronic or geographic treatments (Larsen 1997: 119-60, although see Smith 2003; Boylston 2000; Lambert 1997; Walker 1989).

Many treatments of violence and warfare in pre- and protohistory falter owing to the perceived ambiguous nature of evidence for armed conflict, even in societies with documentary evidence relating to violent encounters on a grand scale. These treatments are often based on the occurrence of defensive sites, weapons, and burials furnished with weapons. Many of these indicators are as much a social statement about the prestige of the builders or buriers - or even a prophetic warning against attack as secure evidence for warfare. Even broken weapons or flint projectiles are not unequivocal evidence of warfare, as breaking of objects is a part of many funerary rites and other ceremonies (Chapman 2000). If warfare is as endemic in recent small-scale societies as Keeley (1996) has demonstrated, then why is there such a seeming paucity of evidence for it in a past dominated by similar social groupings?

Since the publication of Carneiro’s much-cited 1970 paper in which warfare is linked to circumscription of land, population and resources, researchers have often included warfare as an indicator of increasing social inequality and social change, but rarely does its role figure unambiguously in accounts of these developments, often taking a backseat to or forming an extension of ritual concerns (see, for example, Whittle 1996; Whittle et al. 1999; Barrett et al. 2000; Cunliffe 1992; Hill 1995; 1995). In describing enclosures in the Neolithic, for example, Whittle (1996: 268) writes: ‘[t]hough there is evidence not only for continuous barriers, but also for attack, burning, and killing [at some enclosures]… I suggest that the primary role of most ritual enclosures was formally to bound a special space, to separate inside from outside, and to invest the activities associated with various parts of the special space with appropriate significance.’

Physical evidence of wounds, fatal or healed, is the sine qua non of warfare and its results, and it is to the injuries found within skeletal remains that one must turn and, to identify lesions, to the forensic literature and the fracture mechanics of bone. Examples of peri-mortem injuries are presented in an attempt to better substantiate the nature and extent of violence and warfare in the past.

In another example, in his work on the human remains from Cadbury Castle, Somerset, Forbes (in Barrett et al. 2000) identifies a number of instances of cut-marks and chop-marks, burning, and animal gnawing, and the site photograph archive attests to the presence of articulated body parts, yet the sub-heading of ‘Violence’ (ibid.: 235ff.) ushers in a discussion of weapon types. Despite the human remains being described by the original excavator as belonging to ‘massacre levels’ (ibid.: iv), ‘…no comprehensive analysis of the bone assemblages was ever undertaken…’ and Forbes precedes his presentation of the physical evidence of the human remains, based on the site catalogue, with the proviso that ‘… the term massacre is used here… due to its historical appellation, but should not be understood to prejudge the nature of the deposit’s formation.’ (ibid.: 117). Woodward and Hill (ibid.: 114-15) emphasize the complex nature of the human deposits; they attribute these to four possible human ritual behaviours, including cremation, secondary burial of crania, hoarding and structured deposition. A violent encounter is not entertained, despite evidence suggesting the destruction of a gate (through burning), silting around the remains and deliberate covering of the remaining visible deposits after a period of time when the area was rebuilt.

Perspectives on warfare Perhaps the greatest expression of social inequality is the act of premeditated killing, whether ritualized, as in sacrificial acts and executions, or committed in the context of armed conflict. The power of the victor and the powerlessness of the victim take on a much more obvious social significance when performed or reiterated in the presence of a social group. Therefore, it is not without reason that the origins, evolution, and causes of violence and warfare have been a commonplace in anthropology and ethnographic treatments (e.g. Ferguson 2001; Otterbein 1999; Ferguson and Whitehead 1992; Haas 1990) and historical studies (e.g. Connolly 1998; Keen 1999; O’Connell 1995; Keegan 1993; Tallett 1992; Contamine 1980; Allmand 1973). In archaeology, researchers have relied on a vast array of information on fortifications, defences, weaponry and armour, warrior burial, iconography and literature to identify the occurrence of warfare in the past (e.g. Hill and Wileman 2002; Guilaine and Zammit 2001; Harding 2000: chap. 8; Osgood et al. 2000; Carman and Harding 1999). Far fewer studies have concentrated on weaponrelated injuries to human remains and those that do exist tend to be site-specific treatments (see, for example, Novak and Kopp 2003; Fiorato et al. 2000; Cunha and Silva 1997; Liston and Baker 1996; Willey and Emerson

There are others, however, (see, for example, Carneiro 1970; 1990; Haas 1990; Earle 1997; Kristiansen 1998) who have stressed the opposite view that warfare is integral to social change, either precipitating it or as a 49

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY significance of the scale of such evidence can often only be fully appreciated when it is placed in an ethnohistoric context (see, for example, Frayer 1997; Knüsel and Boylston 2000).

complement to changes in the social, political, and economic spheres. Keeley (1996: 117) highlights the importance of archaeological considerations of warfare, believing such considerations to be less coloured by personal perspective than is the information from written accounts. Importantly, though, when he attempts to address the physical evidence for warfare from the archaeological record, he has to rely on anecdotes rather than on synthesized accounts.

Even given ample skeletal analysis, however, multiple and variable definitions of warfare greatly colour perceptions of the prevalence and scale of violent conflict in the past. Warfare can be defined as ‘organised purposeful group action, directed against another group, that may or may not be organised for similar action, involving the actual or potential application of lethal force’ (Ferguson 1984: 5), although Milner (1995: 221) prefers: ‘Warfare refers to purposeful violence calculated to advance the ambitions of separate political factions, regardless of who was involved, the regularity of fighting, the numbers of participants, or specific combat tactics.’ Although these two definitions share an emphasis on group action and an implicit ability to do harm, they differ subtly in the extent to which they weigh the intention or purposefulness of actions. Ferguson’s definition invokes the potential to do harm, even presumably if none was done, while Milner’s appears to reject this notion with his use of the more general term violence as the defining feature of this group action.

Earle (1997: 106) notes that ‘Chiefdoms are continually at war, and war is a critical element in chiefly power strategies.’ It seems clear that even if warfare is deemed important, when researchers turn to the archaeological record they experience difficulty in judging the effects of warfare. In large measure this seems to be because the archaeological indicators used to identify its occurrence cannot easily be separated from those relating to the threat of violent confrontation. As Earle (ibid.: 109) puts it: ‘The threat of violence, where no superstructure exists to mediate it, creates opportunity for local leaders to gain strong authority through offers of protection.’ Therefore, the threat of violence, presumably from both paramount chiefs within the society and from outsiders, is enough to create an atmosphere of warfare without violence actually taking place.

The difference of emphasis between these two definitions is implicit in the debates concerning the frequency and scale and thus the social impact of warfare in the past (see Sharples 1991a, for example). With its fortifications, presence of sling stones and arrows, and its ‘war cemetery’, researchers have frequently cited Maiden Castle, Dorset, as a site documenting protohistoric warfare. Sharples (1991a; 1991b) notes, though, that fewer than half of the burials show signs of violent trauma1, and that the context of this evidence, in an already established cemetery, does not clarify the scale of warfare and leaves interpretation of its social impact contentious (see below). Although this contribution cannot pretend to resolve this debate, it attempts to clarify the importance of thorough integrative analysis of human remains and their context in discussions of violence and warfare in the past.

Even in Hawai’i, where warfare was endemic and characterized by brutal reprisals linked to chiefly status and the conquest of land, Earle (1997: 134) observes that: ‘There is virtually no archaeological evidence for warfare in the settlement pattern and artifact inventory… The only clear archaeological indications of warfare are a few concentrations of sling stones, the odd stone club head, and refuges to which a threatened population might run.’ The answer to this conundrum of recorded violence on a grand scale apparently leaving little archaeological evidence lies in a description of ‘one of the greatest warriors’ who was a close relative of the paramount chief: ‘His body was almost covered with scars, and he was quite a cripple; and to add to his distressing situation, he had entirely lost one eye.’ (Portlock 1789: 177, cited in Earle 1997: 140). This passage indicates that for archaeologists examining evidence of warfare the answer lies in the human remains; the wounds indicated here were not only disfiguring in their soft tissue manifestations but, given their extent and severity, likely also to have affected the underlying skeleton. Yet even those archaeological treatments that specifically include discussion of skeletal remains are often frustrated by a lack of studies upon which to draw, as in the case of the Bronze Age Aegean ‘…because skeletal remains have not always received much attention during excavations…’ (Osgood et al. 2000: 122). The same situation obtains elsewhere, and the extent of the neglect of this source of potentially crucial data only becomes obvious when researchers reexamine previously excavated material (see, for example, Orschiedt 1998; Frayer 1997). Likewise, the full

Human remains In a recent article, Walker (2001: 573) provides one of the most succinct statements of the fundamental value of skeletal analysis to studies of violence and warfare in the past. He writes: ‘Traumatic injuries in ancient human skeletal remains are a direct source of evidence for testing 1 Here, violence is understood to mean violent acts committed against another person whereas warfare is defined on the basis of massed and continuous acts of violence among groups of people. 2 In Wheeler’s (1943) original report, Morant and Goodman (p. 352) make reference to 11 males out of 18 and 3 females out of 7 from the ‘Belgic War Cemetery’ that show signs of ‘mutilation’. This is a high frequency of 56% (14 of 25), with some individuals showing more than one injury. Research undertaken by Rebecca Redfern, (pers. comm.), University of Birmingham, argues for plentiful evidence of violent traumatic injury among these remains.

50

CHRISTOPHER J. KNÜSEL: THE PHYSICAL EVIDENCE OF WARFARE - SUBTLE STIGMATA? start of the Bronze Age.’ It is highly possible that even quite life-threatening or lethal injuries of soft tissue leave only subtle stigmata in the skeleton (see Saul and Saul 2002 for some examples). This may be why Chapman (1999) can find no mention of injuries in analyses of a series of human remains from a variety of Mesolithic and Neolithic sites in the Balkans, even though some were found as multiple burials (see below). This apparent absence of injuries occurs during a time when mass graves appear in central Europe, with a number dating to the Late Mesolithic and Neolithic (see Vencl 1999; Orschiedt 1998; Frayer 1997; Wahl and König 1987).

theories of warfare and violence that are not subject to the interpretative difficulties posed by literary creations such as historical records and ethnographic reports.’ Walker (ibid.) recommends population-based studies of trauma. Population-level studies allow us to gain insight into not only the occurrence but also the intensity of warfare through prevalence data. In combination with an absolute dating programme, one can also begin to address changes in frequency of warfare diachronically. Larsen (1997) also emphasizes the fact that archaeological skeletons provide the only direct evidence of violent encounters in the past. It is the multiple occurrences of weapon injuries healed and unhealed, the patterning of these injuries, and the number of individuals involved, as well as their burial context, that permits association with the massed and sustained violence that defines warfare.

Re-analysis of such human remains can, however, produce evidence for violence, even when disarticulation and commingling of human remains has occurred during secondary burial, as in the case of Neolithic chambered tombs. In their re-analysis of crania from British Neolithic chambered tombs, Schulting and Wysocki (2002) found that 26 of 350 (7.4%) crania show evidence of traumatic injuries. To securely identify such lesions, one must turn to the on forensic anthropological and bone taphonomy literature.

Curiously, the study of violent injuries in skeletal remains has lagged behind the trend to provide prevalence rates for populations of individuals in much of palaeopathology, although this is not universal as Walker’s (1989), Milner’s (1995), and Smith’s (2003) treatments of pre-Columbian warfare in North America demonstrate. Cases of projectiles embedded in the bodies of individuals are, perhaps, the most obvious indicator of wounding with intent. If unaccompanied by other types of evidence, however, such cases can attest only to the occurrence of violence and not to its extent. As in the disputed case of the English King William II (‘Rufus’), who is said to have died as a result of a hunting accident, some of these instances may be considered the result of unfortunate, if politically propitious, events.

Distinguishing injuries indicative of armed conflict The timing of bone breakage and the nature of fractures Bone breaks in predictably different ways depending upon whether force is applied to the bone when it is still fresh, or after a certain amount of drying (but when the organic content is still present in some quantity), or when most of the organic content has been lost. The collagen content determines the elasticity of bone and how it fractures. Changes in this content provide the means by which to distinguish between different types of fractures encountered in archaeological assemblages. Three principal criteria can be used to judge fracture type (Fig. 1). These are:

When such instances begin to mount up, however, they reveal that deliberate killing and wounding were far from uncommon. In a review of embedded French Neolithic projectile points, Guilaine and Zammit (2001: 193-201) indicate the extent of violence in the Neolithic of France. They also provide prevalence rates for the site of Grotte des Baumes-Chaudes (Lozère) where, from an estimated population of 300 or 400 individuals, 17 display evidence of embedded projectile points, while another bears evidence of having been stabbed with a copper dagger blade. The number of embedded projectile points in single individuals would seem to preclude the possibility of accidental wounding. The 25-35 year old Early Bronze Age male found in the Neolithic ditch around Stonehenge seems also to be the victim of deliberate rather than accidental wounding: of his three peri-mortem injuries, two were caused by projectile points shot at close range from both the front and the back (Evans et al. 1984).

1. 2. 3.

the fracture margin with respect to the cortical surface of the bone, the texture of the fracture surface, and fracture outline (i.e. the shape of the fracture pattern).

Using these features, fractures can be grouped along a continuum, from having occurred on fresh bone full of collagen, to bone containing some collagen (which produces ‘dry fractures’), through to bone which was totally mineralized at the time of the fracture (i.e. made up largely of the crystalline hydroxyapatite component of bone), with little or no collagen remaining (Outram 2002; Galloway 1999; Lyman 1994; White 1992; Villa and Mahieu 1991). The importance of distinguishing these manifestations lies in their use in identifying excavation or curatorial damage, distinguishing these from taphonomic effects that have to do with breakage that occurred in the past when bones were disturbed by human, animal, or natural processes (when dry fractures

Osgood et al. (2000: 140) hint at the limitations of depending on this type of evidence alone when they note that ‘Although almost no dagger wounds have been found, or at least none recognised as such, this weapon was also an important part of the warrior panoply at the 51

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY fracture margin is irregular with respect to the cortical bone surface. The fracture surface is rough or corrugated in appearance with spicules of bone disrupting it.

occur), and, in turn, distinguishing each of these from traumatic lesions of bone, most importantly those that occur around the time of death (peri-mortem fractures).

This type of fracture is relatively easy to distinguish from both ante-mortem (i.e. healed) and peri-mortem (i.e. unhealed) trauma, although the means to distinguish the latter has only entered archaeological discourse comparatively recently. Ante-mortem trauma is identifiable as new, woven, or fibre (these terms can be considered synonymous) bone formation (i.e. bearing a fracture callus) with rounded, smooth margins and fracture surfaces, but with sub-perisoteal bone expansion, which is indicative of healing (Fig. 3).

FIGURE 3. A healed (ante-mortem) fracture of the mandible of Towton 16 from the mass grave near the site of the battle of Towton, 29th March 1461. Note the rounding of the fracture margin beneath the roots of the first molar. (Photograph: Jean Brown)

FIGURE 1. Butchered cattle metacarpals from the Bronze Age site of Velim Skalka, Czech Republic, with peri-mortem fractures showing the fracture surface (1), fracture margin (2), and fracture outline (3). (Photograph: Carol Palmer; see Harding et al. forthcoming)

FIGURE 2. Post-mortem breaks of an archaeologicallyrecovered human rib (left) and cranial bone fragment (right). Note the transverse fracture outline, jagged fracture margin with spicules of bone and the lighter colouration of the fracture surface. (Photograph: author)

FIGURE 4. A dry fracture in a right femoral diaphysis from the Late Bronze Age site of Velim Skalka, Czech Republic. Note the spiral or helical fracture outline, uneven appearance of the fracture surface, and the undulating fracture margin. (Photograph: author)

The most immediately eye-catching feature of postmortem damage is a colouration of the fracture surface that is noticeably different from surrounding bone, which has been stained by the soil matrix (Fig. 2). The fracture outline of recently broken bone is transverse and the

Dry fracture characteristics have only recently been recognized in archaeological material. Such bone damage often possesses a helical fracture outline and may resemble peri-mortem fractures in this feature; dry fractures have a relatively sharp fracture margin but one 52

CHRISTOPHER J. KNÜSEL: THE PHYSICAL EVIDENCE OF WARFARE - SUBTLE STIGMATA? that may be more undulating, with a fracture surface that is more uneven than is seen in peri-mortem fractures (Fig. 4). What distinguishes them most from peri-mortem fractures is their rough fracture surface texture, which may have a stepped or nodular appearance. They also lack the transverse fracture outline and spicules of post-mortem fractures. Dry fractures occur some time after death and are produced when the organic component (collagen) has decayed to some extent. They are generally indicative of post-depositional disturbance and may indicate a secondary burial ritual (Valentin and Le Goff 1998; Ubelaker and Adams 1995).

FIGURE 5. A peri-mortem parry fracture of the right ulna of Towton 30 from the mass grave near the site of the battle of Towton, 29th March 1461. Note the sharp fracture margin of the uppermost fracture and the smooth appearance of the splinter or spall present on either side of the fracture, the left one being shorter than the right, the latter of which carries on out of the frame. (Photograph: Jean Brown)

Peri-mortem fractures are breakages of bone that occur around the time of death, with death occurring before any substantial healing can take place. They are referred to as peri-mortem, rather than ante-mortem, owing to the difficulties in determining the cause of death in skeletonized individuals. Such fractures may occur in the immediate post-mortem interval, as may be the case for suspected witches in the prehistoric American Southwest, where the bodies of such individuals were dismembered and struck with stones after death in order to disable the corpse so that it would no longer trouble the living (Ogilvie and Hilton 2000; Darling 1998). The manner of death, as opposed to its cause, can be determined from the patterning and type of fracture trauma (see below). Determining the cause of death in skeletonized remains without knowledge of the damage to soft tissue- is always a matter of probability, estimated from the likelihood of immediate deadly effects arising from any one injury.

Peri-mortem fractures can be distinguished by their helical or spiral fracture outline and by a fracture margin that is sharp (i.e. forming a distinct ridge or crest) with respect to the cortical surface of both post-cranial and cranial bone. They possess a fracture surface that is smooth in texture (Figs 5 and 6). The smooth texture of the fracture surface can be identified just by touch. When a spiral fracture passes through the entire thickness of a long bone, the tip of the broken fragment can itself break off, producing a ‘butterfly’ fracture (Fig. 6; Galloway 1999: 182; Ubelaker 1996: fig. 7; Dandy 1993). The ‘butterfly’ fracture is, in essence, a comminuted fracture (i.e. one with multiple fragments).

Peri-mortem fractures are similar in appearance to those identified in faunal assemblages that have undergone butchery and marrow extraction (see Lyman 1994; Blumenshine et al. 1996; Outram 1999). Essentially, this is what a fracture resembles in a living person, prior to the application of remedial treatment. Peri-mortem trauma may also be inflicted on the deceased by natural processes, including animal scavenging (see Weigelt 1989; Merbs 1989; Haglund et al. 1988; 1989), trampling (Brain 1981), and rockfalls (see Gargett 1989; 1999). These can be differentiated, however, on the basis of animal tooth marks, surface features and specific archaeological contexts.

FIGURE 6. A peri-mortem fracture of the ‘butterfly’ type from a large ungulate long bone from the Late Bronze Age site of Velim Skalka, Czech Republic. Note the smooth texture of the fracture surface and sharpness of the fracture margin. (Photograph: Carol Palmer)

53

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY

FIGURE 7. Peri-mortem sharp force trauma from a heavy bladed weapon: an occipital view of Towton 32 from the mass grave near the site of the battle of Towton, 29th March 1461. Note the smooth margin to the right of the centremost fracture surface that runs along the sagittal suture superiorly and the more irregular fracture outline and irregular fracture surface to the left. Note, as well, that there is a second such injury to the right of the centremost one, which again has the same appearance, with the smooth margin being below the more irregular margin above. Both of these blows have produced radiating fractures at the terminal ends of the apertures. (Photograph: Eric Houlder)

FIGURE 8. A peri-mortem depressed fracture of the left parietal and temporal region of Towton 11 from the mass grave near the site of the battle of Towton, 29th March 1461. Note the radiating fractures that emerge from the margin of the wound, above the squamous suture to the right of the image, a shorter one to the left (above the perforation), as well as a larger one to the bottom left that runs out of the frame. The piece of bone that conjoins within the perforation suggests that this fracture was also comminuted (i.e. made up of a series of broken fragments). (Photograph: Eric Houlder)

54

CHRISTOPHER J. KNÜSEL: THE PHYSICAL EVIDENCE OF WARFARE - SUBTLE STIGMATA? In addition, especially in the long bones of the body, perimortem injuries can be identified by the splintering or spalling of the cortical bone surface. Splinters or spalls are produced when a hinged piece of bone breaks free from the cortical surface when an embedded weapon is withdrawn from the bone (Fig. 5). Sometimes this is referred to as ‘peeling’ because of the resulting fracture’s resemblance to a broken green tree branch (White 1992). In essence, these are longitudinal fractures that run within the cortical bone laminae, those structures that give bone its grain.

Weapon injuries form a sub-group of both ante-mortem and peri-mortem traumatic lesions and can be distinguished by the type of force that created the fracture and, occasionally, through replication studies, to particular types of weapons or tools (see Symes et al. 2002; Novak 2000; Shipman and Rose 1983; Walker and Long 1977). Weapon trauma can be divided into: 1. 2. 3.

sharp force trauma that is produced by a bladed weapon; blunt force trauma that produces crushing and is indicative of club-like weapons; penetrating wounds from projectiles or pointed weapons.

Cranial fractures Recognition of the extent and number of injuries, especially of the more lethal injuries to the cranium, is dependent not only on the researcher’s skill in identifying such injuries but also on the reconstruction of crania from surviving fragments. Reconstructing crania increases the ability to identify such injuries many fold. The lack of such undertakings may be a contributing factor to the apparent absence of violence in past populations because the cranium is often the target of blows, owing to its close association with the identity of the individual. These injuries might have often been overlooked in the past, especially in the absence of cranial reconstruction (Boylston 2000).

Sharp force trauma is characterized by a lesion that has one elliptical or straight fracture outline, a sharp fracture margin, and a smooth fracture surface, and one outline that is more irregular (Fig. 7). The elliptical fracture outline results from the entry of the weapon into the bone, while the irregular outline relates to the removal of the weapon from the bone after it has become embedded in it. Blunt force trauma is characterized by crushing injuries that can produce comminuted fractures, those with multiple pieces (Fig. 8). Sharp weapons may also produce comminution depending on the weight of the weapon. Penetrating injuries are usually small-diameter depressions or perforations of bone (Fig. 9). All of these fractures produce a bevelled fracture surface internally on the inner table of the cranium (and within the medullary canal), and this feature identifies them as having occurred in the peri-mortem interval (see below). If the recipient survives the injury, all of these may show signs of healing in the form of rounded fracture margins.

The unusual structure and composition of the cranium distinguishes it from other bones of the body and influences the way it fractures. Most bones in the body are anisotropic; they respond differently depending on the direction from which force is applied. The adult cranium is isotropic; it resists force transmission with its entire

FIGURE 9. A peri-mortem penetrating wound to the left side of the frontal bone of Towton 40 from the mass grave near the site of the battle of Towton, 29th March 1461. Note the shape margin of this entry hole and the radiating fractures that emerge from it to the right and left, indicating that the blow was a high velocity one. This size of this perforation fits that of an armour-piercing projectile point. (Photograph: Eric Houlder)

55

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Linear fractures, the most common form of cranial fracture, occur under low velocity impacts and may appear at some distance from the point of impact (Fig. 10). Linear fractures, in most instances, extend through both tables of the cranial vault and the intervening diploë layer of cancellous bone and, because they result from cranial deformation and movement of intra-cranial contents, are often accompanied by haematoma (intracranial bleeding). Depressed fractures of the vault occur from in-bending at the point of impact when the cranium is struck by an object moving at high velocity. Depressed fractures are usually circumferential in shape and can be of variable size, from a single depressed piece of bone to a series of concentric or stellate fractures surrounding a depressed focus (Fig. 11), depending on the shape of the object involved.

surface area as a single entity, not as a collection of different bones that make it up. This quality makes it an extremely strong structure but it also means that, when the cranium is fractured, the fracture lines tend to migrate across its entire surface.

FIGURE 11. A peri-mortem blunt force injury to the left occipito-mastoid region of the cranial vault of Towton 8 from the mass grave near the site of the battle of Towton, 29th March 1461. Note the concentric nature of the fractures surrounding the central depression. (Photograph: Eric Houlder)

Penetrating injuries produce a perforation in the cranial vault. It is not the size or weight of the object that determines penetration but the velocity with which the cranium is struck. If struck over a small surface area, the cranium is more likely to be perforated. Thus, projectiles or objects with a small point are most likely to perforate the cranial vault. A sharp fracture margin identifies the entry wound of such an injury, usually, although not always, on the ectocranial (i.e. the external) surface of the vault (Fig. 9), while bevelling of bone occurs at the point of exit, usually, although not always, on the endocranial (i.e. internal) cranial surface.

FIGURE 10. A peri-mortem linear radiating fracture of the occipital of Towton 9 from the mass grave near the site of the battle of Towton, 29th March 1461. This fracture runs to the void created by the foramen magnum, which is out of sight beneath the cranial vault. (Photograph: Jean Brown)

Impact assessments on whole crania by Gurdjian and Lissner (1945; 1946; 1947), Lissner and Gurdjian (1946), Gurdjian et al. (1947; 1950) and Gurdjian (1975) form the basis of our understanding of fracture mechanics of the cranium. These researchers observed several particularly relevant phenomena in their experiments that have been summarized and further developed more recently (see Galloway 1999; Berryman and Symes 1998; Courville 1962). Amongst the earliest observations was that, even when struck in the same way, the crania of different individuals fracture in different ways owing to variations in shape and thickness. In addition, owing to irregularities in cranial thickness, the cranium bends in an irregular manner and consequently fractures are often of an irregular outline. Fractures of the cranium result from tensile strain (i.e. stretching) generated as the cranium deforms (i.e. bends) in response to the force of a blow.

The bevel occurs in most injuries that penetrate the cranial vault and is one of the best indications, along with radiating fractures, that the wound was a peri-mortem one (Fig. 12). At times, an incomplete bevel with an attached fragment characterizes injuries that do not completely penetrate the cranial vault (Figs 13 and 14). Wounds that penetrate the entire cranial vault may have an entrance hole, with its sharp margin, on the endocranial surface and an exit hole, with its bevel, on the ectocranial surface. This appearance is most often associated with high 56

CHRISTOPHER J. KNÜSEL: THE PHYSICAL EVIDENCE OF WARFARE - SUBTLE STIGMATA?

FIGURE 12. The endocranial view (with an internal bevel) of the perimortem blunt force injury shown in Fig. 11. This bevel develops from in-bending caused by the force of the blow and its irregular fracture outline is due to the variable thickness of the cranial bone in the occipito-mastoid region of the cranium. Note the radiating fractures that emerge from the margin of the injury. With the internal bevelling, these radiating fractures attest to the peri-mortem timing of the wound. (Photograph: Jean Brown)

FIGURE 13. An uplifted fragment of the internal table (i.e. endocranial surface) of the cranial vault from a peri-mortem and incompletely perforating penetrating injury delivered to the posterior portion of the left parietal of an individual from the Late Bronze Age site of Velim Skalka, Czech Republic. The irregular fracture outline of this fragment, if completely detached, would create a bevel like that depicted in Fig. 11. (Photograph: Carol Palmer and author)

FIGURE 14. The ectocranial surface of the same specimen as in Fig. 13 with its incomplete penetrating injury. In this image, one can also see the crushing of the diploë layer that is characteristic of the early stages of cranial fracture. (Photograph: Carol Palmer and author)

57

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY extend through the entire thickness of the cranial vault. This bevel will be the same colour as the surrounding bone, and both features assist in the identification of perimortem injury. Secondary linear fractures may extend (they radiate) from the edges of the fracture and, again, attest to the peri-mortem nature of the injury (Figs 7-9, 11-12).

velocity projectiles such as bullets, but could occur with other powerful penetrating injuries. Radiating fractures can develop from a central depression that may cross sutures, sometimes with such force as to actually disrupt the suture, causing a diastatic fracture. Radiating fractures will propagate along the course of least resistance, either toward a void in the bone, as in the case of a normal anatomical feature, such as the foramen magnum (Fig. 10), or toward a pre-existing fracture (e.g. when one fracture line runs into another), which allows sequencing of the injuries (i.e. which came first and so on; Sauer 1998; Madea and Staak 1988).

Healed injuries of the cranium will have rounded margins and, sometimes, re-attachment of even large fragments may occur. When such fragments are removed, leaving a hole in the vault, a central calcification of the inner table alone forms and thin spicules of bone attach this area of calcified bone to the margins of the fracture (Nerlich et al. 2003a; G. Martin, pers. comm.). Context and conflict The potential importance of multiple burials The use of the term ‘multiple burial’ or ‘collective burial’ - the burial of more than one individual in a single grave cut or feature - disguises an array of potential behavioural interpretations, both human and taphonomic, that might have contributed to their formation. Although forensic anthropologists (Mant 1987; Haglund 2002; Schmitt 2002; Skinner et al. 2002) differ on the number of individuals required to define a mass grave, all agree on a set of distinguishing characteristics of such features, including:

FIGURE 15. A wedge of bone removed by sharp force trauma from Towton 10 from the mass grave near the site of the battle of Towton, 29th March 1461. Note the sharp fracture margin uppermost and the more irregular margin below, features shared with a penetrating blade injury. (Photograph: Eric Houlder)

1. 2.

A glancing blow delivered with a heavy bladed weapon may remove a wedge of bone from the ectocranial surface of the cranial vault (Barnes 1870: fig. 5; Kaufman et al. 1997: fig. 28), an injury that can be distinguished from a trepanation, for example, by its smooth crescent-shaped outline opposed by a fracture outline that is more irregular in shape with a sharp fracture margin (Fig. 15). Unlike a trepanation, which heals from a central calcification of the inner table and usually possesses scrape-marks around the area affected by the procedure (see Verano 2003: fig. 2; Nerlich et al. 2003a), this wedge-shaped injury lacks such features. Trepanation may co-occur with cranial injuries for which it was a remedial treatment in the past, as in the present, to relieve intra-cranial pressure from bleeding of the meninges that cover the brain (Nerlich et al. 2003b). Among prehistoric Andean populations, the increased frequency of such treatments appears to be correlated with increased cranial trauma resulting from interpersonal violence (Verano 2003).

3. 4. 5.

the presence of a body mass or masses within a grave cut or cuts; the presence of disorder in the orientation of the bodies indicating an apparent disregard for the manner of deposition that is often outside the bounds of normative practice; bodies that are in contact with one another; the presence of traumatic injuries, and a common pattern of a trait or traits related to cause and manner of death.

In archaeological examples of fully skeletonized remains, Verano (1986) and Knüsel and Boylston (2000) have noted that it is the relatively disorganized and nonnormative burial pattern of such individuals that identifies them as unusual upon first inspection (Fig. 16). Upon further analysis, though, these burials share all of the traits that define a mass grave of fleshed individuals. It is the presence of a high frequency of skeletal trauma of a similar variety that distinguishes the mass grave from burials relating to the occurrence of a mass mortality caused by disease, for example. Burials resulting from the passage of a contagion not only lack such lesions but also often reveal greater care in the interment of the individuals than is seen in a mass grave. The burials excavated at the Royal Mint Site, East Smithfield, London, despite being tiered up to five deep, were neatly ordered in a consistent burial position and orientation, with little inter-cutting of graves, and with soil separating

The functional characteristics of the cranium have implications for the distinguishing features of perimortem injuries. These can best be identified by disruptions of the bone by linear, depressed, and penetrating fractures that create a jagged bevel when they 58

CHRISTOPHER J. KNÜSEL: THE PHYSICAL EVIDENCE OF WARFARE - SUBTLE STIGMATA? reminiscent of the generally high levels of cranial trauma identified among the Towton remains, individuals buried after a Wars of the Roses battle in AD 1461 (Novak 2000).

each from the one above (Hawkins 1990; Grainger and Hawkins 1988). The palaeodemographic profile of these individuals also indicates their origin from a catastrophic passage of disease (Gowland and Chamberlain 2003; Margerison and Knüsel 2002).

A prehistoric example of an organized grave group comes from the Bronze Age site of Wassenaar, where 12 individuals were placed in a single grave in two opposed rows with some overlying skeletal elements between contiguous individuals, seven of which were oriented east-west with their heads to the west and five in the same orientation but with their heads to the east (Louwe Kooijmans 1993). Their apparent uniformity in burial, though, is interrupted by two individuals being buried prone with the remainder being supine. Smits and Maat (1993) provide evidence for a ‘common violent death’ from the evidence of three individuals bearing cut-marks and one found with an arrowhead in the thoracic region, amongst the otherwise poorly preserved remains. Vencl (1999) notes how the accumulating evidence for interments marked by haphazard and commingled human remains has fundamentally altered the interpretation of the Neolithic from being a time of peaceful farmers to one that occasioned extremes of violence (see also Louwe Kooijmans 1993 for a review of similar occurrences in the Bronze Age). Vencl (ibid.) also makes the potentially prophetic observation that some of these features might originally have been interpreted as ritual deposits and, perhaps, as reflective of secondary burial rites. For example, the remains of a group of males from the Iron Age (third century BC) palisaded cult centre at Ribemont-sur-Ancre (Somme) was originally interpreted as evidence to support the existence of funerary rite of exposure at the site (Brunaux 1993; 1996; 1998). Recently, however, Duday’s (1998) preliminary reanalysis of the human remains - some bearing perimortem injuries and all decapitated after death - suggests that the bodies of these individuals were transported to the burial site and deposited as ‘fresh’ cadavers, stacked up to four layers deep with some three to four individuals per square metre, and without regard for order or orientation. This type of pattern would fit the definition of a mass grave.

FIGURE 16. A view of the disordered burials from the mass grave at Towton. Note how the left upper limb of the individual on the right is submerged within a void in the mass of skeletal elements and how an elbow (distal humerus and proximal ulna and radius) from an individual beneath protrudes just to the left of the left ribs of the prone individual on the far right. (Photograph: Malin Holst)

From their experiences in Bosnia-Herzegovina, Skinner et al. (2002) provide evidence for what Haglund (2002: 245) refers to as an ‘organised grave group’, where bodies are placed in rows, a circumstance that distinguishes them from a mass grave with its commingled bodies. Importantly, though, another trait indicative of the mass grave was present - skeletal injuries revealed that these individuals died as a result of violent trauma. Twenty of 22 individuals showed signs of weapon-related injuries, with nine individuals displaying two or more wounds, in this case gunshot wounds. The pattern of traumatic injury - with the majority concentrated in the head and trunk region - is

One of the reasons that the number of mass graves may be under-appreciated is due to the difficulty of recognizing complete skeletons in a limited excavation as appears to have been the case at Ribemont-sur-Ancre (Duday 1998). A common occurrence in mass graves of skeletonized individuals is for limbs of superimposed bodies to pass down through the remains of individuals buried beneath, sometimes ending up as an articulated limb at a lower level or at the base of the grave (Fig. 17). This situation requires a three-dimensional recording system in order to recover complete individuals from entangled skeletons (Sutherland 2000).

59

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY finding difficult a holistic and singular definition to cover adequately the great variety of sacrifices, looked at sacrificial rites as a whole and defined them thus: ‘Le sacrifice est un acte religieux qui, par la conséquence d’une victime, modifie l’état de la personne morale qui l’accomplit ou de certains objets auxquelles elle s’intéresse. (‘Sacrifice is a religious act which, through the victim, modifies the moral standing of the sacrificer or of certain goals that are of interest to him/her’; author’s translation).’ In this definition, motivation appears to concern personal standing or the bringing about of events that may benefit the person making the sacrifice. This is surely an emic definition - one person or group’s sacrificial rite is another person or group’s loss and murder. Moreover, this definition does not indicate how such individuals could be identified from their physical remains. It seems that in order to identify a sacrificial victim several conditions must be met, and these follow the lines suggested by Verano (1986; 2001) for interpreting human remains from the Pacatnamu site in Peru (dating to the Moche V or Chimu phase of the Late Intermediate Period, c. AD 1100-1400) as sacrificed war prisoners. These individuals: x FIGURE 17. A reconstructed body overlay on the skeletal remains of Towton 16 as excavated in the midst of the Towton mass grave. Note the unusual prone burial position with outstretched and partially flexed limbs that is reflective of a non-normative burial rite in the Late Medieval period. Note, too, how the skeletonized forearms appear to disappear in this two-dimensional reconstruction - they descend down through bodies beneath. The apparent incongruity between the reconstructed right lower limb and the underlying skeletonized limb relates to a post-mortem displacement of the right os coxae. (Drawing: Caroline Needham)

x x

were found in a mass grave with little formal deposition and no grave inclusions; possess evidence for having been either hobbled or bound to a structure with ropes, lengths of which were preserved; had sustained multiple wounds – between five and 19 - delivered from a variety of directions.

Furthermore, x

The question of sacrifice The ritualized killing of defeated, captured, and often incarcerated enemies that is very much a part of warfare and its aftermath presents a more intractable problem which requires the greatest integration of the physical analysis of skeletal remains and their context. AldhouseGreen (2001) has linked a number of unusual burial rites and depositions of human remains with sacrificial rites. She (ibid.: 163-76) also outlines some of the intentions that can be ascribed to these acts - as understood by Greek and Roman writers at least - and as laid down in texts. These include aversion, appeasement, reprisal, vengeance and punishment for wrong-doing. As Aldhouse-Green (ibid.: 176) notes, however, the contexts for these acts and their circumstances remains poorly understood. Hubert and Mauss (1899: 41), although

x

x

these wounds seem to have been delivered by more than one assailant and their severity would seem to indicate that these individuals had been restrained because the number of injuries is not consistent with retaining an upright position without support of some kind; the individuals from Pacatnamu also show evidence of body part removal, the left radius in three of them, limbs in others, and the heart through an incision running inferiorly from the manubrium and cutting and fracturing ribs; the remains of insects suggest that these individuals had been left unburied for a time after death, a nonnormative funerary treatment.

Verano (2001) contrasts the evidence for mutilation and lack of formal burial with the careful interment and richly accompanied burials of the capac hucha (‘solemn sacrifice’) of children known from the region (see Benson 2001). The location of the individuals from Pacatnamu -

60

CHRISTOPHER J. KNÜSEL: THE PHYSICAL EVIDENCE OF WARFARE - SUBTLE STIGMATA? at the entrance to a ceremonial complex - and the inclusion of two vulture skeletons in the mass grave link them to contemporary iconography depicting the fate of what may be war captives. The all male composition of the group and their ages at death (between 15 and 35 years) make this a likely possibility, as does the presence of old, healed injuries. In each case the missing skeletal elements are accompanied by cut-marks and peri-mortem trauma which indicate that the absence of skeletal elements is not due to taphonomic factors but relates to forcible removal at or about the time of death. It is clear from this example that the physical evidence for violent treatment must be completely recorded and considered together with the contextual disposition of the remains to make the case for sacrifice.

Conclusion Owing to problems of identification, but also as a consequence of paradigms that are too narrowly drawn, it is likely that much evidence of violence and warfare in the archaeological record has gone unnoticed and/or unrecorded. The difference between intent and capability to do harm, on the one hand, and the commission of violent acts on the other, can be distinguished in the archaeological record, especially given adequate recovery, recording, and analysis of human remains. Both of these types of information are important if we are to identify the circumstances under which the potential for violence becomes the perpetration of violence. In the process, too, we can anticipate not only times and regions with evidence of violent encounters, but also when and where these were absent or, at least, less obvious.

Execution is a symbolic act and, at least on physical grounds, it may be indistinguishable from sacrifice, even given contextual evidence. Both are ritualized killings and thus may be equivalent from the perspective of human remains analysis, based as it is on the patterning and number of injuries sustained. Evidence of what appears to be binding of the hands motivates Lambot (1998) to interpret the remains of a young male from Acy Romance, Ardennes (see Craig et al., this volume), who had sustained a peri-mortem cut to the right temporal and parietal, as a sacrificial victim. The unusual placement of this individual’s forearms, drawn to the right side of the pelvis and lying beneath the vertebral column, suggest that they might have once been bound behind his back, although the organic restraints were not preserved in this instance.

In a sense, this treatment envisions a time when violent conflict can be reliably identified and diachronic comparisons made, as well as ruled out when such evidence is not found. Nystrom’s contribution to this volume on co-operative behaviour and violence avoidance among non-human primates already presages this development and provides us with a research orientation not solely concerned with demagogy, rapine and cruelty. Acknowledgements I thank Mike Parker Pearson and the Sheffield University Archaeological Society for inviting me to present at the conference that spawned this paper. Carol Palmer (Sheffield), as ever, helped to define a more tractable paper from one that was much longer. Alan Outram (Exeter), Shannon Novak (Utah) and Anthea Boylston (Bradford) have helped to refine some of the ideas and identifying features that appear in this paper. Any transmission errors are mine, though. Graham Martin (Wakefield Hospital, Wellington, N.Z.) brought my attention to the ‘seat’ or non-perforating, wedge-type of cranial injury. I thank Jean Brown, Eric Houlder, Carol Palmer and Malin Holst who took photographs that appear in this paper. Caroline D. Needham, of the Unit of Medical Art in Medicine, University of Manchester, drew Figure 15. Her time and expertise was made available through British Academy Grant SG-36744.

Comparison with recent killings in the context of ethnic violence and war crimes suggests, however, that the presence of binding and cranial trauma may also indicate execution (cf Haglund 2002: fig. 12.5). In the context of ethnic violence, these individuals may also be found in mass graves or as ‘satellite remains’ to mass graves (ibid.: 247). For other periods, execution has been primarily by sharp force trauma, using weapons such as swords. The Anglo-Saxon male, aged 28-32 years at death, found buried at Stonehenge may be an example of execution by the sword in the post-Roman period. This individual had sustained a peri-mortem fracture to the gonial angle of the ascending ramus of the mandible (Pitts et al. 2002; cf Smits and Maat 1993: fig. 3 for a similar example from Wassenaar), which is consistent with decapitation by sharp force trauma. The use of a sword, a highly symbolically charged weapon, may indicate a more high-profile and public event that is more in accordance with recent execution methods, when the purpose is to demonstrate that a particular individual is dead and that the death was carried out by decree of a higher authority. Sacrifice is the sometimes emic equivalent of execution from the etic perspective by which anthropologists and archaeologists interpret the past.

Bibliography Allmand, C.T. (ed.) 1973. Society at War: the experience of England and France during the Hundred Years War. Edinburgh: Oliver and Boyd. Barnes, J.K. 1870. The Medical and Surgical History of the War of the Rebellion (1861-1865). Washington DC: Government Printing Office. Barrett, J.C., Freeman, P.W.M. and Woodward, A. 2000. Cadbury Castle, Somerset: the later prehistoric and early historic archaeology. London: English Heritage Archaeological Report 20. 61

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Dandy, D.J. 1993. Essential Orthopaedics and Trauma. Edinburgh: Churchill Livingstone. Second edition. Darling, J.A. 1998. Mass inhumation and the execution of witches in the American Southwest. American Anthropologist 100: 732-52. Duday, H. 1998 Le charnier gaulois de Ribemont-surAncre (Somme). Etudes et Documents Fouilles 4, Les Celtes: rites funéraires en Gaule du nord entre le VIe et le 1er siècle avant Jesus-Christ. Namur: Ministère de la Région Wallone. 113-19. Earle, T. 1997. How Chiefs Come to Power: the political economy in prehistory. Stanford CA: Stanford University Press. Evans, J.G., Atkinson, R.J.C., O’Connor, T. and Green, H.S. 1984. Stonehenge - the environment in the Late Neolithic and Early Bronze Age and a Beaker-Age burial. Wiltshire Archaeological and Natural History Magazine 78: 7-30. Ferguson, R.B. 1984. Introduction: studying war. In R.B. Ferguson (ed.) Warfare, Culture, and Environment. Orlando FL: Academic Press. 1-81. Ferguson, R.B. 2001. Materialist, cultural, biological theories on why Yanomami make war. Anthropological Theory 1: 99-116. Ferguson, R.B. and Whitehead, N.L. (eds) 1992. War in the Tribal Zone: expanding states and indigenous warfare. Santa Fe NM: School of American Research. Fiorato, V., Boylston, A. and Knüsel, C.J. (eds) Blood Red Roses: the archaeology of a mass grave from the battle of Towton AD 1461. Oxford: Oxbow Books. Frayer, D.W. 1997. Ofnet: evidence for a Mesolithic massacre. In D. Martin and D.W. Frayer (eds) Troubled Times: violence and warfare in the past. Amsterdam: Gordon and Breach. 181-216. Galloway, A. 1999. Broken Bones: anthropological analysis of blunt force trauma. Springfield IL: Charles C. Thomas. Gargett, R.H. 1989. Grave shortcomings: the evidence for Neanderthal burial. Current Anthropology 30: 157-90. Gargett, R.H. 1999. Middle Palaeolithic burial is not a dead issue: the view from Qafzeh, Saint-Césaire, Kebara, Amud, and Dederiyeh. Journal of Human Evolution 37: 27-90. Gowland, R.L. and Chamberlain, A.T. 2003. The age of plague: a palaeodemographic study of a catastrophic death assemblage. American Journal Physical Anthropology Supplement 36: 103. Grainger, I, and Hawkins, D. 1988. Excavations at the Royal Mint Site 1986-1988. The London Archaeologist 5: 429-36. Green, M.J.A. 2001. Dying for the Gods: human sacrifice in Iron Age and Roman Europe. Stroud: Tempus Guilaine, J and Zammit, J. 2001. Le Sentier de la Guerre: visages de la violence préhistorique. Paris: Seuil. Gurdjian, E.S. 1975. Impact Head Injury. Springfield IL: Charles C. Thomas. Gurdjian, E.S. and Lissner, H.R.1945. Deformation of the skull in head injury: a study of the ‘Stresscoat’ technique. Surgery, Gynecology and Obstetrics 81: 679-87.

Benson, E.P. 2001. Why sacrifice? In E.P. Benson and A.G. Cook (eds) Ritual Sacrifice in Ancient Peru. Austin TX: University of Texas Press. 1-20. Berryman, H.E. and Symes, S.A. 1998. Recognizing gunshot and blunt cranial trauma through fracture interpretation. In K.D. Reichs and W.M. Bass (eds) Forensic Osteology: advances in the identification of human remains (2nd edition). Springfield IL: Charles C. Thomas. 333-52. Blumenshine, R.J., Marean, C.W. and Capaldo, S.D. 1996. Blind tests of inter-analyst correspondence and accuracy in the identification of cut marks, percussion marks, and carnivore tooth marks. Journal of Archaeological Science 23: 493-507. Boylston, A. 2000. Evidence for weapon related trauma in British archaeological samples, In: M. Cox and S. Mays (eds) Human Osteology in Archaeological and Forensic Science. London: Greenwich Medical Media Ltd. 357-80. Brain, C.K. 1981. The Hunters or the Hunted? Chicago: University of Chicago Press. Brunaux, J.-L. 1993. Les sanctuaires gaulois. La Recherche 256 (24): 820-8. Brunaux, J.-L. 1996. Les Religions Gauloises: rituels celtiques de la Gaule indépendante. Paris: Errance. Brunaux, J.-L. 1998. Un monumental trophée celtique à Ribemont-sur-Ancre (Somme). Etudes et Documents Fouilles 4, Les Celtes: rites funéraires en Gaule du nord entre le VIe et le 1er Siècle avant Jesus-Christ. Namur: Ministère de la Région Wallone. 107-13. Carman, J. and Harding, A. 1999. Introduction. In J. Carman and A. Harding (eds) Ancient Warfare: archaeological perspectives. Stroud: Sutton. 1-9. Carneiro, R.L. 1970. A theory of the origin of the state. Science 169: 733-8. Carneiro, R.L. 1990. Chiefdom-level warfare as exemplified in Fiji and the Cauca Valley. In J. Haas (ed.) The Anthropology of War. Cambridge: Cambridge University Press. 190-211. Chapman, J. 1999. The origins of warfare in the prehistory of central and eastern Europe. In J. Carman and A. Harding (eds) Ancient Warfare: archaeological perspectives. Stroud: Sutton. 101-42. Chapman, J. 2000. Fragmentation in Archaeology: people, places and broken objects in the prehistory of southeastern Europe. London: Taylor and Francis. Connolly, P. 1998. Greece and Rome at War. London: Greenhill Books. Contamine, P. 1980. War in the Middle Ages. (trans. M. Jones). Oxford: Basil Blackwell. Courville, C.B. 1962. Forensic neuropathology IV. Significance of traumatic extracranial and cranial lesions. Journal of Forensic Sciences 7: 303-22. Cunha, E. and Silva, A.M. 1997. War lesions from the famous Portuguese medieval battle of Aljubarrota. International Journal of Osteoarchaeology 7: 595-9. Cunliffe, B. 1992. Pits, preconceptions, and propitiation in the British Iron Age. Oxford Journal of Archaeology 11: 69-83.

62

CHRISTOPHER J. KNÜSEL: THE PHYSICAL EVIDENCE OF WARFARE - SUBTLE STIGMATA? Gurdjian, E.S. and Lissner, H.R.1946. Deformations of the skull in head injury studied by the ‘Stresscoat’ technique, quantitative determinations. Surgery, Gynecology and Obstetrics 83: 219-33. Gurdjian, E.S. and Lissner, H.R. 1947. Deformation of the skull in head injury as studied by the ‘Stresscoat’ technique. American Journal of Surgery 73: 269-81. Gurdjian, E.S., Lissner, H.R. and Webster, J.E. 1947. The mechanism of production of linear skull fracture: further studies on deformation of the skull by the ‘Stresscoat’ technique. Surgery, Gynecology and Obstetrics 85: 195-210. Gurdjian, E.S., Webster, J.E., and Lissner, H.R.1950. The mechanism of skull fracture. Journal of Neurosurgery 7:106-14. Haas, J. 1990. Warfare and the evolution of tribal polities in the prehistoric Southwest, In J. Haas (ed.) The Anthropology of War. Cambridge: Cambridge University Press. 171-89. Haglund, W.D. 2002. Recent mass graves, an introduction. In W.D. Haglund and M.H. Sorg (eds) Advances in Forensic Taphonomy: method, theory, and archaeological perspectives. Boca Raton FL: CRC Press. 243-61. Haglund, W.D., Reay, D.T. and Swindler, D.R. 1988. Tooth mark artifacts and survival of bones in animal scavenged human skeletons. Journal of Forensic Sciences 33: 985-97. Haglund, W.D., Reay, D.T. and Swindler, D.R. 1989. Canid scavenging/disarticulation sequence of human remains in the Pacific Northwest. Journal of Forensic Sciences 34: 587-606. Harding, A.F. 2000. European Societies in the Bronze Age. Cambridge: Cambridge University Press. Harding, A.F., Outram, A.K., Knüsel, C.J., Knight, S. and Palmer, C. forthcoming. The Late Bronze Site of Velim Skalka, Czech Republic. Hawkins, D. 1990. The Black Death and the new London cemeteries of 1348. Antiquity 64: 637-42. Hill, J.D. 1995. Ritual and Rubbish in the Iron Age of Wessex: a study on the formation of a specific archaeological record. Oxford: BAR Publishing, BAR British Series 242. Hill, J.D. 1996. Hillforts and the Iron Age of Wessex. In T. Champion and J. Collis (eds) Recent Trends in the Archaeology of Iron Age Britain. Sheffield: JR Collis Publications. 95-116. Hill, P. and Wileman, J. 2002. Landscapes of War: the archaeology of aggression and defence. Stroud: Tempus. Hubert, H. and Mauss, M. 1899. Essai sur la nature et la fonction du sacrifice. Année Sociologique 2: 29-138. Kaufman, M.H., Whitaker, D. and McTavish, J. 1997. Differential diagnosis of holes in the calvarium: application of modern clinical data to paleopathology. Journal of Archaeological Science 24: 193-218. Keegan, J. 1993. A History of Warfare. London: Pimlico. Keeley, L.H. 1996. War before Civilization: the myth of the peaceful savage. Oxford & New York: Oxford University Press.

Keen, M. (ed.) 1999. Medieval Warfare: a history. Oxford: Oxford University Press. Knüsel, C.J. and Boylston, A. 2000. The human remains in their medieval context, In V. Fiorato, A. Boylston and C.J. Knüsel (eds) Blood Red Roses: the archaeology of a mass grave from the battle of Towton AD 1461. Oxford: Oxbow Books. 169-88. Kristiansen, K. 1998. Europe before History. Cambridge: Cambridge University Press. Lambert, P.M. 1997. Patterns of violence in prehistoric hunter-gatherer societies of coastal southern California. In D.L. Martin and D.W. Frayer (eds) Troubled Times: violence and warfare in the past. Amsterdam: Gordon and Breach. 77-109. Lambot, B. 1998. Les morts d’Acy Romance (Ardennes) à La Tène finale. Etudes et Documents Fouilles 4, Les Celtes: rites funéraires en Gaule du nord entre le VIe et le 1er siècle avant Jesus-Christ. Namur: Ministère de la Région Wallone. 75-87. Larsen, C.S. 1997. Bioarchaeology: interpreting behavior from the human skeleton. Cambridge: Cambridge University Press. Lissner, H.R. and Gurdjian, E.S. 1946. A study of the mechanical behavior of the skull and its contents when subjected to injuring blows. Proceedings of the Society of Experimental Stress Analysis 3: 40-6. Liston, M.A. and Baker, B.J. 1996. Reconstructing the massacre at Fort William Henry, New York. International Journal of Osteoarchaeology 6: 28-41. Louwe Kooijmans, L.P. 1993. An Early/Middle Bronze Age multiple burial at Wassenaar, the Netherlands. Analecta Praehistorica Leidensia 26: 1-20. Lyman, R.L. 1994. Vertebrate Taphonomy. Cambridge: Cambridge University Press. Madea, B. and Staak, M. 1988. Determination of the sequence of gunshot wounds of the skull. Journal of the Forensic Science Society 28: 321-8. Mant, A.K. 1987. Knowledge acquired from post-war exhumations. In A. Boddington, A.N. Garland and R.C. Janaway (eds) Death, Decay, and Reconstruction: approaches to archaeology and forensic science. Manchester: Manchester University Press. 65-78. Margerison, B.J. and Knüsel, C.J. 2002. A comparison of attritional and catastrophic cemeteries: the palaeodemography of the medieval plague cemetery at the Mint site, London. American Journal of Physical Anthropology 119: 134-43. Merbs, C.F. 1989. Trauma. In M.Y. Iscan and K.A.R. Kennedy (eds) Reconstruction of Life from the Skeleton. New York: Wiley-Liss. 161-89. Milner, G.R. 1995. An osteological perspective on prehistoric warfare. In L.A. Beck (ed.) Regional Approaches to Mortuary Analysis. New York: Plenum Press. 221-38. Nerlich, A., Peschel, O., Zink, A. and Rösing, F.W. 2003a. The pathology of trepanation: differential diagnosis, healing and dry bone appearance in modern cases. In R. Arnott, S. Finger and C.U.M. Smith (eds)

63

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY theory, and archaeological perspectives. Boca Raton FL: CRC Press. 277-92. Schulting, R.J. and Wysocki, M. 2002. Cranial trauma in the British Earlier Neolithic. Past 41: 4-6. Sharples, N.M. 1991a. Warfare in the Iron Age of Wessex. Scottish Archaeological Review 8: 79-89. Sharples, N.M. 1991b. Maiden Castle: excavations and field survey 1985-6. London: Historic Buildings and Monuments Commission for England, English Heritage Archaeological Report 19. Shipman, P. and Rose, J. 1983. Early hominid hunting, butchering and carcass processing behaviors: approaches to the fossil record. Journal of Anthropological Archaeology 10: 465-74. Skinner, M.F., York, H.P. and Connor. M.A. 2002. Postburial disturbance of graves in BosniaHerzegovina. In W.D. Haglund and M.H. Sorg (eds) Advances in Forensic Taphonomy: method, theory, and archaeological perspectives. Boca Raton FL: CRC Press. 293-308. Smith, M.O. 2003. Beyond palisades: the nature and frequency of late prehistoric deliberate violent trauma in the Chickamauga reservoir of east Tennessee. American Journal of Physical Anthropology 121: 303-18. Smits, L. and Maat, G. 1993. An Early/Middle Bronze Age common grave at Wassenaar, the Netherlands: the physical anthropological results. Analecta Praehistorica Leidensia 26: 21-8. Sutherland, T. 2000. Recording the grave. In V. Fiorato, A. Boylston, and C.J. Knüsel (eds) Blood Red Roses: the archaeology of a mass grave from the battle of Towton AD 1461. Oxford: Oxbow Books. 36-44. Symes, S.A., Williams, J.A., Murray, E.A., Hoffman, M., Holland, T.D., Saul, J., Saul, F. and Pope, E.J. 2002. Taphonomic context of sharp force trauma in suspected cases of human mutilation and dismemberment. In W.D. Haglund and M.H. Sorg (eds) Advances in Forensic Taphonomy: method, theory, and archaeological perspectives. Boca Raton FL: CRC Press. 403-34. Tallett, F. 1992. War and Society in Early-Modern Europe, 1495-1715. London: Routledge. Ubelaker, D.H. 1996. Skeletons testify: anthropology in forensic science. AAPA luncheon address, April 12, 1996. Yearbook of Physical Anthropology 39: 229-44. Ubelaker, D.H. and Adams, B.J. 1995. Differentiation of peri-mortem and post-mortem trauma using taphonomic indicators. Journal of Forensic Sciences 40: 509-12. Valentin, F. and Le Goff, I. 1998. La sépulture secondaire mésolithique de La Chaussée-Tirancourt: fractures sur os frais ou os secs? L’Anthropologie 102: 91-5. Vencl, S. 1999. Stone Age warfare. In J. Carman and A. Harding (eds) Ancient Warfare: archaeological perspectives. Stroud: Sutton. 57-72. Verano, J.W. 1986. A mass burial of mutilated individuals at Pacatnamu. In C.B. Donnan and J.A. Cook (eds) The Pacatnamu Papers, Vol. 1. Los Angeles: UCLA Museum of Cultural History. 117-38.

Trepanation: history – discovery - theory. Lisse: Swets and Zeitlinger. 43-51. Nerlich, A., Zink, A., Szeimies, U., Hagedorn, H.G. and Rösing, F.W. 2003b. Perforating skull trauma in Ancient Egypt and evidence for early neurosurgical therapy. In R. Arnott, S. Finger and C.U.M. Smith (eds) Trepanation: history – discovery - theory. Lisse: Swets and Zeitlinger. 191-201. Novak, S. 2000. Battle-related trauma. In V. Fiorato, A. Boylston, and C.J. Knüsel (eds) Blood Red Roses: the archaeology of a mass grave from the battle of Towton AD 1461. Oxford: Oxbow Books. 90-102. Novak, S.A. and Kopp, D. 2003. To feed a tree in Zion: osteological analysis of the 1857 mountain massacre. Historical Archaeology 37: 85-118. O’Connell, R.L. 1995. Ride of the Second Horseman: the birth and death of war. Oxford: Oxford University Press. Ogilvie, M.D. and Hilton, C.E. 2000. Ritualized violence in the prehistoric American Southwest. International Journal of Osteoarchaeology 10: 27-48. Orschiedt, J. 1998. Ergebnisse einer neuen Untersuchung der spätmesolithischen Kopfbestattungen aus Süddeutchland. Urgeschichtlichen Materialhefte 12: 147-60. Osgood, R. and Monks, S. with Toms, J. 2000. Bronze Age Warfare. Stroud: Sutton. Otterbein, K.F. 1999. A history of research on warfare in anthropology. American Anthropologist 101: 794805. Outram, A.K. 1999. A comparison of Paleo-Eskimo and medieval Norse bone fat exploitation in western Greenland. Arctic Anthropology 36: 103-17. Outram, A.K. 2002. Bone fracture and within-bone nutrients: an experimentally based method for investigating levels of marrow extraction. In P. Miracle and N. Milner (eds) Consuming Passions and Patterns of Consumption. Cambridge: University of Cambridge McDonald Institute Monograph. 51-63. Pitts, M., Bayliss, A., McKinley, J., Boylston, A., Budd, P., Evans, J., Chenery, C., Reynolds, A. and Semple, S. 2002. An Anglo-Saxon decapitation and burial at Stonehenge. Wiltshire Archaeological and Natural History Magazine 95: 131-46. Sauer, N.J. (1998). The timing of injuries and manner of death: distinguishing among antemortem, perimortem and post-mortem trauma. In K.D. Reichs and W.M. Bass (eds) Forensic Osteology: advances in the identification of human remains. Springfield IL: Charles C. Thomas. Second edition. 321-32. Saul, J.M. and Saul, F.P. 2002. Forensics, archaeology, and taphonomy: the symbiotic relationship. In W.D. Haglund and M.H. Sorg (eds) Advances in Forensic Taphonomy: method, theory, and archaeological perspectives. Boca Raton FL: CRC Press. 293-308. Schmitt, S. 2002. Mass graves and the collection of forensic evidence: genocide, war crimes, and crimes against humanity. In W.D. Haglund and M.H. Sorg (eds) Advances in Forensic Taphonomy: method,

64

CHRISTOPHER J. KNÜSEL: THE PHYSICAL EVIDENCE OF WARFARE - SUBTLE STIGMATA? Walker, P.L. and Long, J.C. 1977. An experimental study of the morphological characteristics of tool marks. American Antiquity 42: 605-15. Weigelt, J. 1989. Recent Vertebrate Carcasses and their Paleobiological Implications. (trans. J. Schaefer). Chicago: University of Chicago Press. Wheeler, R.E.M. 1943. Maiden Castle, Dorset. Oxford: Oxford University Press, Reports of the Research Committee of the Society of Antiquaries of London 12. White, T.D. 1992. Prehistoric Cannibalism at Mancos 5MTUMR-2346. Princeton NJ: Princeton University Press. Whittle, A.W.R. 1996. Europe in the Neolithic: the creation of new worlds. Cambridge: Cambridge University Press. Whittle, A.W.R., Pollard, J. and Grigson, C. 1999. The Harmony of Symbols: the Windmill Hill causewayed enclosure. Oxford: Oxbow Books, Cardiff Studies in Archaeology. Willey, P. and Emerson, T.E. 1993. The osteology and archaeology of the Crow Creek massacre. Plains Anthropologist 38: 27-269.

Verano, J.W. 2001. The physical evidence of human sacrifice in ancient Peru. In E.P. Benson and A.G. Cook (eds) Ritual Sacrifice in Ancient Peru. Austin: University of Texas Press. 165-84. Verano, J.W. 2003. Trepanation in prehistoric South America: geographic and temporal trends over 2,000 years. In R. Arnott, S. Finger and C.U.M. Smith (eds) Trepanation: history - discovery - theory. Lisse: Swets and Zeitlinger. 223-36. Villa, P. and Mahieu, E. 1991. Breakage patterns of human long bones. Journal of Human Evolution 21: 27-48. Wahl, J. and König, H.G. 1987. Anthropologischtraumatologiche Untersuchung der menschlichen Skelettreste aus dem Bandkeramischen Massengrab bei Talheim, Kreis Heilbronn. Fundberichte aus Baden-Württemberg 12: 65-193. Walker, P.L. 1989. Cranial injuries as evidence of violence in prehistoric southern California. American Journal of Physical Anthropology 80: 313-23. Walker, P.L. 2001. A bioarchaeological perspective on the history of violence. Annual Review of Anthropology 30: 573-96.

65

The head burials from Ofnet cave: an example of warlike conflict in the Mesolithic Jörg Orschiedt Archäologisches Institut, University of Hamburg

The absolute dating of the skull clusters has long been disputed but has been resolved by C14 accelerator dates. A date in the Late Mesolithic is confirmed by four dates from Oxford as well as two other dates from Cologne and Los Angeles (Schulte im Walde et al. 1986; Hedges et al. 1989: 211; Orschiedt 1998).

The circumstances of discovery During the excavations by Robert Rudolf Schmidt at Grosser Ofnet in 1908, two ‘skull clusters’ were discovered in the immediate vicinity of the cave entrance (Schmidt 1909; 1912: 34-42; 1913). The heads had been deposited in two pits which were situated within the Magdalenian layer VI. The ‘skull clusters’ came from the Mesolithic layer VII, only 5cm thick, which cut into the underlying layer (Fig. 1). The pits lay only about 1m apart and both were filled by a thick sediment stained with red ochre and interspersed with flecks of charcoal and calcified bone. The large skull cluster had a diameter of 0.76m and, according to new, revised calculations, contained the heads of 28 individuals. The smaller cluster, of unknown diameter, contained six individuals. All the skulls were oriented on the same line of sight, facing west.

The human remains During the examination of the material, the skull and fragments of the mandible of a neonate (designated 9A) were identified amongst the skeletal remains of Individual 9 (a four or five year-old child). The skeletal remains thus consist of 34 crania with mandibles and 82 cervical vertebrae in total. All the individuals were apparently deposited very soon after death occurred, as not only the lower jaw but also the cervical vertebrae were found to be still articulated. The skulls are in a quantatively variable state of preservation, whereas the qualitative preservation is designated as good to very good. In some cases there is major damage to the children’s skulls which can be explained by the lack of robustness of the bone. The age determination (Table 1) shows a preponderance of individuals in the age group ‘infant I’ (ages 1-6; 16 individuals; 46%). Likewise the age band ‘young adult’

Several other human skeletal remains, probably from the Bronze Age or Iron Age, were also found near the skulls within Ofnet cave (Weissmüller 1982: 37, 71, 187, figs 9, 12). The relationship between the skull clusters and one of these other skull fragments (accompanied by a mandible and a fragment of cervical vertebra) from Oskar Fraas’ 1875 excavation is unclear (Fraas 1876; Schlitz 1912: 241).

FIGURE 1. Cross-section of the cave entrance area at Grosser Ofnet, showing the skull clusters (Schmidt 1912: fig. 6).

67

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY (Lithoglypus naticoides, Neritina fluviatis L., Columbella rustica and Carinifex multiformis). The teeth were described as actually cemented together in some cases; some of them were lying on top of the skulls, as were individual shells. In the majority of cases, however, the teeth and the shell ornaments were found lying beneath the neck region (Schmidt 1912: 37). Twenty stone artefacts found from the burial level and within the pit fill are not to be viewed as offerings but rather they denote a component of the culture layer VII. Like the artefacts from layer VII, they belong to the Early and Late Mesolithic (Müller-Beck 1983: 396). Among the artefacts described more specifically were two geometric triangles which are dated to Beuronian A or B (Naber 1974: 78-9).

unknown male

elderly 1

female

older mature adult mature adult

1

older adult

1 1

young adult

3 7

juvenile 2

infant II

In the analysis of the distribution of grave goods and their association with particular individuals, certain difficulties arise. Judging by the excavation methodology, it is in the first place probable that not all grave goods were retrieved. As well as losses during the excavation itself, it is necessary to take into account later museum losses resulting from the very small size of the artefacts. The incomplete recording of the grave goods can be deduced from the fact that, during the recent preparation of the skeletons, a total of eight pieces of shell jewellery, seven flints and 26 fragments of animal bone were identified. Furthermore, uncertainties arise regarding the association of grave goods and individuals made during the original excavation. The tightly packed context of the skulls and the excavation methods of the time leave it open to doubt whether the grave goods can be ascribed to individual skeletons.

17

infant I 1

neonate 0

5

10

15

20

TABLE 1. The age and sex determination of the Ofnet skulls.

(20-30 years) is well represented with 10 individuals (29%). The entire age range extends from neonate (0-1 year old) to elderly (60-70 years old). The sex determination of the skulls (see Table 1) produced a distinct predominance of females (although this analysis is restricted by the presence of nine children for whom sex could not be determined). From this unequal distribution, it can be concluded that, for the skeletal remains at issue here, it is not a question of a complete population in the conventional demographic sense (Esenwein-Rothe 1982: 61). Moreover, the timespan between the isolated C14 dates leaves a sufficient period of time for repeated deposition to have occurred at the site. Accordingly the proposition that the deposition of the skull clusters was a single occurrence is not maintained.

Red deer teeth as grave goods can be assigned securely to 62% of the individuals and shell ornaments to 71%. For only two individuals were there no identified offerings. In contrast to the shell jewellery, the teeth were counted out so their distribution can be more closely differentiated. With respect to the quantity of grave goods, there is a distinct disproportion between adults and children (Table 2). With 19 individuals, children make up half the total number of individuals present, yet only 54 teeth are associated with them, in contrast to 167 items counted for the adult individuals in total. Nonetheless, 60% of children and 64% of adults were provided with red deer teeth as grave goods. Furthermore it is evident that among the adults, it is above all the female individuals who possessed the greater share of the teeth (Table 2).

Furthermore, for Late Mesolithic hunter-gatherers in the early Atlantic, a small group size is to be expected. As the forests advanced, the game animals changed and no longer appeared in herds. With this change in the environment, the mobility ratio would have increased, leading to a reduction in group size which, according to ethnographic data (Müller-Beck 1983: 401; Helbling 1987: 218), comprises an average of 25 people for modern hunter-gatherers. On these grounds it is to be assumed that the skull clusters in the Grosser Ofnet cave represent the repeated use of a burial site by one or more groups.

Less clear are the proportions for the shell ornaments. Here the children are equipped quantatively more abundantly than are the adults. Shell jewellery was detected irrespective of age with 77% of the females, 75% of males and 57% of individuals of indeterminate sex. On further inspection differentiating by age, it is apparent that 65% of children and 85% of adults were provided with shell ornaments as grave goods.

The grave goods Among the grave goods were 215 pierced teeth of red deer (Cervus elaphus) and 4,250 shell ornaments

68

JÖRG ORSCHIEDT: THE HEAD BURIALS FROM OFNET CAVE: AN EXAMPLE OF WARLIKE CONFLICT IN THE MESOLITHIC Schmidt’s data, cut-marks were apparent on two further individuals. These findings could not be verified since the relevant bones could no longer be located and appear to be missing. Nevertheless, a photograph survives of one of these items that allows us to distinguish unmistakable cut-marks (Schmidt 1912: pl. XII.16). From this evidence, it is reasonable to accept Schmidt’s statement. Therefore there appear to be nine individuals in total with cut-marks on the cervical vertebrae.

180 156

160 140 120 100 80 60 40 20

28

20

6

The absence of cut-marks on the cervical vertebrae of the remaining individuals is not to be interpreted as evidence of a distinctive method of decapitation. Removal of the head from the body must unavoidably leave cut-marks on the cervical vertebrae, notwithstanding the fact that this is demonstrable on only a small proportion of the individuals from Grosser Ofnet, so it is to be assumed that the cervical vertebrae bearing the corresponding cutmarks remained with the torso.

11

0 inf (m)

inf (f)

inf (indet.)

ad - eld (m)

ad - eld (f)

TABLE 2. The distribution of red deer teeth according to age group.

It is clear that several of the conclusions advanced by R.R. Schmidt must be re-examined. Contrary to his statements, offerings of red deer teeth were clearly given to male individuals. Moreover, there is no increase in the number of teeth with individuals of advanced age. The highest number of items – 69 in all – was found with a young adult woman (Ofnet 3), whereas 36 teeth (the next highest number) belonged to an old woman (Ofnet 18).

During the reassembling and preparation of the Ofnet skulls by T. Mollison in the 1930s, numerous defects and injuries were recognized which could not be attributed to decay or post-depositional events (Mollison 1936). Mollison documented definite and probable ante-mortem slash wounds on a total of 21 skulls. Only five individuals showed the defects as definite slash wounds. In spite of the detailed descriptions recorded for these bone defects, the findings could not always be reconfirmed. With the group of probable slash wounds, it was a problem to differentiate these injuries from damage caused by taphonomic processes. For this reason, a complete re-investigation of all the skulls was undertaken, which considered differentially diagnostic features (Orschiedt 1998; 1999). Possible ante-mortem skull injuries were analysed according to the conventional criteria of forensic medicine (Brückner and Hinze 1991: 93-5; Polson 1965: 126-38; Maples 1986; Merbs 1989; Sellier 1971; Ubelaker 1991; Wahl and König 1987: 11226). As criteria for determining defects as definite antemortem skull injuries, the following points were applied:

Traces of skeletal injuries The occurrence of cut-marks on the cervical vertebrae of some individuals was first pointed out in the publication by Schmidt of the discoveries from Grosser Ofnet (Schmidt 1912: 37; Bégouen 1912). A complete examination of all cervical vertebrae was not carried out at that time. As a consequence of the search for traces of bone modification or injury, the surfaces of all cervical vertebrae have therefore been re-examined systematically with a binocular microscope (at up to x40 magnification) and with a scanning electron microscope (SEM). In order to make the completion of the investigation easier and to avoid damage to the original surfaces of the bone, the reexamination of recorded cut-marks under the SEM took place with the help of a two-phase method of making casts (Haidle and Orschiedt 1995).

1. 2. 3.

4.

FIGURE 2. Cut-marks on the fourth cervical vertebra (C4) from Individual 20 from Grosser Ofnet.

the shape of the fracture margin; the funnel-shaped expansion formed from the penetration of the inner table of the cranial vault (a bevel); the occurrence of punched-out bone fragments in the centre of the fracture defect, caused by the force of a blow (comminution from blunt force trauma); the occurrence of radiating and concentric fracture lines and/or fissures which lead outwards from the injury, on a regular and rectilinear course.

On the basis of this examination, 14 definite ante-mortem skull injuries were observed on a total of six individuals (Ofnet 1, 2, 11, 21, 24 and 30). On two further individuals (Ofnet 31 and 32) probable ante-mortem skull injuries were diagnosed. A precise conclusion was not

Cut-marks were determined on the body of the cervical vertebrae of seven individuals (Fig. 2). In five cases, the fourth vertebra shows cut-marks. According to 69

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY The location of 64% of all fatal skull injuries in the area of the back of the head (Fig. 4) enables us to infer a situation involving a surprise attack, as can be shown for the Bandkeramik mass grave at Talheim (Wahl and König 1987: 184). Only for the two adult men who exhibit the highest number of fatal skull injuries can an attack from more than one direction be calculated on the basis of the location of their wounds. All the remaining individuals suffered a single fatal injury to the back of the skull (Fig. 5).

possible for these two cases given the state of preservation of the cranial regions in question. The results of Mollison’s examination (1936) contain numerous anomalies. The large group of defects that for him appeared to be probable slash wounds can be identified as, for example, unmistakable post-mortem damage caused by taphonomic processes. The group of individuals with definite slash wounds is made up of two children in the age range ‘infant 1’ (Ofnet 1 and 30) and four adults in the age bands ‘young adult’ (Ofnet 11, 21 and 24) and ‘mature adult’ (Ofnet 2; Table 3). The comparison of age and sex of the individuals with definite slash wounds indicates that it is adult men above all who exhibit most skull injuries. The skulls of individuals 2 and 21 (Fig. 3) show a total of 11 slash wounds. In no case can any signs of the healing process be identified on the fracture edges, which demonstrates that there was an immediately fatal outcome to the wounds through injury to the brain, haemorrhage or cerebral oedema. For two individuals (Ofnet 2 and 24), however, traces can be distinguished of old skull injuries that did not perforate the cranial vault. elderly (m)

antemortem skull injuries

older mature (m)

number of individuals

older mature (f) 1 1

mature (m) mature (f)

2

young ad (m)

3 1

young ad (f)

7

juv (m) juv (f) inf ll (f) inf l (indet)

1

inf l (f)

1

8

FIGURE 4. The location of injuries on the skulls from Grosser Ofnet.

8

neo (f)

0

2

4

6

8

10

TABLE 3. Age and sex of individuals with ante-mortem skull injuries.

FIGURE 3. Ante-mortem skull injuries on the os frontale of Individual 21 from Grosser Ofnet.

FIGURE 5. Ante-mortem slash wound on the os occipitale of Individual 11 from Grosser Ofnet.

70

JÖRG ORSCHIEDT: THE HEAD BURIALS FROM OFNET CAVE: AN EXAMPLE OF WARLIKE CONFLICT IN THE MESOLITHIC

FIGURE 6. Spatial distribution of the individuals with ante-mortem skull injuries (after Schmidt 1912: table XIV).

reconstruction of events in the case of the Ofnet skulls cannot be accomplished since the postcranial skeleton was not buried with the skull. It is, however, conceivable that there were wounds from cutting weapons or from projectile points, as can be substantiated in other

The location within the skull clusters of the individuals with fatal slash wounds reveals that five such individuals were deposited on the left edge of the larger skull cluster (Fig. 6). From this positioning, it is probable that these heads were deposited together and that consequently a single depositional event becomes comprehensible.

examples (Vencl 1991; Wahl and König 1987). Ethnographic research shows intra- and interethnic conflict occurs more rarely amongst hunter-gatherer groups than in sedentary populations (Schmidt 1993). Conflict within the group seldom occurs among huntergatherers and has a fatal outcome in very few cases. It cannot be ruled out that the death of six individuals from Grosser Ofnet is the rare result of a conflict over resources between two groups.

The clearly discernible contours that the weapons have left behind on the crania are consistent only with the use of a chopper-like instrument as the death-dealing weapon. The stratigraphically certain existence of axes in the Early Mesolithic of southwest Germany is already known from two finds of flat axes similar in form to antler axes from the open site of Rottenburg Siebenlinden (Hahn and Kind 1991: 21-8, fig. 5; Hahn et al. 1993: 43-5, figs 1112). The new find of a flake from an amphibolite stone axe from the same findspot confirms the existence of polished stone axes at this time (Kind 2001). For the Late Mesolithic, examples of polished stone axes come from the Falkenstein cave (Peters 1935: pl. III.7) and the as yet unpublished excavation by W. Taute in the Jägerhaus cave (layer 7, in the form of two fragments of stone axes; Taute 1971: 97, pls 15, 19-20; Oeschger and Taute 1980). Furthermore, the richly equipped Late Mesolithic grave of a woman with an infant from Bad Dürrenberg yielded evidence for the existence of polished flat axes for the early Late Mesolithic of middle Germany (Geupel 1977; Hedges et al. 1992: 345-6; Orschiedt 1999: 126-30).

Deposition of heads as a burial rite in the Late Mesolithic The dating of the find from Grosser Ofnet to the Late Mesolithic through C14 analysis allows us to identify a specific, tangible grave site involving head-burial in southern Germany (Orschiedt 1998; 1999: 151). What needs to be highlighted here are the painstaking deposition of heads, the evidence of anatomically informed decapitation, the use of red ochre and the provision of grave goods. Although an anomaly remains with regard to the occurrence of ante-mortem cranial injuries in the skull deposits from Ofnet and on the three contemporaneous skulls from Hohlenstein-Stadel (ibid. 1998; 1999: 131-51), it is not possible to interpret the find from the Ofnet cave as evidence of a Mesolithic

The motivations for warlike behaviour in the Mesolithic are not easy to determine. Moreover a complete 71

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Hahn, J., Kind, C.-J. and Steppan, K. 1993. Mesolithische Rentierjäger in Südwestdeutschland? Der mittelsteinzeitliche Freilandfundplatz Rottenberg – Siebenlinden III (Vorbericht). Fundberichte in BadenWürttemberg 18: 29-52. Haidle, M.N. and Orscheidt, J. 1995. Die Verwendung von Repliken bei der rasterelektronenmikroskopischen Untersuchung von osteologischem Material. Archäologische Korrespondenzblätter 25: 265-73. Hedges, R.E.M., Housely, R.A., Law, J.A. and Bronk Ramsey, C. 1989. Radiocarbon dates from the Oxford AMS system: Archaeometry datelist 9. Archaeometry 31: 207-34. Hedges, R.E.M., Housely, R.A., Bronk Ramsey, C. and van Klinken, G.J. 1992. Radiocarbon dates from the Oxford AMS system: Archaeometry datelist 15. Archaeometry 34: 345-6. Helbling, J. 1987. Theorie der Wildbeutergesellschaft: eine ethnosoziologische Studie. Frankfurt & New York: Campus Forschung 521. Kaulich, B. 1983. Das Paläolithikum des Kaufertsberges bei Lierheim, Gem. Appetshofen, Ldkr. Donau-Ries. Quartär 33/4: 29-97. Kind, C.-J. 2001. The Mesolithic open air sites of Siebelinden, Rottenburg, Kr. Tübingen, BadenWürttemberg, Germany. http://www.landesdenkmalamt-bw.de/english/ archaeol/siebenlinden/ Maples, W.R. 1986. Trauma analysis by the forensic anthropologist. In K.D. Reichs (ed.) Forensic Osteology: advances in the identification of human remains. Springfield IL: Charles C. Thomas. 218-28. Merbs, C.F. 1989. Trauma. In M.Y. Iscan and K.A.R. Kennedy (eds) Reconstruction of Life from the Skeleton. New York: Wiley-Liss. 161-89. Mollison, T. 1936. Zeichen gewaltsamer Verletzungen an den Ofnet Schädeln. Anthropologischer Anzeiger 13: 79-88. Müller-Beck, H. 1983. Die späte Mittelsteinzeit. In H. Müller-Beck (ed.) Urgeschichte in BadenWürttemberg. Stuttgart. 393-404. Naber, F.B. 1974. Das Ende des Ofnet Problems. Quartär (1975) 25: 73-84. Oeschger, H. and Taute, W. 1980. RadiokarbonAltersbestimmungen zum süddeutschen Mesolithikum und deren Vergleich mit der vegetationsgeschichtlichen Datierung. In W. Taute (ed.) Das Mesolithikum in Süddeutschland. Teil 2: Naturwissenschaftliche Untersuchungen. Tübingen: Tübinger Monographien zur Urgeschichte (1978) 5/2: 15-19. Orschiedt, J. 1998. Ergebnisse einer neuen Untersuchung der spätmesolithischen Kopfbestattungen aus Süddeutschland. In N.J. Conard and C.-J. Kind (eds) Aktuelle Forschungen zum Mesolithikum. Tübingen: Urgeschichtliche Materialhefte 11: 147-60. Orschiedt, J. 1999. Manipulationen an menschlichen Skelettresten: taphonomischce Prozesse, Sekundär-

massacre (Frayer 1997: 212) given the absence of traces of injury on the remaining skulls. As another example of the custom of head-burial, we can turn to the find of a young adult male’s skull, dating to the same period but not yet dated by C14, from the Hexenküche (‘Witches’ Kitchen’) at Kaufertsberg near Lierheim (Kr. Nördlingen). This was discovered in 1913 under similar circumstances to the head-burials at Grosser Ofnet and Hohlenstein-Stadel (Birkner 1915:125). This skull was likewise found with the mandible and the first two cervical vertebrae still articulated. Although the stratigraphic position of this find has frequently been discussed (Naber 1974: 79; Kaulich 1983: 29-97; Schröter 1983: 99-101), a Mesolithic classification of the find is likely. On neither the find from Kaufertsberg nor the find from Mannlefelsen at Oberlarg in Alsace are there any discernable ante-mortem skull injuries or cutmarks on the cervical vertebrae. An intentional deposit of the find from Oberlarg is likewise safely assigned to the Mesolithic on the grounds of the find’s circumstances (Thévenin 1980: 12, fig. 8). With this result, the proven practice of head-burial in southern Germany during the Late Mesolithic can also be understood to have occurred in Alsace. From the find of the head-burials from Grosser Ofnet and other similar results, we can prove in consequence not only the existence of a specific burial rite in southern Germany during the Late Mesolithic but also evidence for warlike activity during this period. Bibliography Bégouen, C. 1912. Quelques observations sur la décapitation aux temps préhistoriques. Bulletin de la Société Préhistorique Française 9: 336-40. Birkner, F. 1915. Der Eiszeitmensch in Bayern. Beiträge für Anthropologie und Urgeschichte Bayern 19. Brückner, H. and Hinze, M. 1991. Frakturen, Luxationen, Begleitverletzungen. Berlin. Esenwein-Rothe, I. 1982. Einführung in die Demographie: Bevölkerungsstruktur und Bevölkerungsprozess aus der Sicht der Statistik. Wiesbaden. Fraas, O. 1876. Die Ofnet bei Utzmemmingem im Ries. Korrespondenzblätter der Deutschen Gesellschaft für Anthropologie 7. Frayer, D.W. 1997. Ofnet: evidence of a Mesolithic massacre. In D.L. Martin and D.W. Frayer (eds) Troubled Times. New York: Gordon and Breach. 181216. Geupel, V. 1977. Das Rötelgrab von Dürrenberg, Kr. Merseburg. In J. Herrmann (ed.) Archäologie als Geschichtswissenschaft. Studien und Untersuchungen 1. Öhringen. 101-10. Hahn, J. and Kind, C.-J. 1991. Neue mesolithische Fundstellen in Rottenburg a. N., Kreis Tübingen. Archäologische Ausgrabungen in Baden-Württemberg 1990: 26-9. 72

JÖRG ORSCHIEDT: THE HEAD BURIALS FROM OFNET CAVE: AN EXAMPLE OF WARLIKE CONFLICT IN THE MESOLITHIC Sellier, K. 1971. Das Schädel-Hirn-Trauma: neuere Erkenntnisse und Zusammenstellung von Toleranzwerten von knöchernem Schädel un Gehirn bei mechanischer Gewalteinwirkung. Zeitschrift für Rechtsmedizin 68: 239-52. Taute, W. 1971. Untersuchungen zum Mesolithikum und zum Spätpälaolithikum im südlichen Mitteleuropa. Unpublished inaugural dissertation, University of Tubingen. Thévenin, A. 1980. Paleoenvironment et peuplement de l’Alsace de 1.000.000 d’années à 800 ans avant J.-C. Cahiers Alsaciens d’Archéologie 23: 5-25. Ubelaker, D.H. 1991. Perimortem and postmortem modifications of human bone: lessons from forensic anthropology. Anthropologie (Brno) 29: 171-4. Vencl, S. 1991. Interpretation des blessures causées par les armes au Mésolithique. L’Anthropologie 95: 21928. Wahl, J. and König, H.G. 1987. Anthropologischtraumatologische Untersuchung der menschlichen Skelettreste aus dem bandkeramischen Massengrab bei Talheim, Kr. Heilbronn. Fundberichte aus BadenWürttemberg 12: 65-186. Weissmüller, W. 1982. Postmesolithische Funde aus Höhlen und Abris am Beispiel des südlichen Riesrandgebietes. Unpublished PhD thesis, University of Erlangen.

bestattungen oder Kannibalismus? Tubingen: Urgeschichtliche Materialhefte 13. Peters, E. 1935. Die Falkensteinhöhle bei Tiergarten. Fundberichte aus Schwaben (1933-35) 8 (2): 2-12. Polson, C.J. 1965. The mechanism of head injuries. In C.J. Polson (ed.) The Essentials of Forensic Medicine. Oxford. 126-38. Schlitz, A. 1912. Die diluvialen Menschenreste Deutschlands. In R.R. Schmidt, Die diluviale Vorzeit Deutschlands. Stuttgart. 241-56. Schmidt, R.R. 1909. Die spätpälaolithischen Bestattungen der Ofnet. Mannus (1912) 1 (suppl. vol. 1): 56-62. Schmidt, R.R. 1912. Die diluviale Vorzeit Deutschlands. Stuttgart. Schmidt, R.R. 1913. Die altsteinzeitlichen Schädelgräber der Ofnet und der Bestattungsritus der Diluvialzeit. Stuttgart. Schmidt, S. 1993. Kriege bei rezenten Wildbeutern und Nicht-Wildbeutern: eine explorative interkulturelle Vergleichsstudie mit der Kölner Text-Code Datenbank. Archäologische Informationen 16: 18999. Schröter, P. 1983. Zum Schädel vom Kaufertsberg bei Lierheim, Gem. Appetshofen, Ldkr. Donau-Ries. Quartär 33/4: 99-109. Schulte im Walde, T., Freundlich, J.C., Schwabedissen, H. and Taute, W. 1986. Köln radiocarbon dates 3. Radiocarbon 28: 134-40.

73

Assessing rank and warfare-strategy in prehistoric hunter-gatherer society: a study of representational warrior figures in rock-art from the Spanish Levant, southeastern Spain George Nash Centre for the Historic Environment, University of Bristol

The permanence of rock-art can be considered as reflecting economic, political, social and symbolic stability. Rock-art can also be seen as a dynamic component that manipulates and enhances social cohesion between individuals and neighbouring groups. This manipulation is probably a result of the ongoing development of social identity, be it passive or aggressive.

In contrast, it could be suggested that these scenes represent simulated fighting dances whereby disputes are fought out symbolically (Beltrán 1982: 48). However, the presence of dead and wounded warriors on a limited number of panels suggests this is not the case (Table 1). Site

The art of warfare has been largely ignored, especially within prehistoric hunter-gatherer research. This is mainly due to the limited number of panels worldwide with scenes of violence. However, a social and economic framework has been determined by lithic studies which is usually supported by environmental considerations (e.g. Mithen 1991; Price and Brown 1985; Smith 1992; Zvelebil 1986). Whilst these empiricist studies are essential for understanding frameworks within prehistoric hunter-gatherer societies, the evidence for social and civil unrest is near impossible to quantify. Within the hunter-gatherer rock-art assemblage of Levantine Spain, however, there is a group of representational figures that portrays a society, the social and political framework of which rests, in part, upon violence revealed in scenes of execution, skirmishing and warfare (Beltrán 1968; 1982; Bosch Gimpera 1964; Cabré Aguiló 1915; Dams 1984; Mateu 2002; Nash 2000; Pericot Garcia 1950).1 These painted panels, once witnessed by an audience, can be questioned in several ways. Firstly, are scenes where violence is displayed reflecting a reality within hunter-gatherer Levantine society? In other words, does warfare form part of a way of life? Secondly, do these scenes, in particular the executions, reveal a society in social and political turmoil, and do they counter the more traditional portrayal of hunter-gatherers as societies in harmony with nature and eachother? Furthermore, could a regime advocating capital punishment in this way be considered unstable and insular? The complex images of the panels display a violence that, within our own society, would be immediately repugnant.

Type of scene

No. of participants

Cingle de la Mola Remigia, Castellón IV

wounded warrior

1

Cingle de la Mola Remigia, Castellón VII

dead warrior

1+1

Cingle de la Mola Remigia, Castellón VII

execution scene

1

Cueva Remigia, Castellón I

wounded warrior

1

Cueva Remigia, Castellón III execution scene

1 + 17

Cueva Remigia, Castellón V

execution scene

1+5

Cueva Remigia, Castellón V

execution scene

1 + 10

Cueva Remigia, Castellón V

execution scene

1 + 14

Cuevas de la Araña, Bicorp, Valencia I

wounded warrior

1

Los Dogues, Castellón

wounded warrior

1

Los Trepadores, Teruel

execution scene

1+7

Minateda, Albacete

wounded warriors

2

Polvorin, Castellón

wounded warrior

1

Saltadora, Castellón, XII

wounded warrior

1

Total

69 Table 1. List of execution and gladiatorial scenes

In this paper, I discuss figures from a limited number of Spanish Levantine panels, in particular those from El Cingle de la Mola Remegia (Figs 1a-1c, 1e-1g), Cueva Remigia (Fig. 2), Les Dogues (Fig. 3), Cuevas del El Civil (Fig. 4) and Minateda (Fig. 5). I am particularly interested in the spatial arrangement of certain warrior types in order to assess rank and status. On each of the

1 As part of ongoing research, 39 Levantine panels (2,026 figures) have been included within a structural analysis (Nash 1997 and forthcoming). The Levantine panels have been regionally divided into eight groups. Within Group III (Teruel) and VI (Castellón), warring figures occur at seven sites. Of the 2,026 figures covered within this analysis, 413 are archers. Archers are divided into two groups: hunting archers and warring archers. Both groups are present on 24 panels.

75

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY

FIGURE 1. a. Execution scene from the El Cingle de la Mola Remegia, Gasulla, Castellón (after Beltrán 1982). b. A file of ceremonial archers from El Cingle de la Mola Remegia, Gasulla, Castellón (after Beltrán 1982). c. Execution scene from El Cingle de la Mola Remegia, Gasulla, Castellón (after Beltrán 1982). d. Gladiatorial combat with wounded warrior from Cuvea Remegia (caveat IV), Gasulla, Castellón (after Mateu 2002). e. Execution victim from El Cingle de la Mola Remegia, Gasulla, Castellón (after Beltrán 1982). f. Injured warrior with two arrows through the legs from El Cingle de la Mola Remegia, Gasulla, Castellón (after Ripoll Perello 1963). g. Archer and dead or dying comrade from the El Cingle de la Mola Remegia, Gasulla, Castellón (after Ripoll Perello 1963).

76

GEORGE NASH: ASSESSING RANK AND WARFARE-STRATEGY IN PREHISTORIC HUNTER-GATHERER SOCIETY

FIGURE 2. Cuvea Remegia, Gasulla, Castellón, panel IX (after Ripoll Perello 1963).

FIGURE 3. Les Dogues, Castellón (after Ripoll Perello 1963).

77

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY

FIGURE 4. Cuevas del El Civil, Castellón, panel III (A) (after Mateu 2002).

FIGURE 5. Minateda, Albacete (after Mateu 2002).

78

GEORGE NASH: ASSESSING RANK AND WARFARE-STRATEGY IN PREHISTORIC HUNTER-GATHERER SOCIETY

MAP 1. Distribution of sites showing scenes of violence (after Nash 2000).

the panel (Beltrán 1982; Dams 1984; Mateu 2002; Porcar 1947; 1953). These warrior representations, according to Dams (1984: 303), appear to date to around 6,500 BC (categorized as Style III). Recent reassessment of these figures has placed them within the Neolithic (ibid.: 304). Irrespective of period though, these figures probably represent a hunter-gatherer economy. I base this on other non-violent figures present either on the same panel or on neighbouring panels, including hunted red deer, chamois/ibex and bulls (e.g. Cingle de la Mola Remigia, caveats IV and VII). Although warfare scenes are found elsewhere in Levantine Spain, the greatest concentration is found within the gorges of Gasulla and Valltorta in Castellón (Table 2).2

panels displaying warfare or skirmishing, figures appear to be deliberately placed, suggesting strategic battlefield planning and formation. Some figures possess highly elaborate head-dresses which may represent superior rank, whist some are merely stick figures and represent infantry soldiers. Within the same assemblage, there are a small number of execution and gladiatorial scenes, the victims of which also possess head-dresses (Figs 1f-1g). Could it be that these figures represent warriors from neighbouring territorial groups, suggesting social and political instability? Four sites displaying violence lie within the upland gorges of Gasulla and Valltorta, in the province of Castellón (Map 1). This blatant portrayal of violence is in contrast to the herding and hunting scenes that otherwise dominate this assemblage (Nash 1997). The figures, all representational, are arranged in a deliberate and systematic way; usually in the form of two opposing sets of warriors which have been strategically placed on

2 Other panels within the Levantine region include Abrigo del Molino de las Fuentes, Nerpio (Albacete); Cueva de la Vieja, Alpera (Albacete), Abrigo del Voro, Quesa (Valencia) and Abrigo Grande de Minateda (Albacete) (see Dams 1984; Mateu 2002).

79

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Site

No. of figures*

It is the rock paintings around the Gasulla and Valltota gorges, however, that contain graphic evidence of organized violence, including execution by a squad of archers (Fig. 1c), gladiatorial combat (Fig. 1d) and hanging (Ortego 1948: 3-37; Porcar 1945: 145-52). Here, the artist appears to show complete reverence towards his/her subject.4 These scenes, along with dancing, foodgathering, herding and hunting, suggest the artist is attempting to paint a rational reflection of huntergatherer/warrior life. If so, are we witnessing in the warfare scenes (organized) state/tribal violence, in which the rules of conflict, trial and punishment are considered the norm?

No. of warriors

Cingle de la Mola Remigia, Gasulla, Castellón, IV

50

11

Cingle de la Mola Remigia, Gasulla, Castellón, IX

49

44

Cueva de la Vieja, Castellón

156

22?

El Civil, Castellón

93

44

Galeria del Roure, Teruel

35

12

Los Dogues, Castellón

41

27

Minateda, Albacete

317

11

Molino de las Fuentes I ou Sautuola, Albacete

46

35

Anthropology of social relations In anthropological terms, warfare can be considered a form of armed conflict, usually between neighbours or what may be termed territorial groups. Armed conflict, also referred to as aggression, can range from raiding, skirmishing and feuding to full-scale battle, the latter being particularly evident within the recent past. It has been suggested that warfare is an important mechanism for the growth of centralized political systems (Sahlins 1963; 1972). Above all, within a modern context, warfare creates (ethnic) unity that is driven by anger against a common enemy, usually resulting in military mobilization. The causes of warfare can be characterized by two schools of thought, one of which stresses underlying psychological tension - something that appears to be inherent within the human psyche, arguably as a result of biological evolution – while the other emphasizes ecological rationality (Kock 1974: 52-4).

Total 787 206 * figures are based on the author’s analysis of each panel Table 2. List of battle scenes

European distribution The distribution of hunter-gatherer rock-art, both painted and carved, within Europe is concentrated in three main areas: northern Scandinavia (including Finland and western Russia), the Val Camonica (northern Alpine Italy) and the Spanish Levant (southeastern Spain).3 All these areas are categorized topographically as coastal or upland environments. The siting of rock-art within these areas may be considered part of the sociosymbolic/political process of a rock-art performance.

For the purposes of this paper, and in view of the implications for understanding art, I shall tend to take the view that warfare is basically an expression of human aggression and territoriality. Within the anthropological record there are, however, many types of cultural and ritual forms of conflict within human society (Ingold 1998). Warfare offers just one set of possibilities for resolving stress, overpopulation and (political) territorial disputes (Rubinstein 1998: 983). In many cases, neither warfare nor any other form of aggression is present. Where physical aggression is present, either between individuals or groups, warfare is a tradition and forms the basis for future aggression (Chagnon 1967; 1983). This ecological approach to problem solving revolves around population dynamics, demography, available resources and territoriality. Physical aggression involving these requisites establishes an equilibrium between society, societies and nature (Nash 2000).

The most poignant statement of orchestrated violence involving human versus human is witnessed within the Spanish Levantine assemblage. Arguably, much later, during the Iron Age, there is evidence of warfare and possible execution from the Bohuslän carvings of southwestern Sweden (Coles 1990: 56-8). Nordbladh has linked Iron Age and later hoards and carved armoured figures to a warrior/prestige society, referring to them as ‘self-promoting warriors’ (1989: 323-30). In the same paper, Nordbladh analyzed this assemblage further and noted that similar warrior regalia has been found in a large number of graves that date from roughly the same period. The art of this region depicts warriors engaged in battle, but there appears to be no portrayal of injury or death. It could be argued that these scenes depict a type of ritualized warfare (or war games) rather than conflict in the true sense of the word.

There are within many contemporary non-western huntergatherer and farming societies substitutes for violence,

3

There are several other rock-art areas in Europe, such as southern Italy, Sicily and southern Greece. These areas, however, lack spatial and stylistic continuity with other areas of Europe. Furthermore, the dating of many panels through style, form and cultural content is, as yet, difficult to quantify.

4 It should be noted that the art involves male subjects participating in either violence or simulated violence. One can, therefore, assume the artist might have been male and what is being portrayed is maleness (see Hodder 1990; Nash 1998).

80

GEORGE NASH: ASSESSING RANK AND WARFARE-STRATEGY IN PREHISTORIC HUNTER-GATHERER SOCIETY However, the ecological approach excludes analysis of the symbolic, political and ideological components of warfare. The concept that warfare creates social equilibrium may be misconceived, in that warfare usually flares up as a result of imbalance, a contradiction in terms of internal and external constraints - be it socially, politically or economically induced (or elements of all three). Nevertheless, warfare does reaffirm group identity and, in most cases, strengthens group cohesion, as well as establishing internal class structure. Similarly, according to Marxist principles, warfare is a product of class antagonism. Assuming that a class structure is present (and it usually is) then conflict is an inevitable product. This is certainly evident within present-day Western (and Westernized) society.

but perhaps not for aggression (Leach and Leach 1983; Weiner 1983). Simulated battles between neighbouring groups are known from the central highlands of Papua New Guinea and the Trobriand islands (Brown 1978: 39; Malinowski 1922). The Trobriand example substitutes games such as ‘Trobriand cricket’ for war. Introduced by European missionaries during the late nineteenth century as a substitute for actual conflict, the rules of the game are made so as to create a passive winner; the object being not to physically injure (Rappaport 1999). Within the New Guinea highlands region, the main objective of the warring parties is to injure and maim, but not to kill. Group survival is secured by the consolidation and control of ritual resources, including land tenure and commodities. However, owing to the ebb and flow of village populations, victorious neighbours rarely take land (ibid.: 119).

When discussing the concept of neighbours (both good and bad), one imagines a relationship between people that draws on communal group identity, an identity that relies on social and political contact: contact in the form of exchange, obligation and common ideology, as well as (in most cases) a mutual understanding of territorial space. Usually, these components ensure peaceful coexistence or, at least, tolerance between neighbours. One extreme case where inter-communal tolerance is lacking is that of the Yanomamö of Brazil, who apply aggression at different levels. Here, aggression is both internally and externally driven (see Layton, this volume). According to Chagnon, inter-relational violence appears to be the norm. An added element of this aggression is the need to secure group procreation (Chagnon 1983). This is resolved by the abduction of females from neighbouring clan groups. The act of abduction is, in itself, a provocation that can set off a tit-for-tat process of retaliation, as well as open warfare.

Within the same highland area, the Dani look on warfare as an essential mechanism that is demanded by the ancestors. Skirmishes usually occur in the form of ambushes and, according to Brown (1978: 207), these acts appear to be a part of everyday life. In order to secure continuous conflict, no treaties are accepted. As a result, the ancestral spirits are constantly demanding revenge. Inter-tribal conflict relies entirely on men who are conscripted from a number of villages (and not from a single village). Usually, death occurs as a result of injury sustained through intentional wounding (ibid.: 208). Once the skirmish is over, retaliation from the defeated tribal group is inevitable, thus continuing the cycle of violence. Similarly, the Masai people of central Kenya regard cattle raids and counter-raids against the neighbouring Okiek as ongoing ‘natural sport’ (Blackburn 1982: 295). In this instance, skirmishing reaffirms group identity, group solidarity, power relations, established boundaries and territorial rights.

The need to secure social and political alliances with neighbouring communities is of paramount importance to the formation of successful territories. It is these constraints that may force members of over-populated groups to break away and migrate from established (ancestral) homelands, thereby initially giving rise to the concept of neighbours. Recently, Taçon and Chippindale (1994) suggested that dense human populations can, in a small area, exceed the minimum size necessary for genetic survival. This ‘genetic independence can lead to xenophobia’. A similar view is taken by Tainter with reference to population pressure in pre-Classic Mayan society, which could be solved either by agricultural intensification or conflict (1988). However, inter-group politics can also result in territorial neighbour formation. It may be that population dynamics during the Mesolithic was controlled internally by mechanisms such as infanticide, ritual human sacrifice and execution. Bourdieu (1977: 183-93) suggests that social practices and I would include the above mechanisms - rely on social and political group relations based upon domination and repression, which are the result of repeated and modified social practices. I would add that

Rappaport (1967; 1999), who adopted a systems approach to conflict, argues that warfare among the Tsembaga Maring of New Guinea is one component of an already-established complex society. Here, warfare forms part of a cybernetic inter-relationship between culture and ecosystem. Warfare, along with the raising and ritual sacrifice of pigs, feasting, gift exchange and marriage, forms part of a ritual cycle and preserves equilibrium within society. This cycle of ritual behaviour provides stability to the social order, inter-group relations and redistribution of natural resources and population dynamics.5

5 It is suggested by Rappaport that the ritual cycle of the Tsembaga Maring, starts with warfare, and always between ‘engaging adjacent groups’ (1999: 75). Within houses of the antagonists, ritual fighting stones - mbamp ku - are placed around the central pole. The stones represent the (red) spirits (rawa mugi) of killed men who have fallen in battle. A number of rituals including the preparation of special charcoal and the sacrificing of pigs are undertaken when the fighting stones are hung within the house. The stones, according to Rappaport, change opponents into formal enemies, cenang yu (ibid.: 75).

81

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY heads, displaying almost triumphant jubilation at the execution of a victim within the foreground of the panel. Up to six arrows are embedded in the victim, who lies slumped lifeless on the ground. Arguably, executions could represent internal strife or enforcement of social law and order.

the severity of such mechanisms requires complex tribal organization. Prehistoric images of war Art, although static, draws upon many of the codes of display. The prime reason for display is harmonization: the need to belong and to be identified. Rock-art creates an identity for the artist, the user and the group as a whole. Art also distinguishes the group from other groups and establishes regionality. Cultural regionality is evident in the Spanish Levant, with variations to standardized designs present throughout the area.

Within the Levantine material, warring archers appear to be ranked and are usually recognized by the varying complexity of the head-dresses. The silhouette headdresses appear very similar to those worn by North American Plains Indians. Head-dresses are graded according to size, which may relate to rank. Many also appear to flank archers without head-dresses - a strategy still used in modern warfare. Nearly all archers are also depicted according to a distinct gender coding: many appear to have a phallus, suggesting that the artist is attempting to emphasize maleness, an important component of the Spanish assemblage.

The high concentration of rock-painting panels from both Gasulla and Valltorta gorges may display clan/group identity. Similar images - particularly identical designs of male archers - are replicated throughout. Beltrán lists nine and eight rock painting sites from each gorge respectively. All appear to share a stylistic and sequential structural affinity with neighbouring panels. Four of these panels - El Cingle de la Mola Remigia, Mas Blanc, Les Dogues and Cuevas del El Civil - all show identical fighting warriors. The figures, displaying a standardized design of elongated thin torsos with exaggerated calf muscles, are usually depicted running and shooting arrows. Some have elaborate head-dresses indicating possible rank. The artists’ brush strokes, both simple and well-orchestrated, suggest an impressionistic approach to each panel sequence. But why replication? It could be suggested that panels were commissioned from the same artist. Alternatively, the designs may represent clan identity and their production was shared by a number of artists within the clan/tribe.

On at least three of the panels - Cingle de la Mola Remigia, Les Dogues and Cuevas del El Civil - there is evidence of warriors injured in combat (Figs 1a, 1d-1e). At this point, I should stress that these panels do not depict warfare scenes but are merely a palimpsest; portraying a series of different but possibly interwoven narratives. The warrior figures have distorted bodies suggesting the twisting and turning of the torso soon after arrow impact. In one such case, Cingle de la Mola Remigia, a warrior carries his dead comrade in his arms (Fig. 1g). The body lies slumped over the arms of the grieving warrior. The motif of war, a single longbow, is balanced over the left side of the warrior’s head. Above these two figures, a further warrior lies outstretched with at least three arrows embedded in the torso. These two sets of warriors appear to be the only figures engaged in combat. Either side are scenes of hunting bulls and chamois. It could be that both sets of warriors were engaged in a duel.

Archers are portrayed not only in warring/combat scenes but in hunting and gathering scenes as well. Levantine art shows the most effective weapon for hunting (and warfare) was the longbow. This weapon is shown in the hunting of chamois/ibex/wild goat, red deer, wild cattle (in particular, bulls), as well as in warfare (Fig. 3). The depictions of hunters and archers on the panels at Riparo di Boro, Quesa (Valencia), Cueva Saltadora, Valltorta gorge (Castellón) and Cueva de la Vieja, Alpera (Albacete) all clearly show the use of such weapons. Herbert Kühn (1952) recognizes two types of longbow, which may represent different manufacturing methods over time or regional manufacture.

The extent of war during this period remains to be considered. Was war waged exclusively between neighbouring groups, or are there wider implications? It has been hinted that warring groups depicted at the rock shelters of El Molino de las Fuentes, Nerpio and El Polvorin at La Cenia are racially opposed (Beltrán 1982). A further suggestion is that one of the groups at Cinto de las Letras are Negroid in form, while at Minateda warring figures with distinct triple-curved longbows are considered to be of Asiatic origin (Fig. 5). If these figures do depict Negroid and Asiatic forms, warring during the Mesolithic would have been a world event!

Warriors are identical and structured in a standardized display. Warring scenes usually involve two sets of opposed (energetic) warriors, with bows drawn. Sets of warriors are running towards each other. Interestingly, none of the combat panels portrays dead or injured warriors. Given the absence of injured and killed warriors on painted panels, these scenes may represent simulated war-dancing between rival groups. However, executions scenes do occur elsewhere. In one scene from the Cingle de la Mola Remigia panel, a row of archers (archer squad) is walking away with bows above their

The emergence of state complexity and warfare Warfare, or at least aggressive conflict, forms only part of a complex cybernetic structure within society. Yesner (1980: 727-50) has identified a series of points that characterize the emergence of social complexity, coupled with a shift towards marine/coastal adaptation, especially 82

GEORGE NASH: ASSESSING RANK AND WARFARE-STRATEGY IN PREHISTORIC HUNTER-GATHERER SOCIETY water and hunting rights. Taçon and Chippindale, therefore, suggest that, during times of extreme ecological stress, conflict would have increased (ibid.: 225). This hypothesis is further reinforced by the chronological change from small-scale skirmishing to large-scale battle scenes on a number of rock-painting panels. However, the style and form of both sets of figures appear to remain unchanged.

where there is a need to support large populations. It has been estimated that hunter-gatherer groups number between 45 and 240 people per group (Rowley-Conwy 1981: 55). The basis of Yesner’s arguments for the emergence of social complexity is rooted in the semisedentary populations of the Northwest Coast American Indians. The environment and, arguably, the social organization and structure may be similar to the climax/hiatus palaeoenvironment of Levantine huntergatherers.

Unfortunately, the figures from the Gasulla and Valltorta Gorges do not show any noticeable chronological variation. It would appear that, at one particular time, the expressive nature of the artist towards aggression and conflict was considered important. Whether or not state or inter-tribal violence was an important component of Levantine life prior to and after warfare paintings were executed remains a question of debate. However, Taçon and Chippindale’s idea that conflict is linked to the varying states of social and economic prosperity gives rise to an interesting (Darwinian) paradigm, in that social (or unsociable) dynamics are, in part, controlled by the management of resources. It could be the case, though, that artists from the Gasulla and Valltorta gorges were fixated on an ideal state society with controlled boundaries that were both gender-encoded and politically rigid. A similar state may well have been present around the Tassili n’Ajjer region of north Africa.6

It is apparent that large populations require a greater level of social organization. Clive Gamble (1986: 41) suggests that hunter-gatherers adopt a ‘communal interplay’ between ‘organization and scheduling of work parties to exploit the scattered resources of the environment’. Further, he links organization with the wider implications of alliance through marriage and goods exchange. The basic assumption that hunters and gatherers only hunted and gathered is further denounced by Yesner. The 10 points that he outlines concerning the development of social complexity are particularly dependent upon environmental considerations. They are i. ii. iii. iv. v. vi. vii. viii. ix. x.

higher resource biomass; high resource diversity; lower resource seasonality; unearned (migratory) resources; linear settlement patterns; sedentism; complexity and co-operative socio-economic factors and resource exploitation; high per capita productivity; high population density; territoriality, resource competition and warfare.

Warrior display and battlefield strategies Research undertaken by the author in 1997 (and forthcoming), involving four sites in Castellón, appears to show a number of correlations within the strategic placing of certain warrior figures onto rock-shelter walls. Warrior figure depictions extend to sites outside the gorge area, including Abrigo del Voro (Quesa, Valencia) and Abrigo del Molino de las Fuentes (Nerpio), Cueva de la Vieja (Alpera) and Abrigo Grande de Minateda, all in Albacete (Breuil 1920; 1935; Breuil et al. 1912; Garcia Guina 1963; Mateu 2002; Porcar 1934; Ripoll Perello 1963).

On this latter point, I would suggest that warfare is a result of, and not a pre-requisite for, social development in later Mesolithic and Neolithic society. Warfare, therefore, probably forms part of the upward spiralling effect towards social complexity, where there is a need to consolidate prime hunting and fishing territories. Settlement, therefore, becomes not only a physical statement within the landscape but a (positive) state of mind. This model is certainly relevant when commissioning rock-art, itself another statement of permanency within the landscape.

The research, involving 117 warrior figures, was concerned with the detail and form of each figure (Nash 1997: 348). Figures from the four sites were divided into figures of stick form (Type I), figures with exaggerated calf and thigh muscles (Type II), figures with pantaloons (Type III), figures with phallus (Type IV) and figures possessing elaborate head-dresses (Type V). Some figures possessed more than one of these traits.

Taçon and Chippindale have recognized a significant change in rock-painting complexity concerning scenes of conflict (1994: 225). They argue that, between 20,000 and 10,000 years ago, in Arnhem Land (Australia) hunter-gatherer-foragers would have been highly mobile, working within a collective hunting regime. This strategy was determined by harsh environmental conditions. The art of this era reflects a number of socio-economic and political activities, including small-scale skirmishing. However, in times of stress, the competition for resources may have forced a greater need for installing well-defined territorial boundaries necessitating the legitimization of

The largest group represented at the four Castellón sites is that of the stick figures (Type I). These number 59 and are present on four of the Cingle de la Mola Remigia panels, as well as panels at Cueva Remigia, Les Dogues and Cuevas del El Civil. Of these, four figures possess a phallus. These figures, usually constructed from a series 6 See discussions on warring scenes by Lhote 1962; Beltrán 1978; Muzzolini 1995.

83

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY warriors in their attacks on Roman legions in the midfirst century BC.7

of simple brush stokes, can be regarded as the least complex of all the warrior types and are located within the central section of the panel narrative. They are usually painted in a running stance (either full stride or part stride), holding a longbow with arrows drawn.

The final group of warriors, the most elaborate of all figures, possesses head-dresses (Type V). These figures number eight and are found on three panels: Cingle de la Mola Remigia (panels IV and IX) and Les Dougues. One figure from panel IX of the Cingle de la Mola Remigia also possesses a phallus. From each of the panels displaying conflict, the head-dress designs range from a simple painted blob above the head of the warrior to extremely complex head-dresses probably made of feathers, as seen in the Les Dougues panel. These figures are usually strategically located to the rear of or flanking the battle groups which include Types I to IV and probably represent battle commanders who controlled the various battlefield manoeuvres.

In assessing rank, it is probable that these figures represent infantry as they are usually positioned in the thick of battle. There is little or no difference between stick figures on opposing sides. The second largest group of warriors numbers 43 (Type II). These warriors possess exaggerated thigh and calf muscles, designs that are also used in the construction of hunter- gatherer figures on other panels within the Levant. The greatest number appear on the Cuevas del El Civil panel, especially on caveat A (Fig. 4). The artist would have applied more care in the construction of these figures, using several more brush strokes to construct the body sections of each of the figures (i.e. to include the lower torso and the calf muscles). In nearly all examples from each of the four panels, the central area of the torso is narrow, thus exaggerating the upper thigh and shoulder areas of the body, thereby portraying healthy, virile male warriors. Within several battle scenes, in particular those present on the caveat B panel section at Cuevas del El Civil, Type II figures are integrated with Type I figures, suggesting that they may possess rank and status superior to that of the stick figures, albeit limited (based on form and number). The exaggerated calf muscles and elongated thighs represent strong healthy legs, probably the result of running. The artist appears to be concerned in portraying warriors in this way.

Warfare within an ideal state of mind Throughout the historical periods there has been a need for conflict. These acts of aggression are well documented and their causes are usually political and economic greed, especially for territory. This is certainly true of conflicts during the European colonial period of the eighteenth and nineteenth centuries, as well as during state formation and the rise of extremist political ideologies in the twentieth century. For prehistory, the evidence is less forthcoming. As suggested earlier, there are written texts that portray a fragmentary image of warfare and, as concerns this paper, there is the evidence of rock-art, especially within later prehistory. These images, in particular those from the Spanish Levant and the Bronze Age and Iron Age images from the Val Camonica (Italy) and Bohuslän (southwestern Sweden), suggest a prehistory of conflict. These figures, however, do not portray any strategy pertaining to battlefield formation. However, Levantine images that date to the southern European Neolithic or Mesolithic do show strategic battle organization. Here, scenes depict warriors who are strategically placed.

There are four Type III figures, representing warriors wearing pantaloons, present on the Cingle de la Mola Remigia and Les Dougues panels. The ‘pantaloons’ are single blocks of paint along the length of each of the legs. The pantaloons appear to extend to the ankle area of each of the warriors where they become tufted. Why this style of trouser garment is used for battle is not known. However, it could be the case that the artist was concerned with separating stick and thigh-and-calf figures from pantaloon figures, thus portraying military rank.

This paper has suggested that within hunter-gatherer society conflict forms an integral part of inter-tribal organization, what one might term an ideal state of mind. By this I mean conflict forms part of a socio-political calendar, with aggression relieving tension between neighbouring groups. It is more than probable that conflict - albeit strategically timed - stems, in part, from the pressure on resources. Conflict would have resulted in territorial disputes, cattle raiding or simply a history of bad relations between neighbouring groups.

Type IV figures, those possessing a phallus, number four. These figures may have been present in greater numbers, especially those attached to Type I stick figures. However, paint deterioration over time, recent defacement and the size of each of the figures (sometimes up to 5cm in height) may have obliterated the extremely narrow brush stroke representing the phallus. As a result, phallus figures are only present on the Les Dougues and Cuevas del El Civil panels. There is a probability that these figures show warriors going naked into battle, a tactic used, according to Julius Caesar, by British tribal

7 Caesar describes them thus: ‘Of the islanders most do not sow corn, but live on milk and flesh and clothe themselves in skins. All the Britons, indeed, dye themselves with woad, which produces a blue colour, and makes their appearance in battle more terrible. They wear long hair, and shave every part of the body save the head and the upper lip.’ Gallic Wars V: 253 (trans. H.J. Edwards).

84

GEORGE NASH: ASSESSING RANK AND WARFARE-STRATEGY IN PREHISTORIC HUNTER-GATHERER SOCIETY Portrayed in each of these Levantine scenes, especially those from the Valltorta and Gasulla gorges, is conflict between two defined groups. The numbers in each conflict scene vary but five different types of warrior are present, each type defined according to design complexity (Types I-V). In each of the scenes, design complexity, along with strategic positioning of each of the groups, suggests rank and, more importantly, the concept of organized inter-group conflict.

Chagnon, N. 1983. Yanomamö: the fierce people. New York: Holt, Reinhart and Winston. Coles, J.M. 1990. Images of the Past: a guide to the rock carvings and other ancient monuments of northern Bohuslän. Uddevalla: Hällristningsmuseet, Vitlycke. Skrifter av Bohusläns Museum och Bohusläns Hembygdforbund 32. Dams, L. 1984. Les Peintures Rupestres du Levant Espagnol. Paris: Picard. Gamble, C. 1986. Hunter-gatherers and the origin of states. In J.A. Hall (ed.) States in History. London: Blackwell. 22-47. Garcia Guina, M.A. 1963. Le nouveau foyer de peintures levantines á Nerpio. Bulletin de la Société Préhistorique d’Ariége 18: 17-35. Hodder, I. 1990. The Domestication of Europe. London: Blackwell Press. Ingold, T. (ed.) 1998. Companion Encyclopaedia of Anthropology: humanity, culture and social life. London: Routledge. Kock, K-F. 1974. The Anthropology of Warfare. Reading MA: Addison-Wesley, Addison-Wesley Modules in Anthropology 52. Kühn, H. 1952. Die Felsbilder Europas. Stuttgart: Kohlhammer. Leach, E.R. and Leach, J.W. (eds) 1983. The Kula: new perspectives on Massim exchange. Cambridge: Cambridge University Press. Lhote, H. 1962. Le problème de la datation des peintures rupestres en Espagne et en Afrique. Jahrbuch fur Prähistorische und Ethnographische Kunst (1960-62) 20: 62-8. Malinowski, B. 1922. Argonauts of the Western Pacific. London: Routledge (reprinted edition). Escoriza Mateu, T. 2002. La Representación del Cuerpo Femenino: mujeres y arte rupestre levantino del arco mediterráneo de le península ibérica. Oxford: BAR Publishing, BAR International Series 1082. Mithen, S.J. 1991. A cybernetic wasteland?: rationality, emotion and Mesolithic foraging. Proceedings of the Prehistoric Society 57: 9-14. Muzzolini, A. 1995. Les Images Rupestres du Sahara. Collection: Préhistoire du Sahara 1. Nash, G.H. 1997. Northern European hunter/fisher/ gatherer and Spanish Levantine rock-art: A study in performance, cosmology and belief. Unpublished thesis, University of Wales, Lampeter. Nash, G.H. 1998. Exchange, Status and Mobility: Mesolithic portable art of southern Scandinavia. Oxford: BAR Publishing, BAR International Series 710. Nash, G.H. 2000. The power of the violence of the archers: all is not well down in Ambridge. (A study of representational warriors from the Mesolithic Spanish Levant). In special issue ‘New Approaches to the Palaeolithic and Mesolithic’ (ed. C. Conneller). Archaeological Review from Cambridge 17: 81-98. Nash, G.H. forthcoming. The Mesolithic Art of North Western Europe. London: Routledge. Nordbladh, J. 1989. Armour and fighting in the south Scandinavian Bronze Age, especially in the view of

Acknowledgements I wish to thank the following people for comments. First of all my good friend and colleague George Children. Thanks also to Maria Cruz-Barrocal (Madrid) who took valuable time to make comments. All mistakes are, of course, my own responsibility. Bibliography Beltrán, A. 1968. Arte Rupestre Levantino. Saragossa: Arqueología 4. Beltrán, A. 1978. Los problemas de la investigación de las pinturas y grabados prehistóricos al aire libre: referencía al conjunto del Tassili n'Ajjer y al del arte rupestre levantino. Caesaraugusta 85-86: 5. Saragossa. Beltrán, A. 1982. Rock-art of the Spanish Levant. Cambridge: Cambridge University Press. Blackburn, R.H. 1982. In the land of milk and honey: Okiek adaptions to their forests and neighbours. In E. Leacock and R. Lee (eds) Politics and History in Band Societies. Cambridge: Cambridge University Press. 283-306. Bosch Gimpera, P. 1964. The chronology of the rockpaintings of the Spanish Levant. In L. Pericot García and E. Ripoll Perello (eds) Prehistoric Art of the Western Mediterranean and the Sahara. Chicago: Aldine, Viking Fund Publications in Anthropology 39. 125-32. Bourdieu, P. 1977. Outline of a Theory of Practice. (trans. R. Nice). Cambridge: Cambridge University Press. Breuil, H. 1920. Les peintures rupestres d’Espagne X: les roches peintes de Minateda (Albacete). L’Anthropologie 30: 1-50. Breuil, H. 1935. Les Peintures Schématiques de la Péninsule Ibérique. Lagny. Breuil, H., Cabré, J. and Serrano, P. 1912. Les peintures rupestres d’Espagne V: les abris du Bosque á Alpera. L’Anthropologie 23: 529. Brown, P. 1978. Highland Peoples of New Guinea. Cambridge: Cambridge University Press. Cabré Aguiló, J. 1915. El Arte Rupestre en España. Madrid: Comisión Investigaciones Paleontología y Prehistoría, Mem. 1. Caesar, J. 1951. Gallic Wars (trans. H.J. Edwards). Harmondsworth: Penguin. Chagnon, N. 1967. The Fierce People. Bantam: New York. 85

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Rowley-Conwy, P. 1981. Mesolithic Danish bacon: permanent and temporary sites in the Danish Mesolithic. In A. Sheridan and G. Bailey (eds) Economic Archaeology: towards an integration of ecological and social approaches. Oxford: BAR International Series 96. 51-65. Rubinstein, R.A. 1998. Collective violence and common security. In T. Ingold (ed.) Companion Encyclopaedia of Anthropology: humanity, culture and social life. London: Routledge. 983-1009. Sahlins, M. 1963. Poor man, rich man, big man, chief: political types in Melanesia and Polynesia. Comparative Studies in Society and History 5: 285303. Sahlins, M. 1972. Stone Age Economics. London: Tavistock Press. Smith, C. 1992. Late Stone Age Hunters of the British Isles. London: Routledge. Taçon, P. and Chippindale, C. 1994. Australia’s ancient warriors: changing depictions of fighting in the rockart of Arnhem Land, N.T. Cambridge Archaeological Journal 4: 211-48. Tainter, J.A. 1988. The Collapse of Complex Societies. Cambridge: Cambridge University Press. Weiner, A.B. 1983. A world of made is not of born: doing kula in Kiriwina. In J.W. Leach and E.R. Leach (eds) The Kula: new perspectives on Massim exchange. Cambridge: Cambridge University Press. 147-70. Yesner, D.R. 1980. Marine hunter-gatherers: ecology and prehistory. Current Anthropology 21: 727-50. Zvelebil, M. 1986. Mesolithic societies and the transition to farming: problems of time, scale and organisation. In M. Zvelebil (ed.) Hunters in Transition. Cambridge: Cambridge University Press. 167-88.

rock-art representations. In T.B. Larsson and H. Lundmark (eds) Approaches to Swedish Prehistory: a spectrum of problems and perspectives in contemporary research. Oxford: BAR International Series 500. 323-33. Ortego, T. 1948. Nuevas estaciones de arte rupestre aragonés ‘El Mortero’ y ‘Cerro Felio’ en termino de Alacon (Teruel). Archivo Español de Arqueología 1: 3-37. Pericot Garcia, L. 1950. Arte Rupestre. Madrid. Porcar, J. 1934. Pinturas rupestres del barranc de la Gasulla. Boletin de la Sociedad Castellónense de Cultura, Castellón de la Plana 15: 343-7. Porcar, J. 1945. Iconografía rupestre de Gasulla y Valltorta: danza de arqueros ante figuras humanas sacificadas. Boletin de la Sociedad Castellónense de Cultura, Castellón de la Plana 21: 145-52. Porcar, J. 1947. Iconografía rupestre de Gasulla y Valltorta. (Representaciones pictográficas del toro). Boletin de la Sociedad Castellónense de Cultura, Castellón de la Plana 23: 315-29. Porcar, J. 1953. Las pintures rupestres del barranco de ‘Los Dogues’. Archivo de Prehistoría Levantina 4: 75-80. Price, T.D. and Brown, J.A. (eds) 1985. Prehistoric Hunters-Gatherers: the emergence of cultural complexity. New York: Academic Press. Rappaport, R. 1967. Pigs for the Ancestors. New Haven CT: Yale University Press. Rappaport, R. 1999. Ritual and Religion in the Making of Humanity. Cambridge: Cambridge University Press. Ripoll Perello, E. 1963. Pintures Rupestres de la Gasulla. Barcelona: Monografía de Arte Rupestre Levantino 2.

86

The emergence of warfare in the Early Bronze Age: the Nitra group in Slovakia and Moravia, 2200-1800 BC Andreas Hårde Department of Prehistoric Archaeology, University of Aarhus

Baldrick The thing is - the way I see it, these days there's a war on, right? And ages ago, there wasn't a war on, right? So there must have been a moment when there not being a war went away, right, and there being a war came along, right? So what I want to know is, how did we get from the one case of affairs to the other case of affairs, right? Blackadder You mean, how did the war start? - Goodbyeee. Blackadder Goes Forth.1 distinguished themselves by equipment that changed from the bow and arrow to knives and daggers. The most interesting change is the increase in the occurrence of skeletal trauma, warrior cenotaphs and grave plundering. Thus the expansion of territory and changes in social structure had a high price.

At Mýtna Nová Ves in southwestern Slovakia, five prominent men were buried in large graves with wooden chambers, each beneath his own mortuary house. Several fractures, on both the skull and the post-cranial skeleton, show that they died in combat. The skeletons of these men are not the only ones with signs of having been killed in battle. Skeletal trauma occurs on many burials within sites from the Nitra group and, at a quick glance, it would appear that warfare and conflict were rife during this period. Although warrior graves are present from the onset of the period, the emergence of war cannot, however, be observed until later in the history of this cultural group. In its earliest phase, the Nitra group occupied only a small geographic area along the Nitra river in southwestern Slovakia. Access to fertile soils and copper ore were probably the prime objective for this group. Apart from the occurrence of weapons in graves there are few indications that violent conflict might have been common.

These changes give rise to two burning questions: firstly, why did conflict emerge? Secondly, what effect did this have on society? Warfare in the Nitra group cannot be understood by studying hunter-warrior graves, skeletal trauma and the spread of weapons alone. Settlement structure and economy must also be taken into consideration. This paper will focus on the emergence of violent conflict in this cultural group and will discuss how warfare caused changes in both economy and social structure. Warfare in the Early Bronze Age: a point of departure

Signs of belligerent activity emerge as the Nitra group started to spread to the neighbouring area near the Váh river during the Classical Phase. In this period the importance of hunter-warriors becomes evident. They can be defined as an elite fraction of society associated with warfare and hunting. Their importance is evident when considering the occurrence of prestige ornaments made of materials deriving from hunting. Thus, the social status of these men was linked to the prestige economy of the community. Furthermore, the amount of weaponry in their graves suggests that hunter-warriors were responsible for defending their people, as well as being charged with securing the wealth of the society.

The Nitra group is especially interesting when it comes to the study of warfare as extensive research on this group has been carried out during the last few decades, casting new light upon the subject. The major burial sites are well documented and several osteological analyses have provided valuable insights into the nature of violence. Anton Toþík (1963; 1979) and Jozef Bátora (1990; 1991; 2000) have presented overviews that provide detailed information on settlement structure, burial customs and social organization. Recently Mária Novotná (2001) and Bátora (1999a) have presented outlines of warfare in the Early Bronze Age in Slovakia, discussing the role of warfare in the Nitra, ÚnČtice and Maćarovce groups by focusing on skeletal trauma and weaponry. Whereas Novotná investigates the development of metal weapons, Bátora’s paper is the first attempt to emphasize warfare as an important feature in the development of Early Bronze Age cultures, by drawing attention to differences between age and sex groups in the frequency of trauma during the various

Contacts with the Proto-ÚnČtice culture in Lower Austria and Moravia in the Classical and Late Phases brought several changes. Noteworthy is the change in social structure leading to distinct social differentiation within these communities. High-ranking hunter-warriors 1

Atkinson, R., Curtis, R. and Lloyd, J. 1998. Blackadder: the whole damn dynasty, 1485-1917. London: M. Joseph.

87

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY cultural stages as well as discussing the significance of weapons and their display. Unfortunately, there is no discussion about the significance of war and how it relates to society in either paper. The subject is alluded to, however, in an earlier paper by Bátora (1991) where he mentions that skeletal trauma occurs mainly during the transitional phase between the Nitra group and the ÚnČtice culture. Furthermore, cenotaphs are synchronous with the occurrence of skeletal trauma, supporting the theory that warfare was common in this transitional phase. Bátora also notes the frequency of weapons and suggests that communities living near the society’s borders were more involved than others in bellicose action. These observations intimate that warfare is derived from conflicts with neighbouring regions during periods of cross-cultural contacts, notably with the ÚnČtice culture.

The Nitra group The location of burial sites and settlements The Nitra group is located within relatively clear boundaries in Slovakia and Moravia (Toþík 1963; Bátora 1991; Furmánek et al. 1999; Stuchlík 2001). Its remains occur between the Žitava river in southwestern Slovakia and the Morava river in eastern Moravia. To the south the area is defined by the Little Danube and, to the north, by the beginning of the central Slovakian highlands, near the modern cities of TopoĐþany and Trenþín (Fig. 1). This region can in turn be divided into three major areas: (1) the Nitra basin and (2) the Váh basin in Slovakia, and (3) the Rusava and Olšava valleys in Moravia. Of these areas, the Nitra basin is the most important with many large burial sites such as Branþ, Nitra-ýermáĖ, Jelšovce, Výþapy-Opatovce and Mýtna Nova Vés. Cemeteries from this region usually comprise 200-500 graves. Most burial sites in this area lie on small terraces close to the river and a very few are located on sand and loess dunes. Although traces of settlements are scarce, the fact that many burial sites lie close to each other

Although I concur with Bátora’s remarks, warfare in the Nitra period is a complicated matter that requires close attention. My proposition is that warfare was a key element in the maintenance of control over the flow of prestige goods as well as in expansion - the acquisition of new resources.

FIGURE 1. The distribution of Nitra group sites in Moravia and Slovakia (after Toþík 1963, Furmánek et al. 1999 and Stuchlík 2001).

88

ANDREAS HÅRDE: THE EMERGENCE OF WARFARE IN THE EARLY BRONZE AGE: THE NITRA GROUP suggests that settlements must have been close to the cemeteries. The Nitra basin is divided geographically into two regions: a southern lowland (the lower Nitra basin) and a hilly northern area (the middle Nitra basin). The border between these two regions can be drawn at the modern town of Nitra, where the lowland changes into a landscape of rolling hills. This division marks not only different geographical areas but also socially distinct areas. As regards burial customs there are differences between burial sites from these regions, in particular when it comes to the sacrifice of animals during funerals. In the northern area sheep and goats prevail whereas cattle are found in the south (Bátora 1991: 96).

Chronology of burial sites and settlements The Early Bronze Age in eastern Moravia and southwestern Slovakia is divided into two successive ages: the Nitra and ÚnČtice periods. The Nitra period can be divided into three chronological phases (Fig. 2; Toþík and Vladár 1971: 418; Bátora 2000): 1. An Early Phase distinguished by influences from the earlier Cháopice-Veslé and Kosihy-ýaka groups, and the Bell Beaker/Corded Ware culture; 2. the Classical Phase during which the Nitra group develops a distinct material culture. Later in this phase the ÚnČtice tradition from Moravia and Lower Austria begins to influence the material; 3. the Late Phase, or Nitra-ÚnČtice Phase, which marks a transition to a ÚnČticeinfluenced society where stone, bone and copper artefacts are entirely replaced by counterparts in bronze.

The western extension of the Nitra group covers the Váh basin. In contrast to the locations around the Nitra river, the burial sites here are few and most finds derive from scattered locations on loess and sand dunes with few remains. Important sites are Abrahám and VeĐký Grob (Chropovský 1960), medium-sized burial sites with fewer than 200 individuals interred at each. In eastern Moravia the River Morava runs past lowland near KromČĜíž, set between the Moravian highlands and the White Carpathians, where the third division of the Nitra group is located. Although there is generally a limited number of finds from this area, the cemetery at Holešov is an important one with its several hundred graves and a high frequency of weapons (Ondráþek 1972; Ondráþek and Šebela 1985). The location of burial sites is by no means random but follows a very strict pattern. There is clear congruence between the geographical environment, soil type and size of cemeteries (Bátora 1991: 93). The largest cemeteries are located on river or brook terraces with black soil, suggesting that optimal agricultural land was preferred. Not surprisingly, these sites were in use for long periods. Branþ, for instance, is estimated to have been in use for 120 years and Výþapy-Opatovce for 150 years. Sites situated on black soil are mainly located in the middle Nitra basin but also occur in the Váh basin at VeĐký Grob and Abrahám, in Moravia at Holešov and in the lower Žitava basin although here they are few in number. Second to black soil is a preference for the river soil in the lower valleys of the Nitra, Váh and Žitava rivers. The least preferred soil type was brown soil, found mainly in the middle and northern part of the Váh basin region. Occasional sites on brown soil also occur in the middle Nitra basin. Following my earlier argument that the location of burial sites indicates the presence of settlements, it is evident that the people of the Nitra group preferred to live on fertile black soil in river valleys. Settlements on inferior soil elsewhere can be explained by the chronological sequence, as sites in the Váh basin are more recent than most in the Nitra basin. Thus, there was apparently a territorial expansion when the Nitra basin became overpopulated.

FIGURE 2. Comparative chronological sequence in the Early Bronze Age in south-western Slovakia, Moravia and Lower Austria (cf. Bátora 2000).

89

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY The later ÚnČtice period represents a complete transition to a material culture generally known as the ÚnČtice culture in central Europe. The Mad’arovce (in Slovakia) and VČteĜov culture (in Moravia) is the closing stage of this period.

difference indicates that cattle were preferred in the lowlands and sheep/goats dominated the hills and highlands. Along with domesticated livestock the hunting of wild boar, roe deer and red deer was important. They might not have been hunted primarily to provide meat and hide, as the domesticated animals must have met these needs. I should like to draw attention to the fact that resources from hunted animals were used to produce prestige goods,2 thereby making hunting an essential means of competing for the accumulation of wealth and prestige.

The oldest settlements are situated in the middle Nitra basin, which was inhabited by the ýaka group during the Eneolithic (Pavúk 1981). During the Classical Phase settlements spread into the lower Nitra basin and, in the later part of this phase, it is here that we find the earliest ÚnČtice settlements from the Hurbanovo group (Dušek 1969; Vladár and Romsauer 1999).

The most evident of the prestige goods associated with hunting are the ornaments and pendants made from wild boar tusks. These occur mainly in male burials, often together with weapons, an important association whose significance I shall return to below. Since the wild boar is a fierce and dangerous animal only the most brave must have dared to hunt it. The sizes of the tusks, with a mean length of 15cm, indicate that the boars were large and, no doubt, difficult opponents to bring down. The wearing of these tusks was probably a declaration of bravery and courage.

Contemporary with the earliest settlements along the Nitra river are some settlements in the Váh basin. Although these can be attested to derive from settlements of the Cháopice-Veselé group, the Nitra group does not properly populate this area until the end of the Classical Phase (Pavúk 1981: 168). The same is true for the Nitra settlements in Moravia, which date to the Classical and Late Phases (Ondráþek and Šebela 1985; Stuchlík 2001). Bátora (1991: 92ff) shows that settlements on terraces prevailed in the Early and Classical Phases and were gradually replaced by settlements on dunes during the Nitra-ÚnČtice period. Settlements on hilltops coincide with the emergence of the ÚnČtice culture.

The other important item fashioned from wild resources was the antler bead: these beads were central to the manufacture of large and eye-catching necklaces. More than 25,000 antler beads have been found in Slovakia, primarily in the cemeteries of the Nitra basin and only seldom at sites outside this region. Shells, from both riverine and marine shellfish, were also used in prestige necklaces (Bátora 1991: 58; 2000: 343).

In summary, the organization of settlements and burial sites suggests an expansion from the middle Nitra basin into neighbouring regions, notably the Váh basin. The apparent cause seems to be a need for new agricultural land. These new settlements, however, were smaller than the older ones, implying that the communities of the Nitra basin maintained economic power.

Commodities used as prestigious items were numerous in the Nitra group, consisting mainly of female ornaments and metal weapons. Apart from necklaces of antler and shell beads, which were evidently in great demand, most prestige goods were made of copper. Spiral rod necklaces, ear- and arm-rings, diadems and pins, of many different forms, predominate (Toþík 1963; Bátora 1991). Copper ornament production was conservative in its design throughout the period but, with the advent of the

This expansion took place during the late Classical Phase and coincided with closer contacts with the ÚnČtice culture in Moravia and Lower Austria. The expansion is important in terms of warfare since it can be associated with other signs of troubled times in this phase, a subject to which I shall return. First, it is important to understand how the economy of the Nitra society was connected to warfare and the emergence of conflict. Economy

2 It is difficult to draw a general distinction between commodities and prestige goods in the Nitra group as they are context dependent. Basically, imported or rare artefacts can be seen as prestige goods if they were used to command position in people’s minds by being restricted to a certain fraction in the society. Hunter-warrior ornaments, copper daggers and imported trinkets from the ÚnČtice culture are some examples of prestigious items. Common and non-foreign artefacts should then be labelled as commodities as most of them are widespread and frequent. However, certain artefacts, notably antler beads and copper rings, vary in sheer number between the different burial sites and among the burials within the same site to such a degree that it can be assumed that the possession of a large number of these items was an indication of wealth and prestige. These objects are not necessarily restricted to a certain fraction in the community although they are most frequent amongst women. As single objects or as a group in general these common artefacts will be referred to as commodities, but when used to command position and signify wealth they will be referred to as prestige goods in order to denote their specific use.

Pastoralism and arable farming were the basis of the economy. In the previous period, the main economic activity of the Cháopice-Veselé group was stockbreeding, with small and mobile settlements (Bátora 1991: 94). The Nitra group’s settlement on black soil made possible the extensive use of cultivated cereals to support a large population. Breeding of cattle, pigs and sheep/goats was also essential, as indicated by several burial sites where animal sacrifice was part of funerary rituals. Ribs of cattle are frequent in graves at Branþ and Holešov whereas sheep/goats are totally dominant at VýþapyOpatovce, Jelšovce and Mýtna Nová Ves (Vladár 1973; Ondráþek and Šebela 1985; Bátora 1990: 172). This 90

ANDREAS HÅRDE: THE EMERGENCE OF WARFARE IN THE EARLY BRONZE AGE: THE NITRA GROUP ÚnČtice culture, much of the metallurgy changed as bronze superseded copper. Old forms of prestige goods ceased to circulate and were replaced by ÚnČtice counterparts, mainly pins and smaller rings.

This expansion might have been driven by an elite yearning for more power but it is difficult to demonstrate. It is striking that contemporary burial sites such as Mýtna Nová Ves and Výþapy-Opatovce, located close to each other in the middle Nitra basin, have a distinctly different proportion of elite graves. Whereas the former has many such graves the latter has only a few.3 Weapons occur in 10-25% of the graves at Mýtna Nová Ves, in striking contrast to Výþapy-Opatovce where the proportion is only 3.47%.

The single most valuable resource the Nitra group had at its disposal was copper. Although this alone cannot explain the growth of this cultural group, access to metal provided an advantage over other societies in neighbouring regions. The mountains of central Slovakia are rich in copper ore and many extraction sites were in use from the Bronze Age to medieval times. Although the Nitra group occupied flat country to a large extent, copper was available in the neighbouring highland massif of the Little Carpathians (Toþík and Vladár 1971: 418; Bátora 1991: 106ff).

Even if it is difficult to connect a warring elite to territorial expansion it is clear that one major strategy for the acquisition of power and status was the control of the production and distribution of commodities and prestige goods. To further understand their importance in the Nitra society, mortuary practices must be considered.

In summary, the economy of the Nitra society was based on agriculture and hunting with the latter pursuit being important in providing prestige goods. The abundance of copper ornaments together with necklaces of antler beads and pendants of boars’ tusks attests to prestige competition having been a crucial part of the economy. To have both access to resources and control over the flow of commodities and prestige objects was probably a source of tension and conflict. Land acquisition and prestige emergence of a warring elite?

competition:

Hunter-warriors - mortuary practices and social organization Mortuary practices and an outline of Nitra social organization To understand social organization it is important to study not only grave goods but also the construction and arrangement of graves within cemeteries. In fact, the construction of a grave sometimes reveals more about the social position of the deceased than the accompanying grave goods. In short, burial practices in the Nitra group conform to a standard whereby graves have similar construction whilst grave goods vary with the sex and social position of the deceased.

the

The connection between hunters and status is central to understanding the emergence of warfare. Burials that include hunting equipment are better provided than those of other men, suggesting that hunters therefore belonged to an elite fraction in the society. Hunting seems to have been limited to distinct groups who provided resources not available to the larger population. The larger grave good assemblages suggest that members of these groups also controlled the manufacture and distribution of commodities and prestige goods. Furthermore, since both hunting and warfare are emphasized in their burials, these individuals will henceforth be referred to as hunterwarriors.

All graves are inhumations and are placed in flat cemeteries which are oblong in plan and usually extend in a south-north or north-south direction (cf Toþík 1979; Bátora 1991; 2000). The graves are usually oblong with sharp or rounded corners and the deceased is placed in a contracted position on their side - males lie west-east on their right side and females east-west on their left side, both with their face to the south. Some graves deviate from this pattern in terms of size and construction. These graves indicate a differentiation in social structure, as they were reserved for hunterwarriors and the social elite. They differ from other graves to a greater or lesser extent in the following features:

Taking both settlement location and economy into account, it is apparent that some regions and communities had an advantage over their neighbours, leading to inequality and privilege, and consequent tension between neighbouring groups. Areas with black soil were needed to sustain a large and well-nourished population. Settlements on black soil in the Nitra basin probably became overpopulated and subsequently their inhabitants expanded into neighbouring areas. These new settlements were, however, located on poorer soil and could not support similarly large populations. This is indicated by the small size of the cemeteries and the absence of wealth in the graves.

1. 2.

3

space – each contains a large burial chamber; the special position of the body, either a) skeleton lying on its back, or b) displacement of body;

The exact number of weapons for Mýtna Nová Ves cannot be estimated at present since the site is not published but preliminary reports by Bátora (1991: 94; 1999a) claim that the proportion is above normal for the Nitra group.

To take control of neighbouring communities was probably one strategy for expanding the subsistence base. 91

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY 3.

4.

the construction of the grave pit, consisting of one or more of a) a wooden chamber or wood lining, b) stone lining, c) whitewash, d) the use of red ochre; the construction above the grave pit, consisting of both or either a) a mortuary house - with (1) post holes at the corners or (2) oblong ditch, b) a burial mound - with (1) concentric ditch or (2) without ditch.

From Branþ, Jelšovce and Mýtna Nová Ves there are burials which indicate that we are dealing with two fractions: high-ranking and common hunter-warriors. They are recognized by a differentiation in grave goods and by grave construction. Common hunter-warriors are buried in ordinary graves with weapons and perhaps ornaments of bone or copper. Those men belonging to the higher rank group are often buried with several weapons, an array of ornaments, and sometimes also pots and amphorae. Occasionally the grave itself has exclusive features, such as a wooden lining, and was sometimes covered by a mound or a mortuary house. Position on the burial site is also an indication of social status. At Branþ, certain hunter-warriors had war-gear and grave constructions different to that of other hunterwarriors, although both can be classified as belonging to the commoners as I shall show below.

Many features of the Nitra burial customs were new to the area and have more in common with funerary practices in eastern Europe (i.e. eastern Slovakia and western Ukraine) than with contemporary cultures in Austria, Moravia and Hungary. The use of whitewashed walls in grave pits, red ochre and wooden crypts are also of eastern origin but disappear during the Late Phase, giving way to the more informal burial fashion of the ÚnČtice culture. On the other hand, the custom of building mortuary houses spread to various regions of Austria, Germany, Moravia and Poland from the ÚnČtice to the Lausitz and Urnfield periods (Bátora 1999a).

Hunter-warriors were equipped in a uniform fashion and their grave goods consist of weapons, tools and ornaments. Amongst their arms, equipment for archery prevails with arrowheads and bracers. Other arms present are daggers, knives and axes. Archery dominates in the Early and Classical Phases, whereas daggers and knives were preferred from the transition to the Nitra-ÚnČtice Phase. Axes are few and belong to this Late Phase; as for bladed weapons, this is an influence from the burial custom of the ÚnČtice culture. Tools connected with hunter-warriors include whetstones, presumably used to sharpen daggers and knives. Raw material of stone found in these male graves is not considered here as hunterwarrior equipment since it is difficult to assess its precise use (although it might have been used to manufacture arrowheads).

Considering the distribution of graves, their construction and wealth in grave goods, it is possible to distinguish three social categories in the society for men, women and children alike (Bátora 1991: 135ff; 2000; cf Shennan 1975). At the top of the social ladder stood high-ranking hunter-warriors with ostentatious graves rich in grave goods. Below them were the commoners who made up the majority of the population and whose graves were less elaborate, but not poor in grave goods. Women are predominant in this category and are followed by the rest of the hunter-warriors not belonging to the higher-ranking fraction. At the bottom of the ladder were those who were either not provided with grave goods or had just a few items. Children and men without weapons are overrepresented in this category.

Ornaments associated with hunter-warrior burials are boars’ tusks, bone discs, decorated bone tubes and belt fittings of copper. These ornaments belong to the Early and Classical Phases and occur in contemporary Eneolithic cultures in the region and in the Mierzanowice and KošĢany cultures (Bátora 1999c: 35-44). With the exception of copper belt fittings, these ornaments disappear during the Late Phase.

The division between the last two categories does not necessarily indicate two separate social levels as females in general had more grave goods than males. Male grave goods consisted mainly of raw materials for manufacturing tools and only occasionally were men presented in death with ornaments such as copper rings and beads. In other words, while burial customs prescribed ornate grave goods for women, men were not likely to get anything special.

Boars’ tusks occur at many burial sites and their use was widespread far beyond southwestern Slovakia. They often have drilled holes at the proximal end and, in some cases, also at the distal tip: their position in graves near the chest or head indicates that they were worn either as ornaments on the body or as hair decoration. A hunterwarrior usually wore one to three boars’ tusks but some had as many as eight (Branþ, grave 104). The numbers of tusks might have been a sign of status amongst hunterwarriors, perhaps a token of bravery.

Hunter-warriors as a social group The various burial sites make it possible to distinguish a category of individuals who can be defined as hunterwarriors owing to the presence of arms and trappings associated with hunting and warfare in the graves. In death hunter-warriors were entitled to a certain burial rite with specific grave goods.

Bone discs with several drilled holes are usually found at the back of the head, suggesting that their use was connected with the hunter-warrior’s hairstyle. These discs are mainly found at sites in the Nitra basin, such as

92

ANDREAS HÅRDE: THE EMERGENCE OF WARFARE IN THE EARLY BRONZE AGE: THE NITRA GROUP Mýtna Nová Ves, Branþ and Nitra-ýermáĖ, and only on one occasion in the Váh basin at VeĐký Grob.

It is difficult to draw any clear chronological boundaries amongst the graves. Although the burial site appears to have been divided into distinct groups, it is not possible to ascertain whether these divisions derive from family groups or from some other social ordering. Cemeteries in the Nitra groups are usually organized into several rows of graves and in many cases it is possible to observe that men, women and children are buried in separate clusters within these rows. Also, hunter-warriors’ graves are often organized in similar patterns. At Výþapy-Opatovce and Tvrdošovce they form parallel rows, at Jelšovce and Mýtna Nová Ves they lie in clusters and at Branþ both styles are observed (Fig. 3). I should like to argue that this indicates that rank and probably also descent played a crucial part in determining the location for individual burials.

Ornamented bone tubes occur on a number of sites, such as Alekšince, Branþ, ýerný Brod, Jelšovce, Mýtna Nová Ves, Nitra-ýermáĖ, ŠaĐa I and Tvrdošovce. They are made of the tibia of sheep or the humerus of large birds and are associated with high-ranking hunter-warriors. The use of such artefacts is debatable: Bátora (1999c: 39) suggests that they were some kind of whistle used by hunters but they could also have been cases for awls or pins, or simply ornaments of unknown kind. Hunter-warriors in the context of Branþ In order to clarify the social status of hunter-warriors and their importance within society I have chosen the Branþ cemetery as a case study. Being one of the largest burial sites from the Nitra group and located in the very heartland of the Nitra basin, it is an ideal subject for study. The cemetery at Branþ was in use through most of the Early Bronze Age - from the Early Phase in the Nitra period to the ÚnČtice-Maćarovce horizon (Vladár 1973; Bátora 2000: fig. 690). The cemetery seems to have been abandoned at some point during the Nitra-ÚnČtice Phase, but further to the north a short revival of the cemetery took place during the ÚnČtice-Maćarovce Phase.

The cemetery at Branþ extends in a south to north direction both chronologically and spatially. Later graves respect the old ones as no graves cut each other. This indicates that graves were marked above ground and were visible for a very long time. When the later Maćarovce cemetery was established it was placed further away from the previous burial site indicating that the old graves were still known, and were perhaps also visible. Out of the 237 graves from Branþ, 40 graves (16.9%) can

FIGURE 3. Hunter-warrior graves. Branþ. Nitra group.

93

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY be considered as belonging to hunter-warriors. With one double grave, the total number of hunter-warriors was 41. The majority of the hunter-warriors were adult men (80.5%), while adult women (4.9%; two individuals) and children (14.6%) make up only a small fraction. Hunterwarrior equipment in child graves - which at Branþ consists of knives alone - should be seen as an indication that these children were expected to become hunterwarriors when they grew older and it is of note that one child was a girl (grave 168).

bone disc. The mortuary houses at Mýtna Nová Ves contained similar grave goods. Most spectacular was grave 262 with its many arrowheads, a stone axe, a boar’s tusk, pots and bear claws from the funerary shroud (Bátora 1999c: 5ff). These mortuary houses belong to the Early and Classical Phases of the Nitra group and resemble the chamber graves found in the Ukraine and Russia. The custom of building mortuary houses was taken up by the ÚnČtice culture and spread further to the west during the Middle Bronze Age (Bátora 1999c).

It is interesting that some women were buried as hunterwarriors. At Branþ the two women were each provided with a boar’s tusk. A boar’s tusk also occurs in a woman’s grave at Tvrdošovce (grave 42; Toþík 1979) and a female hunter-warrior buried with several arrowheads is known from Mýtna Nová Ves (grave 177; Bátora 1991: 125). Should these grave goods be seen as a mark of respect given to the dead as a part of the funeral ritual, rather than as items worn or used life? Not necessarily. Women’s graves with hunter-warrior equipment are known from other contemporary cultures: in the Mierzanowice culture some women were provided with arrowheads as grave goods and in the KošĢanyer culture some were buried with daggers (Bátora 1991: 125). Thus, we should consider that some women were indeed hunter-warriors or at least acquired this status.

There are other grave constructions worth mentioning in relation to hunter-warriors and their kin (Fig. 4). These are graves with mounds, whitewashed walls, and wood linings or sprinkled with red ochre. At Branþ these graves are found almost exclusively in the southeastern part of the cemetery. Considering that these graves are located within a group with the members of the hunterwarrior elite, it seems as if the practice of using whitewash, red ochre and wood lining in graves was a privilege reserved for high-ranking families and their kin (see below). Regarding the hunter-warriors’ grave goods there are distinct synchronous and chronological variations. During the Early and Classical Phases the traditional Nitra group hunter-warrior was characterized as an archer, with ornaments of boars’ tusks, decorated bone tubes and bone discs. This archer’s set more or less disappears in the Late Phase as the burial practice became influenced by the ÚnČtice culture. In this phase daggers and knives became the main indicators of a hunterwarrior. From a synchronous perspective there are reasons to believe that the burial custom for hunterwarriors in the Nitra group follows a distinct structure, given their arrangement in the cemetery and grave goods indicative of rank and perhaps also of descent.

As mentioned earlier it is possible to distinguish between common and high-ranking hunter-warriors. The division can be determined by the position and construction of the grave as well as by the range of grave goods. In Branþ there are three such graves of high-ranking individuals (graves 31, 62 and 88) located in the eastern part of the burial site. Most conspicuous is grave 31, with a construction of a mortuary house above the grave pit. This kind of construction is rare and elsewhere is found only at Mýtna Nová Ves (graves 206, 262, 305, 509 and 513) and Moravská Nová Ves (grave 19; Bátora 1999c: 32ff). In Jelšovce a similar construction is observed (grave 444). Here a high-ranking hunter-warrior was buried in a grave surrounded by an oblong ditch with remains of timber indicating a wooden construction above the ground (Bátora 2000).

A closer look at the southeastern part of the cemetery at Branþ, for example, reveals that it is divided by two larger clusters of hunter-warriors, separating the burial area into a western and an eastern section. The graves in the two sections are quite distinct and the differences cannot be dismissed as being chronological. As I shall argue below, we are dealing either with two different families or two separate ranks within the same kin group in these sections.

The construction of the grave was similar on all these sites. A large pit was lined with wood which sometimes also covered the opening, creating a wooden crypt for the dead. Wooden posts were erected at the corners of the pit supporting a roof or, in the case of Jelšovce, a construction of logs dovetailed at the corners. Inside the mortuary house, and above the grave, a large amphora was placed as a funerary offering. Since the area in the immediate vicinity of the mortuary houses was free from burials, these graves must have been accessed regularly, perhaps for ritual activity or offerings to the ancestor.

Graves in the western section usually contain boars’ tusks whereas these artefacts are conspicuously few in the eastern section. Within the western section a further division can be made. The southwestern division is characterized by boars’ tusks and weapons are almost absent whilst in the western central division, knives and daggers appear in the graves. In the northwestern division, arrowheads prevail, together with boars’ tusks. In the southeastern division there are graves with boars’ tusks, arrowheads, knives and daggers, whereas the eastern-central and northern divisions are mainly characterized by knives and daggers.

Hunter-warriors in these graves were often given many grave goods. At Branþ the grave beneath the mortuary house contained a knife, belt fittings, boars’ tusks and a 94

ANDREAS HÅRDE: THE EMERGENCE OF WARFARE IN THE EARLY BRONZE AGE: THE NITRA GROUP

FIGURE 4. Special grave features. Branþ. Nitra group.

The manner in which the graves were constructed also suggests a distinction between the western and eastern parts of Branþ (Fig. 4). Graves with boars’ tusks prevail in the western half of the cemetery and knives and daggers mark graves in the eastern half. It is interesting that axes are found close together in the central western part since this weapon is rarely used in funerary contexts.

them to distinguish themselves by adopting new rites and customs. It is my suggestion that the southeastern part of the cemetery at Branþ (and probably the later northeastern part as well) was shared by a distinct kin group which was divided into two fractions: one common and one elite. This kin group is characterized by the use of specific grave features, such as whitewashed walls, wood lining and red ochre. Since these special features are foreign to the region they might have been used to commemorate an ancestry or heritage in the western Ukraine where these features are common. The combination of old-fashioned mortuary elements together with new ones from the ÚnČtice culture suggests that this group enforced both traditional ways of living as well as taking advantage of the new way. Perhaps this was a way to strengthen an old social structure that legitimatized their power, yet at the same time adopting a ‘new’ social structure in order to control new commodities.

The use of daggers and knives at Branþ is noteworthy. Initially the hunter-warrior graves in the southeastern part (southern and central sections) combined the traditional archer’s set with daggers and knives. Later, this practice changes: knives and daggers disappear from the northwestern division and the archer’s set is abandoned in the northeastern division. This separation becomes accentuated in the later stages of Branþ when the hunterwarrior graves in the northeastern part become almost pure dagger and knife burials whereas the contemporary graves in the western-central and northwestern parts follow the traditional burial practice, without daggers and knives (except grave 2).

Whereas the common hunter-warriors in the western section of the southeastern part were not distinguished, those in the eastern section were. The use of daggers and knives and the absence of traditional ornaments, together with the presence of high-ranking hunter-warrior graves, strengthen the hypothesis that we are here dealing with a fraction of common hunter-warriors who must have had a position higher than their fellows in the western section and also higher than the rest of the individuals interred in

I would like to argue that the reason why these hunterwarriors were distinguished by the use or possession of daggers earlier than the rest of the hunter-warriors is that they were high-ranking with close connections with the ÚnČtice culture. As this culture influenced the Nitra group, close connections to it must have been an advantage for the ruling elite, giving them the upper hand as they controlled the flow of prestige goods and allowing 95

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY contemporary hunter-warriors’ graves elsewhere in the cemetery. Thus, they were a part of an elite connected to the high-ranking hunter-warriors.

slightly with the advent of the ÚnČtice culture. Least evidence of violence was found in the Maćarovce horizon, according to Bátora (ibid.), and this is also a time when there are fewest weapons in the graves.

The warring elite Apart from evidence of skeletal trauma and weaponry there are several other sources that attest to the reality of warfare in the Nitra group. Both cenotaphs and grave robbery can also be related to hostilities during this period.

To summarize, hunter-warriors played an important role in Nitra society. Their burials show that they were separated from other social groups in terms of grave goods and, in some cases, by grave construction. Furthermore, it is possible to divide this social group into ranks with commoner and high-ranking members.

Skeletal trauma

The social organization of hunter-warriors is illustrated by graves at Branþ. Here it is possible to distinguish between hunter-warriors both synchronically and through time. From a chronological viewpoint the hunter-warrior outfit changed from ornaments of boars’ tusks and archery gear to knives and daggers. This change may be due to influence from the ÚnČtice culture, suggesting transformation of both social and martial values over time. This must be compared with the differentiation in burial practice between various parts of the cemetery at any one time.

Traces of trauma caused by intentional violence are common on skeletons from Nitra period burials. Both the presence and the variety of fractures indicate serious clashes between groups - the grim reality of deliberate conflict - as opposed to occasional scuffles and accidents. From the many osteological analyses conducted, it is possible to reconstruct the nature of the harmful encounters, outlining the strategy behind the combat as well as fighting techniques. There are at least 59 occurrences of skeletal trauma (Tables 1 and 2): 31 cases of cranial fractures on 28 individuals (16 men, 8 women and 4 children) and 14 cases of post-cranial fractures (10 men, 3 women and 1 child; Bátora 1999a: figs 6-8, 11).4

At Branþ it is the eastern part that draws attention with its special grave constructions and an early transformation in the hunter-warrior’s outfit from a traditional Nitra style to ÚnČtice fashion in the Classical Phase. This change is noteworthy as it emphasizes that the warring elite adopts the new way of living ahead of the rest of society. This, in conjunction with earlier remarks on the flow of prestige goods, as well as implications of settlement expansion, suggests that bellicose activities were key elements in sustaining Nitra society.

The majority of the traumatic traces are of blunt- and sharp-edged weapon injuries to the left side of the body, mainly occurring on the skull. The nature of the wounds as well as their position indicates that they were caused by clashes between armed persons. Some fractures such as depressions on skulls, broken clavicles or fractured arms might have been caused by accidents unrelated to warfare but these are considered to be far fewer than those resulting from confrontation in close combat.

Thus far, warfare may only be inferred from its representation, not from evidence of its reality: the existence of an armed and prominent element of society does not necessarily imply that war was actually pursued. Various signs of troubled times in this period, however, accord with this warlike image presented by the grave goods of Nitra society.

Blows to the head illustrate most varieties of wounds and battle techniques. There are basically three kinds of skull fractures from the Nitra group: 1, sharp force trauma; 2, blunt force trauma; and 3, depressions. Sharp force trauma injuries are characterized by breaking through the bone wall, often creating an opening that takes the shape of the weapon (see Knüsel this volume for detailed information on the appearance of fractures). The edges are sharp and the impact of the weapon often flakes the bone as it strikes and sometimes also when it is pulled out. These wounds are often fatal since the impact, as well as the splinters of bone from the flaking, causes immense blood loss or damage to the brain.

Troubled times: evidence and implications Sources of evidence for conflict and violence Various sources attest to the occurrence of war and violence in the Nitra period. To understand the development of warfare Bátora (1999a) studied the frequency of weapons as well as the frequency of skeletal trauma in burials of the period, concluding that the bow and arrow together with axes gave way to daggers during the course of the Nitra sequence.

Breaking and shattering of the bone, often with several fracture lines, distinguish injuries from blunt force 4 Owing to difficulties in getting access to some of the osteological reports mentioned by Bátora (1991; 1999a) it has not been possible for me at present to display all known evidence of skeletal trauma from the Nitra group in Tables 1 and 2. Also, osteological data from some sites, such as Mýtna Nova Ves and Jelšovce, are not yet compiled (J. Bátora pers. comm.).

Although it may be assumed that conflict would be less deadly with the coming of a new style of armament, the skeletal material shows otherwise. Trauma increased during the end of the Classical Phase and then declined 96

ANDREAS HÅRDE: THE EMERGENCE OF WARFARE IN THE EARLY BRONZE AGE: THE NITRA GROUP

TABLE 1. Blunt and sharp force trauma. Nitra group.

TABLE 2. Impaling force trauma. Nitra group.

trauma. These fractures do not necessarily break through the bone wall and are not automatically fatal if they are not too extensive.

the left elbow and left lower arm (Bátora 1991: 124). The fractures are similar to each other, indicating either a blow to the elbow that fractured the distal humerus and proximal ulna or a stroke to the lower arm breaking the ulna and radius in two. This kind of injury is attributed to attempts to parry an incoming blow to the head. Although a broken arm would make the victim an easy target of further attack, many of these parry injuries have healed, indicating that the victim survived the incident. The same can be said for leg injuries although these are rather few in number.

The last category of cranial injuries consists of small round depressions. Unless these wounds are clearly distinct they are often overlooked since they can be hard to determine as fractures. They are either the result of minor blunt fractures failing to break the bone or wellhealed fractures. In this last case the fracture may only be visible as a small dent in the skull and thus may easily be overlooked if the skeleton is not well preserved.

Injuries to the torso are almost absent. The wounds in this area consist of broken ribs and arrowheads lodged in vertebrae and ribs. Again, poor preservation of the skeleton inhibits the identification of wounds as it is difficult to separate cut-marks from scratches and dents caused by taphonomic processes. Also, it can be debated to what extent penetrating weapons, such as daggers and arrows, would leave marks on the skeleton when they hit the upper body. However, there are two plausible cases from Jelšovce of injuries caused by arrowheads in the torso (Table 3).

In appearance the cranial fractures seen on individuals dating to the Nitra period are either round or long, narrow and quadrilateral, with marked edges indicating that they were caused by blunt or heavy weapons with a crude edge, such as stone clubs or axes. Sling shots might have caused the smaller depressions. This seems a paradox since these kinds of weapons are very rare in the burials of the Nitra group, to which I shall return below. The second largest area of injuries is the arm, especially

TABLE 3. Plausible impaling force trauma. Nitra group.

97

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Both skeletons are male and in each a single arrowhead was located inside the ribcage. In burials of the Nitra group arrowheads are usually placed near the pelvis, as if they were carried in a quiver. This can be seen in grave 473 from Jelšovce where four projectile points were placed behind the knees. However, a fifth arrowhead was stuck between the ribs, suggesting that it was the cause of death rather than a grave gift. The same can be said of the arrowhead in grave 436 which was located in the upper part of the torso near the spine. Although it cannot be ascertained whether the arrows had killed these men, the position of the arrowheads is striking.

and fatal aggression. The multiple fractures on each skeleton show that the victim was severely mauled by his or her antagonist although the wounded must have died instantly in some cases. It is noteworthy that blows were often directed towards the head, smashing the skull several times while leaving the post-cranial body alone (graves 226, 262 and 509; Jakab 1999). In one case the whole body was mauled, resulting in a smashed shoulder blade and broken ribs (grave 305; ibid.). That hunter-warriors of high rank at Mýtna Nová Ves were prime targets for assailants is evident not only from the presence of multiple injuries, but also from the occurrence of cenotaphs for the same hunter-warrior rank. The fact that bodies were so badly injured on the battleground (and that, given the existence of the cenotaph graves, others were perhaps carried away by the enemy) indicates that it must have been important to desecrate the body of the fallen hunter-warrior as far as possible. The death of a high-ranking hunter-warrior must have been a severe blow to a small community, thereby altering the balance of power in the region. To kill a hunter-warrior of high rank was perhaps seen as especially courageous, granting the killer special status. Whether a similar strategy was employed in other areas with sites containing cenotaphs for high-ranking hunterwarriors is difficult to say but it seems likely.

When looking at the trauma pattern a division into sex and age is apparent (see Bátora 1991: 124; 1999a: fig. 8). Cranial fractures occur mainly on mature male skeletons, with an exception at Mýtna Nová Ves where young male adults with skull injuries prevail. Broken arms, however, are more frequent amongst women. It is interesting that trauma from intentional violence occurs on those individuals who would not be expected to participate in a violent confrontation, i.e. older men and women. The high frequency of skeletal trauma on these persons indicates that they were less skilled in defending themselves, either because they were unarmed or simply because they were not agile enough to dodge an attack. It should be pointed out that although skeletal traumas occur throughout the Nitra phases they are most frequent in the Classical and Late Phases. Bátora (1991: 124; 1999a) considers that most of the evidence for trauma dates to the transition between these two phases as well as to the period during the change to the ÚnČtice horizon. When looking at the occurrence of trauma in the Nitra group, it is important to recognize that each site has a distinct ‘trauma history’ of its own although a general pattern can be outlined for the group as a whole. These distinctions are important in understanding the role of violence in the community and its consequences.

Weapons Arms in the Nitra group are rather diverse in character. Their presence is attested by the occurrence of both weapons and skeletal trauma. It is important to take both circumstances into account as they give different information about how weapons were employed in both combat and funerary circumstances. Thus it is necessary to distinguish between occasions of actual war and the social display of martial values, as I shall return to below. Most weapons in the Nitra group are the same as those found in the preceding Eneolithic cultures in central Europe, notably Corded Ware and Bell Beaker cultures, and there is in general no unity in style for these that could be regarded as distinct for this society. Only copper daggers clearly distinguish the group. All weapons have been found in graves as the custom of hoarding has not been recorded for the Nitra group.

At Branþ it is surprising that cranial fractures have actually healed in all cases. The man in grave 56 is especially interesting as he has five healed fractures on the skull indicating that he had been in combat several times and survived the encounters (Vyhnánek and Hanulík 1971). Healed fractures indicate that the individual somehow got away from his antagonist quickly. This is interesting since a blow to the head causing a fracture stuns the victim, turning him or her into an easy kill. Either the enemy did not have time to finish his victim off or it was not the strategy to do so.

Long-range weapons Arrowheads are by far the most common piece of weaponry. The majority of the arrowheads are made of flint but a few are bone or antler (Bátora 1991: 98; 1999a: 43). Flint arrowheads can be divided into two groups: (1) short heart-shaped arrowheads and (2) elongated triangular arrowheads. Both of these forms stem from the preceding Eneolithic cultures in this area and there are no indications that one or other of the two types should be related to an older or later phase within the Nitra group. In fact, in many cases they are all placed together in

The situation is completely different at Mýtna Nová Ves. Contrary to other burial sites, evidence of trauma is most common amongst younger men. Those most prone to violence were the young chieftains buried in mortuary houses and it is worth noting that they also show traces of multiple fractures, which is otherwise a rare phenomenon. Whereas the men at Branþ survived their fights, the combatants at Mýtna Nová Ves were victims of intense 98

ANDREAS HÅRDE: THE EMERGENCE OF WARFARE IN THE EARLY BRONZE AGE: THE NITRA GROUP graves, suggesting that their shape may be connected to their function. The projectile points were all made in a similar manner with serrated edges formed into wings at the base to create barbs. When hitting their target the edges would have caused blood loss and the barbs would have prevented the arrow from falling out or, in case of human enemies, made removal difficult and likely to cause yet more loss of blood.

resemblance to the Eneolithic copper blades such as those of the Bell Beaker culture (Toþík 1963). The knives in the Nitra group are all willowleaf-shaped with a tang to the handle. Although daggers to a large extent replaced knives as prestige goods, the knife never went out of use as can be seen at Branþ.

Other indications of archery are bracers. Bracers were strapped onto the wrist of the arm holding the bow. All bracers in the Nitra group were made of fine-grained sandstone although archers might also have used leather bracers, as are known from later times. Bracers are scarce in Slovakia and occur only at Branþ and Mýtna Nová Ves (Bátora 1991: 102). In contrast, many are found in eastern Moravia, especially at Holešov (Ondráþek and Šebela 1985).

When comparing the arms present with the evidence of skeletal trauma there appears to be a paradox. Impaling weapons, such as daggers and arrows, are predominant in graves, but are barely attested to by skeletal injuries. Conversely, blunt force trauma is common in the skeletal material and yet blunt instruments are either few, as in the case of axes, hammers and hammer-axes, or absent, as with clubs and sling shots. This divergence is due not only to taphonomic factors but also to distinctions in the social display of martial qualities.

Social display and the reality of war

Close-combat weapons The presence of shallow depression fractures suggests that blunt instruments such as wooden clubs and slings were most likely used in warfare although there is no tangible evidence to confirm the existence of these weapons. The use of perishable materials for such weapons explains their absence in the burial record. This does not, however, explain the lack of axes and hammeraxes. Less than 0.5% of the hunter-warrior graves at Branþ and Mýtna Nová Ves, and 1.25% at VýþapyOpatovce contained axes or hammer-axes, indicating that they were not associated with the hunter-warrior grave in general. However, the axes from Branþ and VýþapyOpatovce do show an interesting pattern as they are clustered together. Could this occurrence be a matter of social identity rather than a reflection of armaments used in life?

Within the category of close-combat weapons there are blunt and edged weapons which can be labelled as axes, hammers and hammer-axes. This group is uniform neither in form nor deposition since these arms seldom occur in graves and none can be regarded as an artefact type belonging specifically to the Nitra group. The origin of their form can be found in the Bell Beaker, Corded Ware and Cháopice-Veselé cultures (Bátora 1991: 102). Axes, hammers and hammer-axes were made almost exclusively of stone, with exceptions of two axes of bone at Branþ (grave 85) and Výþapy-Opatovce (grave 264). Trauma on skeletons shows that both the sharp edge and blunt neck of axes and hammer-axes were used in combat.

The role of daggers in warfare is debatable but there can be no doubt that they were used (cf Novotná 2001). Many daggers, as well as knives, are heavily worn and show signs of having been sharpened by whetstones. Some even show damage at the blade's proximal end with broken tangs or damaged rivet holes.

I would suggest that this is the case. As I have shown at Branþ, the displays of hunter-warrior attributes not only change over time but also show distinct variations between contemporary parts of the cemetery. Graves in eastern section of the southeastern area show a change to the use of daggers, while contemporary graves elsewhere in the cemetery indicate the continuation of traditional funerary practice. The daggers cannot be explained as indicative of a change in warfare but rather as a change in the social presentation of values and ideals. In fact, the hunter-warrior items of boars’ tusks and copper blades are most frequent and these were apparently the basic insignia for the hunter-warriors. The other attributes are used more sparingly and conform to no clear structure although some graves with matching combinations of objects are sometimes clustered together. Furthermore, boars’ tusks prevail in graves dating to the early period of the Nitra group and copper blades in the later, suggesting a sequential differentiation.

Before the advent of daggers, copper blade production in the Nitra group is characterized by knives. These blades were foreign to central Europe at this time and have no

Consequently, the occurrence of weapons in graves does not necessarily represent the hunter-warriors’ armament in life. This accounts for the divergence between the

Close-quarter weapons The dagger is a widespread weapon in the Early Bronze Age. All daggers of the Nitra group were made of copper, with an exception of one bone dagger from Branþ (grave 247). Bronze daggers do not appear until the beginning of the ÚnČtice culture. No other weapon in the earliest Bronze Age shows such a variety in form as the copper dagger. It was developed by merging the common willowleaf-shaped knife with the style of triangular daggers imported from cultures in the Carpathian basin (Vladár 1974).

99

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY

TABLE 4. Cenotaphs. Nitra group. E=eastern part, W=western part, C=at the centre, N=northern part, S=southern part

funerary display of hunter-warriors and the skeletal trauma deriving from warfare.

these graves can be classified as belonging to common hunter-warriors as they were of ordinary construction. However, the occurrence of pots, either in the grave pit or on top of it, may be an indication that these deposits signalled high rank, as this practice is associated with mortuary houses and mounds. Thus, they were most likely kinsmen to the chieftains in the mortuary houses in the same way as those buried in the eastern part of Branþ.

Cenotaphs It is difficult to discuss cenotaphs of the Nitra group as a general category since the majority are from the same burial site, Mýtna Nová Ves. However, since most cenotaphs contain hunter-warrior equipment, their origin may lie in situations of armed conflict in which the body of a fallen comrade was never retrieved. Hence they potentially give us valuable information on the character of warfare as related to the different sites. Furthermore it is of note that cenotaphs occur in the Early and Late Phase, but not in the Classical one (Table 4; Bátora 1990: 171). This could be an indication of hostilities during transitional phases.

A similar pattern is visible amongst the cenotaphs in the Nitra-ÚnČtice Phase, but these are different from the early cenotaphs. Firstly, animal parts replace the absent human and secondly, bronze-founder’s equipment is placed in many cenotaphs. These new features derive from the ÚnČtice culture where many such cenotaphs are known. Of special interest is grave 6 from Mýtna Nová Ves where the carcass of a sheep or goat was buried lying on its right side with its face to the south, as if it were a man (Bátora 1999b: 67). The grave goods were placed near the head and a copper dagger was stuck between the cervical vertebrae. At Branþ a similar grave dating to the beginning of the Maćarovce culture was found to contain pig bones and bronze rings (grave 4; Vladár 1973).

To receive a proper burial was important for hunterwarriors and their kin, particularly those of high rank. The presence of the dead was important in the community's ritual and social life as the occurrence of high-ranking hunter-warriors’ mortuary houses attests. The connection between cenotaphs and warfare is elucidated in Mýtna Nová Ves where five cenotaphs of the Early Phase contained grave sets of high-ranking hunter-warriors. Given that high-ranking hunter-warriors from Mýtna Nová Ves were subject to considerable aggression, as mentioned above, it seems that the bodies of fallen hunter-warriors were not recovered for burial after falling into the hands of the enemy.

Grave robbery Another feature that may be linked with warfare is the occurrence of plundered graves. Grave-robbing is known from the Early Phase of the Nitra group at Branþ, Jelšovce, Mýtna Nová Ves and Tvrdošovce, but the majority of the robbed graves are dated to the Late Phase of the Nitra group. This increase may reflect ÚnČtice influence: whereas plundered graves are exceptions in the Nitra group they are almost the norm in the ÚnČtice culture in central Europe (Stuchlík 1990; Bátora 2000: 469).

The Early Phase cenotaphs at Mýtna Nová Ves (graves 116, 117, 308 and 458) are all located in the vicinity of the five mortuary houses (Bátora 1999b: fig. 1). Apart from grave 458, which had a wooden burial chamber, 100

ANDREAS HÅRDE: THE EMERGENCE OF WARFARE IN THE EARLY BRONZE AGE: THE NITRA GROUP

FIGURE 5. Plundered graves. Branþ. Nitra group.

The change in proportion of this practice is illustrated at Jelšovce where 64 out of 103 graves from the ÚnČtice period were plundered, whereas only a few graves were touched during the Nitra period (Bátora 2000). The practice of grave-robbing follows a similar pattern throughout the Early Bronze Age, but its cause and effect on society appear to have been different. The reason for plundering the graves was primarily to retrieve the metal objects, but taking body parts, such as skulls, was also on the agenda.

boars’ tusks, which enable us to identify the deceased as hunter-warriors. There are two indications that plundered graves at Branþ were the result of troubled times. Firstly, it must be remembered that the people of the Nitra group paid great respect to their dead. Looting a grave must have been an act of desecration - a means of insulting and bringing harm to an enemy. Secondly, the cemetery fell into disuse during the Nitra-ÚnČtice Phase and was not used again until the Maćarovce period, when the new graves were placed further to the north. With these two factors in mind I conclude that the community burying at Branþ in the Late Phase was most likely involved in a conflict that they lost, in which their cemetery was plundered by the victors.

The plundered graves in Branþ are of special interest since they can be related to warfare. In this cemetery 10.5% of the graves have been looted and they are mainly located at the northern and southern fringes (Fig. 5). Graves of both men and women were affected and there are no indications that hunter-warrior graves were the main focus. It must, however, be noted that we cannot tell whether any of the plundered male graves were hunter-warriors, particularly those in the central eastern part, since here status would have been indicated by a dagger or a knife and such an object would have been stolen by the grave robbers.

Signs of warfare There are several events and phenomena related to warfare and violence in the Nitra period. Although most of them occur sporadically throughout the period they seem to coalesce in the transition to the Nitra-ÚnČtice Phase, a time of increased contacts with the ÚnČtice culture. Occurrence of skeletal trauma is more frequent in this phase as are as incidences of grave-robbing and the construction of cenotaphs.

Such may be the case for grave 56: apart from five healed cranial fractures, indicating that the buried man had been in close combat several times, there were no grave goods left that could indicate whether he had been a hunterwarrior. In some hunter-warrior graves (graves 74, 173 and 192) the looters overlooked or left the arrowheads or

Looking at the phenomena of cenotaphs and the plundering of graves, these occur only in the Early and 101

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Late Phases, with emphasis on the latter stage. Their absence in the Classical Phase is difficult to explain. Could it be an indication of a period of peace? It should be noted that cenotaphs occur almost exclusively at Mýtna Nová Ves and that the incidence of grave plundering in the Early Phase is low. The relationship between violence and cenotaphs is linked to the particular kind of aggression associated with high-ranking hunterwarriors at Mýtna Nová Ves, as I discuss later.

2. 3.

control of copper production, and control of the manufacture and dispersal of prestige goods.

As can be seen at Branþ, Jelšovce and Mýtna Nová Ves, society was stratified as certain (kin?) groups had a higher status which entitled them to rich grave goods and specially constructed graves. These high-ranking families certainly controlled the distribution of prestige goods and presumably also controlled access to land. Settlement expansion may have been an attempt to incorporate neighbouring communities, either to increase subsistence returns or simply to control social networks, including the flow of prestige goods and metal.

Regarding weaponry there are no changes that point directly to an increase or decrease in belligerent activities. However, the move to a ÚnČtice style in burial practice hints at growing social tensions. This is evident at Branþ where one fraction of the community changed style earlier than the others. The change in weaponry from the bow and arrow to knives and daggers marks changes in martial display, potentially undermining traditional values and authority.

By the time of the Classical Phase the Nitra basin was densely populated and expansion westwards was necessary. Incipient contacts with the ÚnČtice culture further fuelled this expansion. It is noteworthy that this expansion coincides with the indicators of war cenotaphs, grave plundering and skeletal trauma - as I have discussed earlier.

Warfare and social change The emergence of war

Comparison between burial sites in the Nitra basin and those in the neighbouring Váh basin shows that commodities and prestige goods are scarce on the latter sites. Burials of high-ranking individuals, such as those with mortuary houses, are also absent. A likely explanation may be that the conquering party forced the conquered population to pay tribute to them in the form of food, prestige goods and metal. The absence of wealth and the lack of a high-ranking stratum, or of any hunterwarriors, on these burial sites may be an indication of their subordinated status. This contrast is most apparent in the northern part of the Nitra basin where Mýtna Nová Ves has many high-ranking graves and a high frequency of weapons whereas Výþapy-Opatovce, a few kilometres away, has no elaborate graves and almost no weapons.

Clear evidence for war does not appear until the Classical Phase and the advent of the Nitra-ÚnČtice Phase. During this period settlements spread from the core area in the middle Nitra basin to the regions in the south and to less fertile areas to the southwest and the northwest. This expansion coincided with an increase in trauma on skeletons as well as a change in weaponry, and in the development of cenotaphs and grave-robbing. It seems as if a shift in both economy and social structure, related to ÚnČtice influence, was preceded by the development of conflict and troubled times. One key to understanding this increased competition and hostility is the social change by which the elite began to adopt ÚnČtice material style. The acquisition of bronze was presumably important since this alloy was superior to copper for tools. It must be noted, however, that in its earlier phase this contact coincided with a realignment of power as the elite from the Nitra group began to identify themselves with those from the ÚnČtice culture. There are no bronze artefacts from the Classical Phase but the grave goods of the highest ranking members of society began to change bows and arrows to daggers, as was common in Moravia and Austria at this time. When bronze first appeared in southwestern Slovakia, its distribution was restricted to this elite.

As pointed out earlier, the Nitra group expanded in a southerly and westerly direction towards the ÚnČtice area in Moravia and Austria. Whether any expansion was directed eastward is not possible to say. There are surprisingly few finds east of the Nitra river. Although it is difficult to talk about major economic and political centres in the Early Phase of the Nitra horizon, sites such as Nitra, Mýtna Nová Ves and Holešov were settled early and grew quickly in size. With the advent of the Classical Phase it is possible to discern sites that held an important position judging by the wealth invested in graves, the number of hunterwarriors interred and the elaboration of graves for the elite. In the north there is Holešov in Moravia and Mýtna Nová Ves in Slovakia, which is situated close to the copper ores in that area. In the middle Nitra basin lie Jelšovce and, further south, Branþ which presumably based its wealth on access to trade and arable land.

The dynamics of change are to be found within Nitra society itself, although its local power structure might have been challenged or enhanced by contacts with foreign regions. As outlined earlier, society rested upon three economic foundations: 1.

domestic production and control of arable land,

102

ANDREAS HÅRDE: THE EMERGENCE OF WARFARE IN THE EARLY BRONZE AGE: THE NITRA GROUP The enemy

Close combat might have been relatively uncommon, initiated by a daring individual during a fight rather than a concerted attack. This may be the reason why so many appear to have survived close combat. The assailant would be on dangerous ground while launching such an assault with little time to finish the attack once the foe's comrades came to the rescue. Although this war strategy is unlikely often to have been lethal, there were periods when people were driven away from their homes by force. This can be seen at Branþ, demonstrated by the occurrence of grave plundering. War was always a power struggle even if it was not always deadly.

Bátora (1991: 94; 1999a: 46) suggests that the greatest threat to the Nitra group came from neighbouring regions. The highest frequencies of weaponry are found at ŠaĐa I, Tvrdošovce, Holešov and Mýtna Nová Ves, which are all located on the borders of the area occupied by the Nitra group until the late Classical Phase. This should be compared to Branþ and Jelšovce where weapons are relatively fewer. At Výþapy-Opatovce weapons are almost absent. Since there is little knowledge about the people living in neighbouring areas at this time I assume that war was a driving factor within the Nitra group itself.

At Holešov and Mýtna Nová Ves the many weapons indicate that hunter-warriors were important to an extent not recognized elsewhere in this cultural group. Together with multiple traces of deadly wounds, their presence suggests that the reality of war was different in this northern part of the Nitra group. Whom the inhabitants of Mýtna Nová Ves fought is not known.

Furthermore, the frequency of hunter-warrior graves and weaponry on many burial sites indicates internal conflict between different areas over the control of resources rather than a concerted defence against external forces. It must also be emphasized that the occurrence of weapons alone does not indicate a warlike population. At Branþ some hunter-warrior graves do not contain weapons but the high frequency of injuries indicates that this population suffered violence. The presence of cenotaphs and evidence of grave-robbing may also indicate conflict. The political nature of warfare must also be taken into account, such as the organization of hunter-warriors and their relationship to high-ranking groups.

On an overall basis war tactics in the north must have been similar to those in the south. Raids and ambushes by archers were also standard practice but it seems that struggles between high-ranking individuals were more hazardous and that control of land and resources was very important. It is striking that high-ranking hunter-warriors show multiple peri-mortem wounds and that this trend occurs together with the emergence of cenotaphs. As mentioned earlier, the death of a high-ranking hunterwarrior might have been a principal aim of the fighting.

Survival or death: strategies of war When it comes to strategies of war, two radically different phenomena can be discerned. On the one hand there are sites with a high frequency of weapons as well as an increased number of individuals who suffered deadly violence. On the other hand there are a number of sites where arms are relatively few and where the frequency of trauma is high, but not of a deadly nature. Mýtna Nová Ves belongs to the former category and Branþ to the latter.

This northern part of the Nitra region with its access to copper ore was an important area and it is not surprising that the burial sites here are both large and wealthy. The differences between regions in the number of highranking graves and the frequency of hunter-warrior burials should also be seen as indications of power centres and their subordinated peripheries.

The trauma pattern from Branþ indicates that killing an enemy was not the prime goal in war. Rather, raiding and ambushing must have been the strategy in order to steal livestock or other resources. Skirmishes between hunterwarriors alone might also have been on the agenda but the many fractures amongst women and older men are perhaps a sign of attacks on settlements rather than encounters between hunter-warriors on battlefields away from home.

Conclusion Although it is difficult to isolate the specific causes of war in the Nitra group it is clear that they are connected with socio-political changes. The environment of the Nitra river enabled people in this region to develop a rich society in comparison to neighbouring areas. The river served as a connection for distributing prestige goods and commodities between the Hungarian plain and Little Poland whilst the fertile soils and access to copper ore made it possible to accumulate wealth and sustain a large population.

Since arrowheads are the most common armaments found in burials it may be assumed that war was waged mainly at a distance and that close combat was rare. Yet this does not accord with the fact that the majority of all traumatic injuries come from weapons such as axes and clubs. These were used in order to close in upon the enemy but were never identified as a proper weapon to be displayed in the burial ritual. Warriors in the Nitra group were displayed in death as hunters and not as closecombat fighters.

The rise of warfare took place during a period of social transformation and change in settlement pattern. It can be said that the presence of a hunter-warrior group within society instigated these changes. In the beginning this group would have operated within their own society, building power through control of the flow of goods and resources that passed through or were directed from the 103

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY in Mýtna Nová Ves. Praehistorische Zeitschrift 74: 58-67. Novotná, M. 2001. Angriffs- und Symbolwaffen der älteren Bronzezeit in der Slowakei (Offensive and symbolic weapons of the Early Bronze Age in Slovakia). In A. Lippert, M. Schultz, S. Shennan and M. Teschler-Nicola (eds) Mensch und Umwelt während des Neolithikums und der Frühbronzezeit in Mitteleuropa: Ergebnisse interdisziplinärer Zusammenarbeit zwischen Archäologie, Klimatologie, Biologie und Medizin. Rahden (Westf.): Verlag Marie Leidorf GmbH. 317-23. Ondráþek, J. 1972. PohĜebištČ nitranské skupiny HolešovČ (Das Gräberfeld der Nitra-Gruppe in Holešov). Archeologické rozhledy 24: 168-72. Ondráþek, J. and Šebela, L. 1985. PohĜebištČ nitranské skupiny v HolešovČ (katalog nálesĤ). Studie Muzea KromČĜíška 1985: 2-130. Pavúk, J. 1981. Die ersten Siedlungsfunde der Gruppe Cháopice-Veselé aus der Slowakei. Slovenská Archeológia 29: 163-75. Shennan, S.E. 1975. The social organisation at Branþ. Antiquity 49: 279-88. Stuchlík, S. 1990. Die sekundären Eingriffe in den Gräbern der ÚnČticer Kultur. Anthropologie 28: 15967. Stuchlík, S. 2001. Die Besiedlung Ostmährens durch die Aunjetitzer Kultur und den epischnurkeramischen Komplex zu Beginn der Bronzezeit (The settlement of eastern Moravia by the ÚnČtice culture and the Epicorded Ware complex at the beginning of the Bronze Age). In A. Lippert, M. Schultz, S. Shennan and M. Teschler-Nicola (eds) Mensch und Umwelt während des Neolithikums und der Frühbronzezeit in Mitteleuropa: Ergebnisse interdisziplinärer Zusammenarbeit zwischen Archäologie, Klimatologie, Biologie und Medizin. Rahden (Westf.): Verlag Marie Leidorf GmbH. 221-30. Toþík, A. 1963. Die Nitra-Gruppe. Archeologické rozhledy 15: 716-74. Toþík, A. 1979. Výþapy-Opatovce - a ćalšie pohrebiská zo staršej dobz bronyovej na juhozápadnom Slovensku - und weitere altbronzezeitliche Gräberfelder in der Südwestslowakei. Nitra: Archeologický ústav Slovenskej Akadémie Vied. Toþík, A. and Vladár, J. 1971. PrehĐad bádania v problematike vývoja slovenska dobe bronzovej (Übersicht der Forschung in der Problematik der bronzezeitlichen Entwicklung der Slowakei). Slovenská Archeológia 19: 365-422. Vladár, J. 1973. Pohrebiská zo staršej doby bronzovej v Branþi (Gräberfelder aus der älteren Bronzezeit in Branþ). Bratislava: VydavateĐstvo Slovenskej Akadémie Vied. Vladár, J. 1974. Die Dolche in der Slowakei. München: C.H.Beck'sche Verlagsbuchhandlung. Vladár, J. and Romsauer, P. 1999. Zur Problematik der Aunjetitzer Kultur in der Slowakei. In J. Bátora and J. Peška (eds) Aktuelle Probleme der Erforschung der Frühbronzezeit in Böhmen und Mähren und in der

Nitra basin. As communities grew larger and new goods appeared from the ÚnČtice area in Moravia and Austria, a change was at hand. Affiliation with the ÚnČtice people was necessary to maintain dominance over the spread of trade goods and perhaps also to recruit allies in wars against neighbouring elite families. It was important for the elite to strengthen the existing social structure which legitimized their power but at the same time they adopted new social strategies to control new commodities, as can be seen at Branþ. By adopting these new prestige goods and new ways of living the elite could consolidate their power yet further. Bibliography Bátora, J. 1990. The latest knowledge on the burial rite of the people of the Nitra group. Anthropologie 28: 16974. Bátora, J. 1991. The reflection of economy and social structure in the cemeteries of the Cháopice-Veselé and Nitra cultures. Slovenská Archeológia 39: 91-142. Bátora, J. 1999a. Waffen und Belege von Kamptreffen während der Frühbronzezeit in der Südwestslowakei. In J. Bátora and J. Peška (eds) Aktuelle Probleme der Erforschung der Frühbronzezeit in Böhmen und Mähren und in der Slowakei. Nitra: Archäologisches Institut der Slowakischen Akademie der Wissenschaften Nitra. 41-52. Bátora, J. 1999b. Symbolische(?) Gräber in der älteren Bronzezeit in der Slowakei. In J. Bátora and J. Peška (eds) Aktuelle Probleme der Erforschung der Frühbronzezeit in Böhmen und Mähren und in der Slowakei. Nitra: Archäologisches Institut der Slowakischen Akademie der Wissenschaften Nitra. 63-73. Bátora, J. 1999c. Gräber mit Totenhäusern auf frühbronzezeitlichen Gräberfeldern in der Slowakei (Beitrag zu Kulturverbindungen zwischen Mittel-, West- und Osteuropa). Praehistorische Zeitschrift 74: 1-57. Bátora, J. 2000. Das Gräberfeld von Jelšovce / Slowake: ein Beitrag zur Frühbronzezeit im nordwestlichen Karpatenbecken. Kiel: Verlag Oetker/Voges. Chropovský, B. 1960. Gräberfeld aus der älteren Bronzezeit in VeĐký Grob. In B. Chropovský, M. Dušek and B. Polla (eds) Pohrebiská zo staršej doby bronzovej na Slovensku I - Gräberfelder aus der älteren Bronzezeit in der Slowakei I. Bratislava: VydavateĐstvo slovenskej akadémievied Bratislava (Verlag der slowakischen Akademie der Wissenschaften). 11-136. Dušek, M. 1969. Bronzezeitliche Gräberfelder in der Südwestslowakei. Bratislava: Verlag der slowakischen Akademie der Wissenschaften. Furmánek, V., Veliaþik, L. and Vladár, J. 1999. Die Bronzezeit im slowakischen Raum. Rahden: Verlag Marie-Leidorf. Jakab, J. 1999. Anthropologische Analyse der Gräber mit Totenhäusern des frühbronzezeitlichen Gräberfeldes

104

ANDREAS HÅRDE: THE EMERGENCE OF WARFARE IN THE EARLY BRONZE AGE: THE NITRA GROUP Slowakei. Nitra: Archäologisches Institut der Slowakischen Akademie der Wissenschaften Nitra. 127-36.

Vyhnánek, L. and Hanulík, M. 1971. Multiple cutting wounds on the calva. Anthropologia: Acta Facultatis Rerum Naturalium Universitatis Commenianae 17: 199-203.

105

Warfare, redistribution and society in western Iberia Eduardo Sánchez-Moreno Departamento de Historia Antigua, Universidad Autónoma de Madrid

procedure enacted within a broad socio-cultural system, never an isolated act. Regardless of the perspective of analysis (whether anthropological, psychological, historical or military), there can be no doubts that war is a key element in the construction processes of complex societies, as attested to by the literature (Fried et al. 1968; Carneiro 1970; Webster 1975; Lewis 1981; Ferguson 1984; Ferguson and Farragher 1988; Earle 1997: 106-10; Haas 1990; 1999; Keeley 1996).

Among many possible perspectives, warfare is reviewed here primarily as a mechanism of contact generating a series of social and economic effects. Taking into account written sources and certain anthropological models, together with the archaeological evidence for wealth distribution in Late Iron Age cemeteries, the role played by redistribution of booty and military tribute (i.e. the acquisition of goods as result of violent exchanges) is considered in relation to the social and political articulation of the peoples of western Iberia during the Iron Age. The way in which the warrior chiefs, who usually led these acts, organized beneficial distribution among their peoples is reflected in anecdotes in the literary record. Yet at the same time, because these acts overlay a process of social ordering, their presence is a useful testimony to the existence of a highly hierarchical society. The image of Viriathus, the famous Lusitanian chief presented to us in the Classical literature, is used as a paradigm of ‘redistributive chief’ in our reflections on warfare and its meaning in the pre-Roman culture of the Iberian peninsula.

Through war people make contact. There are many examples, all of them historical constants that bring together two different entities in a violent way: the assault of a caravan, the stealing of flocks, the siege of a city, the conquest of agricultural lands, the clash of armies, and so on. Important outcomes derive from these interactions. So, for instance, war spoils, military triumphs, the imposition of tributes, and the acquisition of fame and prestige by warriors can all be understood as commodities resulting from negative or unilateral exchange, under the control of the victorious element. Another example would be employment of mercenaries, an activity generating important cultural changes in terms of acculturation, as is well known for different Iron Age societies, in Iberia (Quesada 1994a), Italy (Tagliamonte 1994) and the Celtic world (Nash 1985; Szabó 1991; 1995).

Introduction The purpose of this study is to apply three interrelated aspects to protohistoric Iberia: (1) the notion of warfare as cross-cultural mechanism, (2) its economic effects, and (3) its further impact on the social dynamics of Iron Age communities in western Spain. In short, I shall focus on warfare as a means of economic enrichment and as a strategy for social articulation by examining two sets of actions: the acquisition of booty or war spoils, and the way they are divided amongst local groups by the ruling élites. In terms of the regional and chronological framework, I shall concentrate on the pre-Roman cultures of central and western Iberia (where ancient writers locate peoples such as the Celtiberi, Vettones, Vacceai and Lusitani) dating from approximately the fourth to the second centuries BC, just before the Roman conquest of the Iberian peninsula (Fig. 1). In terms of the methodology, certain written sources (very scarce, to be honest, but otherwise a useful testimony when used critically) can be used to augment the archaeological evidence, principally data from Iron Age cemeteries. Furthermore, anthropological models help to provide a background for contrasting behaviours and ideas.

These interactions often explain the presence of imported items (such as weapons, jewels and coins), as well as the adoption of new activities and behaviour (whether social, technical or aesthetic) and certain ideological representations of power, authority and status by local communities. Such innovations can result, of course, from other interactions such as the exchange of diplomatic gifts or trading relations (Sánchez-Moreno 1998a: 580-90; forthcoming). The multifarious faces of hostility There is huge variety in the conflicts of the protohistoric period. As a general rule all of them share the aim of obtaining booty and related benefits, together with certain other objectives. As noted above, the outcome is a negative or one-sided interchange in which the defeated lose their possessions, suffer from pillage and a reduction in territory and may be obliged to pay tribute to the victors by surrendering prisoners, cattle, crops, arms, etc. One of the main reasons for human aggression is control of natural resources. The goals for attackers, as well as the priorities for the defenders, are thus usually infrastructural (livestock, harvest fields, towns, supply centres and frontiers) in conjunction with other intimate values such as family protection or religious fervour.

Warfare as a distinctive cross-cultural mechanism Among many possible meanings, warfare is considered here primarily as mechanism of intersocial interaction: ‘the most solemn means of contact among societies’, in the majestic definition of the German writer E. Jünger writing about a century ago. It is obvious that war is a 107

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY

FIGURE 1. Pre-Roman peoples of the Iberian peninsula (after Almagro Gorbea and Ruiz Zapatero 1992)

The need to protect animals and drove roads is a key factor for understanding the internal organization of Iron Age communities in central and western Iberia. There is evidence to prove the crucial role played by livestock: not only was this a very suitable environment for stockraising but abundant information is available from archaeology, written sources and ethnography. Graeco-Roman writers underline the very large numbers of animals in the Iberian economy (Blázquez 1957; 1969; Salinas 1993). This is so much the case that some groups often had to resist raids from other peripheral tribes in need of crops and flocks (Livy 21, 43, 8-9; 25, 1; Virgil, Georg. 3, 406-8; Florus II 33, 46-7; Orosius VI 21, 3), even under the form of ritualized customs such as cattleraiding, well documented both ethnographically and in literary sources (Lincoln 1976; 1981). It can be connected to the formation of groups of young warriors or iuvenes who, under the protection of a divinity and sacred symbols (e.g. animals), are obliged to perform an initiation process or rite of passage including the theft of cattle. By means of this rite of passage, they could be

FIGURE 2. Celtiberian pottery representing warriors (late first century BC) (Museo Numantino de Soria)

inducted as male citizens into their communities (or, eventually, found new settlements), and could enjoy having the benefit of juridical and economic rights (Peralta 1990; Ciprés 1993: 136-58; García Moreno 1993: 349-50; Sopeña 1995: 75-86; García Quintela 108

EDUARDO SÁNCHEZ-MORENO: WARFARE, REDISTRIBUTION AND SOCIETY IN WESTERN IBERIA

FIGURE 3. Zoomorphic sculpture (verraco) from Las Cogotas (Cardeñosa, Ávila), one of the most characteristic cultural indicators of the Vettonian territory

FIGURE 4. The so-called ‘Bulls of Guisando’ (El Tiemblo, Ávila)

1999: 278-87). In a similar way to the role played by the ver sacrum among the tribes of archaic Italy (Heurgon 1957; Caro 2000), this custom may constituted an ideological mechanism behind social practices so widely reported by Classical authors: brigandage, raids, military expeditions and migrations, especially when talking of the Lusitanians (Diodorus of Sicily 5, 34, 6-7; Strabo 3, 3, 5).

Raids were probably the most characteristic form of aggression among the peoples of inner Iberia, at least in the view of ancient historians. It is clear that GraecoRoman literature contrasts two opposing categories of warfare: on one hand the ‘primitive model’ (raids, improvised guerrilla attacks and sudden assaults - the proper thing for barbarians such as Celtiberians and Lusitanians in our case) and, on the other hand, the ‘civilized model’ represented by Mediterranean powers (Quesada 1997a: 45-50), namely the rational way of fighting employed by the Greek, Punic and Roman armies recognizable to us as akin to the western mode of set-piece battles (Dawson 1998; Hanson 1999).

Faunal remains recovered from Iron Age sites also indicate high levels of consumption of cattle and, above all, sheep and goats (Castaños 1998; Morales and Liesau 1995; Liesau 1998; Blasco 1999) whose secondary products (wool, meat and milk) may have been subject to specialized exploitation. In addition, many of the widewalled sectors of major settlements seem to be enclosures for keeping livestock, if not wholly at least partially as is indicated in the Vettonian oppidum of Las Cogotas (Cardeñosa, Ávila; Ruiz and Álvarez-Sanchís 1995a: 220-2). Finally, in the iconography of material culture of Celtic Spain, domestic animals are represented widely in the motifs used in the decoration of pottery, terracotta, weapons, fibulae, belt-buckles and stone sculptures (Figs 3 and 4; Blanco Freijeiro1988; López Monteagudo 1989; Galán 1989-90; Alonso and Benito-López 1991-92; Romero and Sanz 1992; Álvarez-Sanchís 1994; Blanco García 1997; Cerdeño et al. 1999).

We should, however, beware of oversimplifying this kind of comparison of modes of warfare because it is little more than a stereotype contrasting the unreasonable bellicosity of western barbarians with the idea of bellum iustum (just war) so well illustrated by Rome as to justify the military conquest of the Iberian peninsula (García Quintela 1990; 1999: 29-72; Ciprés 1993: 35-50; Salinas 1999). It is easy to find in nineteenth-century colonial expansion very similar comparisons made between ‘native’ belligerence and western formalized war, in this case understood as an excuse to legitimize colonialism (Ferguson and Whitehead 1992). The construction of defensive systems in hillforts and oppida comprising walls, towers and bastions, chevaux de frise and moats (Fig. 5; Moret 1991; Almagro 1994; Cerdeño 1997), the presence of weapons in tombs (Cabré and Baquedano 1997; Lorrio 1993; Quesada 1997b) and certain literary references about the Iberian warlike character (García Huerta 1997), confirm - I would say partially - the existence of a significant level of military confrontation. The risk of stereotyping these Celtic societies means that we must avoid uncritical readings, since warfare is an ideologically loaded subject and overestimated in the historiography, both ancient and modern. Methodologically, we should not overlook three key aspects:

All this evidence underlines the centrality of cattle in the modus operandi of these Iron Age communities. Indeed livestock was a valuable item in a double sense: for external groups who appreciated herds and flocks as goods to be captured, and for herding and farming communities who used their surplus to open trading exchanges by means of practices such transhumance, the seasonal displacements of flocks between complementary regions (Sánchez-Moreno 1998b). Cattle were the foremost item of value: their possession conferred social status, provided for élite expressions of power and underlay the need to safeguard herds and drovers’ roads (Earle 1997: 100-2; Almagro 1997: 208; García Quintela 1999: 278-82). 109

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY

FIGURE 5. Walls and chevaux de frise from the Vettonian oppidum of Yecla la Vieja (Yecla de Yeltes, Salamanca)

Fortified sites of monumental character as ideological indicators of power and as landmarks revealing control over territories (Parcero 1995: 135-7). The symbolic meaning of arms in funerary contexts. It is true that weaponry gives the idea of a warlike reality but, as well as being functional, these are selected items indicating status and rulership (Baquedano and Cabré 1997). The Classical construction of the ethno-type of the barbarian, what Latin historians designated feritas celtica (Rankin 1987; Marco 1993; Sopeña 1995: 80-5; Clavel-Lévêque 1996), needs to be appreciated in order to decode the aggressive image of the Iberians reported by Classical texts, as argued above.

Celtic regions (Brunaux and Lambot 1987: 27-8; Cunliffe 1999: 89-99).

The result of the struggle: booty, tribute, and prestige

On the other hand, ideological consequences of war can be equally significant, for instance the fame and prestige gained by victorious leaders which thereby increases their power and legitimacy. At this point warfare becomes a channel for socio-political promotion, as can be observed from ethnographic contexts all over the world (Mishkin 1992; Carneiro 1990; Earle 1991; 1997).

1.

2.

3.

Another sign of success was the imposition of regular taxation on the defeated party, understood as the price of peace (Orosius, Hist. 5, 1, 10-1). The winners could guarantee this by employing coercive means, such as the taking of hostages, to compel the enemy into becoming a tributary community (García Riaza 1997). In protohistoric times these taxes were paid in local units of wealth: horses, cattle, grain, salt, manufactured items, prisoners, metals, etc. In fact this remained the way in which indigenous polities paid taxes to Roman governors during the following decades (Pitillas 1996; García Riaza 1999), as in the case of the Celtiberians presenting thousands of woollen cloaks as military tribute (Appian, Iber. 48-54, 77-9 and 98; Diodorus of Sicily 33, 16).

Warfare is an endemic, stabilized factor in ancient cultures and its outcomes are extremely interesting for the light they throw on the socio-political dynamics of preRoman Iberia. A variety of achievements result from warlike practices and, what is more, they produce important repercussions for local communities.

Wealth distribution and society

The distribution of the spoils of war amongst the population and its social impact offer an extraordinary insight into social organization. On the one hand there were the material profits, for example the capture of cattle, crops and treasures. These items of wealth were the target of ordinary attacks and determined in most cases the military strategy displayed in the ancient Mediterranean, as is vividly confirmed by many episodes from the Greek world (Garlan 1989: 50-5; Ducrey 1999: 201-18; Corvisier 1999: 148-54; Rouveret 2000) and the

The introduction of foreign merchandise - whether as plunder from war or from other sources - into local communities and the way in which it is shared create certain kinds of interaction. One of these which has been much discussed is redistribution from a central authority (Fig. 6). In the definition given by Karl Polanyi (1957: 250-4) redistribution designates appropriational movements toward a centre and out of it again; it is dependent upon the presence of some measure of centricity in the group. 110

EDUARDO SÁNCHEZ-MORENO: WARFARE, REDISTRIBUTION AND SOCIETY IN WESTERN IBERIA

FIGURE 6. Graphic reconstructions of the redistribution model (after Renfrew 1975 and Renfrew and Bahn 1991)

settlement detected in some ethnic territories of western Iberia (Fig. 7), such as those of the Lusitanians (Rodríguez Díaz 1995; Ortiz and Rodríguez 1998), Vettones (Álvarez-Sanchís 1999: 111-64; 334-6; 2000: 77-85; Sánchez-Moreno 2000: 75-87) and Vacceans (Sacristán 1994; 1995; San Miguel 1993; Delibes et al. 1995: 105-6). So oppida can be regarded in these regions as political capitals and, at the same time, redistributive cores.

In contrast to other more balanced mechanisms of exchange, such as barter or reciprocity, redistribution shapes a hierarchical formulation because it presupposes the existence of an allocative centre in the community. In other words, the allocation of goods is manipulated by an existing authority - it takes place by virtue of might, custom, law or ad hoc central decision (Sahlins 1972: 190-5; Renfrew 1975: 8; 11-2; Pryar 1977; Service 1975: 75-9). From this centre, whether a place or a personal powerbase, resources are controlled, production is organized, wealth is distributed, and exchange networks with external units are set in place. In short, that central referent is usually the head of the community through which economic, social and political structures are displayed.

Moving from a territorial scale to a human level, it could be stated that the leaders of pre-Roman administrative units (identifiable as the oppida, forming central places) are the ones who both ruled and had access to restricted resources, including war benefits. From the privileged position of that warring aristocracy (Fig. 2), society was assembled with an increasingly hierarchical disposition as a result of an unequal distribution of wealth and status. This can be understood clearly through explanatory models such as redistribution.

This model has been properly connected to concepts such as central place theory (Grant 1986), archaic complex societies (according to Polanyi), developed chiefdoms (in line with Sahlins, Service, Harris and Earle) and the socalled early state module (after Renfrew). Likewise, it has been extensively applied to Late Iron Age oppida from temperate Europe (Nash 1978; Collis 1984; Crumley 1987; Wells 1984: 135-42; Büchsenschütz 1995: 61-3; Cunliffe 1999: 223-34) but criticized by Woolf (1993a; 1993b).

The role of the leader and social articulation: Viriathus the Lusitanian as paradigm There is a historical figure, shrouded in myth and legend, who provides an instance of how important the allocation of spoils was in western Iberia. Again the idea of distribution is that of an act led by a military chief (the ultimate representation of the previous notion of a central place) who employed it as a means of controlling the social order. I am referring to the famous Viriathus, commander of the Lusitanians in their fight against the Romans from 150 to 139 BC (de Francisco 1989: 57-75; Pastor 2000). He may be called a ‘redistributive chief’ in this context.

Turning to our region, similar concepts can be applied to the interpretation of the major settlements of the Meseta in central Spain (Almagro 1994; 1996; 1999). From the fourth century BC onwards some of these oppida are focal sites for the settlement and defence of populations, as well as focal places for political and religious institutions. In addition they are locations for production and specialization, controlling and servicing wide economic territories and their minor dependent settlements. This picture of central sites spaced regularly across regional zones corresponds well with the pattern of

Viriathus, whose fame has spread far and wide, has always attracted the attention of scholarship (Schulten 111

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY manipulated at different times in the history of Portugal and Spain (Garcia 1985; García Moreno 1989; Guerra and Fabiâo 1992), a late example of this being during Franco’s rule. His image has been an inspiration in literature and the plastic arts, especially in nineteenthcentury historical painting (Fig. 8; Quesada 1994b: 44-6; 1996: 225-8), as well as being used as a symbol in teaching. The portrait of Viriathus in heroic attitude constituted a celebrated image in Spanish schoolbooks for decades (Ruiz and Álvarez-Sanchís 1995b). Leaving all this aside, one of the questions less studied in works devoted to Viriathus is that of the distribution of booty. This is because of the prominence given to other subjects: his battles and triumphs over Roman armies (Gundel 1968; García y Bellido 1977: 26-8, 45-9; Rubinsohn 1981; Santos and Montero 1983), his geographical provenance (García Moreno 1988; Pérez 1989a), the nature of his power (López Melero 1988; García Quintela 1999: 177-222), his movements across the Iberian heartland (Kindelan 1958; Pérez Vilatela 1989b), his natural magnanimity in the context of the Stoic philosophical convictions concerning the noble savage (Lens 1968) or his exemplification of an IndoEuropean ideological triad (García Quintela 1993). More anecdotally, particularly well-known aspects of his story are the circumstances of his wedding and of his murder by treacherous friends who accepted a bribe from the Romans. However, to return to the distribution of spoils, Classical historians do refer to Viriathus providing for his followers.

FIGURE 7. Pattern of settlement in the Vettonian territory (Amblés valley, Ávila province): oppida as central places, minor sites and zoomorphic sculptures throughout the territory (from Álvarez-Sanchís 1999: 118, fig.35)

‘Viriathus, the Lusitanian robber-captain, was scrupulous in the division of spoils: he based his rewards on merit, making special gifts to those of his men who distinguished themselves for bravery, and took for his own use not one thing belonging to the common store. In consequence the Lusitanians followed him wholeheartedly into battle, and honoured him as their common benefactor and saviour’ Diodorus of Sicily (33, 1, 5) ‘So great was the longing which Viriathus left behind him - a man who, for a barbarian, had the highest qualities of a commander, and was always foremost in facing danger and most exact in dividing the spoils. He never consented to take the lion’s share, although always asked to do so, and even the share which he did receive he divided among the bravest. Thus it came about (a most difficult task and one never achieved easily by any commander) that in the eight years of this war, in an army composed of various tribes, there never was any sedition, and the soldiers were always obedient and ready for danger.’ Appian (Iber. 75)

FIGURE 8. ‘The death of Viriathus, commander of the Lusitanians’, oil painting by J. de Madrazo (c. 1818) (Casón del Buen Retiro/Museo del Prado, Madrid)

1917). His historiographical treatment has changed according to ideological and political influences, both past and present (Alvar 1997). In the same way as Vercingetorix in France, Boudica in Britain and other well-known characters from antiquity traditionally associated with the idea of independence and the roots of nationalism, the story of Viriathus has often been

Behind this idealized account it is possible to observe, I think, how military conquests generate a redistribution 112

EDUARDO SÁNCHEZ-MORENO: WARFARE, REDISTRIBUTION AND SOCIETY IN WESTERN IBERIA (1) Military clients were so dedicated to their leaders or commanders that they were prepared to die for them. This is the famous fides or devotio (Rodríguez Adrados 1946; Prieto 1978; Ciprés 1993: 126-9; Dopico 1994), an Iberian custom that astonished Graeco-Roman historians (Strabo 3, 4, 18; Valerius Maximus 2, 6, 11; Dio Cassius 52, 20, 2; Plutarch, Sert. 14, 5-6). We have an echo of this in the picture of the Viriathus’ splendid funeral (Diodorus of Sicily 31, 21a; Appian, Iber. 74-5; Livy, Perioch. 54) where two hundred pairs of warriors (presumably attached to him as clients or devotii) had gladiatorial combats in his honour in front of the funerary pyre (see more generally Blázquez and Montero 1993). (2) The occurrence of feasts of merit and related acts of social exhibition and rivalry (Fig. 2), including exchanges of gifts and deliberate destructions of wealth as the supreme prerogative of power (comparable to the potlatch ceremony well known in the ethnographic literature). These seem to be practices amply developed by the aristocracy in the protohistoric period, among Iberians (Quesada 1995), Celtiberians (García Moreno 1993: 331-6) and, especially, Celts (Feuvrier-Prévotat 1978; Lewuillon 1992; Champion 1995: 93-4; Cunliffe 1999: 105-7; Poux 2000).

system affecting the hierarchical ordering of warrior groups and other members of the community. Simultaneously, we can see how leaders emphasize their power in accordance with military success: the more distant the battlefield, the stronger the enemy, the more powerful the symbolism to be derived from one’s exploits, the greater the prestige and fame for the victorious chiefs. My proposal is no more than a suspicion, a practical hypothesis on the Viriathus described (I would say metaphorically) by ancient writers. In fact his portrait is close to that of an ‘Iron Age Robin Hood’ (Hobsbawn 1985: 41-56), a bandit vindicated by reason of his egalitarian and honest behaviour, following the archetype of the noble barbarian (Lens 1968). Yet, aside from anecdotes and moral judgements, these and other literary references allow us to ask about the nature of power in Iberian protohistory and to seek new evidence concerning the social implications of warfare. Before turning to archaeology, I wish to keep in mind the presence of western Iberian commanders fighting against Roman armies from the beginning of the second century B.C. onwards. They are qualified by Graeco-Roman terms such as dux, imperator, hégemon or dinastés. Their powers seem to be superior to that of an ordinary and elective magistracy, but lower than the royalty of the Iberian kings (reges and reguli) from the southern regions (López Domech 1986-87; Muñiz 1994a: 286-9; 1994b; Pitillas 1997).

There is a final, valuable piece of evidence recoverable from the Viriathus story: the management of fortunes made by the élite as a means of gaining followers and guaranteeing their fidelity and discipline (Appian, Iber. 75; Diodorus of Sicily 33, 1, 3-5; 33, 21a). Here we find more pieces for our particular puzzle:

Written sources emphasize the warlike character of the indigenous peoples because they are reporting a military situation, the time of the Roman conquest. Yet it is probable that local leaders had not only military prerogatives but economic power as well, which gave them access to surplus. Furthermore they appropriated symbols of authority in order to legitimize their power, making use of ideological strategies of manipulation over the population when necessary.

x x x

economic fortune: the spoils of war leader’s actions: the redistributive chief. social dialectics: the establishment of subordinate lines radiating from the top of the hierarchy.

What emerges is the functional value of gifts and the construction of social links through the exchange of presents and reciprocal obligations (Mauss 1990). From my point of view, since Mauss published his essay in 1925 this has been one of the most useful anthropological theories for historians and archaeologists (more recently see van Wees 1998; Godelier 1999). As a rule the ‘economy of the gift’ can be identified with certain political aims and thereby associated with aristocratic groups, princely contexts and complex chiefdoms; this harmonizes well with what we know of Iron Age communities (Lewuillon 1992; Sánchez-Moreno 1998a: 581-4, 697-707; Muñiz 1998). In those circles parade weapons, trophies and other spoils from the enemy torcs, cauldrons, luxury pottery such as Greek vases, horses, women in exogamy - were exchanged as prestige items within a selective and clientary circulation system.

As Timothy Earle has recently stated, by controlling the production and distribution of staples and prestige goods, chiefs invest surplus so as to control military might and ideological right. To the degree that leaders control staple production that supports warriors and priests and control the specialized manufactured of their weapons and symbolic objects, military intimidation and religious sanctity belong to the rulers (Earle 1997: 207). To sum up, chiefly control over critical nodes of distribution in the flow of the economy’s material goods (including the spoils of war) translates into control over the many fields of political action. In the image of Viriathus dividing war spoils one can sense a deeply regulated social order. It fits the other clues about hierarchy and dependence that we find in preRoman Iberia. It also shows how society was developing in an increasily ritualized atmosphere. Two patterns of behaviour well attested in western Iberia can be mentioned in this context:

It is unnecessary to stress the old adage that ‘the more presents you give, the stronger you are’. Thus, the 113

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY and quality of the funerary finds. Thus, the most conspicuous graves including weapons (conventionally and perhaps inappropriately called ‘warrior graves’) are underneath these tumuli.

accumulation of many of these prized goods meant a spreading of social relations and a practical means of consolidating the rank of the possessors. Even more valuable were objects of external provenance (for example war booty obtained abroad, something Viriathus and certain Vettonian commanders acquired frequently when fighting and/or interacting in foreign regions; Sánchez-Moreno 2000: 220-3), because the commodities coming from a distance and the knowledge derived from outside explorations further reaffirmed their excellence and prestige (Helms 1988). Examining the funerary record: warrior graves and social structure Finally let us turn to the funerary record of western Iberia’s Iron Age with the aim of confronting some of the previous ideas. To be exact, the well-studied cremation cemeteries from one of the principal pre-Roman tribes of this region, the Vettones, will be considered (ÁlvarezSanchís 1999: 169-98, 295-303; see Sánchez-Moreno 2000: 87-106, 235-40 for earlier references). They date from the early fourth century BC to the middle of the second century BC and are quite extensive in terms of the spatial dispersal and number of tombs: some of the cemeteries contain more than 1,500 graves. Although cremation was the prevalent rite, other funerary rites which left no material traces were probably practised, such as the expositio of warriors’ corpses on stone platforms as an offering to vultures and other birds of prey (recorded by Classical authors with regard to Celtiberians and Vacceans; Silius Italicus, Pun. 3, 340-3; Elianus, De Nat. An. 10, 22), or the dropping of the dead into water sources following ancestral Atlantic traditions. One of the most interesting features in Vettonian cemeteries is the gathering of graves into different funerary sectors separated by open land (Fig. 9). This is likely a post mortem reflection of a kind of lineage or gentilician system, which seems to have been a traditional affiliation for these groups (Sánchez-Moreno 1996a).

FIGURE 9. Funerary sectors of the Vettonian cemeteries of La Osera (Chamartín, Ávila) and Las Cogotas (Cardeñosa, Ávila) (from Álvarez-Sanchís 1999: 296, fig. 131)

In relation to grave goods it should be said that their distribution is not homogeneous: in some cemeteries they appear in just 15% of cases (Las Cogotas), elsewhere in 50% (La Osera), and even in 80% of graves (El Raso, El Mercadillo). Since there are different levels of grave goods in each of the funerary sectors, a grading in wealth and social status has been put forward. This has been applied to the well-studied necropolises of La Osera, Las Cogotas and El Raso in Ávila province, where hundreds of graves have been excavated, and the different zones have been interpreted as indicating a society stratified into diverse groups and ruled by warrior chieftains (Martín 1985: 121-3; González-Tablas 1985; Castro 1986; Kurtz 1987; Fernández 1986: 529-877; 1997; Sánchez-Moreno 1996b; Baquedano and Martín 1997).

The funerary deposits (burnt bones and ashes within ceramic urns, sometimes also enclosing grave goods) were contained in small pits or cavities in the ground, often sealed by levels of soil and stones. Above some graves was placed an external structure: stelae and stone bust barrows of circular or rectangular shape beneath which there may be one or more graves (Fig. 10). Occasionally some of the stone structures have no human cremated remains even though they contain grave goods. In zone I within the cemetery of La Osera (Chamartín, Ávila; Baquedano and Martín 1997: 179-84) what may be interpreted as cenotaphs might have been erected for warrior chiefs or mercenaries who died abroad and, around these tumuli, are dense concentrations of graves. The connection between these tumuli and social élites seems to be confirmed not only by the fact that they are prestige monuments in themselves but also by the fact that they cover the richest burials in terms of the quantity

Among the offerings placed with the burials in these cemeteries there is a wide variety of pots and containers besides the cinerary urn, arms (swords, spears, shields, 114

EDUARDO SÁNCHEZ-MORENO: WARFARE, REDISTRIBUTION AND SOCIETY IN WESTERN IBERIA

FIGURE 10. Tumuli structures from the cemetery of La Osera, zone VI (Chamartín, Ávila) (after Cabré et al. 1950)

selected by the mourners in response to certain symbolic devices and ritual behaviours. The inclusion of grave goods in tombs and their changing significances depend on the specific rules of family inheritance or on broader political and ideological currents in society (Parker Pearson 1999: 72-94). An interesting instance in this context is the hypothesis that weapons in funerary and ritual contexts are ‘man-makers’ revealing the acquisition of warrior status by initiated young men. This provides an alternative perspective to the traditional interpretation of their inclusion in graves as the prestige goods of elites (Fokkens 2001).

cuirasses, etc.), horse-gear, fibulae, ornaments made in bronze and glass, domestic or economic tools such as loom-weights, and a few ritual objects associated with feasts, hospitality and similar social practices (cauldrons, spits, thymiateria [censers]). Among these grave goods, weaponry is probably the clearest token of social distinction (Figs 11-13). Weapons appear in low proportions although not at exactly the same rate in all cemeteries, varying from being present in 3% of the total number of tombs (Las Cogotas, La Coraja) to 15% (El Raso, El Romazal I) and 25% (La Osera, Las Ruedas) (Fig. 14). Obviously ‘warrior graves’ - or rather, graves of individuals of warrior status - are not uniform and the military equipment varies from just one or two spears (the most frequent) to a few of the fullest panoplies belonging to élite warriors and including swords, spears, shields, daggers, horse-trappings and other prestige items (Figs 11-13; Álvarez-Sanchís 1999: 177).

Nevertheless, it does seem to hold true for the Iberian Iron Age that the weapons, together with horse-gear, which accompanied the burials of certain privileged members of society formed the accoutrements of rank and authority (Quesada 1997b). As a very rough estimation - and considering the previous warnings with regard to social readings based on the funerary record - it can be said that this aristocratic group represents approximately 15-20% of the whole population in each cemetery community of western Iberia. It is possible to discern a small group of individuals, around 5-10% occupying the most spectacular tombs (tumulus structures), who were leaders buried with exceptional panoplies which included parade weapons.

It has to be remembered here that funerary material culture does not always form a direct reflection of social status that can simply be reconstructed by pragmatic analysis of inequalities between funerary deposits. Grave goods are not always personal possessions advertising the position and wealth of the deceased, but are offerings

115

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY

FIGURE 11. Grave with weapons, number 78bis, from the Vettonian cemetery of El Raso, zone Las Guijas B (Candeleda, Ávila) (after Fernández Gómez 1997)

116

EDUARDO SÁNCHEZ-MORENO: WARFARE, REDISTRIBUTION AND SOCIETY IN WESTERN IBERIA

FIGURE 12. Warrior grave goods from Vettonian cemeteries in Ávila province: A) El Raso (grave 13), B) La Osera (grave 388; zone VI), C) La Osera (grave 417; zone VI), D) La Osera (grave 200; zone VI), E) La Osera (grave 228; zone VI) (from Álvarez-Sanchís 1999: 185, fig. 74)

117

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY

FIGURE 13. Warrior grave goods from Vettonian cemeteries in Ávila province: A) Las Cogotas (grave 513), B) La Osera (grave 509; zone VI), C) La Osera (grave 514; zone VI) (from Álvarez-Sanchís 1999: 188, fig. 75)

118

EDUARDO SÁNCHEZ-MORENO: WARFARE, REDISTRIBUTION AND SOCIETY IN WESTERN IBERIA

FIGURE 14. Burials including weaponry of the total excavated graves from Vettonian cemeteries (Late Iron Age) (from Álvarez-Sanchís 1999: 177, fig. 70)

They seem to have been an equestrian nobility who used horses as emblems of their power (Piggott 1986; Sánchez-Moreno 1995-96; Quesada 1997c; Almagro 1998: 112-3; Almagro and Torres 1999). In addition, they were dignitaries (shall I say ‘redistributive chiefs’?) who gave and received luxury and prestige goods. Occasionally they were buried with exotic items (such as the famous Iberian falcata or curved sword, cuirasses and belts, Greek vases, glass ornaments and ritual bronze vessels). These proclaim the extent to which external relations, whether warlike or pacific, were an incalculable source of power (Sánchez-Moreno 1998a: 397-445, 52538, 696-707). FIGURE 15. Proposed social organization according to gravegood distribution seen in the Vettonian cemeteries of Las Cogotas (Cardeñosa, Ávila) and La Osera, zone VI (Chamartín, Ávila). A) Warriors. B) Craftsmen. C) Women. D) Others. E) Burials with no grave goods (from Álvarez-Sanchís 1999: 298, fig. 132)

It is easy to detect status segregation in terms of access to necropolis burial. Behind the ruling class it is possible to distinguish an intermediate social layer constituted of relatives and clients (Figs 15 and 16). The middleranking burials, with more common grave goods, are interpreted as the graves of this segment of society. These people may be considered to be free men or citizens with full rights whose lives were dedicated to economic activities (stock raising, agriculture, craftsmanship and trade), rendering military services when appropriate. These obligations could eventually be rewarded by gifts and awards from the ruling aristocracy. Although we know nothing of the relationship between this group and the upper class, it is reasonable to think of clientship and other warlike submissions as illuminated by Classical texts. Something of this relationship seems to be corroborated archaeologically through the dispersal of middle-level burials in concentric patterns around a central princely tumulus.

FIGURE 16. Theoretical model of social organization in western Iberia (Lusitania) during the Late Iron Age (after Martín Bravo 1999: 251, fig. 109)

The lives of the rest of the society are very poorly documented. Approximately three-quarters of the population in each community probably constituted a mass of people or ‘commoners’ living in poverty with

low social status. It is a wide and scattered body composed of different families and clans. In the funerary

119

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY ‘Warfare, violence and slavery in prehistory and protohistory’ (University of Sheffield, 2-3 February 2001). This paper is presented here as it was delivered at the Sheffield conference, with minor modifications and the addition of necessary references. I should also like to thank Professor Barry Cunliffe for his advice and help during my stay as academic visitor at the Institute of Archaeology, Oxford University (1999-2001). This paper was produced during my stay at Oxford by means of a postdoctoral research project on ‘Cultural contact in European protohistory’, funded by the Spanish Ministry of Education and Culture.

record it could correspond to the simplest grave pits with no grave goods, although this assumption must be questioned. It is more likely that some individuals might not have been buried in cemeteries, perhaps because they would not have been of sufficient status even for that (Figs 15 and 16). At this point it should be noted that, as a rule, the number of graves discovered in Iron Age cemeteries does not match the population believed to have lived in the associated major settlements. We do not know whether the condition of these people as a whole was servitude. It is quite possible that both free and subjugated populations coexisted although, in my view, prisoners of war can be estimated to be the most likely subjects of slavery in these pre-Roman societies. Apart from that, the lower orders would be involved in the primary labours of herding, farming and mining, not as owners but as dependent workers or servants. At the same time they might have been employed in services for the community such as the building of the oppida’s defensive walls and ditches or recruited into warrior units as ‘cannon fodder’.

Bibliography Almagro Gorbea, M. 1994. El urbanismo en la Hispania céltica: castros y oppida del centro y occidente de la península ibérica. In M. Almagro Gorbea and A.Mª. Martín Bravo (eds) Castros y Oppida en Extremadura. Madrid: Editorial Complutense. 13-75. Almagro Gorbea, M. 1996. Los castros de la meseta. Gallaecia 14-15: 261-308. Almagro Gorbea, M. 1997. Guerra y sociedad en la Hispania céltica. In La Guerra en la Antigüedad: una aproximación al origen de los ejércitos en Hispania. Madrid: Comunidad Autónoma de Madrid. 207-21. Almagro Gorbea, M. 1998. Signa Equitum de la Hispania céltica. Complutum 9: 101-15. Almagro Gorbea, M. 1999. Estructura socio-ideológica de los oppida celtibéricos. In F. Villar and F. Beltrán (eds) Pueblos, Lenguas y Escrituras en la Hispania prerromana: actas del VII Coloquio sobre Lenguas y Culturas Paleohispánicas. Zaragoza: Institución Fernando el Católico. 35-55. Almagro Gorbea, M. and Ruiz Zapatero, G. 1992. Paleoetnología de la península ibérica: reflexiones y perspectives de futuro. In special issue ‘Paleoetnología de la Península Ibérica’ (eds M. Almagro Gorbea and G. Ruiz Zapatero). Complutum 2-3: 469-99. Almagro Gorbea, M. and Torres Ortiz, M. 1999. Las Fíbulas de Jinete y de Caballito: aproximación a las élites ecuestres y su expansión en la Hispania céltica. Zaragoza: Institución Fernando el Católico. Alonso Hernández, P. and Benito-López, J.E. 1991-92. Figuras zoomorfas de barro de la edad del hierro en la meseta norte. Zephyrus 44-45: 525-36. Alvar Ezquerra, J. 1997. Héroes ajenos: Aníbal y Viriato. In J. Alvar and J.Mª. Blázquez (eds) Héroes y Antihéroes en la Antigüedad Clásica. Madrid: Cátedra.137-53. Álvarez-Sanchís, J.R. 1994. Zoomorphic Iron Age sculpture in western Iberia: symbols of social and cultural identity? Proceedings of the Prehistoric Society 60: 403-16. Álvarez-Sanchís, J.R. 1999. Los Vettones. Madrid: Gabinete de Antigüedades de la Real Academia de la Historia.

Conclusion To recapitulate, let me stress the idea of warfare in the context of western Iberia protohistory as a complex form of behaviour conferring economic income, social promotion and political prestige. Bearing in mind that warrior chiefs were the great beneficiaries of the system, as leaders and protectors of the community, wealth flowed to them. The major rewards were the appropriation of new land and the fame and authority inherent in successful military leadership. Associated with these were the procurement and control of mobile merchandise derived from warfare, booty and tribute. New property and captured goods were then distributed to the population by means of a social circulation akin to the redistributive pattern from a central power. One particular form of division adopted by rulers was the giving of presents within a hierarchical (and even ritualized) setting, thereby regularizing bonds of dependency and reciprocity. This was even more pronounced among societies in transformation that were strongly stratified, as can be inferred from the literary and archaeological evidence relating to Iron Age Iberia. All of this meant war. This critical perspective on war and its outcomes may be the key to deciphering Dio Cassius’s statement on Viriathus, the archetypal ‘redistributive chief’ of this paper: ‘And, in fine, he carried on the war not for the sake of personal gain or power nor through anger, but for the sake of warlike deeds in themselves; hence he was accounted at once a lover of war and a master of war’ (Dio Cassius, 22). Acknowledgements I wish to express my thanks to Mike Parker Pearson for his kind invitation to participate at the conference 120

EDUARDO SÁNCHEZ-MORENO: WARFARE, REDISTRIBUTION AND SOCIETY IN WESTERN IBERIA Caro Roldán, J.Mª. 2000. Una aproximación a la naturaleza del ver sacrum. Gerion 18: 159-90. Castaños Ugarte, P.Mª. 1998. Evolución de las faunas protohistóricas en Extremadura. In A. Rodríguez Díaz (ed.) Extremadura Protohistórica: paleoambiente, economía y poblamiento. Cáceres: Universidad de Extremadura. 63-72. Castro Martínez, P.V. 1986. Organización espacial y jerarquización social en la necrópolis de Las Cogotas (Ávila). Arqueología Espacial. Coloquio sobre el Micro-espacio 3: 127-38. Cerdeño Serrano, Mª.L. 1997. Sistemas defensivos en el ámbito celta peninsular. In La Guerra en la Antigüedad: una aproximación al origen de los ejércitos en Hispania. Madrid: Comunidad Autónoma de Madrid. 231-9. Cerdeño Serrano, Mª.L., Cabanes Miró, E. and Fernández Sabugo, M., 1999. Representaciones animales en la meseta prerromana. In R. de Balbín and P. Bueno (eds) II Congreso de Arqueología Peninsular. Tomo III: Primer milenio y metodología. Zamora: Fundación Rei Afonso Henriques. 325-33. Champion, T.C. 1995. Power, politics and status. In M. Green (ed.) The Celtic World. London: Routledge. 8594. Ciprés Torres, P. 1993. Guerra y Sociedad en la Hispania Indoeuropea. Vitoria: Universidad del País Vasco. Clavel-Lévêque, M. 1996. Codage, norme, marginalité, exclusion: le guerrier, la pleureuse et la forte femme dans la barbarie gauloise. Dialogues d´Histoire Anciènne 22: 223-51. Collis, J. 1984. Oppida: earliest towns north of the Alps. Sheffield: Sheffield University Press. Corvisier, J.-N. 1999. Guerre et Société dans les Mondes Grecs (490-323 av. J.-C.). Paris: Armand Colin. Crumley, C.L. 1987. Celtic settlement before the conquest: the dialectics of landscape and power. In C.L. Crumley and W.H. Marquardt (eds) Regional Dynamics: Burgundian landscapes in historical perspectives. New York: Academic Press. 403-30. Cunliffe, B.W. 1999. The Ancient Celts. London: Penguin. Dawson, D. 1998. The Origins of Western Warfare. Oxford: Westview Press. Delibes de Castro, G., Romero Carnicero, F., Sanz Mínguez, C., Escudero Navarro, Z. and San Miguel Maté, L.C. 1995. Panorama arqueológico de la edad del hierro en el Duero medio. In G. Delibes, F. Romero and A. Morales (eds) Arqueología y Medio Ambiente: el primer milenio a.C. en el Duero medio. Valladolid: Junta de Castilla y León. 49-146. Dio Cassius. 1962. Roman History. Vol. I, frags. Bks 111. (trans. E. Cary and H.B. Foster). London: Harvard University Press, Loeb Classical Library. Diodorus Siculus. 1939. Books 4-8. (trans. C.H. Oldfather). London & Cambridge MA: Heinemann and Harvard University Press, Loeb Classical Library. Dopico Caínzos, Mª.D. 1994. La devotio ibérica: una revisión crítica. In J. Mangas and J. Alvar (eds)

Álvarez-Sanchís, J.R. 2000. The Iron Age in western Spain (800 B.C.-AD 50): an overview. Oxford Journal of Archaeology 19: 65-89. Appian. 1912. Roman History. Vol. I, Bks 1-8. (trans. H. White). London & New York: Heinemann and Macmillan, Loeb Classical Library. Baquedano Beltrán, I. and Cabré de Morán, E. 1997. Caudillos celtas y armamento de parada. In La Guerra en la Antigüedad: una aproximación al origen de los ejércitos en Hispania. Madrid: Comunidad Autónoma de Madrid. 261-9. Baquedano Beltrán, I. and Martín Escorza, C. 1997. Distribución espacial de una necrópolis de la II edad del hierro: la zona I de La Osera en Chamartín de la Sierra, Ávila. Complutum 7: 175-94. Blanco Freijeiro, A. 1988. Las estatuas de verracos y las fíbulas zoomorfas celtibéricas. Espacio, Tiempo y Forma. Serie II, Historia Antigua 1: 69-78. Blanco García, J.F. 1997. Zoomorfos celtibéricos en perspectiva cenital: a propósito de los hallazgos de Cauca y el castro Cuesta del Mercado (Coca, Segovia). Complutum 8: 183-203. Blasco Sancho, Mª.F. 1999. Factores condicionantes de la composición de la cabaña ganadera de la II edad del hierro en la mitad norte de la península ibérica. In F. Burillo (ed.) IV Simposio sobre los Celtíberos: economía. Homenaje a J.L. Argente Oliver (Daroca, Zaragoza; Septiembre 1997). Zaragoza: Institución Fernando el Católico. 149-56. Blázquez Martínez, J.Mª. 1957. La economía ganadera de España antigua a la luz de las fuentes literarias griegas y romanas. Emerita 25: 159-84. Blázquez Martínez, J.Mª. 1969. Economía de los pueblos prerromanos del área no ibérica. In M. Tarradell (ed.) Estudios de Economía Antigua de la Península Ibérica. Barcelona: Vicens Vives. 225-30. Blázquez Martínez, J.Mª. and Montero, S. 1993. Ritual funerario y status social: los combates gladiatorios perromanos en la península ibérica. Veleia 10: 71-84. Brunaux, J.L. and Lambot, B. 1987. Guerre et Armement chez les Gaulois (450-52 a.C.). Paris. Büchsenschütz, O. 1995. The significance of major settlements in European Iron Age society. In B. Arnold and D.B. Gibson (eds) Celtic Chiefdom, Celtic State. Cambridge. 53-63. Cabré Aguiló, J., Cabré de Morán, Mª.E. and Molinero Pérez, A. 1950. El Castro y la Necrópolis del Hierro Céltica de Chamartín de la Sierra (Ávila). Madrid. Cabré de Morán, Mª.E. and Baquedano Beltrán, Mª.I. 1997. El armamento céltico de la II edad del hierro. In La Guerra en la Antigüedad: una aproximación al origen de los ejércitos en Hispania. Madrid: Comunidad Autónoma de Madrid. 240-59. Carneiro, R. 1970. A theory of the origin of the State. Science 169: 733-9. Carneiro, R. 1990. Chiefdom-level warfare as exemplified in Fiji and the Cauca valley. In J. Haas (ed.) The Anthropology of War. New York & Cambridge: Cambridge University Press. 190-211.

121

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Santiago de Compostela: Universidad de Santiago de Compostela. 373-82. García Moreno, L.A. 1989. La Hispania anterior a nuestra era: verdad, ficción y prejuicio en la historiografía antigua y moderna. In Actas del VII Congreso Español de Estudios Clásicos, vol. III. Madrid: Universidad Complutense. 17-43. García Moreno, L.A. 1993. Organización sociopolítica de los celtas en la península ibérica. In M. Almagro Gorbea (ed.) Los Celtas: Hispania y Europa. Madrid: Actas. 327-55. García Quintela, M.V. 1990. Les peuples indigènes et la conquéte romaine d’Hispanie: essai de critique historiographique. Dialogues d’Histoire Ancienne 16: 181-220. García Quintela, M.V. 1993. Viriato y la ideología trifuncional indoeuropea. Polis 5: 111-38. García Quintela, M.V. 1999. Mitología y mitos de la Hispania prerromana, III. Madrid: Akal. García Riaza, E. 1997. La función de los rehenes en la diplomacia hispano-romana. Memorias de Historia Antigua 18: 81-107. García Riaza, E. 1999. Las claúsulas económicas en las negociaciones de paz romano-celtibéricas. In F. Burillo (ed.) IV Simposio sobre los Celtíberos: economía. Homenaje a J.L. Argente Oliver. Zaragoza: Institución Fernando el Católico. 515-20. García y Bellido, A. 1977. Bandas y guerrillas en las luchas con Roma. In A. Prieto (ed.) Conflictos y Estructuras Sociales en la Hispania Antigua. Madrid: Akal.13-60. Garlan, Y. 1989. Guerre et Economie en Grèce Ancienne. Paris: Editions La Découverte. Godelier, M. 1999. The Enigma of the Gift. Cambridge: Polity Press. González-Tablas Sastre, F.J. 1985. La necrópolis de Trasguija: aproximación al estudio de la estructura social de Las Cogotas. Norba 6: 43-51. Grant, E. (ed.) 1986. Central Places, Archaeology and History. Sheffield: University of Sheffield Department of Archaeology and Prehistory. Guerra, A. and Fabiâo, C. 1992. Viriato: genealogia de um mito. Penélope: fazer e desfazer a história 8: 923. Gundel, A. 1968. Viriato, lusitano, caudillo en las luchas contra los romanos, 147-139 a.C. Caesaraugusta 3132: 175-98. (Also published in R.E. Pauly-Wisowa IX, A 1 [1961, Stuttgart]; s.v. ‘Viriatus’, cols. 20330). Haas, J. (ed). 1990. The Anthropology of War. New York & Cambridge: Cambridge University Press. Haas, J. 1999. The origins of war and ethnic violence. In J. Carman and A. Harding (eds) Ancient Warfare: archaeological perspectives. Stroud: Sutton. 11-24. Hanson, V.D. 1999. The Wars of Ancient Greeks and Their Invention of Western Military Culture. London: Cassell. Helms, M.W. 1988. Ulysses’ Sail: an ethnographic odyssey of power, knowledge and geographical

Homenaje a José Mª Blázquez, vol. II. Madrid: Ediciones Clásicas. 181-93. Ducrey, P. 1999. Guerre et Guerriers dans la Grèce Antique. Paris: Hachette Litératures. Earle, T. (ed.) 1991. Chiefdoms: power, economy, and ideology. Cambridge: Cambridge University Press. Earle, T.K. 1997. How Chiefs Come to Power: the political economy in prehistory. Stanford: Stanford University Press. Elianus Claudius. 1958. On the Characteristics of Animals. Vol. II. (trans. A.F. Scholfield). London: Harvard University Press. Ferguson, R.B. (ed.) 1984. Warfare, Culture and Environment. New York: Academic Press. Ferguson, R.B. and Farragher, L.E. (eds) 1988. The Anthropology of War: a bibliography. New York: Occasional Papers of the Harry Frank Guggenheim Foundation, 1. Ferguson, R.B. and Whitehead, N.L. 1992. The violent edge of empire. In R.B. Ferguson and N.L. Whitehead (eds) War in the Tribal Zone: expanding states and indigenous warfare. Santa Fe: School of American Research Press. 1-30. Fernández Gómez, F. 1986. Excavaciones Arqueológicas en El Raso de Candeleda (Ávila), Vol. II. Ávila: Institución Gran Duque de Alba. Fernández Gómez, F. 1997. La Necrópolis de la Edad del Hierro de ‘El Raso’ (Candeleda, Ávila): ‘Las Guijas, B’. Zamora: Arqueología en Castilla y León, Memorias 4. Feuvrier-Prévotat, C. 1978. Echanges et sociétés en Gaule indépendante: à propos d´un texte de Poseidonios d´Apamée. Ktema 3: 243-59. Florus. 1973. Epitome of Roman History (trans. E.S. Forster). London: Harvard University Press, Loeb Classical Library. Fokkens, H. 2001. Cows, wars and wisdom: martiality and the constitution of man. Paper presented at conference ‘Warfare, violence and slavery in prehistory and protohistory’, University of Sheffield, 2-3 Feb 2001. Francisco Martín, J de. 1989. Conquista y Romanización de Lusitania. Salamanca: Ediciones Universidad de Salamanca. 57-75. Fried, M., Harris, M. and Murphy, R. (eds). 1968. War: the anthropology of armed conflict and aggression. Garden City: Natural History Press. Galán Domingo, E. 1989-90. Naturaleza y cultura en el mundo celtibérico. Kalathos 9-10: 175-204. Garcia, J.M. 1985. Viriato: uma realidade entre o mito e história. Prelo 9: 59-70. García Huerta, R. 1997. La guerra entre los pueblos célticos: las fuentes literarias greco-latinas. In La Guerra en la Antigüedad: una aproximación al origen de los ejércitos en Hispania. Madrid: Comunidad Autónoma de Madrid. 223-9. García Moreno, L.A. 1988. Infancia, juventud y primeras aventuras de Viriato, caudillo lusitano. In Actas I Congreso Peninsular de Historia Antigua, vol. II.

122

EDUARDO SÁNCHEZ-MORENO: WARFARE, REDISTRIBUTION AND SOCIETY IN WESTERN IBERIA Mauss, M. 1990. The Gift: the form and reason for exchange in archaic societies. London & New York: Routledge. (First published as ‘Essai sur le don: forme et raison de l’echánge dans les sociétés archaïques’, L´Année Sociologique (1925) 1: 30-186). Mishkin, B. 1992. Rank and Warfare among the Plains Indians. Lincoln & London: University of Nebraska. Morales Muñiz, A. and Liesau von Lettow-Vorbeck, C. 1995. Análisis comparado de las faunas arqueológicas en el valle medio del Duero (provincia de Valladolid) durante la edad del hierro. In G. Delibes, F. Romero and A. Morales (eds) Arqueología y Medio Ambiente: el primer milenio a.C. en el Duero medio. Valladolid: Junta de Castilla y León. 455-514. Moret, P. 1991. Les fortifications de l’âge du fer dans la Meseta espagnole: origine et diffusion des techniques de construction. Mélanges de la Casa de Velázquez 27: 5-42. Muñiz Coello, J. 1994a. Monarquías y sistemas de poder entre los pueblos prerromanos de la península ibérica. In P. Sáez and S. Ordóñez (eds) Homenaje al Profesor Presedo. Sevilla: Universidad de Sevilla. 283-96. Muñiz Coello, J. 1994b. Instituciones políticas celtas e ibéricas: un análisis de las fuentes literarias. Habis 25: 91-105. Muñiz Coello, J. 1998. Riqueza y pobreza en la España prerromana: notas sobre la función social de los objetos suntuarios. Habis 29: 23-36. Nash, D. 1978. Territory and state formation in central Gaul. In D. Greene, C. Haselgrove and M. Spriggs (eds) Social Organisation and Settlement: contributions from anthropology, archaeology and geography. Oxford: BAR International Series (suppl.) 47. 455-75. Nash, D. 1985. Celtic territorial expansion and the Mediterranean world. In T.C. Champion and J.V.S. Megaw (eds) Settlement and Society: aspects of west European prehistory in the first millennium BC. Leicester: Leicester University Press. 45-67. Orosius. 1982. Historias (Historiarum adversus paganos). (trans. E. Sánchez Salor). Madrid: Gredos. Ortiz Romero, P. and Rodríguez Díaz, A. 1998. Culturas indígenas y romanización en Extremadura. In A. Rodríguez Díaz (ed.) Extremadura Protohistórica: paleoambiente, economía y poblamiento. Cáceres: Universidad de Extremadura. 256-63. Parcero Oubiña, C. 1995. Elementos para el estudio de los paisajes castreños del noroeste peninsular. Trabajos de Prehistoria 52: 127-44. Parker Pearson, M. 1999. The Archaeology of Death and Burial. Stroud: Sutton. Pastor Muñoz, M. 2000. Viriato: la lucha por la libertad. Madrid: Alderabán. Peralta Labrador, E. 1990. Las cofradías guerreras indoeuropeas en la España antigua. El Basilisco 3: 4960. Pérez Vilatela, L. 1989a. Procedencia geográfica de los lusitanos de las guerras del s.II a.C. en los autores clásicos (154-139 a.C.). In Actas VII Congreso

distance. Princeton & New Jersey: Princeton University Press. Heurgon, J. 1957. Trois Etudes sur le Ver Sacrum. (Coll. Latomus XXVI). Brussels. Hobsbawm, E.J. 1985. Bandits. Harmondsworth: Penguin. Keeley, L.H. 1996. War before Civilization: the myth of the peaceful savage. Oxford & New York: Oxford University Press. Kindelan Duany, A. 1958. Viriato: sus teatros de operaciones. Revista de Historia Militar 2: 9-21. Kurtz, W.S. 1987. La Necrópolis de Las Cogotas. Vol. I: Ajuares. Revisión de los materiales de la necrópolis de la segunda edad del hierro en la cuenca del Duero (España). Oxford: BAR International Series 344. Lens Tuero, J. 1968. Viriato, héroe y rey cínico. Estudios de Filología Griega 2: 253-72. Lewis, H. 1981. Warfare and the origin of the State: another formulation. In H. Claessen (ed.) The Study of the State. Leiden: Mouton. 206-17. Lewuillon, S. 1992. Contre le don: remarques sur le sens de la reciprocité et de la compensation sociale en Gaule. Dialogues d’Histoire Ancienne 18: 105-56. Liesau von Lettow-Vorbeck, C. 1998. El Soto de Medinilla: faunas de mamíferos de la edad del hierro en el valle del Duero (Valladolid, España). Archaeofauna: revista de la Asociación Española de Arqueozoología 7: 7-210. Lincoln, B. 1976. The Indo-European cattle-raiding myth. History of Religions 16: 42-65. Lincoln, B. 1981. Priests, Warriors and Cattle: a study in the ecology of religions. Berkeley & London: University of California Press. Livy. 1960. Ab Urbe Condita (History of Rome). Vols V, VI, XIV. London: Heinemann, Loeb Classical Library. López Domech, R. 1986-87. Sobre reyes, reyezuelos y caudillos militares en la protohistoria hispana. Studia Historica: historia antigua 4-5: 19-22. López Melero, R. 1988. Viriatus Hispaniae Romulus. Espacio, Tiempo y Forma, serie II 1: 247-61. López Monteagudo, G. 1989. Esculturas Zoomorfas Celtas de la Península Ibérica. Madrid: Anegos de Archivo Español de Arqueología, CSIC. Lorrio Alvarado, A.J. 1993. El armamento de los celtas hispanos. In M. Almagro Gorbea (ed.) Los Celtas: Hispania y Europa. Madrid: Actas. 285-326. Marco Simón, F. 1993. Feritas Celtica: imagen y realidad del bárbaro clásico. In E. Falque and F. Gascó (eds) Modelos Ideales y Prácticas de Vida en la Antigüedad Clásica. Sevilla: Universidad de Sevilla. 141-66. Martín Bravo, A. Ma. 1999. Los Orígenes de Lusitania: el I milenio a.C. en la Alta Extremadura. Madrid: Gabinete de Antigüedades de la Real Academia de la Historia. Martín Valls, R. 1985. Segunda edad del hierro: las culturas prerromanas. In J. Valdeón (ed.) Historia de Castilla y León, vol. 1. Valladolid: Ambito. 104-31.

123

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY ejércitos en Hispania. Madrid: Comunidad Autónoma de Madrid. 33-52. Quesada Sanz, F. 1997b. El Armamento Ibérico: estudio tipológico, geográfico, funcional, social y simbólico de las armas en la cultura ibérica (siglos VI-I a.C.). Montagnac: Editions Monique Mergoil. Quesada Sanz, F. 1997c. ¿Jinetes o caballeros? En torno al empleo del caballo en la edad del hierro peninsular. In La Guerra en la Antigüedad: una aproximación al origen de los ejércitos en Hispania. Madrid: Comunidad Autónoma de Madrid. 185-94. Rankin, H.D. 1987. Celts and the Classical World. London: Croom Helm. Renfrew, C. 1975. Trade as action in distance. In J.A. Sabloff and C.C. Lamberg-Karlovsky (eds) Ancient Civilization and Trade. Alburquerque: University of New Mexico Press. 3-59. Renfrew, C. and Bahn, P. 1991. Archaeology: theories, methods and practices. London: Thames and Hudson. Rodríguez Adrados, F. 1946. La fides ibérica. Emerita 14: 128-209. Rodríguez Díaz, A. 1995. Extremadura prerromana. In Extremadura Arqueológica IV. Arqueología en Extremadura: 10 años de descubrimientos. Mérida. 106-12. Romero Carnicero, F. and Sanz Mínguez, C. 1992. Representaciones zoomorfas prerromanas en perspectiva cenital: iconografía, cronología y dispersión geográfica. In II Symposium de Arqueología Soriana. Homenaje a D. Teógenes Ortego y Frías. Soria: Diputación Provincial de Soria. 453-71. Rouveret, A. 2000. Captiua arma: guerre, butin, économie dans les cités de Grande Grèce et de Campanie du Ve siècle à l´expédition de Pyrrhus. In J. Andreau, P. Briant and R. Descat (eds) Économie Antique: la guerre dans les économies antiques. Saint-Bertrand-de-Comminges: Musée Archéologique Départemental de Saint-Bertrand-de-Comminges. Rubinsohn, Z.W. 1981. The Viriatic war and its Roman repercussions. Rivista Storica dell’Antichitá 11: 1724. Ruiz Zapatero, G. and Álvarez-Sanchís, J.R. 1995a. Las Cogotas: oppida and the roots of urbanism in the Spanish Meseta. In B. Cunliffe and S. Keay (eds) Social Complexity and the Development of Towns in Iberia: from the Copper Age to the second century A.D. London: British Academy, 86. 209-35. Ruiz Zapatero, G. and Álvarez-Sanchís, J.R. 1995b. Prehistory, story-telling, and illustrations: the Spanish past in school textbooks (1880-1994). Journal of European Archaeology 3: 213-32. Sacristán de Lama, J.D. 1994. Apuntes sobre la geografía poblacional vaccea. Boletín del Seminario de Arte y Arqueología 60: 139-52. Sacristán de Lama, J.D. 1995. Reflexiones en torno al modelo de poblamiento de época celtibérica en la cuenca media del Duero. In F. Burillo (ed.) El Poblamiento Celtibérico: III Simposio sobre

Español de Estudios Clásicos, vol. III. Madrid: Universidad Complutense. 257-62. Pérez Vilatela, L. 1989b. Notas sobre la jefatura de Viriato en relación con la ulterior. Archivo de Prehistoria Levantina: homenaje a D. Fletcher Valls 19: 191-204. Piggott, S. 1986. Horse and chariot: the price of prestige. In D. Ellis Evans and J.G. Griffith (eds) Proceedings of the Seventh International Congress of Celtic Studies. Oxford: Cranham Press. 25-30. Pitillas Salañer, E. 1996. Una aproximación a las reacciones indígenas frente al expansionismo romano en Hispania (205 al 133 a.n.e.). Memorias de Historia Antigua 17: 133-55. Pitillas Salañer, E. 1997. Jefaturas indígenas en el marco de la conquista romana en Hispania y la Galia. Hispania Antiqua 21: 93-108. Plutarch. 1961. Parallel Lives. Vol. VIII: Sertorius and Eumenes. Phocion and Cato the Younger. (trans. B. Perrin). London: Harvard University Press, Loeb Classical Library. Polanyi, K. 1957. The economy as instituted process. In K. Polanyi, C.M. Arensberg and M.W. Pearson (eds) Trade and Market in the Early Empires: economies in history and theory. Chicago: Free Press. 243-70. Poux, M. 2000. Festins sacrés, ivresse collective et cultes guerriers en Gaule celtique: traces littéraires, perspectives archéologiques. In S. Verger (ed.) Rites et Espaces en Pays Celte et Méditerranéen: etude comparée a partir du sanctuaire d´Acy-Romance (Ardennes, France). Rome: Ecole Française de Rome. 108-21. Prieto Arciniega, A. 1978. La devotio ibérica como forma de dependencia en la Hispania prerromana. Memorias de Historia Antigua 2: 131-5. Pryar, F.L. 1977. The Origins of the Economy: a comparative study of distribution in primitive and peasant economies. New York: Academic Press. Quesada Sanz, F. 1994a. Vías de contacto entre Magna Grecia e Iberia: la cuestión del mercenariado. In D. Vaquerizo (ed.) Arqueología de la Magna Grecia, Sicilia y la Península Ibérica. Córdoba: Diputación Provincial de Córdoba. 191-246. Quesada Sanz, F. 1994b. La imagen del héroe: los antiguos iberos en la plástica española del XIX. Revista de Arqueología 162: 36-47. Quesada Sanz, F. 1995. Vino y guerreros: banquete, valores aristocráticos y alcohol en Iberia. In S. Celestino (ed.) Arqueología del Vino: los orígenes del vino en Occidente. Jerez de la Frontera: Consejo Regulador de las Denominaciones de Origen Jerez. 273-96. Quesada Sanz, F. 1996. La imagen de la antigüedad hispana en la plástica española del siglo XIX. In R. Olmos (ed.) Al Otro Lado del Espejo: aproximación a la imagen ibérica. Madrid: Lynx. 211-38. Quesada Sanz, F. 1997a. Aspectos de la guerra en el Mediterráneo antiguo. In La Guerra en la Antigüedad: una aproximación al origen de los

124

EDUARDO SÁNCHEZ-MORENO: WARFARE, REDISTRIBUTION AND SOCIETY IN WESTERN IBERIA Schulten, A. 1917. Viriatus. Neue Jahrbücher 39: 209-37. Service, E.R. 1975. Origins of the State and Civilization: the process of cultural evolution. New York: W.W. Norton. Silius Italicus. 1978. Punica. Vol. I, Bks 1-8. (trans. J.D. Duff). London: Harvard University Press, Loeb Classical Library. Sopeña Genzor, G. 1995. Ética y Ritual: aproximación al estudio de la religiosidad de los pueblos celtibéricos. Zaragoza: Institución Fernando el Católico. Strabo. 1960. Geography. Vol. II, Bks 3-5. (trans. H.S. Jones). London: Harvard University Press, Loeb Classical Library. Szabó, M. 1991. Mercenary activity. In S. Moscati, O.H. Frey, V. Kruta, B. Raftery and M. Szabó (eds) The Celts (I Celti). New York & Milan: Grupo Editoriale Fabbri. 333-6. Szabó, M. 1995. Guerriers celtiques avant et après Delphes: contribution à une période critique du monde celtique. In J.-L. Charpy (ed.) L’Europe Celtique du Ve au IIIe Siècle avant J.C.: contacts, echanges et mouvements de populations. (Actes du deuxième symposium international d´Hautvillers, octobre 1992). Epernay: Kronos Editions. 49-67. Tagliamonte, G.1994. I Figli di Marte: mobilitá, mercenari e mercenariato italici in Magna Grecia e Sicilia. Rome: Giorgiu Bretschneider Editore. Valerius Maximus. 1983. Memorable Doings and Sayings. Vol. I, Bks 1-5. (trans. D.R. Shackleton Bailey). London: Harvard University Press, Loeb Classical Library. van Wees, H. 1998. The law of gratitude: reciprocity in anthropological theory. In C. Gill, N. Postlethwaite and R. Seaford (eds) Reciprocity in Ancient Greece. Oxford: Oxford University Press. 13-49. Virgil. 1963. The Works of Virgil. Vol. I: Eclogues and Georgics. (revised by F. Haverfield). Hildesheim: Olms. Webster, D. 1975. Warfare and the evolution of the State: a reconsideration. American Anthropology 40: 46470. Wells, P.S. 1984. Farms, Villages and Cities: commerce and urban origins in late prehistoric Europe. Ithaca: Cornell University Press. Woolf, G. 1993a. Rethinking the oppida. Oxford Journal of Archaeology 12: 223-34. Woolf, G. 1993b. The social significance of trade in Late Iron Age Europe. In C. Scarre and H. Frances (eds) Trade and Exchange in Prehistoric Europe. Oxford: Oxbow. 211-18.

Celtíberos. Zaragoza: Institución Fernando el Católico. 369-80. Sahlins, M. 1972. Stone Age Economics. Chicago: Aldine-Atherton. Salinas de Frías, M. 1993. Problemática social y económica del mundo indígena lusitano. In El proceso histórico de la Lusitania oriental en época prerromana y romana. Cuadernos Emeritenses, 7. Mérida: Museo Nacional de Arte Romano. 9-36. Salinas de Frías, M. 1999. De Polibio a Estrabón: los celtas hispanos en la historiografía clásica. In A. Alonso, S. Crespo, T. Garabito and Mª.E. Solovera (eds) Homenaje al Profesor Montenegro: estudios de historia antigua. Valladolid: Universidad de Valladolid. 191-203. San Miguel Maté, L.C. 1993. El poblamiento de la edad del hierro al occidente del valle medio del Duero. In F. Romero, C. Sanz and Z. Escudero (eds) Arqueología Vaccea: estudios sobre el mundo prerromano en la cuenca media del Duero. Valladolid: Junta de Castilloa y León. 21-66. Sánchez-Moreno, E. 1995-96. El caballo entre los pueblos prerromanos de la meseta occidental. Studia Historica: historia antigua 13-14: 207-29. Sánchez-Moreno, E. 1996a. A propósito de las gentilitates: los grupos familiares del área vetona y su adecuación para la interpretación de la organización social prerromana. Veleia 13: 115-42. Sánchez-Moreno, E. 1996b. Aproximación social a la meseta occidental prerromana: riqueza y jerarquización en la necrópolis de El Raso (sector El Arenal). Candeleda, Ávila. Cuadernos de Prehistoria y Arqueología de la Universidad Autónoma de Madrid 23: 164-90. Sánchez-Moreno, E. 1998a. Meseta occidental e Iberia exterior: contacto cultural y relaciones comerciales en época prerromana. Unpublished PhD thesis, Universidad Autónoma de Madrid. Sánchez-Moreno, E. 1998b. De ganados, movimientos y contactos. Revisando la cuestión trashumante en la protohistoria hispana: la meseta occidental. Studia Historica: historia antigua 16: 53-84 (On-line: www.ffil.uam.es./antigua/piberica/ganado/ganado1.ht ml). Sánchez-Moreno, E. 2000. Vetones: historia y arqueología de un pueblo prerromano. Madrid: Ediciones UAM. Sánchez-Moreno, E. forthcoming. Cross-cultural links in ancient Iberia: archaeological traces on interregional diplomacy. Oxford Journal of Archaeology. Santos Yanguas, N. and Montero Honorato, Mª.P. 1983. Viriato y las guerras lusitanas. Bracara Augusta 37: 153-81.

125

Warfare, violence and the construction of masculinity in the Iron Age rock art of Valcamonica, northern Italy Lynne Bevan Living History Co-ordinator, Balsall Heath Local History Society, Birmingham and independent artefact specialist

Introduction

Pre-Iron Age carvings

The Camonica valley – Valcamonica – is a deep fault almost 50 miles long, situated north of the city of Brescia in the Italian Alps, between Lake Iseo and the Swiss frontier (Fig. 1). Here, images were carved on the natural bedrock, large outcrops of sandstone and schist at the valley sides, from the Epi-Palaeolithic (or Italian Mesolithic) period onwards, with the majority of carvings occurring during the Iron Age. Many of the bestpreserved carvings are in Naquane National Park, near the modern village of Capo di Ponte (Fig. 2).

During the Epi-Palaeolithic, when the first carvings were made by mobile hunting bands, the subject matter comprised carvings of elks and large deer, often pierced by spears (Anati 1995: figs 17-18, 20, 24-26). The first anthropomorphs were produced during the Neolithic. These figures were male, female or sexually undifferentiated, and were often associated with dogs (ibid.: 40, fig. 35) and possible solar motifs (ibid.: 104, figs 74-75). Some groups of oranti or worshipper figures with upraised arms date to the Chalcolithic or Bronze Age periods (see Fossati 2002 for dating). These have been interpreted as images of the participants in initiation ceremonies or funerals (Anati 1964: 179). Motifs such as maps, ploughs, deer and other animals, and a variety of weapons appeared from the Neolithic onwards. Statue stele bearing a similar repertoire of motifs were created during the Chalcolithic.

At Naquane the massive size of many of the carved stones contrasts with the relatively small size of individual carvings. Some stones have a single image or composition on them, while others are almost completely covered with carvings and have been subsequently recarved during later periods, with Iron Age carvings being applied to rocks that already had Neolithic, Chalcolithic and Bronze Age carvings on them. On the so-called Great Rock earlier carvings have become surrounded by, and incorporated into, predominantly Iron Age scenes.

Art by men for men? The dominance of obviously male images among the stick-figure rock art anthropomorphs during the Iron Age

FIGURE 1. General location map.

127

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY campaigns of 29-25BC and 14BC. The names of the tribes are inscribed on the Alpine Trophy of Augustus at La Turbie in France between two carvings showing captives chained to a trophy. The Camuni were sometimes confused by classical authors such as Strabo with the Raeti (Salmon 1982: 192, n.198), a blanket term given to a wide variety of Alpine people, but it is possible that, then as now, the inhabitants of the valley maintained their own identity. Miranda Green has suggested that the Camuni were ‘very probably Celtic’ (Green 1986: 195). Chronology of the battle scenes Armed anthropomorphs, often shown in apparent battle scenes, did not appear in the rock art until at least the Late Bronze Age (Fossati 2002: 98); they subsequently appear in quantity during various phases of the Iron Age, primarily the Early Iron Age, when most of the rock art of Valcamonica was produced (A. Fossati pers. comm.). Recently, Judith Toms has included the martial carvings in her study of Bronze Age warfare, attributing them to the Bronze Age/Early Iron Age (Toms 2000), although such carvings continued to be made during the Middle and Late Iron Ages. Certainly, as Toms states, rock art is difficult to date (ibid.: 111), and although there is some debate among Italian rock art scholars regarding the dating of certain images, some reliable and detailed dating schemes are available. For example, a good general chronology of the principal images by period is provided by Fossati (1996: 54-5, fig. 29; 2002). Anati also provides a chronological summary of key images by period (1994: 57).

FIGURE 2. Location of rock art sites of Valcamonica, including Naquane and other sites mentioned in the text (after Fossati et al. 1991: fig. 2).

Battle scenes has led some archaeologists to assume that the rock art was being made exclusively by men for men (e.g. Barfield 1998). Emmanuel Anati has suggested that an apparent absence of identifiable female figures after the Chalcolithic and Bronze Age periods indicates a relatively lower status for females whose activities must have been of lesser importance because they were apparently not being represented (Anati 1964: 240). Although most identifiably female representations do precede the Iron Age, I have challenged these simplistic views which owe more to late twentieth-century biased preconceptions of gender roles than to any real understanding of the rock art or the society it represents (Bevan 1999; 2000; 2001; forthcoming a and b).

Depictions of warriors, often apparently engaged in battle (Fig. 3), proliferate in the Iron Age rock art, as also do scenes interpreted as ‘jousting tournaments’ or duelling between young warriors in training (Fig. 4), possibly connected with initiation ceremonies and the attainment of warrior status (Fossati 1996; 2002). The emergence of a warrior elite has also been suggested (ibid.). While it is not always possible to separate the carvings chronologically, the subject matter tells us a great deal about prehistoric life, particularly during the Iron Age when warfare and violence appear to have been important elements in both everyday life and in the construction of masculinity. Despite this emphasis upon warfare and violence, there is comparatively little evidence of actual killing shown in the carvings. One exception, on the Great Rock at Naquane, features a large warrior, wearing a possibly Etruscan helmet, who is shown in the act of running a smaller figure through with a spear (Fig. 5). More often, the emphasis appears to have been on armed display rather than on actual killing, a good example of which, comprising a row of seven warriors, some of whom are wearing helmets and/or carrying oversized spears, appears on Rock 27 at Naquane (Fig. 6). These stick-

The Camuni The rock art of Valcamonica is often studied in a cultural vacuum, despite the fact that many external cultural influences can be discerned in the styles of art and elements of material culture portrayed in the carvings. This is partly because the identity of the valley’s inhabitants is uncertain. Very little is known about the Camuni, an Iron Age tribe who lived in the Valcamonica, beyond the fact that they were one of 45 Alpine tribes defeated and conquered by the Romans in the two major 128

LYNNE BEVAN: WARFARE, VIOLENCE AND THE CONSTRUCTION OF MASCULINITY IN THE IRON AGE ROCK ART

FIGURE 4. Warriors from Rock 50, Naquane (after Fossati et al. 1991: fig. 42).

FIGURE 3. Large warrior, Rock 50 (after Fusco and Galbiati 1990: 49).

FIGURE 5. A warrior spearing a smaller figure, detail from Rock 1, Naquane (after Anati 1964: 127).

FIGURE 6. Row of warriors on Rock 27, Naquane (after Fusco and Galbiati 1990: 35).

129

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY figure anthropomorphs, which date to the Late Bronze Age, also occur in other rock art assemblages, including the one at nearby Seradina, where an almost identical carving appears on Rock 1 (Anati 1987: 36, fig. 32). Similar scenes involving broader-bodied anthropomorphs dating to various phases of the Iron Age appear on other rocks at Naquane, including Rocks 47, 50 and the Great Rock. It is often difficult to determine whether the figures are depicted in combat or training, or whether they are participating in dances or armed displays. A scene on the Great Rock has been described by Emmanuel Anati as a ‘procession’ (Anati 1987: 21, fig. 14). Were these scenes intended to record myths and legends, or were they even war memorials honouring the dead and providing a narrative of their campaigns? War memorials are not restricted to the modern age but were also known in diverse cultures during antiquity, including those of Greece, Rome, Egypt and Mesopotamia (see Borg 1991).

FIGURE 8. Hostage scene from Carpene (after Sansoni 1987: fig. 63).

Hostage scenes The threatened figures, whose arms are not visible (suggesting that they are bound behind their backs or at their sides), are always represented without a phallus, whereas the aggressors are always represented with one. Does the phallic image define manhood, warriorhood, or both? If the prisoner is male, does being held captive cancel manhood (i.e. a captive is unmanned, therefore non-phallic and not a warrior)? Conversely the phallic image may portray phallic enhancement, a warrior’s penis sheath. Such an enhancement might have been made of an organic material such as leather or wood, neither of which is likely to survive archaeologically. Another potential hostage scene interpreted as guards escorting a row of bound captives appears on Rock 4 of In Valle, Paspardo (Fossati et al. 1991: 28, fig. 45). Neither the possible captives nor the armed guards appear to have been phallic, although the carving is very indistinct. On the topic of the emasculation of captives, although in this case in Roman art, see Ferris (2000: 378).

Perhaps the most sinister and threatening scenes identified in Valcamonica involve a phallic figure holding a weapon above the head of a possibly bound captive, two examples of which come from Naquane (one of which is illustrated here; Fig. 7), and one from nearby Carpene (Fig. 8). Similar scenes have been identified in the rock art of northern Europe (Sansoni 1987: 71, fig. 65), but these date to the Bronze Age, whereas the Italian scenes are later, dating to the Iron Age, in common with the majority of the Valcamonica carvings, particularly those featuring weapons and fighting.

In a violent scene at Fossum, Bohuslän in Sweden, the weapon held over the captive is an axe (Sansoni 1987: 71, fig. 65), while at Carpene and Naquane in Valcamonica the weapons are swords or heavy sticks (Sansoni 1987: 71, figs 63-64). These scenes have been interpreted as, in one instance, human sacrifice (Anati 1964: 177), or probable executions (Sansoni 1987: 71). Execution becomes a less convincing interpretation, however, when the Valcamonica carvings are compared to the more graphic scenes of execution conducted by groups of archers in the Neolithic rock art of the Spanish Levant (Guilaine and Zammit 1998: figs 32-33, 35). These scenes portray the end result of the violence in the form of bodies shot through with multiple arrows, overlooked by groups of archers (ibid.), in contrast to the sinister scenes from Valcamonica and northern Europe which hint, perhaps more disturbingly, at impending violence.

FIGURE 7. Hostage scene from Naquane (after Anati 1964: 177).

130

LYNNE BEVAN: WARFARE, VIOLENCE AND THE CONSTRUCTION OF MASCULINITY IN THE IRON AGE ROCK ART The size of the captive in the Fossum scene suggests strongly that he was a male of large build, even larger than his armed opponent, whereas size contrast in the Italian scenes is less marked. Emanuel Anati interpreted the threatened figure in one of the scenes from Naquane as a man ‘enclosed in a sack or a cloth’ (Anati 1964: 177) but this figure actually appears to be wearing a long cloak-like garment. The figure’s slightly smaller stature suggests that it could be female, especially when the clothing is compared to that of Etruscan women, for example, on the Certosa situla (Museo Civico Archeologico di Bologna 1988: 30, 295). Was this unfortunate hostage a woman? The common occurrence of rape in warfare is well known in the present age, as well as throughout history (e.g. Brownmiller 1975; Scott 2001). Does the threat of rape or capture perhaps explain the virtual absence of females in the Iron Age carvings? Perhaps women were confined to the settlements, especially after dark, for fear of rape, murder or capture for slavery by marauding bands or episodic acts of violence from neighbouring tribes. As is so often the case in the rock art of Valcamonica, the sex of this unfortunate person remains ambiguous. Carpene, a few kilometres up the valley from Naquane, where one of the hostage scenes is located, was a focus for carvings of a predominantly martial nature. The areas of carvings are less accessible and the carvings themselves are generally less well-preserved than those at Naquane. In contrast to Naquane, the areas of rock art are more open, but harder to find, as well as being at a higher altitude. The main focus of carvings is an area of large stones which appears to have formed a natural arena

PLATE 1. View of main concentration of rock carvings at Carpene (author’s photograph).

PLATE 2. Pair of combatants from Carpene (author’s photograph).

131

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY from Naquane, shows a devil with beak and horns, equipped with a trident, engaged in hacking limbs from other figures (Anati 1987: 46, fig. 38). Deer hunting The other principal male activity represented in the Iron Age carvings at Naquane is hunting. The prey animal is almost exclusively the deer, or, to be more precise, stags. Just as over-sized weapons and, often, the phallus express masculinity within the context of fighting and hunting, so the antlers appear to identify the stag, and in many instances they are much larger than normal. This is especially true on the Great Rock at Naquane where the carvings include many representations of stags, one of which is of enormous size in relation to the other stags and human figures in the scene, and equipped with correspondingly over-sized antlers (Anati 1964: 28-9). This stag is associated with other scenes of stags being killed, as well as with a large Iron Age house. Most hunting scenes, however, are straightforward, such as an example, also on the Great Rock, featuring a phallic male spearing a stag with a large spear, aided by a dog which bites the stag’s hindquarters (ibid.: 119). Toms has suggested that hunting may be ‘as important an activity as warfare’ in the rock art imagery, and that similar qualities are required for both activities, including ‘bravery, intelligence, determination, physical strength and dexterity, skill in the use of weapons, a respect for and understanding of the prey and opponent, and a willingness to take life’ (2000: 111). PLATE 3. Giant wielding an axe from Carpene (author’s photograph).

Ethnographic research, such as Eugenia Herbert’s West African study Iron, Gender and Power (1993), suggests that constructions of masculinity in non-western societies are extremely complex, and that hunting (as well as other male domains such as warriorhood, metallurgy and sacred kingship) is connected with many other social and ritual concerns, including fertility, ancestral appeasement and gender opposition (ibid.: chap. 7). Hunting is the preserve of male societies and its secrets are revealed through initiation (ibid.: chap. 7). Moreover, hunting is often regarded as equivalent to, as well as inextricably linked to, female childbearing, and the forest spirits, who must be properly appeased by performing rituals and observing certain taboos, can affect the success of either the hunt or human reproduction (ibid.: 168). Herbert concludes that ‘the fecundity of the hunter-warrior [is] manifested in killing animals and men’ (ibid.: 186).

(Plate 1). Here, the theory that the carvings related to military training, perhaps for juveniles, seems justified. The subject matter is almost exclusively connected with fighting, with scenes involving paired combatants and large, potentially mythological figures brandishing weapons (Plates 2 and 3). The so-called Camunian rose motif appears at Carpene, and has been interpreted as a derivation of the swastika, the earliest examples of which date to the Bronze Age in Italy. Whether or not the rose was a solar symbol, as has been suggested, it seems to have been a recurring motif in Iron Age carvings associated with warriors. Carvings also appear to have been male-orientated and militaristic in nature at nearby Sellero (Fig. 2). Similar themes appear in the rock art of Paspardo, which lies across the valley from Naquane (Fig. 2), where many of the carvings, including a scene showing oversized warriors (Fig. 9), date from the Middle to Late Iron Age. Luine, some 20 kilometres further down the valley (Fig. 2), near the spa town of Boario, is also a focus for carvings of weaponry and warfare and of other occurrences of the Camunian rose. An unusual scene from the so-called Rock of the Devil at Bedolina, located just above Capo di Ponte, on the other side of the valley

A scene on Rock 4 at In Valle, Paspardo dating to the Late Iron Age seems to link pairs of competing stags with armed combatants, suggesting an ideological symmetry between the two (Fig. 10). The man-stag relationship is even closer in a scene in which man and stag are merged into one body and accompanied by armed males who seem to be involved in ritual action (Green 2001: 214, fig. 7). This inevitably leads to discussion of the Iron Age stag god Cernunnos who appears at Naquane as a massive phallic figure with a snake, accompanied by a 132

LYNNE BEVAN: WARFARE, VIOLENCE AND THE CONSTRUCTION OF MASCULINITY IN THE IRON AGE ROCK ART

FIGURE 9. Large warriors on Rock 4 at In Valle, Paspardo (after Fossati et al. 1991: fig. 27).

FIGURE 10. Warriors and stags on Rock 4 at In Valle, Paspardo (from the author’s photograph).

at Naquane would have been conducive to such rites which might have been centred on male cult houses and hunting grounds.

small, presumably mortal, follower (ibid.: 213, fig. 6a). A famous, more detailed representation from the Gundestrup cauldron shows Cernunnos holding a snake and accompanied by a stag and other wild creatures (Green 1996: 135, fig. 101). Other manifestations of Cernunnos in the rock art of Valcamonica consist of a badly preserved carving from Paspardo (Green 2001: 213, fig. 6b) and a much later and sketchier Cernunnos from Pian Cogno (Priuli 1996: 29, fig. 51).

The palette, paddle or shovel symbol Some scenes of stags and fighting warriors are associated with the mysterious ‘palette’ symbol. The identification and meaning of this symbol are uncertain but it might have been intended to represent a razor, a mirror, or a paddle. Sometimes this image is referred to as a ‘shovel’ (Fossati 2002: 98). Morphologically all of the objects are similar. Occasionally, small figures are shown holding the palette, often equipped with a looped handle, suggesting that it was more likely to have been a razor or mirror (Fig. 11), either of which might have been used for signalling purposes by pointing the object at the sun. Some of the high-ground locations where rock art is often found would have been ideal vantage points where large areas of the valley could be viewed, but perhaps this is

Despite the problems of relating these images chronologically and ideologically, a very close ritual relationship is suggested between men and deer. However these beliefs were expressed, they seem to have been restricted to males, and possibly involved shamanism and perceived shape-shifting, as suggested by Miranda Green (Green 2001). These images and associated beliefs might have been connected with male initiation rites, sacrifices to the ancestors and other rituals, held in sacred groves. Certainly the wooded areas

133

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY The frequent linking of this symbol with fighting or death suggests that it may be a metaphor for death itself. The stele di S. Michele in Bosco in Bologna Museum shows the demon Charun holding a small paddle while leading a female figure to her death (Museo Civico Archeologico di Bologna 1988: 262). The demon’s paddle is very similar in shape to the Valcamonica palette symbol and the death symbolism is obvious. Another possibility is that the palette was a mortuary shovel used in cremation rituals for removing ashes and burnt bone from pyres and placing them in a cinerary urn, a rite which was known in Lombardy during the first millennium AD (Fusco and Galbiati 1990: 22). Conversely, Emmanuel Anati has suggested that the palette (or ‘paddle’ as he terms it) was an apotropaic symbol or ‘talisman’ (Anati 1964: 202-10).

stretching the evidence too far. There are examples of palette-shaped Villanovan razors in Bologna Museum (Museo Civico Archeologico di Bologna 1988: 228, 232) where the collection also includes a bone spatulate object of a similar palette shape (ibid.: 102). There are also other palette-shaped objects of organic materials from archaeological contexts, including a wooden object from Polada (Anati 1995: 78, fig. 76). If the palettes represented wooden implements, perhaps even battle standards, the original objects might not have survived archaeologically.

The construction of Iron Age masculinity Other recurring motifs in Iron Age fighting scenes, particularly on Rock 50 at Naquane, include water birds, the swan boat and sun motif which seems to represent a barque of the dead, footprints (often shown in pairs) and writing of Etruscan or Latin origin (Fusco and Galbiati 1990: 48). Rock 50, the so-called Rock of the Oranti, referring to a frieze of earlier dancing or praying figures which occupies one area of the rock, was extensively recarved during the Iron Age. Many of the carved Iron Age scenes are reminiscent of the subject repertoire of Etruscan tomb paintings, including scenes of hunting, fowling, and gladiatorial combat. Similar scenes also appear on contemporary metalwork (Fossati et al. 1991: 26, fig. 43; Barfield 1998). These are not just artistic scenes expressing favoured pastimes of the living, but seem to express concepts of masculinity in both this world and the next. It has been suggested by some Italian archaeologists, notably Angelo Fossati, that small footprints represent the feet of young boys undergoing initiation rites and taking their first steps towards achieving adult male status (Fossati 1996; 2002). Fossati has also suggested that the initiates were members of an aristocratic warrior class and that duelling, deer hunting and riding must have played an important part in their training (Fossati 1996). Scenes such as one on Rock 27 at Foppe di Nadro featuring a mounted rider and a smaller figure interpreted as a squire seem to support this Medievalized view of prehistory (Anati 1994: 178, fig. 140). A contrasting view has been put forward by John Robb who sees Valcamonica more as a world akin to the competing Big Men of Papua New Guinea, with their ritualized feasting and fighting (Robb 1994). The elaborate armed posturing practised by the tribesmen seldom erupts into real warfare and killing. Certainly some of the scenes from Valcamonica could be interpreted as armed men showing themselves at their most fierce, hoping to repel potential invaders by psychological means rather than by actually fighting them. The reality was probably somewhere between these two viewpoints.

FIGURE 11. Figure holding palette, Paspardo (after Fossati et al. 1991: fig. 29).

Palettes are usually represented as being of a much larger size than any associated images, as, for example, in a scene on the Great Rock at Naquane of warriors fighting around a labyrinth (Anati 1987: fig. 19). In another scene they are associated with a trapped stag which appears to be falling to its death (ibid.: 16, fig. 9). An interesting manifestation of the palette symbol occurs on Rock 47 at Naquane where two palettes have been attached to a spear and shield in the hands of a large phallic warrior who appears to be heading a contingent of much smaller warriors (Fig. 12). Above this scene on Rock 47, another phallic figure is holding a palette above his head (Fig. 13). While these could be battle standards, as suggested above, they could equally indicate, in the case of the large warrior, the death-dealing capabilities of the warrior and his men. Conversely, they could also indicate that the warrior had died in action, and in this way the rock art becomes a memorial recording the brave deeds of past ancestors.

134

LYNNE BEVAN: WARFARE, VIOLENCE AND THE CONSTRUCTION OF MASCULINITY IN THE IRON AGE ROCK ART

FIGURE 12. Warrior with two palettes, detail, Rock 47, Naquane (after Fusco and Galbiati 1990: 46).

FIGURE 13. Figure holding a palette, detail, Rock 47, Naquane (after Fusco and Galbiati 1990: 46).

around a male identity or suite of identities constructed upon warfare, bravery and hunting, together with the recurring house motif, a place for either dwelling, storage or cultic rites. The carvings and the activities they represent linked men with their territory and located them within society and a context of gender relations. The areas selected for the carvings were marginal, liminal zones, places where men went hunting, where they acted as watchmen, surveying their villages in the valleys below. Here, perhaps, were their hunting lodges or cult houses where male identities were forged and where male

Conclusions In conclusion, based upon the subject matter and the apparent male identity of most of the protagonists, the argument that the Iron Age rock carvings of the Valcamonica were the work of men is persuasive. Logically their function was to record myths, legends and triumphs, and express male identity, as well as revealing male preoccupations and providing a glimpse into the male religious psyche. The scenes clearly express concepts of masculinity during the Iron Age, based

135

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY division between female beauty and male violence, with social roles involving women located in the domestic and men in the public sphere (especially warfare)’ (2001: 90). Whitehouse does, however, suggest regional variation and ‘hints of more complex and subtle ideologies recognizable within the system’ (ibid.: 90).

rituals were celebrated. In contrast to earlier periods when Naquane and Foppe di Nadro were used for mixed sexual and exclusively female rites centred upon female cult houses, during the Iron Age these areas were used by men (Bevan forthcoming a and b). Other areas such as Carpene and Sellero appear to have been used almost exclusively during the Iron Age for militaristic carvings.

Robb raises a question which is pivotal to the iconography discussed here - ‘was male violence practical protection or symbolic repression or women?’ (Robb 1997: 57). I would argue that the Iron Age rock art, laden with male symbolism and with an emphasis upon warfare and violence, was not specifically intended for the female gaze but it was indeed intended for the men it portrayed, or for their descendants, perhaps as a way of establishing or maintaining land ownership. Recalling the ancestors might also have been a factor (Toms 2000: 113). It could also be the case that, in view of its warlike nature, it acted as a warning to other men from neighbouring groups, or to opportunistic war parties from further afield, groups of men dispossessed of their own lands who might have been desperate to plunder winter stores or steal animals. The prime purpose of the rock art was not to oppress women.

This is not to say, as is often claimed in anthropological works, that males deliberately excluded females and monopolized ritual. Access to the lightly-wooded valley sides might have been denied to the unarmed members of the community, especially in times of crisis, when warfare, violence and pressure on resources increased. Even if female rituals were no longer conducted in the rock art zones, and during the Iron Age the emphasis was on masculinity, warriorhood and the construction of male identity, there is still evidence of females in the rock art at this time. Non-phallic, presumably female figures with digging sticks can be seen in ploughing scenes, attesting to the continuing female role in agriculture (e.g. Anati 1994: 97, fig. 68). It is also safe to assume that females were weavers, based upon the occurrence of weaving equipment in female graves, a good example being provided by the cemetery of Osteria dell’Osa, near Rome (Bietti-Sestieri 1992). Weaving equipment and, occasionally, models of looms are found in richer female graves of the Este and Villanovan cultures, while poorer females had spindle whorls (Chieco Bianchi and Tombolani 1988). There are also representations of weavers in Etruscan art, for example, on a rattle from the Tombi del Ori, Bologna (Barfield 1998: 149, fig.5). There are also many representations of looms on the Great Rock at Naquane (Anati 1994: 159, figs 120-121), images that probably date to the Late Bronze Age (ibid.: 159 and A. Fossati pers. comm.), and perhaps express female gender, or female production, in a symbolic way.

It should be remembered that rock art is only one aspect of the artistic expression of a society and that female art might have been expressed in other ways, perhaps through pottery decoration. House decoration is another possibility, perhaps like an artist’s impression of an Etruscan woman decorating the exterior of a house with geometric symbols (Soprintendenza Archeologica di Roma 1998: 18, fig. 29). Ruth Whitehouse claims that ‘the Iron Age has the greatest potential for further work on gender, with rich burial and iconographic data as yet barely discussed from this perspective’ (2001: 90). In the future, it may be possible to compare various sources of evidence to formulate a view of changing gender relations in Valcamonica on similar lines to Gibbs’ work on in prehistoric Zealand, Denmark, in which she assessed the comparative visibility of males and females in hoards, burials, settlements, rock art and figurines (Gibbs 1987).

Again, on the Great Rock at Naquane, earlier female ritual scenes (Anati 1964: 178) and phallic male figures (ibid.: 223-4) have been deliberately incorporated within a larger assemblage of Iron Age carvings showing hunting and warfare. This suggests an ideological continuity, perhaps the recognition of a debt to an earlier spiritual or mythological tradition, rather than an erasure or re-writing of the past in favour of a new, exclusively male religious orthodoxy (Bevan forthcoming a and b). It is too simplistic to say that an apparent overrepresentation of male figures and the accoutrements of war in the Iron Age rock art at Valcamonica mean that women were undervalued. Nor can it be claimed, as Robb has suggested, that Italian Iron Age society was ‘a firm precursor to Mediterranean patriarchy’ (Robb 1997: 56). Robb suggests that women were ‘economically and politically dependant’ upon men for ‘status, security, and the means of pursuing their goals’, a thesis which he admits is based upon ‘scrappy’ Iron Age material (ibid.: 54). Ruth Whitehouse suggests that the available Iron Age evidence ‘on one hand supports Robb’s generalizing view of a widespread basic dichotomous symbolic

The rock art which formed the ritual landscape of Valcamonica, and which in turn was formed by it, represents a complex and layered ideological palimpsest. Warfare, despite being interpreted solely as the business of fighting men, whether of chiefly lineages, armed professionals, or farmers who took up arms in times of crisis, destroyed settlements. Then, as now, warfare created refugees. Warfare is as much about women and children and the elderly, as the fighting men whom we see portrayed in the rock art. Although I would disagree with Barfield’s assertion that rock art in general in Valcamonica was ‘largely done by men for men’ (1998: 143), with regard to the Iron Age images, it would seem that rock art does appear to have been an arena for male expression and competition. Then, as now, the interests

136

LYNNE BEVAN: WARFARE, VIOLENCE AND THE CONSTRUCTION OF MASCULINITY IN THE IRON AGE ROCK ART of other sections of society do not appear to have been represented.

Fossati, A. 1996. Southern Europe: rock art in the Alps, Italy and the Balkans 1990-1994. In P. Bahn and A. Fossati (eds) Rock Art Studies: news of the world 1. Oxford: Oxbow Books, Oxbow Monograph 72. 4158. Fossati, A. 2002. Landscape representations on boulders and menhirs in the Valcamonica-Valtellina area, alpine Italy. In G. Nash and C. Chippindale (eds) European Landscapes of Rock Art. London: Routledge. 93-115. Fossati, A., Jaffe, L. and de Abreu, M.S. 1991. Etched in Time: the petroglyphs of Val Camonica. Val Camonica, Brescia: Edizioni della Cooperativa Archeologica ‘Le Orme dell’Uomo’, Valcamonica Preistorica 3. Fusco, V. and Galbiati, A. 1990. Naquane Parco Nazionale delle Incisioni Rupestri. Guida itineraria. Lombardia: Libreria Del Parco. Gibbs, L. 1987. Identifying gender representation in the archaeological record: a contextual study. In I. Hodder (ed.) The Archaeology of Contextual Meanings. Cambridge: Cambridge University Press. 79-89. Green, M.A. 1986. The Gods of the Celts. England: Bramley Books. Green, M.A. 1996. Celtic Art. Everyman Art Library. London: Calman and King Ltd. Green, M.A. 2001. Cosmovision and metaphor. European Journal of Archaeology 4: 203-31. Guilaine, J and Zammit, J. 1998. Le Sentier de la Guerre: visages de la violence préhistorique. Paris: Éditions du Seuil. Herbert, E.W. 1993. Iron, Gender and Power: rituals of transformation in African societies. Bloomington & Indianapolis: Indiana University Press. Museo Civico Archeologico di Bologna. 1988. Guida al Museo Civico Archeologico di Bologna. Bologna: University Press. Priuli, A. 1996. Le Più Antiche Manifestazioni Spirituali: arte rupestre. Paleoiconografia camuna e delle genti alpine. Torino: Priuli & Verlucca Editori. Robb, J. 1994. Gender contradictions: moral coalitions and inequality in early prehistoric Italy. Journal of European Archaeology 2: 20-49. Robb, J. 1997. Female beauty and male violence in early Italian society. In A.O. Koloski-Ostrow and C.L. Lyons (eds) Naked Truths: women, sexuality and gender in Classical art and archaeology. London & New York: Routledge. 43-65. Salmon, E.T. 1982. The Making of Roman Italy. London: Thames and Hudson. Sansoni, U. 1987. L’arte rupestre di Sellero. Capo di Ponte, Brescia: Edizioni del Centro Camuno di Studi Preistorici, Studi Camuni 9. Scott, E. 2001. The use and misuse of rape in prehistory. In L. Bevan (ed.) Indecent Exposure: sexuality, society and the archaeological record. Glasgow: Cruithne Press. 1-18. Soprintendenza Archeologica di Roma. 1998. Fidene: una casa dell’eta del ferro. Milano: Electa.

Acknowledgements Thanks are due to Angelo Fossati for guiding me through the rock art of Naquane on my last visit there in 2002, and also to my husband Iain Ferris and my supervisor Martin Stringer, for all their help and advice. Bibliography Anati, E. 1964. Camonica Valley. London: Jonathan Cape. Anati, E. 1987. Capo di Ponte. Capo di Ponte, Brescia: Centro Camuno di Studi Preistorici. Anati, E. 1994. Valcamonica Rock Art. Capo di Ponte, Brescia: Centro Camuno di Studi Preistorici. Anati, E. 1995. Brescia Preistoria: 300 mila anni di presenza umana nel territorio bresciana. Capo di Ponte, Brescia: Centro Camuno di Studi Preistorici, Studi Camuni 16. Barfield, L.H. 1998. Gender issues in north Italian prehistory. In R.D. Whitehouse (ed.) Gender and Italian Archaeology: challenging the stereotypes. London: Institute of Archaeology, Accordia Specialist Studies on Italy 7. 143-56. Bevan, L. 1999. The art of assumption: identifying gender in the rock art of Valcamonica. 3rd Stone Magazine 34: 35-7. Bevan, L. 2000. Women’s art, men’s art: gender specific image selection. In G.H. Nash (ed.) Signifying Place and Space: world perspectives of rock art and landscape. Oxford: BAR Publishing, BAR International Series 902. 103-9. Bevan, L. 2001. Gender bias or biased agenda? Identifying phallic imagery, sexual scenes and initiation in rock art. In L. Bevan (ed.) Indecent Exposure: sexuality, society and the archaeological record. Glasgow: Cruithne Press. 64-88. Bevan, L. forthcoming (a). Worshippers and warriors: rock art and changing ritual landscapes of the Valcamonica, northern Italy from the Neolithic to the Iron Age. In G.H. Nash and C. Chippindale (eds). Assessing Time and Fluidity in Rock Art: recent trends and perspectives in European rock art. Bevan, L. forthcoming (b). Worshippers and warriors: changing gender relations in the prehistoric rock art of Naquane Park, Valcamonica, northern Italy. PhD thesis, University of Birmingham. Bietti-Sestieri, A.M. 1992. The Iron Age Community of Osteria dell’Osa. Cambridge: Cambridge University Press. Borg, A. 1991. War Memorials. London: Leo Cooper. Brownmiller, S. 1975. Against Our Will: men, women and rape. New York: Simon and Schuster. Chieco Bianchi, A.M. and Tombolani, M. (eds) 1988. I Paleoveneti. Padova: Editoriale Programma. Ferris, I.M. 2000. Enemies of Rome: barbarians through Roman eyes. Stroud: Sutton. 137

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Toms, J. 2000. Italy. In R. Osgood and S. Monks with J. Toms (eds) Bronze Age Warfare. Stroud: Sutton. 89114.

Whitehouse, R. 2001. Exploring gender in prehistoric Italy. Papers of the British School in Rome 69: 49-96.

138

The dead of Tormarton - Middle Bronze Age combat victims? Richard Osgood Environmental Advisor (Archaeology), Defence Estates

A brief report tucked away in the Transactions of the Bristol and Gloucestershire Archaeology Society (Knight et al. 1972) discusses the chance discovery of the remains of two human skeletons in a ditch or a pit at Tormarton in south Gloucestershire (ST 76737667). Some of these bones bore the trace of a seemingly violent end – with spearheads embedded in at least two of the victims.

burials or, indeed, discover whether there were more bodies to be recovered? This paper presents the initial findings of that work, notwithstanding the fact that much of the material has already been given some publicity by an episode of the BBC’s Meet the Ancestors.

The discovery in 1968

An initial magnetometer and resistivity survey in the immediate vicinity of the location of the burials gave no relevant information, the signals being entirely distorted by the presence of the gas pipeline. Aerial photographs also proved unhelpful and thus, to gain any further information, it was necessary to insert a couple of small trenches.

The excavations in 1999 and 2000

The skeletons had been discovered by a local farmer, Mr Dick Knight, who had a keen interest in local archaeology and had been following the progress of the insertion of a new high-pressure mains gas pipeline for southwest England. When the bones were revealed, he enlisted the support of his friend, Mr Charles Browne, and together they recovered as much material as possible, publishing their findings with the help of the late Leslie Grinsell.

Initially, a 10m x 10m grid square was cleared and extensions were then added along the line of the pipe trench until an archaeological feature was uncovered. It was soon apparent that the only feature in the vicinity was a length of linear ditch which had been cut through the solid limestone (Fig. 1). The ditch was V-shaped and was around 1.4m deep. What was also clear was that the ditch had been filled in one, quick, event; there was only a very small amount of primary silting and much of the rest of the ditch was filled with limestone slabs with voids around them. From the angle at which they lay,

This report contained much to tantalize, particularly the fact that the deposition context was not established, and a remark that ‘the feet of the skeleton were not exhumed, because the trench in which they lay was here covered by a spoil-heap which could not then be removed’ (Knight et al. 1972: 14). The site thus seemed ripe for further investigation. Could we establish a context for the

FIGURE 1. A segment of the Bronze Age linear ditch excavated in 1999, facing south.

139

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY The 2000 season began with a more extensive magnetometer survey (undertaken by Stratascan and funded by the BBC). This survey avoided the immediate proximity of the pipeline for reasons illustrated earlier (Fig. 2). The results were quite dramatic and very clear – the ditch was part of a linear feature and not a segment of an enclosure, defended or otherwise. This ruled out one of the initial hypotheses, that these bodies were the victims of an attack upon their settlement. Interestingly, the ditch clearly did not continue far beyond the pipe trench and the possibility remained open that bodies had been thrown into the ditch terminal. Terminals of linear ditches have, of course, huge importance on many Bronze Age sites, Springfield Lyons in Essex being a classic example, where moulds for casting bronze weapons were placed in its main ditch terminals (Buckley and Hedges 1987).

these appeared to have been tipped in from the east side of the feature. This was a pleasing initial result as we had established the context of the bones straightaway – a ditch rather than a pit. However, no additional skeletal material was recovered from the 2m slot to the south of the pipeline trench. We thus excavated a further 2m segment of the ditch, to the north of the gas pipe and it was here that further human remains were located. These were around 1m from the top of the ditch and were covered by the limestone slabs visible in the other segment (and indeed noted by Knight and Browne in 1968). The bones appeared to be quite crushed and jumbled in places, articulated in others. Several jaws and skull fragments were recovered and it was apparent that the condition of the bone had deteriorated from its state of preservation as recorded in 1968.

FIGURE 2. Magnetometry survey conducted by Stratoscan in 2000.

140

RICHARD OSGOOD: THE DEAD OF TORMARTON - MIDDLE BRONZE AGE COMBAT VICTIMS? remains excavated in 1999 and 2000 displayed no weapons injuries. Of course this need not mean that they did not come to a violent end; after all, their bodies were cast into the ditch at the same time as those with clear evidence for violence. A violent death need leave no trace on the skeleton.

In 2000 we thus excavated the ditch terminal closest to the gas pipe. This had been neatly cut down through the solid bedrock of Jurassic limestone and was spatulate in plan, half-sectioned. We also cut a further section through the ditch joining the segment excavated in 1999, in order to establish the extent of the human remains. Absolutely nothing was recovered from the ditch terminal and only a couple of patellae and ribs were found in the ditch cutting. So, although they were close to the terminal, the bodies had not been placed in a ritually significant position at the end of the ditch. Indeed, there was a lack of material culture from any of the ditch cuttings, including the ones we put through the ditch in the field to the south. The process of the ditch’s filling-in seems to have been the same throughout its length, as is also shown by the assemblage of molluscs (see below). The lack of pottery was perhaps unsurprising given that no known Bronze Age settlements have been located nearby and that there can often be a paucity of finds from excavations of linear ditches, an experience I had had whilst working on the excavation of a Bronze Age linear ditch at Windy Dido in Hampshire with Barry Cunliffe (Cunliffe and Poole 2000). In stripping the topsoil we did, however, find the opposite terminal in the field to the south and were able to establish that the feature was only c. 70m in length, a far cry from the Bronze Age linear ditches which run for many kilometres across much of southern England (Bradley et al. 1994). The ditch neither begins nor ends at any surviving notable topographic features – to modern eyes at least – although it does point roughly towards a surviving tumulus on the crest of the ridge some 400m to the northeast, and there remains the possibility that trees or woodland were formerly present, to provide a natural boundary at either end of the ditch.

FIGURE 3. The 1968 pelvis with a lozenge-shaped perforation from a spear wound.

The clearest wounds visible on the 1968 bones are lozenge-shaped spear perforations on the pelvis of one individual, together with a spearpoint which had broken off in the bone (Fig. 3). Another individual, an older man, had received a blow to the skull (Figs. 4 and 5). He also had a spear wound which seems especially nasty, piercing the spinal cord (Figs. 4 and 6). This would have led to instant paralysis of the victim if it was not inflicted whilst the individual was already prone.

So what then of the results of the excavation? We had established that the bodies had been thrown into a short linear ditch, close to its terminal, and that they had been covered without ceremony. The deposition of these bodies contrasts strongly with the usual evidence of ritual and ceremony characteristic of the disposal of the dead in this period (Parker Pearson 1993; Brück 1995). No apparent provision has been made for the afterlife of the individuals buried here and it may be that those who disposed of these bodies in a manner so different from the norm were hoping to deny the victims their afterlife.

The spearpoint that was embedded in the spine of the older man (Fig. 7) indicates a form and metallurgical composition which points towards an origin in Austria or Switzerland, at least in terms of the alloying of the metal (high nickel and arsenic content); it dates to around 1400 BC.2 This is slightly earlier than the radiocarbon accelerator date for a humerus of the oldest male skeleton of 2970+30 bp (1320-1050 cal BC at 2 sigma; OxA13092). Spears of this type have traditionally been

The ditch produced the remains of at least four - and very probably five - young men ranging in age from their early teens to their early or mid thirties.1 All the skeletons displayed Schmorl’s nodes which are indicative of a hard physical lifestyle – perhaps farming – though the skeletal 1 Dr Joy Langston examined all the bones from the excavations at Tormarton in 1968, 1999 and 2000 and her interpretations and conclusions are summarized here.

2 Dr Peter Northover of the Department of Materials, University of Oxford examined the spearpoint and his conclusions are summarized here.

141

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY described as ‘hunting spears’ because of their small size, though the context of this find shows that Bronze Age people were prepared to use any available weapon. Conventional distinctions between ‘throwing’ and ‘thrusting’ types of spear may well be unsustainable.

FIGURE 6. The older male’s spine with the spearpoint injury.

FIGURE 4. The older male with the injury on the side of his head and the spear wound to his lumbar vertebra. FIGURE 7. The Middle Bronze Age spearpoint found in the older male’s lumbar vertebra.

As we excavated the ditch, we took a series of bulk soil samples from around the burials, and also column samples from the ditch sections themselves in the hope that these would provide further clues to the circumstances which led to the deposition of the bodies, an event which had clearly been undertaken without any ceremony. No macrobotanical carbonized material was recovered, not surprisingly given the lack of any other evidence for occupation or agricultural processing in the vicinity. The results of the analysis of the molluscs present in the column samples were certainly of greater interest.3 Snails from the soil used to fill the ditch represent an almost entirely woodland fauna. Species present include Acicula FIGURE 5. The older male’s skull with trauma on its side. 3 The samples were taken and analysed by Dr Mark Robinson of the University Museum in Oxford; the results of his analysis are presented here.

142

RICHARD OSGOOD: THE DEAD OF TORMARTON - MIDDLE BRONZE AGE COMBAT VICTIMS? fusca, quite a rare mollusc which is characteristic of relatively old, undisturbed woodland. Only small quantities of grassland species such as Vallonia were found. The ditch appears to have been cut through recently cleared woodland and had not been open for long before it was refilled – fitting in with our evidence of a rapid covering over of the bodies, as shown by the limestone slabs and voids. The voids and the presence of human remains below them explain the presence of another mollusc species, Oxychilus cellarius. This is a rock-rubble genus which is carnivorous and found frequently in the skull cavities of human bodies in chambered tombs, such as Ascott-under-Wychwood in Oxfordshire (Chesterman 1977).

Ditches as boundaries In many ways it is the linear ditch or boundary which is of central importance as a context of deposition. Increasingly, the study of these features is being deemed important with the work of Barry Cunliffe at Windy Dido in Hampshire (Cunliffe and Poole 2000) and Bradley, Entwhistle and Cleal on Salisbury Plain in Wiltshire (Bradley et al. 1994). The former excavations revealed some human remains of Iron Age date that had been inserted into one of the Windy Dido Bronze Age linear ditch segments, whilst Bradley also found Iron Age human skeletal material placed deliberately close to the terminal of a Bronze Age linear ditch at Sidbury hillfort (Wiltshire).

Comparisons Pre-existing Bronze Age monuments were certainly important to later societies and the insertion of human remains into them as special deposits might have been vital to groups staking claims to ancestors. As Bradley states, ‘If we were to correlate this [deposition of bone] sequence with what is known about the linear ditches, it would suggest that they originally acted in rather the same way as earthworks around settlements and hillforts. Human identities may have been dispersed among the bone deposits associated with both kinds of boundary.’ (Bradley 2000: 150-1).

What then can we make of this incident in a south Gloucestershire field in the Middle Bronze Age? Certainly there are parallels in other parts of Europe for the discovery of Bronze Age bodies with bronze weaponry embedded in them. Notable examples come from grave 122 at Hernádkak in Hungary where a spear is fixed in the pelvis of a male individual. This is particularly interesting because it comes from an established cemetery site in which other burials were provided with grave goods, including boar’s tusks (Bóna 1975: 150). A further Tumulus Culture burial from Klings in Germany (Feustel 1958: 8) also revealed a projectile wound, a bronze arrowhead embedded in a vertebra.

The fact remains, however, that the Tormarton bodies were Middle Bronze Age, and were placed in a Bronze Age ditch: the two were contemporary and - what is more – the bodies were deposited without ceremony. At Tormarton, the linear ditch is central to events or, at least, the culminating event. It might well have been a boundary, one element of the landscape clearance of the period, connected to pressure for land and resources and the emergence of group identities. Intriguingly, the sense of a border or boundary has long been important here. Tormarton lies on the present Wiltshire/Gloucestershire border and was also on the Wessex/Mercia territorial edge. Tormarton itself means ‘thorn tree on the boundary’, neighbouring Marshfield ‘field on the Marches (borders)’, and the names of nearby Rodmarton and Didmarton have similar connotations of boundaries. Although I am not suggesting that these names are connected to Bronze Age boundaries, there is nevertheless a rather neat indication of the importance of boundaries in the region’s past.

The only other example in Britain of a bronze weapon embedded in a human body comes from Dorchester on Thames in Oxfordshire (Osgood et al. 2000: 22). Rather like the case of the corpse from Hernádkak, in this instance also a spear pierced the pelvis of the victim. A radiocarbon date of 1260-990 cal BC has been obtained for this individual. Pre-mortem or post-mortem injuries? One should perhaps sound a note of caution against automatically considering projectile points embedded in human bones as an indication of a death in combat. The Iron Age cemeteries of East Yorkshire at Rudstone and Garton Slack both have several graves in which the interred body had been pierced by several spearheads (Stead 1991). Stead’s interprets these events as having taken place after the body had been laid in the grave. One grave at Garton Slack (GS10) contained 14 spearheads, six of which had been driven into the corpse after death. Grave GS7 had 11 spearheads, with five of them transfixing the body. Certainly there seems to have been an element of deliberate ‘ritual’ killing of an already deceased individual, in much the same manner as the destruction of shields (South Cadbury; Coles et al. 1999), swords (Flag Fen; Pryor 2001) and phalerae (Melksham; Osgood 1995) which we find in Late Bronze Age Britain.

Another example of possible antiquity in territorial boundaries comes from nearby. Excavations at Ashton Keynes on the current Wiltshire/Gloucestershire border have revealed an extensive alignment of Middle and Late Bronze Age pits (with some special deposition of human skeletal material) which ran directly along the current county and parish boundaries, boundaries known to have been in place by Domesday in 1087 (G. Hey, pers. comm.). Rather than forming a genuine physical barrier, this pit alignment was a symbolic border close to

143

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY prehistoric field systems. As the Tormarton ditch was only 70m in length, it too probably formed no real barrier.

Bibliography Bóna, I. 1975. Die Mittlere Bronzezeit Ungarns und ihre südöstlichen Beziehungen. Archaeologia Hungarica 49. Bradley, R. 2000. The Archaeology of Natural Places. London: Routledge. Bradley, R., Entwistle, R. and Raymond, F. 1994. Prehistoric Land Divisions on Salisbury Plain: the work of the Wessex Linear Ditches Project. London: English Heritage Archaeology Report 2. Brück, J. 1995. A place for the dead: the role of human remains in the Late Bronze Age. Proceedings of the Prehistoric Society 61: 245-77. Buckley, D. and Hedges, J. 1987. The Bronze Age and Saxon Settlements at Springfield Lyons, Essex: an interim report. Chelmsford: Essex County Council Occasional Paper 5. Chesterman, J.T. 1977. Burial rites in a Cotswold long barrow. Man (NS) 12: 22-32. Coles, J.M., Leach, P., Minnitt, S.C., Tabor, R. and Wilson, A.S. 1999. A Later Bronze Age shield from South Cadbury, Somerset, England. Antiquity 73: 3349. Cunliffe, B.W. and Poole, C.E. 2000. The Danebury Environs Programme. The Prehistory of a Wessex Landscape. Vol. 2 Part 7: Windy Dido, Cholderton, Hampshire 1995. Oxford: English Heritage and OUCA Monograph 49. Feustel, R. 1958. Bronzezeitliche Hügelgräberkultur in Gebeit von Schwarza (Sudthüringen). Weimar: Herman Böhlaus Nachfolger. Knight, R.W., Browne, C. and Grinsell, L.V. 1972. Prehistoric skeletons from Tormarton. Transactions of the Bristol and Gloucester Archaeological Society 91: 14-17. Osgood, R.H. 1995. Three bronze phalerae from the River Avon near Melksham. Wiltshire Archaeological and Natural History Magazine 88: 50-9. Osgood, R. and Monks, S. with Toms, J. 2000. Bronze Age Warfare. Stroud: Sutton. Parker Pearson, M. 1993. Bronze Age Britain. London: Batsford and English Heritage. Pryor, F. 2001. The Flag Fen Basin: archaeology and environment of a fenland landscape. Swindon: English Heritage Archaeological Report. Stead, I.M. 1991. Iron Age Cemeteries in East Yorkshire: excavations at Burton Fleming, Rudston, Garton-onthe-Wolds and Kirkburn. London: English Heritage Archaeological Report 22 with British Museum Press.

The Middle Bronze Age was a period of great change in terms of settlement practice, crop regimes, burial modes and the division of the landscape. Boundaries were very real and had practical as well as ideological elements, backing up claims to land. As yet Tormarton is perhaps the only site which reveals the threat implicit in the creation and maintenance of boundaries. Possibly the ditch was dug by a group to lay claim to land. It was a symbolic territorial marker, a demarcation of land though, at only 70m, it might easily have been bypassed. I would postulate that it was those connected with the construction of this ditch who ended up within it, when it was filled quickly soon after the vast effort of cutting it. The limestone slabs dug out in order to form the ditch and perhaps used as an upcast bank - were thrown in to cover the corpses. This does not appear to have been a votive deposit as the bodies were thrown into the ditch, rather than placed in it. These four or five young men were killed in a violent action surrounding the establishment of territory – the retained spearheads and other wounds show the ferocity of the event, with the bronze possibly snapping off as the bodies were stabbed once on the ground. In many societies (such as the Zulus in the nineteenth century) it has been important for a male to ‘wash his spear in blood’ in order to achieve full adult male status. On the deaths of the victims the ditch was filled and perhaps the claims to land crushed. Potentially others in the group were taken as captives, wives or suchlike. In some senses, the exact nature of the various spear blows is unimportant. A deed of violence occurred here, although Hobbes’ famous claim that early humans lived a life which was ‘nasty, brutish and short’ may, perhaps, be a little strong. Whatever the exact scenario, it is clear that Tormarton shows that certain small-scale yet very nasty actions did occur in the British Bronze Age and life was not always a peaceful rural idyll.

144

Giving up weapons David Fontijn Faculty of Archaeology, University of Leiden

Some years ago excavations in Wassenaar, in the west of the Netherlands, uncovered a remarkable site: a multiple burial dated to the transition from the Early to the Middle Bronze Age. The grave contained the remains of twelve individuals - men, women and children - who must have died in a violent conflict and were buried together (Fig. 1; Louwe Kooijmans 1993a; Smits and Maat 1993). Both in archaeological circles and in the popular press the find created quite a stir since it provided indications of violent conflict on a scale that was unexpected in the Dutch Bronze Age. The find astonished Dutch archaeologists because all available evidence had hitherto supported the notion of a relatively peaceful society during this period (Louwe Kooijmans 1993a: 18). There is indeed no evidence at all for fortified/defended residences at this time and warrior’s graves are also extremely rare among the more than a thousand Bronze Age burials known from the Netherlands (ibid.: 14-18).

support the idea of a Bronze Age that was more violent than we have realized. A closer inspection of the weapons themselves and their depositional context, however, indicates that things are more complex than that. In this paper, I shall argue that this new evidence does not so much inform us of the behavioural reality of warfare but rather reveals the growing significance of an ideology of martiality. The discussion will be based on the metalwork evidence from one of the richest regions, the southern Netherlands and the adjacent part of northern Belgium (Fontijn 2002), which is defined by the river Rhine in the north, the Scheldt in the west, the Meuse in the east and the Demer in the south. The weapons Before exploring the problem of weapon deposition it must be made clear what is understood by the term ‘weapon’. If we define weapons as those objects that are used to enact violence on other human beings, then almost any tool with a heavy weight or a sharp cutting edge can be considered as such. Skeletal evidence from European prehistory shows that people were indeed attacked with bows and arrows, axes and daggers (for examples see Smits and Maat 1993; Osgood et al. 2000), but we know that these objects were used for other activities as well. It is only from the Middle Bronze Age onwards that we have convincing evidence that a particular artefact must have been designed primarily for battle: the bronze sword (Treherne 1995: 109). The following discussion will therefore be based mainly on the evidence provided by swords, although we shall occasionally turn our attention to spears as well. After all, most bronze spears must have been designed for and used in battle as well. Although in theory spears might have been useful for hunting purposes, they are mainly thrusting weapons, in contrast to javelins. Archaeozoological studies of Bronze Age settlements, moreover, show that the role of hunting was peripheral by the Middle Bronze Age.2 This suggests that the presence of large numbers of spears in the archaeological record must first and foremost be related to their role as weapon.

The Wassenaar find neatly illustrates that the conceptualization of Bronze Age communities as peaceful ones had been somewhat naïve and idealistic and yet the evidence we have for this period does not support the bellicose Hobbesian alternative either (cf Vandkilde 2003). We are dealing with largely egalitarian communities, with no indications at all of a warrior aristocracy as is claimed to have existed in other European regions (Harding 1999: 169; Kristiansen 2000). Still, both Louwe Kooijmans (1993a; 1998) and Fokkens (1999) have argued that armed conflict might well have been a regular, even endemic, practice among the Dutch and Belgian Bronze Age communities. Such conflicts must have been small-scale ones, however, involving modest war-like encounters possibly organized around prestigious cattle raids. In this paper, I wish to pursue Louwe Kooijmans’ and Fokkens’ analysis of Bronze Age warfare in the Low Countries by focusing on a category of evidence that has hitherto been neglected: the weapons themselves. When publishing their ideas, Louwe Kooijmans and Fokkens were able to refer to only a small amount of published Bronze Age weaponry from the Low Countries (Butler 1963; O’Connor 1980). Recent surveys of all Bronze Age metalwork finds1 from specific regions in the Netherlands and Belgium indicate that weaponry was a much more important category in material culture than was once thought, particularly in the regions south of the Rhine. On the face of it, this new evidence seems to

Some 68 bronze swords and over 150 spears are known from the southern Netherlands. The number of swords is modest when compared to those from other regions. In Britain, for example, an average of 2.87 swords have been found per 1000 sq. km. (Harding 2000: table 8.1); in the southern Netherlands, the average density of swords is as low as 0.23 swords per 1000 sq. km. (Fontijn 2002: table 11.1). The number of weapon finds from the

1

Butler has published several important surveys of bronze artefacts in the Netherlands; for the present discussion his 1987 publication is the most interesting. See Fontijn 2002 for the southern Netherlands and adjacent part of northern Belgium. See Warmenbol 1992 and Verlaeckt 1996 for the Belgian province of Oost-Vlaanderen.

2 See Clason 1999; Louwe Kooijmans 1993b; Ijzereef 1981; Van Dijk et al. 2002: 607-11; Schoneveld 2001: 187-8.

145

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY

FIGURE 1. The multiple burial at Wassenaar (courtesy of Faculty of Archaeology, Leiden).

‘swords’ are no more than lengthened daggers, either with a broad blade (‘dirk’) or a narrow one (‘rapier’). They cannot have been very versatile objects and, because of the vulnerable hilt-blade connection, their use in battle must have been restricted: we are dealing with bronze blades that are connected to an organic hilt with rivets or notches (cf the Wohlde sword in Fig. 2). The vulnerability of this construction method is verified by recurrent damaged and reworked sword butts (Fontijn 2002: appendices 5.1-5.2). These swords were practical only for stabbing at (very) close quarters and not for parrying blows or hacking. Harding (1999: 166) even questions whether a rapier thrust could cause a fatal wound.

southern Netherlands is, however, much higher than was originally thought and far outnumbers finds in the adjacent regions (ibid.: table 11.2). The earliest swords from the Low Countries are dirks and rapiers of the Sögel and Wohlde type; these had hilts of an organic material and date to 1600-1500 BC (Fontijn 2002: 100-3). Bronze thrusting spears were probably introduced in this same period (cf Harding 2000: 281). Fierce ‘Homeric’ warriors carrying swords figure in many popular accounts of the European Bronze Age, and are to some extent the Leitmotiv of the bellicose version of this period’s history (cf Vandkilde 2003). Such views blind us to the fact that, for a long time in their history, swords were quite impractical weapons. The early 146

DAVID FONTIJN: GIVING UP WEAPONS It is only with the introduction of metal-hilted swords at the end of the Middle Bronze Age (around c. 1000 BC) that swords became less vulnerable weapons. True, versatile and practical ‘cut-andthrust’ swords are not introduced until the last centuries of the Bronze Age (Bronze final IIIb / Ha B 3; c. 925-800 BC; Fontijn 2002: 170-1). The fact that swords became an inextricable element of material culture from the moment of their introduction should, therefore, not be interpreted as a progressive development towards a more effective, rational kind of warfare. Rather the contrary: these swords indicate the significance of a particular kind of close-range fighting, which was probably guided by rules and codes. In the case of the long rapiers of the later part of the Middle Bronze Age (1300-1000 BC), we may even infer that effective rapier fighting required special training (Osgood et al. 2000: 23). However ineffective and impractical the early swords may seem to us, the impact marks, torn rivet holes and extensively resharpened and refashioned blades and butt ends all indicate that many of these were nevertheless used in battle. The same is true for most spearheads (Fontijn 2002: appendices 5.1-5.5, 6.1-6.3).

FIGURE 2. Contents of the Overloon hoard: Wohlde rapiers (1-2), spearheads of Torsted type (3) and Bagterp type (4), nick-flanged axe (5) and Bargloy pin/needle (6) (after Butler 1990: fig.15; copyright Groningen Institute of Archaeology [formerly BAI], University of Groningen).

of bodily adornment was part of this warrior appearance (ibid.: 227-9, appendix 5.6).

In spite of their limited use value, there are clear indications that swords were held in high esteem. Some of them bear fine decoration (Sögel dirks), in contrast to the plain character of contemporary metalwork in the region. Virtually all swords known from the southern Netherlands must have been imports from distant regions (Fontijn 2002: 222), or – if locally produced - modelled according to a supra-regional style. Some sword finds are associated with warrior equipment that recalls weaponry from remote regions. Such finds illustrate how some sword-bearing individuals were defined by their membership of non-local, ‘imagined’ elite communities (Isbell 2000; Fontijn 2002: 229). This is most clear in the case of the warrior sets with Wohlde swords from Overloon in the southern Netherlands (Fig. 2), and in the case of the rich Sögel grave from Drouwen in the north of the Netherlands (Fontijn 2002: 103, 228). From the occasional presence of razors, tweezers and ornaments in such warrior sets we may deduce that a particular mode

Above all, the special significance of swords in the Low Countries is illustrated by the fact that they are one of the few bronze artefact types for which ceremonial versions existed. For the Middle Bronze Age, the giant, aggrandized ‘swords’ of the Plougrescant-Ommerschans type and of the Caistor St. Edmunds-Melle series can be mentioned (Butler 1990; Fontijn 2001; Needham 1990). For the Late Bronze Age, one can consider the decorated solid-hilted swords of the Mörigen type and – especiallythe Dreiwulstschwert from Buggenum (Fontijn 2002: 166-9, fig. 8.12). The fact that a high-quality ceremonial item celebrates the sword indicates the high symbolic value attributed to this peculiar weapon.3 3 Spears must have been much less significant than swords but a ceremonial version of a spearhead is known: the giant example from Tollebeek in the north of the Netherlands (Butler and Hogestijn 1988).

147

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Middle (1800-1050 BC) and the Late Bronze Age (1050800 BC) in one particular micro-region of the southern Netherlands (Midden-Limburg). It neatly illustrates that most bronzes, including weapons, come from a variety of environments which are predominantly wet. On the dry parts of the landscape bronzes can survive, as is shown by the small finds from the urnfield sites. However, in spite of intensive surveys and excavations, bronze finds are extremely scarce and this seems to represent a prehistoric reality.

To sum up, swords may be the earliest specialized weapon in our region but they do not indicate a move to a more effective or intensive kind of warfare. Rather, they present us with a peculiar, apparently highly valued, commitment to close-range fighting which was probably governed by rules and codes. That high quality ceremonial objects take the form of swords illustrates their high symbolic value and significance in ritual practices. The selective deposition of weaponry

The preference for deposition in watery environments does not imply that bronzes were placed there arbitrarily. This is particularly clear in the case of swords. As Figure 3 shows, these were almost exclusively deposited in the region’s major river and hardly ever in the adjacent inland swamps and streams. This preference for placing swords in major rivers is true for the entire study area. Comparison between the depositional patterns of the Middle Bronze Age and those of the Late Bronze Age (Fig. 3) reveals barely any difference. In both periods, swords were placed in rivers, with very few exceptions. This implies that the ideas behind selective deposition were extremely traditional, remaining stable over hundreds of years. Apparently, there was some common understanding of both swords and rivers that made it logical that a sword should be sacrificed in a major river and not – for example – on a high plateau or in a grave. It may be suggested that rivers had specific martial connotations, although the other river finds show that it was certainly not exclusively ‘martial’.

The next step in the argument comes from the depositional context of weapon finds. A recent investigation of the depositional contexts of bronze finds from the southern Netherlands showed that contextual information could be found for 69 % of all finds.4 An overwhelming majority of these come from watery places (rivers, marshes, stream valleys; Fontijn 2002). This is also true for swords and spears. In contrast to northern regions, weapons are conspicuously absent from burial sites, in spite of the fact that both Middle and Late Bronze Age burials are known in large numbers and have been intensively studied.5 Analysis of all metalwork finds shows that the predominance of bronze weaponry in such watery places cannot be the product of research factors or post-depositional preservation (ibid.: chap. 4). Objects were deliberately given up by prehistoric communities and deposited at particular places in the landscape. We seem to be dealing with a highly traditional system of selective deposition: particular kinds of objects were deposited in particular places only and not in others (ibid.: chap.15). This is particularly clear in the deposition of weapons, which seem to have been rigidly kept out of burial contexts.

It is particularly for the last phase of the Late Bronze Age (Ha B3 / Bronze final IIIb) that the outlines of a structured, specialized sacrificial landscape become clear. For this period a considerable number of multiple-object hoards are known from the southern Netherlands and Belgium, predominantly consisting of Plainseau axes and ornaments. Many of them are situated close to major rivers in which swords were deposited. The content of such hoards, however, differs considerably from the river finds. The hoards often contain lavish ornaments; some might have been part of (female?) costume and can be considered to be the paraphernalia of specific personal identities (Fontijn 2002: fig. 8.22). Swords and spears, however, are absent. The Plainseau ornaments, on the other hand, are hardly known from the rivers where numerous weapons and other small ornaments have been found (ibid.: chap. 12).

Estimates of the amount of bronze in circulation indicate that it was only a small proportion of bronze artefacts that did not end up in the melting pot (Huth 2003: 48-9). The southern Netherlands experienced a thriving tradition of regional bronze production from the later part of the Middle Bronze Age (from approximately the fifteenth century BC onwards). The most frequent way in which bronzes ended their use-life, therefore, must have been in a pool of scrap, awaiting recycling. Consequently, deposition of bronzes must have been very rare. During the Late Bronze Age, when the highest numbers of objects were deposited, the average rate of deposition in the southern Netherlands can be estimated at almost one deposit per year (ibid.: 214).

How are we to make sense of such long-term patterns in weapon deposition, and what can be learnt from them on the role and meaning of weaponry? Metalwork deposition in the Bronze Age is often explained as a way of upholding the prestigious value of metalwork by creating scarcity (Kristiansen 1978; Rowlands 1980). This explanation, however, relates deposition to politicoeconomic processes alone. It does not make clear why the system of deposition was structured as selective

Figure 3 offers a more detailed illustration of the selective deposition of bronzes in general and weaponry in particular. The distribution of swords, spears, axes and ornaments with known finds contexts are shown for the 4

The large (contextualized) number of metalwork finds from urnfields is not included in this calculation. 5 For Early and Middle Bronze Age barrows, see Theunissen 1999; for Late Bronze Age urnfields, see Roymans 1991; for the metalwork from those graves, see Fontijn 2002: appendices 7.1-7.4.

148

DAVID FONTIJN: GIVING UP WEAPONS

FIGURE 3. Metalwork deposition in the river Meuse and in the adjacent inland marshes in Midden-Limburg for the Middle and Late Bronze Age. Only contextualised finds are mapped (after Fontijn 2002: fig.14.1).

impose their will on others (Claessen 1988: 7-8). One can easily imagine that, in the largely egalitarian society we are dealing with, the possession of weapons by some individuals might have presented a potential threat to social cohesion (Fokkens 1997; Roymans 1996: 14). Yet the issue is probably more complex than that. According to the anthropologist Blok (1994: 34) the use of violence is not just ambiguous in a socio-political way. It is also considered ‘polluting’ as it transgresses the boundary between the categories of life and death. This is the reason why the use of violence is often embedded in ritual practices.

deposits, with different objects ending up in different places in such a traditional way. It brings us to the question of what weapons and depositional zones such as rivers actually meant to people. People and weapons: removing weapons from society The possession of weapons involves more than the possession of any other tool. In essence, a weapon is a tool that enables one to enact violence on others. In a very direct way, weapons are related to power. In sociopolitical terms, weapons tend to be regarded as ambiguous and dangerous because their presence in society implies ‘haves’ and ‘have-nots’. Potentially, it signals the presence of a category of people who can

In the case of swords we have already seen that they must have had important symbolic connotations which 149

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Knowledge of the appropriate place to deposit weapons must therefore have been transferred orally from generation to generation. It thus relates a specific group of people to a specific zone or place in the landscape. It is not inconceivable that knowledge of the ‘proper way of doing things’ was an important ‘social resource’ in the sense of Giddens (1984: 33, 373). The handing-over of such knowledge might have served to create insiders and outsiders and thus to define a sacrificial community, although it is hard to find out which actors would have been involved.

bypassed their practical use-value. Indeed they seem to signal a specific kind of personhood. The fact that swords are sometimes one component of an entire set of objects, alongside items for self-adornment (razors, tweezers and ornaments), is an indication for this. Apparently warriorhood is about a specific martial identity; it is about a transformed self (Treherne 1995). By depositing weapons in a river, people removed the paraphernalia of this identity from society in such a way that they could never again be seen, or touched, or used. Deposition must mark, therefore, the deconstruction of a martial identity in a ritual act.

Next, watery places must have been meaningful for cosmological reasons. Apparently, certain environments were considered more suitable for placing weapons in than others and we may expect that they had martial connotations, whereas this was not the case for other environments (for example, burial grounds). It is hard to find out how this was conceptualized by the sacrificial communities themselves. Weapon depositions have often been explained as sacrifices made to (personified) gods, who were supposed to live in rivers or marshes, an idea which finds support in Roman historical sources describing Celtic/Germanic offering practices hundreds of years later. The possibility that the deposited weapons were considered to be gifts may accord with the fact that these objects were hardly ever destroyed, in contrast to what happened to many of the weapons deposited in the Thames, for example (York 2002). Many are in excellent shape and must have been sharpened not long before deposition, as if they were still to be used.

The meanings of depositional zones People were not just removing weapons from society. They were doing this in a structured manner. Particular objects ended up in particular kinds of places only. The rigid practice of depositing swords in major rivers, continuing for hundreds of years, is a case in point. This implies that such places were considered meaningful in themselves. I do not claim here that we are able to uncover such meanings in a detailed sense but some points can be made. Firstly, we must be dealing with zones in the landscape that were meaningful in social terms. There seems to be an emphasis on deposition in strictly delimited zones that represent boundaries and transitions in the landscape (Fig. 3). It is generally assumed that many of the streams, marshes and rivers that saw deposition represent the boundaries between different social groups. As such, they are somewhat ambiguous themselves, as it is hard to see where in a marsh or river a dividing line may be perceived between ‘us’ and ‘them’. At any rate, archaeology shows that objects were deposited in a diffuse pattern in watery places. In the case of both the major river and swamps, depositions took place on both sides (Fig. 3). Interestingly, the river is not only about separation; it is also a line which relates the local communities to the outside world and, as such, is the common point of reference for inhabitants of the river valley, whether they know each other or not. In the social landscape, the major river must have been of the utmost importance. Indeed, most of the bronze valuables that were imported from far away must have reached this region via this same river.

On the other hand, several methodological and archaeological objections can be made to the historic parallel with the Germanic/Celtic period (Fontijn 2002: 267-8), including the point that Bronze Age depositions deviate in certain aspects from those of the Late Iron Age. The martial connotations of some environments might also have been regarded in a more general way. Pálsson (1996: 74) and Gurevich (1992: 178) give ethnographic and historic examples of cosmologies in which personal qualities and values are regarded as inextricably linked to land. Zegwaard’s ethnography (1956) of the Asmat (Irian Jaya) is also interesting in this context, as it provides an extreme example of a cosmology in which the landscape is directly interpreted in terms of martial values. Archaeology by itself is probably insufficient for shedding more light on the particular way in which Bronze Age people perceived such places while depositing weapons. Still, the persistent identification between weapons and particular environmental zones suggests strongly that warfare and martial values were, in one way or another. seen as related to a wider cosmology of the land.

Finally, as socially meaningful environments, depositional zones might have played an important role in the way in which the local communities defined themselves. Although there was a widespread preference in Atlantic Europe for depositing swords in rivers (Bradley 1990; Torbrügge 1970/1971), some river zones saw much more deposition than others. The fact that such specific zones – and not others - were selected time and time again for weapon deposition implies specific historical knowledge on the part of the participants. After all, there is no evidence at all for the presence of lasting markers indicating such a preferred depositional zone.

Why were weapons deposited? In attempting to understand why people practiced weapon deposition, our starting point will be the socially and culturally ambiguous character of weapons. 150

DAVID FONTIJN: GIVING UP WEAPONS southern Netherlands, there are also large numbers of human bones known from rivers (Ter Schegget 1999: fig. 2, tables 1-2). Of the 27 available radiocarbon dates, only two can be attributed to the Bronze Age, and these finds are not from locations known to have yielded metalwork (Fontijn 2002: 230). From other watery places (marshes and swamps) there is so far not a single shred of evidence which links metalwork finds to human bodies. All in all, there is no basis for refutation of this theory, but there is not much to support it either.

Deposition related to rites of passage during life One interpretation is that weaponry was surrendered and deposited at a certain stage, probably when the warrior became an elder. This has been suggested previously by me (Fontijn 1999) and by Fokkens (1999) as the most viable explanation. Given the low proportion of weapons that ended their life in deposition, it is clear that it must have been the exception rather than the rule. For example, a kit of weapons might have been given up only when there was no suitable heir available. This theory may still be attractive but I am now of the opinion that it cannot be the only explanation. It presupposes that warriorhood was a regular stage in the life-cycle of males, but this may well be an overestimation of the role of warriors in the largely egalitarian Dutch-Belgian Bronze Age communities. On top of that, in Middle and Late Bronze Age burials we find the remains of both older and younger males (potentially individuals who had a warrior status). None, however, carries weapons. The avoidance of weaponry in the burials known to us seems absolute and cannot be explained by the theory that all these people had already passed the ‘warrior stage of life’. We still need an additional theory to account for the non-martial character of graves.

Deposition as a means of removing ‘polluted’ weaponry: the sacrifice of booty? We should also consider the possibility that weaponry was disposed of not because it signalled one’s own identity, but rather because it derived its ambiguity from the fact that it belonged to enemy groups. Deposition of booty is generally thought to have existed in the Nordic Germanic Iron Age (Randsborg 1995; Wenskus 1976). For the Bronze Age in the southern Netherlands, however, such an explanation is unlikely for most of the deposited weapons. Most finds are single deposits, or loosely associated groups in a spatially confined area. Only in the case of finds dredged from rivers (a circumstance which often yields large collections of weapons), might the situation have been different. As already remarked, most of the objects seem to have been resharpened before deposition and were undamaged, which suggests that people took care of them until they were deposited (Fontijn 2002: 212). Only in the case of the remarkable Late Bronze Age weapon hoard from a marshy stream valley of Pulle is the situation different (ibid.: 169-70; Van Impe 1973). Here fragments were found of at least five swords, eight spears and one axe: all were intentionally broken and some had been burnt before deposition. The cutting edges of the axe and the tip of five spears had been intentionally bent downwards after they were heated. This example of ostentatious destruction is completely exceptional in our region. As such, the Pulle hoard may be the closest to a situation in which ‘polluted’ weaponry was really intentionally destroyed and removed from society.

Deposition at death: weapon hoards as funeral hoards This explanation is a variation on the above, but with deposition presumed to have taken place at death, the final rite of passage. It would be in line with a ‘taboo’ on weaponry in graves if these items were buried separately. This Totenschätze explanation has been suggested earlier by Roymans and Kortlang (1999) and Warmenbol (1996) and has its roots in studies by Bradley (1990), Torbrügge (1970/1971) and Wegner (1976). The basis for the theory is the argument that weapon sets were once deposited in graves, with a change at a later period to deposition in watery places. For that reason, a relationship between the later hoards and funerary rites seems viable. In the case of the southern Netherlands, however, this interpretation fits less well, because weapons were excluded from graves in this region from the very beginning. The funeral-hoard theory cannot be dismissed on this basis, but the links between the burial ritual and deposition of valuables elsewhere are less close than in other regions.

Martial identity as a temporary state of the self that is terminated by deposition As a general explanation, I think a more nuanced version of ‘life-cycle related deposition’ is needed, one which is more in accordance with the modest rate at which weaponry was deposited, and with the simple, small-scale and largely egalitarian nature of the Dutch-Belgian Bronze Age communities. The assumption that warriorhood was a fixed identity during a particular stage in the life-cycle may well overemphasize the significance, nature and organization of ‘warriors’ in such small-scale societies (cf Vandkilde 2003: 138). It seems more appropriate that ‘warriorhood’ was only a temporary, context-dependent identity.

Depositions as the remains of graves The most extreme version of the funeral-hoard theory is that weapon depositions are themselves the remains of warrior graves, which were kept separate from the barrows or urnfields known to us. In southern Britain it is indeed true that the Thames has yielded not only a large number of Bronze Age weapons but also many human skulls. Radiocarbon dating has shown that these are from the same period (Bradley and Gordon 1988). Although a true association between weapons and human bones cannot be proven, the theory is attractive. For the

151

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY must have been designed as weapons in the first place (swords in particular). Furthermore, these ‘tools of warfare’ are now more visible archaeologically because they figure in depositional practices that have a greater potential for preservation (they are placed in waterlogged environments). That we are dealing here with an intense commitment to ideological aspects of warfare rather than an increased intensity of warfare can be deduced from the following observations:

An interesting ethnographic study which elucidates how warrior identities were constructed and de-constructed in small-scale tribal societies is Harrison’s study of endemic warfare for the Sepik of Papua New Guinea (1995). He shows how people living in neighbouring areas and who are even related by descent may still fight each other from time to time. Their assumption seems to be that there are not distinct categories of people (such as friend or enemy, warrior or non-warrior). Rather, the same people can be violent or sociable towards one another in different contexts. For neighbours to meet each other as enemies, a temporary shift in identity is needed. Such shifts are ritualized and the identities are acquired by a transformation of the self through self-decoration (the wearing of specific fighting regalia). By thus dressing themselves, people take on a temporary, martial group identity. True warrior bands or retinues do not exist; nor is the role of warrior a pre-ordained element in the process of becoming a person. On occasion people can still be ‘warriors’, but the adoption of such a dangerous and ambiguous identity can take place only in a special, ritualized context. The same goes for the transformation of warriors into ‘normal’ individuals after the time of conflict has passed. Warriorhood seems to be regarded as something on the outer surface of the self ‘which could be worn or shed’.

1. The invention, adoption and emergence of swordfighting and the special significance attached to it. 2. The fact that high-quality ceremonial versions of weapons were in circulation, 3. The context in which weapons are found. The majority of weapons seem to have been intentionally removed from society and ‘ritually’ deposited in watery places. In a way, finds of deposited weapons represent disarmament rather than warfare. I have argued that these are part and parcel of a well-established and highly traditional system of selective deposition of valuables. Particular kinds of objects were deliberately placed in particular places in the land and not in others. Amongst other things, it implies that both the deposited weapons themselves and the locations in which they were placed had special meaning. Certain zones in the landscape must for some reason have been regarded as places with ‘martial’ connotations.

This example is interesting for several reasons. It illustrates how it is possible for people in simple, smallscale societies to fight each other in one context and be sociable in another, owing to the ritual construction and deconstruction of warrior identities. In the Sepik example, the significance of specific bodily ornamentation in this temporary shift of identity recalls my observation that Bronze Age warrior equipment sometimes includes items which seem to imply a transformed self (razors, tweezers, ornaments and dressfittings). In particular, it is the role of ritual in the deconstruction of martial identities which may remind us of the practice of deposition: the deliberate removal of warrior paraphernalia in a ritualized way.

These sacrificed weapons are indicative not so much of the importance of warfare, but rather of the central role of weapons in sacrificial practices. Weapons are more than just tools; they are the paraphernalia of martial identities. I have argued that such identities must have been to some extent ‘ambiguous’ and ‘dangerous’. This may be in line with the archaeological evidence which, in the first place, informs us of the ‘ritual removal’ of so much weaponry from society. Deposition of martial paraphernalia may then imply that a deconstruction of personal or group identities took place. Several scenarios have been suggested which relate such deposition to particular events in the life-cycle of individuals or groups. Given the relatively small-scale and egalitarian nature of the communities in question it seems appropriate to consider warriorhood as an essentially temporary identity, reserved for special situations and contexts only.

Conclusion Central to this paper is the debate as to whether the Bronze Age of the Low Countries was a violent rather than peaceful period. In particular, it focuses on the recent observation that the number of weapons from this period appears to be much larger than was originally thought. Surprisingly, a detailed study of these weapons and their depositional context shows that the larger number of weapon finds does not simply reflect an increase in warfare. It does show an unequivocal commitment to the social values behind the weaponry, an ideology of martiality. Whether such an ideological concern was greater than in the previous period is a question that must remain unanswered. We can only say that it is more visible archaeologically, owing to the fact that we are now in a better position to recognize tools that

Reviewing the argument, we can see that an attempt to relate weaponry in a direct way to the organization, intensity and practical nature of warfare would fail. Instead, the weapon evidence confronts us first and foremost with its ideological and ritual dimensions. We should not see this as an impediment to a further understanding of Bronze Age warfare, but rather as a starting point for grasping what it really was: a multifaceted field of practice which was just as much about ideological values as it was about killing people.

152

DAVID FONTIJN: GIVING UP WEAPONS the structure of metalwork exchange. In W.H. Metz, B.L. van Beek and H. Steegstra (eds) Patina: essays presented to Jay Jordan Butler on the occasion of his 80th birthday. Groningen & Amsterdam. 263-80. Fontijn, D.R. 2002. Sacrificial Landscapes: cultural biographies of persons, objects, and ‘natural’ places in the Bronze Age of the southern Netherlands, 2300600 BC.. Leiden: Faculty of Archaeology, Analecta Praehistorica Leidensia 33/34. Giddens, A. 1984. The Constitution of Society: outline of a theory of structuration. Cambridge: Polity Press. Gurevich, A. 1992. Historical Anthropology of the Middle Ages (ed. J. Howlett). Cambridge: Polity Press. Harding, A.F. 1999. Warfare: a defining characteristic of Bronze Age Europe? In J. Carman and A. Harding (eds) Ancient Warfare: archaeological perspectives. Stroud: Sutton. 157-73. Harding, A.F. 2000. European Societies in the Bronze Age. Cambridge: Cambridge University Press. Harrison, S. 1995. Transformations of identity in Sepik warfare. In M. Strathern (ed.) Shifting Contexts: transformations in anthropological knowledge. London & New York: Routledge. 81-97. Huth, C. 2003. Poor Belgium, rich Belgium: some reflections on the nature of metalwork depositions in the Late Bronze Age and Early Iron Age. In J. Bourgeois, I. Bourgeois and B. Cherretté (eds) Bronze Age and Iron Age Communities in NorthWestern Europe. Brussels: Koninklijke Vlaamse Academie van België voor Wetenschappen en Kunsten. 39-60. IJzereef, G.F. 1981. Bronze Age animal bones from Bovenkarspel: the excavation at Het Valkje. Nederlandse Oudheden 10. Impe, L. van. 1973. Enkele wapens uit de Bronstijd te Pulle. Archaeologica Belgica 150. Isbell, W. 2000. What we should be studying: the ‘imagined community’ and the ‘natural community’. In M.A. Canuto and J. Yaeger (eds) The Archaeology of Communities: a New World perspective. London & New York: Routledge. 243-66. Kristiansen, K. 1978. The consumption of wealth in Bronze Age Denmark: a study in the dynamics of economic processes in tribal societies. In K. Kristiansen and C. Paludan-Müller (eds) New Directions in Scandinavian Archaeology. Copenhagen: National Museum of Denmark. 158-90. Kristiansen, K. 2000. Europe Before History. Cambridge: Cambridge University Press. Louwe Kooijmans, L.P. 1993a. An Early/Middle Bronze Age multiple burial at Wassenaar, the Netherlands. Analecta Praehistorica Leidensia 26: 1-20. Louwe Kooijmans, L.P. 1993b. The Mesolithic/Neolithic transformation in the Lower Rhine basin. In P. Bogucki (ed.) Case Studies in European Prehistory. Boca Raton FL: CRC Press. 95-145. Louwe Kooijmans, L.P. 1998. Bronzezeitlichen Bauern in und um die niederländische Delta-Niederung. In B.

Acknowledgements I am much obliged to Alexander Verpoorte (University of Leiden) and Ruurd Kok (Archaeological heritage manager for the province of Utrecht) for their critical comments on earlier drafts. Bibliography Blok, A. 1994. Zinloos en zinvol geweld. In H. Driessen and H. de Jonge (eds) In de Ban van Betekenis: proeven van symbolische antropologie. Nijmegen: SUN. 27-45. Bradley, R. 1990. The Passage of Arms: an archaeological analysis of prehistoric hoards and votive deposits. Cambridge: Cambridge University Press. Bradley, R. and Gordon, K. 1988. Human skulls from the River Thames: their dating and significance. Antiquity 62: 503-9. Butler, J.J. 1963. Bronze Age connections across the North Sea. A study in prehistoric trade and industrial relations between the British Isles, the Netherlands, north Germany and Scandinavia, c.1700-700 B.C. Palaeohistoria 9: 1-286. Butler, J.J. 1987. Bronze Age connections: France and the Netherlands. Palaeohistoria 29: 9-34. Butler, J.J. 1990. Bronze Age metal and amber in the Netherlands, 1. Palaeohistoria 32: 47-110. Butler, J.J. and Hogestijn, J.W.H. 1988. The Tollebeek spearhead. Palaeohistoria 30: 109-23. Claessen, H.J.M. 1988. Over de Politiek Denkende en Handelende Mens: Een inleiding tot de politieke antropologie. Assen & Maastricht: Van Gorcum. Clason, A.T. 1999. What’s new in the Bronze Age? The livestock and gamebag of the Bronze Age farmers in the western and central Netherlands. In H. Sarfatij, W.J.H. Verwers and P.J. Woltering (eds) In Discussion with the Past: archaeological studies presented to W.A. van Es. Zwolle: SPA/ROB. 35-40. Dijk, J. van, Esser E. and Zeiler, J.T. 2002. Archeozoölogie. In B.H.F.M. Meijlink and P. Kranendonk (eds) Boeren, Erven, Graven: de boerengemeenschappen van De Bogen bij Meteren (2450-1250 v. Chr.). Amersfoort: Rapportage Archeologische Monumentenzorg 87. 547-665. Fokkens, H. 1997. The genesis of urnfields: economic crisis or ideological change? Antiquity 71: 360-73. Fokkens, H. 1999. Cattle and martiality: changing relations between man and landscape in the Late Neolithic and the Bronze Age. In C. Fabech and J. Ringtved (eds) Settlement and Landscape: proceedings of a conference in Århus, Denmark, May 4-7 1998. Aarhus: Jutland Archaeological Society. 35-43. Fontijn, D.R. 1999. Wie het zwaard opneemt… Het krijgersideaal in de Europese bronstijd. Spiegel Historiae 34: 143-7. Fontijn, D.R. 2001. Rethinking ceremonial dirks of the Plougrescant-Ommerschans type. Some thoughts on 153

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Smits, L. and Maat, G. 1993. An Early/Middle Bronze Age common grave at Wassenaar, the Netherlands: the physical anthropological results. Analecta Praehistorica Leidensia 26: 21-8. Theunissen, E.M. 1999. Midden-Bronstijdsamenlevingen in het Zuiden van de Lage Landen: een evaluatie van het begrip ‘Hilversum-cultuur’. Unpublished PhD thesis, University of Leiden. Torbrügge, W. 1970/1971. Vor- und Frühgeschichtliche Flussfunde: zur Ordnung und Bestimmung einer Denkmälergruppe. Berichte der RömischGermanischen Kommission 51/52: 1-146. Treherne, P. 1995. The warrior's beauty: the masculine body and self-identity in Bronze-Age Europe. Journal of European Archaeology 3: 105-44. Vandkilde, H. 2003. Commemorative tales: archaeological responses to modern myth, politics, and war. World Archaeology 35: 126-44. Verlaeckt, K. (ed.) 1996. Between River and Barrow: a reappraisal of Bronze Age metalwork found in the province of East-Flanders (Belgium). Oxford: BAR International Series 632. Warmenbol, E. 1992. Le matériel de l’âge du bronze: le seau de la drague et le casque du héros. In E. Warmenbol, Y. Cabuy, V. Hurt and N. Cauwe (eds) La Collection Edouard Bernays: néolithique et âge du bronze, époque gallo-romaine et médievale. Brussels: Monographies d'Archéologie Nationale 6. 67-122. Warmenbol, E. 1996. L’or, la mort et les hyperboréens.: la bouche des enfers ou le trou de Han à Han-surLesse. In P. Schauer (ed.) Archäologische Forschungen zum Kultgeschehen in der jüngeren Bronzezeit und frühen Eisenzeit Alteuropas: Ergebnisse eines Kolloquiums in Regensburg 4.-7. Oktober 1993. Regensburg: Universitäts-Verlag and Bonn: Habelt. 202-35. Wenskus, R. 1976. Beute. In H. Beck, H. Jankuhn, K. Ranke and R. Wenskus (eds) Reallexikon der Germanischen Altertumskunde. Berlin & New York: Walter de Gruyter. 323-30. Wegner, G. 1976. Die vorgeschichtlichen Flußfunde aus dem Main und aus dem Rhein bei Mainz. Kallmünz: Materialhefte zur bayerischen Vorgeschichte. Reihe A, Fundinventare und Ausgrabungsbefunde 30. York, J. 2002. The life cycle of Bronze Age metalwork from the Thames. Oxford Journal of Archaeology 21: 77-92. Zegwaard, G.A. 1956. Koppensnellerspraktijken van de Asmatters van Nederlands Nieuw-Guinea. Tilburg: Missionarissen van het Heilig Hart. [Also published in American Anthropologist 1959 (61)].

Hänsel (ed.) Mensch und Umwelt in der Bronzezeit Europas. Kiel: Oetker-Voges Verlag. 327-39. Needham, S. 1990. Middle Bronze Age ceremonial weapons: new finds from Oxborough, Norfolk and Essex/Kent (with a contribution from D. Hook). Antiquaries Journal 70: 239-52. O'Connor, B. 1980. Cross-Channel Relations in the Later Bronze Age: relations between Britain, north-eastern France and the Low Countries during the Later Bronze Age and the Early Iron Age, with particular reference to the metalwork. Oxford: BAR International Series 91. Osgood, R. and Monks, S. with Toms, J. 2000. Bronze Age Warfare. Stroud: Sutton. Pálsson, G. 1996. Human-environmental relations: orientalism, paternalism and communalism. In P. Descola and G. Pálsson (eds) Nature and Society: anthropological perspectives. London: Routledge. 6381. Randsborg, K. 1995. Hjortspring: warfare and sacrifice in early Europe. Aarhus: Aarhus University Press. Rowlands, M.J. 1980. Kinship, alliance and exchange in the European Bronze Age. In J. Barrett and R. Bradley (eds) Settlement and Society in the British Later Bronze Age. Oxford: BAR British Series 83. 1555. Roymans, N. 1991. Late urnfield societies in the northwest European plain and the expanding networks of central European Hallstatt groups. In N. Roymans and F. Theuws (eds) Images of the Past: studies on ancient societies in northwestern Europe. Amsterdam: Studies in Prae- en Protohistorie 7. 9-89. Roymans, N. 1996. The sword or the plough: regional dynamics in the Romanisation of Belgic Gaul and the Rhineland area. In N. Roymans (ed.) From the Sword to the Plough: three studies on the earliest Romanisation of northern Gaul. Amsterdam: Amsterdam Archaeological Studies 1. 9-126. Roymans, N. and Kortlang, F. 1999. Urnfield symbolism, ancestors, and the land in the Lower Rhine region. In F. Theuws and N. Roymans (eds) Land and Ancestors: cultural dynamics in the Urnfield period and the Middle Ages in the southern Netherlands. Amsterdam: Amsterdam Archaeological Studies 4. 33-61. Schegget, M.E. ter. 1999. Late Iron Age skeletal remains from the river Meuse at Kessel: a river cultplace? In F. Theuws and N. Roymans (eds) Land and Ancestors: cultural dynamics in the Urnfield period and the Middle Ages in the southern Netherlands. Amsterdam: Amsterdam Archaeological Studies 4. 99-240. Schoneveld, J. 2001. Synthese. In J. Schoneveld and E. Gehasse (eds) Boog C-Noord: een vindplaats bij Meteren op de overgang van Neolithicum naar Bronstijd. Amersfoort: Rapportage Archeologische Monumentenzorg 84. 181-94.

154

Ritual bondage, violence, slavery and sacrifice in later European prehistory Miranda Aldhouse Green Department of Archaeology, University College Newport

the sallekhana (Dundas 1992: 155-6; Anon. 1981; Kamdar 1993). Generally speaking, however, a more comfortable substitute is sought, often involving a victim deemed to be of lesser value than the sacrificer but, at the same time, regarded as sufficiently similar. In his Alcestis, Euripides explores the theme of substitutive ritual death, that of a wife for her husband, within a context (Athens in the late fifth century BC) in which it was considered unquestionably appropriate for women to sacrifice themselves for their men (Hughes 1999: 1). Both people and animals – usually domestic beasts – were legitimate surrogates (Green 2001), and both may sometimes be perceived as scapegoats for the evils of the community, in a manner essentially similar to the Levitical scapegoat of the Old Testament, a he-goat selected from the herd and driven out into the desert laden with the sins of the Israelites (Leviticus 16, vv. 212).

My setting is the pre-Roman Iron Age and early Roman periods in temperate and northern Europe, where there is sporadic but recurrent archaeological and (for the later phases) literary evidence for the use of physical restraint within socio-ritual contexts. It is fitting to begin with a short passage from the Roman writer Tacitus. In chapter 39 of his Germania, the author alludes to a curious practice taking place in a locus consecratus of the Semnones: ‘In another way, too, reverence is paid to the grove. No one may enter it unless he is bound with a cord. By this, he acknowledges his own inferiority and the power of the deity. Should he chance to fall, he must not get up on his feet again. He must roll out over the ground. All this complex of superstition reflects the belief that in that grove the nation had its birth, and there dwells the god who rules over all, while the rest of the world is subject to his sway.’ (Germania 39; 1948: 132-3)

Several Classical authors refer to the topos of the sacrificial pharmakos dedicated by townsfolk in ritual cleansing ceremonies (Hughes 1991: 139-65): though those of ancient Greek cities were merely cursed and driven out, there are repeated references to a southern Gaulish scapegoat, a lowly being who is fêted for a year before being heaped with imprecations and either stoned to death or cast into the sea (Petronius Servius on Virgil’s Aeneid III, 57; Petronius Frag. 1; Lactantius Placidus, commentary on Statius’s Thebais 10, 793 cited in Hughes 1991: 139-65; Green 1998; R. Garland 1995: 23). In describing Gallic cult-practices taking place at the close of the Iron Age, Julius Caesar illustrates the principle of surrogacy:

The testimony of Tacitus is interesting in so far as it clearly affirms the presence of cognitive linkages between bondage and communication with the sacred. In another passage of the same work, the Roman historian records a close association between captivity and the sacred. In speaking of Germanic treatment of malefactors (Germania 7; 1948: 106), Tacitus comments that ‘… capital punishment, imprisonment and even flogging are allowed to none but the priests, and are not inflicted merely as punishment or on the leaders’ orders, but in obedience to the gods whom they believe to preside over battle’. In this essay, I should like to present arguments that low status, violence, bondage and sacrifice might have been closely integrated in the cosmological perceptions of some European communities during the later first millennium BC. Issues that may profitably be explored include principles of substitution (or surrogacy), marginality and exclusion, foreign-ness and value, and the metaphoric meaning of physical subjugation, including restraint, abuse and denial of identity.

‘The Gauls believe that the gods prefer it if the people executed have been caught in the act of theft or some other crime, but when the supply of such victims runs out, they even go to the extent of sacrificing innocent men.’ (De Bello Gallico VI, 16) In a contiguous passage, Caesar comments on the social organization of mid first-century Gaul, remarking that ‘the common folk are treated almost as slaves’ (DBG VI, 13). The clear implication from the first-quoted comment is that marginal and low-ranking persons were preferentially selected as sacrificial victims. We are perhaps glimpsing a pattern in which ‘unempowerment’ was a factor in choice.

Surrogacy and status A principle underlying the sacrifice of people or animals is that of substitution. The ideal sacrifice to the divine is oneself. In extreme circumstances, such auto-sacrifice still occurs (for instance among some particularly devout Jain groups in India, where certain individuals starve themselves to death in a ritual bodily ‘scouring’ known as

Apart from indications of the sacrifice of those of low or servile status, there is a body of Classical literary testimony on the sacrifice of foreigners, another variety

155

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY reprisals and partly because even execution could be hedged about with ritual and sanctity, hence Tacitus’s reference to the involvement of priests in punishment.

of substitution based, in part at least, I suspect, upon the element of risk posed by ‘other’ people. In Euripides’s drama Andromache, Hecuba - wife of King Priam of Troy – is a legitimate sacrificial scapegoat because she is not Greek (Girard 1977: 80). Herodotus records how the Egyptian ruler Bousiris regularly sacrificed strangers (Histories II, 45; van Straten 1995: 40-7). The most readily available supply of foreigners would be warcaptives, and a wide range of ancient texts make reference to the ritual killing of foreign prisoners, some of them of high rank in their own lands. (Indeed, the ambiguity of status might have played a part in the symbolic efficacy of such victims). Thus, in ‘reprisal’ rituals, Virgil’s Trojan hero Aeneas sacrificed ‘captives he had bound, hands lashed behind’ (Aeneid X, lines 5993; 1957: 243), belonging to his enemy Turnus, prince of Latium, in conscious imitation of Homer’s Greek Iliad hero Achilles, who sacrificed twelve Trojan boys (Iliad XXIII, lines 175-184; 1963: 507-9) in reparation for the death of Patroclus (interestingly a ratio of twelve to one).

In seeking archaeological evidence for the sacrifice of enemy battle-victims, the so-called ‘war sanctuaries’ of Middle Iron Age Gaul, such as Gournay-sur-Aronde (Oise) and Ribemont-sur-Ancre (Somme) provide some of the most likely candidates, with evidence for human bodies and smashed weapons apparently contributing to an exhibition of victory and trophy-taking although, of course, the corpses found at these shrines might have met their end on the battlefield rather than have been sacrificed in the temples. From the third century BC, sacrificial rituals appear to have taken place at Gournay, involving the dedication of weapons, animals and human bodies, most of the latter being those of young men of prime fighting age. The heads of the slain were selected for suspension at the shrine’s entrance, along with those of cattle; the limbbones were carefully deposited in the enclosure ditch (Brunaux 1996: 69-97; Brunaux et al. 1985). The forms of the weapons and defensive armour deposited here suggest that their owners were not local to the community and it may, therefore, be permissible to infer that, as outsiders, the soldiers represented were enemy warriors whose bodies and martial accoutrements were offered up in thanksgiving to the gods presiding over Gournay and its environs.

Plutarch’s Life of Themistocles (XIII, 2) narrates an episode of aversion-sacrifice, in which the eponymous Greek war-leader sacrificed young Persian prisoners of war before the battle of Salamis in 480 BC, though the practice was repugnant to him as a ‘civilized’ Athenian. Some centuries later, Diodorus (History V, 32, 6) and Strabo (Geography VII, 2, 3) each mention the sacrifice of POWs to the gods, respectively among the Gauls and the Cimbri (cited in Tierney 1959/60) and - not to be outdone - Tacitus alludes to the blood and entrails of prisoners (presumably war-victims) in the sacred grove on Anglesey in the mid first century AD (Annals XIV, 30-31).

The huge sacred precinct at Ribemont-sur-Ancre was broadly coeval with Gournay; it consisted of a quadrangular enclosure, in each of the two surviving corners of which an ‘ossuary’ had been constructed of human leg-bones with those of horses carefully placed around the outside. These bone-altars were built logcabin fashion. The human remains were, again, of those in the prime of life, each structure representing about 200 people, and close examination revealed a complex postmortem ‘biography’ involving decapitation, the heads being removed from the shrine, and dismemberment while flesh and connective tissue had not yet decomposed. More significantly still, the cut-marks on the bones, made by a thin, sword-like implement, suggest that butchery took place while the bodies hung upsidedown (Cadoux 1984; Brunaux 1988; 1996: 77-90; Brunaux et al. 1985), surely the ultimate in degradation.

War-captives and sacrifice ‘To Mars, when they [the Gauls] have determined on a decisive battle, they dedicate as a rule whatever spoil they may take. After a victory they sacrifice such living things as they have taken, and all the other effects they gather into one place. In many states heaps of such objects are to be seen piled up in hallowed spots, and it has not often happened that a man, in defiance of religious scruple, has dared to conceal such spoils in his house or to remove them from their place, and the most grievous punishment, with torture, is ordained for such an offence’ (Caesar, De Bello Gallico VI, 17; 1980: 123-4)

Bearing in mind the treatment of human remains recorded at Ribemont, it is worth noting that both dismemberment and inversion are depicted on Bronze Age Scandinavian rock-art centred in the southern Swedish region of northern Bohuslän: a scene at Aspeberget depicts lines of armless people who may represent prisoners (Coles 1990: 27), and carved on a rock at Hamn is a group of human images apparently suspended upside down from a horizontal pole (Coles 1990: 27, 61, fig. 48).

The sacrifice of prisoners of war was perhaps quite complex: there is the ambiguity of synchronous high/low rank, victory and reparation, foreignness and humanitas. Perhaps, like slaves, foreign war-captives could be chosen because of their social disempowerment and the abrupt transition of status. But interpreting the killing of enemy prisoners in terms of sacrifice is problematical, partly because of complicating factors associated with

156

MIRANDA ALDHOUSE GREEN: RITUAL BONDAGE, VIOLENCE, SLAVERY AND SACRIFICE IN LATER EUROPEAN PREHISTORY The bondage of human ritual victims is sometimes unequivocally illustrated in the material culture and iconography of communities far away from the arena of ancient Europe. Peruvian Moche communities of the first millennium AD adorned ceramic vessels with images of sacrificial victims; two basic variations of this theme were depicted: those of naked women, lying spreadeagled, with vultures pecking at their faces and genitals, and bound male prisoners with cords round their necks, also depicted nude (Hill 2000: 317-26). It appears that both the nakedness of the victims depicted and their bound necks are charged with symbolism associated with an emphasis on generalized corporeality, as opposed to individuality, and consonant eschewance of individuality.

Bondage and restraint ‘Hailing the moon… they prepare a ritual sacrifice and banquet beneath a tree and bring up two white bulls, whose horns are bound for the first time on this occasion.’ (Pliny Natural History XVI, 95) The quotation above is from Pliny the Elder’s wellknown description of a druidic healing ceremony involving sacred oaks, mistletoe and golden sickles. But, in the present context, the interest of the passage lies in its reference to the binding of bulls’ horns preparatory to their sacrifice. Such an act may be associated with ritual subjugation and might even have been associated with this particular Gaulish cult-practice. Although Pliny may, in fact, be referring simply to the Classical pre-sacrificial rituals in which the animal-victims were sometimes decorated with garlands wound round the horns and fillets round their bodies but, in this respect, it is interesting to note that the iconography of Graeco-Roman bull sacrifice does not tend to display bound horns: thus, neither the Roman military scene depicting the suovetaurilia (the sacrifice of a bull, sheep and pig) on Trajan’s Column (Beard et al. 1998: 327, fig. 7.1), nor the Roman sestertius of the Emperor Gaius showing the sacrifice of an ox outside the temple of the divine Augustus (Adkins and Adkins 1996: 196, fig. 88; Price 1984: 207-33), nor Greek vase-paintings of cattlesacrifice (Durand 1989: figs 8, 13, 18-19) show hornbinding as part of sacrificial ceremonial.

Half a world away, in ancient Egypt c. 1900 BC, clay ‘curse’ or ‘execration’ figurines, fashioned with hands lashed behind their backs and inscribed with the names of the community’s enemies (sometimes subsequently obliterated), acted as scapegoat sacrifices to avert evil and channel supernatural energy towards inimical forces, as a means of maintaining both earthly and cosmic order (Posener 1987; Vila 1963: 135-60). The representation of bondage – whether physical or symbolic – may be invested with meaning, in terms of ritual praxis, perhaps associated with humiliation, subjugation, restriction of movement (and maybe metaphoric lifelessness), and denial of personal identity. Neck-restraints, for instance, tend to force the head down, concealing the face and directing vision downwards, away from the world.

FIGURE 1. Experimental archaeology: one of the Llyn Cerrig Bach (Anglesey) slave-chains in use. ¤ National Museum of Wales.

157

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY with bonds that also encircled his throat, and his legs were bound with cloth before he was cast into a marsh to die, perhaps in a sacrificial act (Fig. 4; van der Sanden 1996: 93) . It is tempting to interpret similarly a small copper alloy amulet, of Roman date, from Brough-underStainmore, in the form of a tiny human figure whose wrists and neck are bound together (Fig. 5; Green 1978: pl. 138).

Multiple neckrings, or gang-chains, cause all prisoners to assume the same anonymous position and thus become a single entity, presenting an image of mass humiliation and degradation. This phenomenon can be seen equally clearly in ancient Egyptian representations of Nubian prisoners engraved on the temple walls at Abu Simbel (Shinnie 1996: pl. 19b) and in the modern experimental use of an Iron Age slave-chain from Llyn Cerrig Bach, Anglesey by Cardiff University students (Fig. 1; Manning 2001: pl. opp. 70; Green 2001: col. pl. 21).

Several Iron Age bog-bodies show signs of garrotting: the hangman’s noose remained in place after the people who were killed at, for instance, Tollund in Denmark (van der Sanden 1996: 17, 125), Yde in the Netherlands (ibid.: 83), Lindow Moss in Cheshire (Stead et al. 1986; Turner 1999: 227-34) and Worsley, near Manchester (Turner 1999: 231; A.N. Garland 1995: 104-7) were consigned to their boggy graves. These aforementioned bodies belong to the middle-later Iron Age and Roman periods. Other marsh-victims could have been bound with noose materials that might themselves have been charged with meaning: the male bodies from Gallagh, in Ireland, with a neckring of willow wands (van der Sanden 1996: 73) and from Windeby, with one of hazel (ibid.: 81), come to mind. A new post-mortem by a forensic team, from the University of Århus, on the Haraldskaer woman (who has a recent 14C date of c. 490 BC) has revealed faint traces of a thin line around her throat as if she, too, had been subjected to some kind of neck restraint (Hvass 1998; Green 2001); another Danish female body, from Huldremose, was found with her long hair wound tightly round her throat (Coles and Coles 1989: 188).

Archaeological evidence may allow us to identify ritual fettering in European antiquity. Two elements can be distinguished: neck restraints and bound limbs. Some of the Danebury pit-bodies are in sufficiently tightly flexed positions to suggest bondage; sometimes the hands are crossed at the wrists, consistent with the body having been bound (Figs 2 and 3; Cunliffe 1983: 161, fig. 90). One of the young men buried at Acy-Romance in the Ardennes in the second century BC had his hands apparently lashed behind him; he had died violently, from an axe-blow to his head (Lambot 1998; 2000). An Iron Age grave from the Upper Palatinate region of Germany contained the decapitated bodies of four men, one with his hands still tied behind his back (Simón 1999: 6; Maringer 1942-43: 80). Sometimes there is evidence that people were ‘hogtied’, with limbs and necks restrained, sometimes by a single rope, a particularly uncomfortable and humiliating form of restraint which virtually freezes the body, in so far as any movement of the limbs would cause the noose around the throat to tighten: this occurred at Kayhausen in Schleswig-Holstein where, sometime in the last few centuries BC, a young crippled boy’s hands were tied

The recurrence of neck-ropes in bog-deaths may lead us to re-examine the meaning of so-called ‘slave gangchains’ on later Iron Age sites. They turn up in ritualized

FIGURES 2 and 3. Human bodies, perhaps with bound limbs, buried in grain silos at Danebury, Hampshire. ¤ Danebury Trust.

158

MIRANDA ALDHOUSE GREEN: RITUAL BONDAGE, VIOLENCE, SLAVERY AND SACRIFICE IN LATER EUROPEAN PREHISTORY

FIGURE 6. Iron slave-chain from the later Iron Age votive deposit at Llyn Cerrig Bach, Anglesey. ¤ National Museum of Wales. FIGURE 4. The teenage boy from Kayhausen, SchleswigHolstein deposited bound hand, foot and neck, in a peat bog in the second or first century BC. ¤ Anne Leaver (after van der Sanden 1996).

contexts, sometimes associated with martial regalia, from the second century BC onwards. Such is the context of the two chains from Llyn Cerrig Bach on Anglesey (Fig. 6; Fox 1946) and the example from Verdun-le-Doubs in central France (Musée de Bibracte 2000). Vincent Guichard, of the Centre de Recherches at Bibracte, suggests that such complex, intricate and finely crafted objects should not be explained simply in terms of transporting prisoners or slaves, but might, instead, have been prestigious items of regalia, associated with the sacrifice of battle-prisoners. In this context, it is worth bearing in mind that human remains were allegedly discovered by Cyril Fox at Llyn Cerrig, but went unrecorded, perhaps because of political correctness in the 1940s (Macdonald 1996; Parker Pearson 2000) which did not allow for admittance of barbaric behaviour among Britons even 2000 years earlier. It is certainly true that these elaborate wrought-iron chains represent the apogee of the blacksmith’s skill and a huge investment of time and resources (Hingley 1997; Keller and Keller 1996; Gillies 1981; Green 2002). Violence, abuse and humiliation From the evidence at our disposal, it appears that care was taken to present certain ‘sacrificial’ victims according to a particular ‘grammar’ of treatment, designed to humiliate, to deny status, freedom, identity and dignity. It is possible to discern what may be seen as additional attempts at degradation, including non-

FIGURE 5. Romano-British copper alloy figurine of a bound man, from Brough-under-Stainmore in Cumbria. ¤ The British Museum.

159

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY 1996: 138) and Kayhausen in Schleswig-Holstein (ibid.: 93) both suffered from defects affecting their mobility; the Yde girl was afflicted with idiopathic scoliosis (or extreme curvature of the spine); the Kayhausen boy’s malformed hip would have caused him to walk with a pronounced limp. Similarly, three women suggested as attendant sacrifices at the grave of the nobleman cremated in c. AD 55 at Folly Lane (Niblett 1999: 20) all had hip deformities; they might have all been related, as is argued as being the case for the 12 individuals, all with speech-inhibiting congenital jaw defects, found interred in a fire-pit at Leonding in Austria, and dating to the third century BC (Simón 1999: 10). The Huldremose woman, too, had suffered a broken leg which would have caused her to walk with a pronounced limp for the rest of her life.

normative burial positions and systematic somatic abuse, sometimes apparently enacted over a considerable length of time before death, sometimes probably post-mortem. Recent excavations at the Middle Iron Age settlement of Acy-Romance (Lambot 1998; 2000) produced the remains of some 20 young adult males, all identically treated; after death, they were boxed up in a seated position, lowered into a deep pit to dry out and then reinterred around a cult-enclosure in the earlier second century BC. In contrast to the high-status cremation burials on the site, these inhumed people were buried without grave goods. Similar treatment of bodies has been identified at Geneva (Lambot 2000). The seated position of these victims may be significant (some Late Iron Age images are similarly depicted), perhaps reflective of burial alive or, at any rate, special, nonnormative states (Klein 1999: 86, no. 42). The fatal head injury to another victim, this time bound and buried extended, is highly suggestive of sacrificial ritual: the axe that killed him had an unusual blade and a weapon fitting the wound has been found in one of the rich cremation graves on the site. Bernard Lambot, the excavator of Acy, has suggested that this individual might have met his death sometime after regular human sacrifice here had become obsolete, perhaps in about 100 BC.

Ritual abuse might, in extreme cases, have included anthropophagy: the pelvis of a male at Danebury was systematically butchered (Hooper 1984) and recent finds from a swallet hole at Alveston, near Bristol suggest that human bones were being cracked open for the marrow in the late first century BC or early first century AD. Interestingly, one person from Alveston, though not the same individual whose remains showed signs of cannibalism, had been a victim of Paget’s disease, a degenerative geriatric condition causing the sufferer to shuffle with a ‘simian’ walk (Horton and Cox pers. comm.); another of the Alveston dead was a young woman whose skull displayed a savage wound. Certainly ritual skinning took place: the head of a Romano-British boy at Folly Lane is just one of several exhibiting such deliberate defleshing (Niblett 1999: 83-8; Mays and Steele 1996: 155-61); we have seen (above) that the bodies from Ribemont seem to have been hung upside down before being skinned and the leg-bones used to build the bone-altars. Are these examples of ritual degradation? The same might be argued for the headshaving noted at Yde, Windeby and elsewhere.

The bodies at Acy resonate strongly with some of the human remains at Danebury, also without grave furniture and with signs of extensive head injury (Walker 1984: 442-63), though it is not clear at either site whether we are dealing with sacrificial abuse, war deaths or some other form of untimely and violent death. Some of the Danebury bodies were subjected to ‘overkill’ abuse, such as weighting down or crushing (Cunliffe 1993: 12-13; Hooper 1984: 463-74), identical to treatment meted out to many of the north European bog-victims, such as those at Windeby (van der Sanden 1996: 98, 112) and Haraldskaer (though her knee injury is now interpreted as occurring post-mortem and the swelling in the joint caused by bog acids rather than trauma: Hvass 1998). The woman from Huldremose suffered several mutilations, including a severed arm, before she was killed (van der Sanden 1996: 160), though Pia Bennike is of the view that the arm injury might have occurred postmortem (Bennike 2001); Lindow Man was likewise subjected to a range of injuries: blows to his head, throatslitting and strangulation (Stead et al. 1986: 46-51; Turner 1999: 227-34). Far away, at Eleutherna on the island of Crete, a young man had his legs damaged before dying as an attendant sacrifice at the tomb of a high-born hero, in the seventh century BC (Stampholidis 1996).

Sometimes, there is evidence that the faces of ‘sacrificial’ victims were deliberately obliterated, as if to humiliate the victims after their deaths and rob them of their identity, in the same way that statues of unpopular ancient rulers might be defaced by their communities as an expression of opprobrium (Lange 1997: 167-73). The face of a Danish female bog-body from Borremose might have been so treated (van der Sanden 1996: 163), as were those of Romano-British women buried in graves at Dunstable, Bedfordshire and Lowbury Hill, Oxfordshire (Matthews 1981; Philpott 1991). Andreas Hårde, of Århus University (pers. comm.) has drawn my attention to other examples of deliberately smashed faces in the later Bronze Age sepulchral record of northwestern Bohemia (Chochol 1971: 324-63).

The disabling abuse suffered by some of these victims of violent death may be linked to another feature of many such bodies: that of deformity (R. Garland 1995; Molleson 1999: 69-77). The bodies of many suspected ritual murder victims exhibit signs of physical abnormalities, mostly affecting mobility. The adolescents from Yde in the Netherlands (van der Sanden

Conclusion: status and sacrifice Many, though by no means all, of the Iron Age bogbodies from northern Europe were apparently consigned 160

MIRANDA ALDHOUSE GREEN: RITUAL BONDAGE, VIOLENCE, SLAVERY AND SACRIFICE IN LATER EUROPEAN PREHISTORY 1984). Paris: Errance, Revue Archéologique de Picardie, numéro special. Cadoux, J.-L. 1984. L’ossuaire gaulois de Ribemont-surAncre: premières observations, premières questions. Gallia 42: 53-78. Caesar, Julius. 1980. The Battle for Gaul. (trans. P. Wiseman and A. Wiseman). London: Chatto and Windus. Chochol, J. 1971. Antropologická problematica kostrových hrobu knovíské kultury v Cechách (s archeologickým komentárem J. Hraly) [Zur anthropologischen Problematik der Körpergräber der Knovízer Kultur im Böhmen (mit einem archäologischen Kommentar von J. Hrala)]. Památky Archeologické 62: 324-63. Coles, B. and Coles, J. 1989. People of the Wetlands: bogs, bodies and lake-dwellers. London: Thames and Hudson. Coles, J. 1990. Images of the Past: a guide to the rock carvings and other ancient monuments of northern Bohuslän. Uddevalla: Hällristningsmuseet, Vitlycke. Skrifter av Bohusläns Museum och Bohusläns Hembygdforbund 32. Cunliffe, B. 1983. Danebury: anatomy of an Iron Age hillfort. London: Batsford. Cunliffe, B. 1993. Fertility, Propitiation and the Gods in the British Iron Age. Vijftiende Kroon-voordracht gehouden voor de Stichting Nederlands Museum voor Anthropologie en Praehistorie te Amsterdam op 26 November 1993. Amsterdam: Instituut voor Pre- en Protohistorische Archeologie Albert Egges van Giffen, University of Amsterdam. Dundas, P. 1992. The Jains. London: Routledge. Durand, J.-L. 1989. Greek animals: towards a topology of edible bodies. In M. Detienne and J.-P. Vernant (eds) The Cuisine of Sacrifice among the Greeks. Chicago: University of Chicago Press. 87-118. Fox, C. 1946. A Find of the Early Iron Age from Llyn Cerrig Bach, Anglesey. Cardiff: National Museum of Wales. Garland, A.N. 1995. Worsley Man, England. In R.C. Turner and R.G. Scaife (eds) Bog Bodies: new discoveries and new perspectives. London: British Museum Press. 104-7. Garland, R. 1995. The Eye of the Beholder: deformity and disability in the Graeco-Roman world. Ithaca NY: Cornell University Press. Gillies, W. 1981. The craftsman in early Celtic literature. In special issue ‘Early technology in North Britain’. Scottish Archaeological Forum 11: 70-85. Girard, R. 1977. Violence and the Sacred. Baltimore: Johns Hopkins University Press. Green, M.J. 1978. A Corpus of Small Cult-Objects from the Military Areas of Roman Britain. Oxford: BAR British Series 52. Green, M.J. 1998. Humans as ritual victims in the later prehistory of western Europe. Oxford Journal of Archaeology 17: 169-89. Green, M.J.A. 2001. Dying for the Gods: human sacrifice in Iron Age and Roman Europe. Stroud: Tempus.

naked to the marsh. The Haraldskaer woman went to her watery grave without her clothes but, significantly perhaps, the garments she had worn were placed nearby (Hvass 1998). Nakedness can be read as symbolic of liminality (the skin being perceived as a barrier between inside and outside the body), just as the bog itself is a liminal place between dryland and waterworld, but nakedness could also express paradoxes of honesty (Tilley 1999: 257) and contemptuous denial of identity (Hill 2000). At sites like Acy-Romance, there seems to have been an important distinction between the low-ranking inhumations and the high-status cremations; the same is true for Folly Lane, and even as far away as Archaic Eleutherna. No grave goods accompanied these buried bodies whilst the synchronous cremations were sumptuously furnished. The discrepant treatment of buried bodies may reflect actual status or simply rank relative to the prestigious dead whose attendants the impoverished bodies might have represented. Caesar actually mentions the custom of attendant sacrifice as having only recently become obsolete when he assumed the governorship of Gaul in the mid first century BC. Most importantly, the sacrificial dead, if they are correctly interpreted, seem often to have been symbolically disempowered or unempowered. I end with another reference to a passage in Tacitus’ Germania, where he refers to the sacrifice of the slave-attendants of the goddess Nerthus, who were drowned after her annual ceremony because they had witnessed secret rites, perhaps as expendable surrogates for the priests themselves. ‘The cart, the cloth and, believe it if you will, the goddess herself, are washed clean in a secluded lake. This service is performed by slaves who are immediately drowned’ (Germania 40l; 1948: 134). Bibliography Adkins, L. and Adkins, R. 1996. Dictionary of Roman Religion. New York: Facts on File. Anon. 1981. A Guide to Sravana-belgola. Mysore: Department of Archaeology/Deputy Director Karnataka Government Textbook Press. Beard, M., North, J. and Price, S. 1998. Religions of Rome. Vol. 1: A History. Cambridge: Cambridge University Press. Bennike, P. 2001. Osteological evidence of warfare, accidents and treatment. A question of interpretation. Paper presented at seminar ‘Warfare and sacrificial rituals’, University of Århus, 10 May 2001. Brunaux, J.-L. 1988. The Celtic Gauls: gods, rites and sanctuaries. (trans. D. Nash). London: Seaby . Brunaux, J.-L. 1996. Les Religions Gauloises: rituels celtiques de la Gaule indépendente. Paris: Errance. Brunaux, J.-L., Meniel, P. and Poplin, F. 1985. Gournay I: les fouilles sur le sanctuaire et l'oppidum (1975-

161

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Green, M.J.A. 2002. Any old iron! Symbolism and ironworking in Iron Age Europe. In M.J. AldhouseGreen and P.V. Webster (eds) Artefacts and Archaeology: aspects of the Celtic and Roman world. Cardiff: University of Wales Press. 8-19. Herodotus. 1965. Histories. (trans. A. de Sélincourt). Harmondsworth: Penguin. Hill, E. 2000. The embodied sacrifice. Cambridge Archaeological Journal 10: 317-26. Hingley, R. 1997. Iron, ironworking and regeneration: a study of the symbolic meaning of metalworking in Iron Age Britain. In A. Gwilt and C. Haselgrove (eds) Reconstructing Iron Age Societies. Oxford: Oxbow Monograph 71. 9-18. Homer. 1963. The Iliad. (trans. A.T. Murray). London: Heinemann, Loeb Classical Library. Hooper, B. Anatomical considerations. In B. Cunliffe, Danebury: an Iron Age hillfort in Hampshire. Volume 2: The Excavations, 1969-1978: The Finds. London: CBA Research Report 52. 463-74. Hughes, D.D. 1991. Human Sacrifice in Ancient Greece. London: Routledge. Hughes, T. 1999. Alcestis: Euripides in a new version. London: Faber and Faber. Hvass, L. 1998. Dronning Gunhild: etv moselig fra jernalderen. Vejle: Sesam. Kamdar, S. 1993. Shuffling off this mortal coil. The Times of India, 30 October. Keller, C.M. and Keller, J. D. 1996. Cognition and Tool Use: the blacksmith at work. Cambridge: Cambridge University Press. Klein, M.J. 1999. Dieu en pose accroupie. In S. Deyts (ed.) A la Rencontre des Dieux Gaulois un Défi à César. Dijon: Musée Archéologique. 86. Lambot, B. 1998. Les morts d’Acy Romance (Ardennes) à La Tène Finale: pratiques funéraires, aspects religieuses et hiérarchie sociale. In Etudes et Documents Fouillés 4, Les Celtes: rites funéraires en Gaule du nord entre le VIe et le 1er siècle avant JesusChrist. Namur: Ministère de la Région Wallone. 7587. Lambot, B. 2000. Victimes, sacrificateurs et dieux. In V. Guichard and F. Perrin (eds.) Les Druides. Paris: Errance, L’Archéologue (hors série) 2. 30-6. Lange, C. 1997. Violence and the face. In J. Carman (ed.) Material Harm: archaeological studies of war and violence. Glasgow: Cruithne Press. 167-73. Macdonald P. 1996. Llyn Cerrig Bach: an Iron Age votive assemblage. In S. Aldhouse-Green (ed.) Art, Ritual and Death in Prehistory. Cardiff: National Museum of Wales. 32-3. Manning, W.H. 2001. Roman Wales: a pocket guide. Cardiff: University of Wales Press. Maringer, J. 1942-43. Menschnopfer im Bestattungsbrauch Alteuropas: eine Untersuchung über die Doppel- und Mehrbestattungen im vor- und frühgeschichtlichen Europa, besonders Mitteleuropa. Anthropos 37/8: 1-112.

Matthews, C.L. 1981. A Romano-British inhumation cemetery at Dunstable, Bedfordshire. Bedfordshire Archaeological Journal 15: passim. Mays, S. and Steele, J. 1996. A mutilated human skull from Roman Saint Albans, Herts., England. Antiquity 70: 155-61. Molleson, T. 1999. Archaeological evidence for attitudes to disability in the past. In special issue ‘Disability and archaeology’ (ed. N. Finlay). Archaeological Review from Cambridge 15: 69-77. Musée de Bibracte 2000. Les Druides Gaulois: catalogue d’exposition temporaire du 29 avril au 5 novembre 2000. Glux-en-Glenne: Centre Archéologique Européen du Mont Beuvray. Niblett, R. 1999. The Excavation of a Ceremonial Site at Folly Lane, Verulamium. London: Society for the Promotion of Roman Studies, Britannia Monograph 14. Parker Pearson, M. 2000. Great sites: Llyn Cerrig Bach. British Archaeology 53: 8-11. Petronius. 1969. Translation of Petronius 368-87. (trans. M. Heseltine). Cambridge MA: Harvard University Press, Loeb Classical Library. Philpott, R. 1991. Burial Practices in Roman Britain: a survey of grave treatment and furnishing AD 43-410. Oxford: BAR Publishing, BAR British Series 219. Pliny. 1945. Natural History. (trans. H. Rackham). London: Heinemann, Loeb Classical Library IV. Plutarch. 1960. The Rise and Fall of Athens: nine Greek lives. (trans. I. Scott-Kilvert). Harmondsworth: Penguin. Posener, G. 1987. Cinq Figurines d’Envoûtement. Cairo: Bibliothèque de l’Etude and I.F.A.O. 101. Price, S. 1984. Rituals and Power: the Roman imperial cult in Asia Minor. Cambridge: Cambridge University Press. Shinnie, P.L. 1996. Ancient Nubia. London: Kegan Paul International. Simón, F. 1999. Sacrificos humanos en la Céltica antigua: entre el estereotypico literario y la evidencia interna. Archiv für Religionsgeschichte 1: 1-17. Stampholidis, N.C. 1996. Reprisals: contribution to the study of customs of the Geometric-Archaic Period. Eleuthern, Sector III, 3. Rethymnon: University of Crete. Stead, I., Bourke, J.B. and Brothwell, D. 1986. Lindow Man: the body in the bog. London: British Museum Press. Tacitus. 1948. Tacitus on Britain and Germany. (trans. H. Mattingly). Harmondsworth: Penguin. Tacitus. 1961. The Annals of Imperial Rome. (trans. M. Grant). Harmondsworth: Penguin. Tierney, J.J. 1959/60. The Celtic ethnography of Posidonius. Proceedings of the Royal Irish Academy 60: 189-275. Tilley, C. 1999. Metaphor and Material Culture. Oxford: Blackwell. Turner, R.C. 1999. Dating the Lindow Moss and other British bog-bodies and the problems of assigning their cultural context. In J. Coles, B. Coles and M.S. 162

MIRANDA ALDHOUSE GREEN: RITUAL BONDAGE, VIOLENCE, SLAVERY AND SACRIFICE IN LATER EUROPEAN PREHISTORY Jøgensen (eds) Bog Bodies, Sacred Sites and Wetland Archaeology. Exeter: University of Exeter and WARP. 227-34. van der Sanden, W. 1996. Through Nature to Eternity: the bog bodies of northwest Europe. Amsterdam: Batavian Lion International. van Straten, F.T. 1995. Hierà Kalá: images of animal sacrifice in Archaic and Classical Greece. Leiden: E.J. Brill.

Vila, A. 1963. Un dépôt de textes d’envoûtement au moyen empire. Journal des Savants 135-60. Virgil. 1957. The Poems of Virgil. (trans. J. Rhoades). London: Oxford University Press. Walker, L. 1984. The deposition of the human remains. In B. Cunliffe, Danebury: an Iron Age hillfort in Hampshire. Volume 2: The Excavations, 1969-1978: The Finds. London: CBA Research Report 52. 44263.

163

Fragmentation, mutilation and dismemberment: an interpretation of human remains on Iron Age sites C. Rebecca Craig¹, Christopher J. Knüsel¹ and Gillian C. Carr² ¹Department of Archaeological Sciences, University of Bradford ²Department of Archaeology, University of Cambridge

Harding’s discussion [2000: 76-85] of flat burial inhumations of the Bronze Age of central Europe).

Previous explanations of the apparent scarcity of human remains from Iron Age contexts in central southern Britain have centred on a funerary rite involving excarnation by exposure, followed by deposition of human remains in pits. This ‘Pit Ritual Tradition’, which is most famously observed at the Iron Age hillfort of Danebury in Hampshire, has been linked by Cunliffe (1992) to ancestral rites involving propitiation of deities, and by Hill (1995) to a system of structured, ritual deposition not just limited to pits. Having made an initial re-examination of the human remains from the site, we argue that the presence of weapon-related injuries (some previously described and others identified as a result of a re-analysis), together with evidence of fragmentation, is more likely to be the result of violent deaths and mutilation and exhibition of the dead.

Hill (1995: 105) argues that it is misleading to treat human remains buried in pits and ditches as reflecting a minority mortuary rite, as the range of human remains in Iron Age features does not necessarily fit into our conventional understanding of mortuary ritual. He argues that they were part of a wider ‘pit ritual tradition’ (Hill 1992), where parallels exist between the treatment of human remains and those of animals (and, indeed, pottery and other small finds) in similar deposits, and where sacrifice may be the origin for some of the remains. He further asserts that it may be necessary to ‘stop approaching prehistoric remains from a perspective which expects the majority of these populations to be archeologically visible’ and that ‘from the Neolithic onwards, archaeological deposits of human remains are never simply to do with the treatment of the dead’ (ibid.: 106). Hill’s recognition of ritual patterning and structure in Iron Age pits was inspired by similar observations in Neolithic studies, most notably by Richards and Thomas (1984) and by Maltby (1985) who argued that the deliberate burial of large quantities of butchery waste in pits and ditches was probably associated with feasting and special occasions.

When British Iron Age mortuary remains are placed within the context of the more numerous human remains recovered from contemporary cult sites in northern France, the picture revealed is of a violent society, the practices of which might have included endemic warfare and reprisal raiding. The record of human remains in central southern Britain

The frequent association of human remains with animal bones, pottery and small finds was once considered to be due to the casual deposition of domestic rubbish (e.g. Wilson 1981: 127, who also saw the burials as, perhaps, the disposal of enemies and social outcasts), or as a minority rite of ‘pit burial’ (e.g. Whimster 1981) in pits that had been abandoned after a primary use for storage. The question remains, though: who were these people? Who, from the entire population, was selected for incorporation and deposition in the pits and ditches? Wait (1985) considered that they were, perhaps, the élite, the unclean and the sacrificed. Cunliffe (1992: fig. 6) suggested that the different parts of the bodies deposited reflected different types of death, such as death in warfare demonstrated by whole or partial bodies, head hunting reflected by the presence of skulls, and unclean death represented by complete adults.

The precise nature of mortuary rites for the majority of the Iron Age population of central southern Britain is unclear; for part of the population, however, an excarnation rite by exposure has been suggested (e.g. Ellison and Drewett 1971; Wilson 1981; Walker 1984; Wait 1985; Cunliffe 1992), now thought to have originated in the Late Bronze Age (Brück 1995). The ritual framework of exposure excarnation has been discussed in detail by two of the authors of the present contribution (Carr and Knüsel 1997) who suggested that, of those who were excarnated in this manner, the bodies of the majority were likely to have been dispersed by natural bone modification agencies such as weathering and animal trampling and gnawing, depending on the location of exposure. The implication is that some people were excarnated by exposure and, of those, perhaps only a portion were deposited in pits and ditches. This mortuary treatment might have complemented another rite if, as Hey et al. (1999) suggest, a crouched inhumation rite in which burials were located in the periphery of settlements, as at Yarnton, Oxfordshire, existed as well in the Middle Iron Age (fourth-third centuries BC). These crouched inhumations resemble Bronze Age traditions elsewhere (see, for example,

A number of researchers have argued that the people chosen for deposition should be viewed as undifferentiated, generic ancestors (e.g. Brück 1995: 262; Fitzpatrick 1997: 83; Parker Pearson 1996: 123), who were deposited in features once used as agricultural installations and intimately linked with fertility and

165

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY or splintered bones; (3) disarticulated skeletons; and (4) incomplete skeletons lacking phalanges, a limb or other parts (Carr and Knüsel 1997).

ancestral land claims. More recently, however, Whitley (2002) has noted that ‘the ancestors’ seem to be omnipresent in British prehistory, and are invoked too frequently and readily as the object of veneration or the legitimators of land claims or group rights. He argues that archaeologists need to be more contextual in their interpretations. It would seem that the time is ripe for a reinterpretation of the ‘ancestors’. Human bodies placed in unusual or liminal places - perhaps separated from others or treated in a manner different from the norm – may represent social outcasts or those who died in unusual circumstances, rather than ancestors (see, for example, papers in Hubert 2000). As will be seen, their identity is a key issue in this paper.

Further support for this excarnation hypothesis comes from the four-post structures that are found all over central southern English sites. These are primarily interpreted as structures for drying grain but it has also been suggested that these four-post structures were scaffolds or frames for the exposure of the dead (Bowen and Wood 1967; Ellison and Drewett 1971; Carr and Knüsel 1997), as nearly all sites that have yielded evidence of four-post structures have also produced whole human skeletons or fragmented human bones (Ellison and Drewett 1971: 192). Following decay (partial or otherwise) of the corpse, above or below ground, the skeletal remains were deposited in pits and ditches (although not entirely restricted to these contexts). Before proceeding to open the debate further, it is necessary to explore several more of Cunliffe’s hypotheses: human sacrifice, insult cannibalism, and ritual dismemberment.

Danebury and its skeletal record: opening up the debate Approximately 15%-30% of all pits on most Wessex sites contain human or animal remains and, given that these deposits were made perhaps once every ten to twenty years (Hill 1995: 75, table 1.1), Hill suggests that Danebury is distinctive, having perhaps one to three human or animal bone group depositions every year throughout all of its phases. Cunliffe (1992: 75) estimated that approximately 40% of Danebury’s pits contained ‘special deposits’, although not all of these contained human remains.

In the past, archaeologists have been wary of including human sacrifice in their interpretations. This is changing, as shown by recent research (e.g. Hill 1995: 105; Green 1998; Carr 2002; cf Isserlin 1997). Because butchered animal remains are found intermingled with deposits containing human remains and in similar patterns of completeness, Hill (1995: 106) argues that there may be a metaphorical link between humans and animal sacrifice, although he believes that it would be incorrect to put forward sacrifice to explain the presence of all human remains.

The human remains at Danebury fall into a number of depositional categories: whole bodies; incomplete skeletons (individual depositions); multiple, partial semiarticulated skeletons; crania or parts of crania (with and without mandibles); pelvic girdles; and individual bones (cf Cunliffe and Poole 1991: 418). While the interpretation of human remains at Danebury poses many problems, it should be remembered that they are not unique, but comprise one of the most comprehensive samples for analysis revealed to date (see Haselgrove et al. 2001: 15).

According to Wilson (1981: 147) cannibalism is largely a mythical practice for which there is little secure documentation from ethnohistorical sources. Wilson’s reluctance to accept cannibalism reflects modern opinions of cannibalism as taboo (see discussion in White 1992: 730). However, since Wilson wrote, cannibalism has more frequently been proposed to explain the archaeological patterning of some human remains. Recent research has found strong evidence to support the occurrence of cannibalism at the Middle Palaeolithic site of MoulaGuercy, Ardèche, France (Defleur et al. 1999), the Neolithic Fontbrégoua cave, France (Villa et al. 1986), in the Anasazi culture in the American Southwest (Turner and Turner 1990; White 1992; Marler et al. 2000, although see Darling 1998 and Ogilvie and Hilton 2000 for tempering interpretations of some contexts) and in Fiji (Degusta 2000). However, as far as Danebury is concerned, Cunliffe (1995) mentions but does not consider cannibalism as a cause for the dismemberment noted at Danebury. He (ibid.: 76) writes:

Although Cunliffe and Poole (1991: 425) advocated a variety of interpretations for these remains depending on their degree of articulation, they also interpreted the deposition of partial bodies in pits as the result of ‘clearing up after a massacre’, ‘excarnation’, ‘the remnants of ritual dismemberment’, and ‘the effects of scavengers on bodies originally deposited complete’, to which was later added human sacrifice and insult cannibalism (Cunliffe 1992: 77). Despite their multiple explanations, the most popular interpretation has remained that of excarnation by exposure with some human sacrifice. Excarnation involves the exposure of a corpse until it becomes skeletonized. The bones or selected bones may then be retrieved for burial elsewhere and thus form part of a secondary rite. Archaeological indicators for exposed bodies have been defined as: (1) animal gnawing on bones; (2) scattered, isolated, fragmentary, weathered

‘The emphasis we have given here to the partial bodies deriving from the rite of excarnation should not cause us to overlook other possible explanations – ritual killings and dismemberment, the display of 166

C. REBECCA CRAIG, CHRISTOPHER J. KNÜSEL AND GILLIAN C. CARR: FRAGMENTATION, MUTILATION… bodies of enemies until they fell apart, and could indeed account for some, if not all, of the partial depositions.’

Taphonomic markers Advances in the identification of taphonomic markers now provide a framework by which to distinguish features of bone assemblages that relate not only to site formation processes/accumulation history, but also to the pre-depositional/peri-mortem history of human remains. Bone is susceptible to change while exposed on the surface prior to burial, during the process of burial, while in the burial environment, during erosion and redeposition, and even during excavation, shipment to the laboratory, analysis or storage (Morlan 1984: 161). Taphonomic research relies on the assumption that bone has responded to the impacts of different forces uniformly over time, thereby allowing us to infer which processes were involved in the past (Gifford-Gonzalez 1989: 43).

No single explanation is likely to account for the appearance and distribution of the entire skeletal record (Hill 1995: 106; Cunliffe and Poole 1991: 425). In this paper, we apply the interpretation of ritual killing, dismemberment, and the display of bodies of enemies, obtained as a result of conflict, to explain some of the skeletal record of Danebury. The paucity of discussion on warfare in general has been highlighted by Sharples (1991), who reviewed the evidence for Iron Age warfare but argued that there was little evidence to support widespread warfare in the period. James (forthcoming) has also brought discussion of Iron Age warfare back onto the agenda, arguing that the omission of such discussions is remarkable given that conflict and violence are very widespread aspects of human behaviour (see below). He highlights human remains, although few in number to date for the period, as perhaps the least ambiguous evidence for identifying physical violence.

Other than general post-mortem damage, bones may sustain alterations owing to weathering (Behrensmeyer 1978), fluvial transport (Behrensmeyer 1982; Marshall 1989: 10), animal activity (Shipman 1981; Hill 1989; Haglund et al. 1988; 1989; Berger and Clarke 1995), butchery/fracturing (Shipman and Rose 1983; Olsen and Shipman 1988) and burning (Buikstra and Swegle 1989; Novak and Kollman 2000). The ability to distinguish these taphonomic manifestations from trauma resulting from violence is critical if we are to identify warfare in past societies and, moreover, if we are to distinguish funerary rites involving intentional fragmentation as suggested by Chapman (2000), as well as distinguish each of these cases from the other.

Evidence for violent deaths is not unknown from Iron Age sites. Ritual decapitations have been identified at Bredon Hill, Gloucestershire (Hencken 1938), Stanwick, Yorkshire (Wheeler 1954), and South Cadbury (Woodward and Hill 2000: 111), where many instances of ‘injuries which might have been the result of violent trauma’ (ibid.) were also identified in the ‘massacre deposit’. The cranium found near the entrance to Stanwick, severed at the level of the fourth cervical vertebra (Wheeler 1954: plate XXVIIIb), is evidence of a decapitation. Importantly for the discussion below, this decapitated cranium also bore a series of peri-mortem blade injuries and was found near a sword and scabbard, which Wheeler (ibid.: 53) argued were displayed with the cranium on a pole at the gate to the hillfort. Given the suspected Late Iron Age date of this material, it may represent a late occurrence of what was, by then, an old practice.

One of the most relevant types of taphonomic marker indicating violence is the peri-mortem fracture, which can be found in both post-cranial (Fig. 1) and cranial (Fig. 2) remains. Peri-mortem fractures are breakages of bone that occur around the time of death, and may be distinguished from those occurring post-mortem in the burial environment (see Knüsel, this volume). The criteria are similar to those used to differentiate between

FIGURE 1. A blunt force peri-mortem fracture of the right ulna of Towton 30. Note the fractured surface from which a spall has been removed by the force of the blow (arrowed) (courtesy of the Towton archive, Calvin Wells Laboratory, University of Bradford).

167

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY

The butchery process of animal remains provides an analogy for human dismemberment in the form of element frequency counts and cutand chop-marks. Examination of the human remains from Danebury has revealed some such humanly modified features. The evidence from Danebury

FIGURE 2. A blunt force peri-mortem cranial injury to the left side of Towton 11. Part of the depressed fragment is still attached and radiating fractures are evident in the parietal bone surrounding the depressed fragment (courtesy of the Towton archive, Calvin Wells Laboratory, University of Bradford).

The Danebury remains represent at least 91 individuals (in addition to an uncertain number represented by 45 occurrences of skull fragments). These remains come from 181 storage pits, 14 ‘post-holes’ and 44 stratified layers within the hillfort (Cunliffe 1995: 72), and were divided by Walker (1984: 442) into six categories of deposition:

periand post-mortem fracture in forensic anthropological analyses (Ubelaker and Adams 1995; Frayer 1997; Walker 2001; Saul and Saul 2002). Assessment of the following features aids the identification of individuals who have undergone perimortem mortuary treatment: 1. 2. 3.

4. 5.

6.

7.

Widely regarded as one of the best excavations of a British Iron Age hillfort, and having the greatest volume of evidence, the human remains from the site of Danebury were selected for study. Owing to time constraints, only a sample of the remains could be analyzed. However, it is hoped the following results will illustrate the potential of the human remains to provide us with new information and will demonstrate the necessity of such research in the future.

A. B. C. D. E. F.

the fragment size of bone remains; the presence of cutmarks in the areas of muscle attachments; spalling or uplifting of cortical bone laminae with or without the uplifted flake or spall still attached (also referred to as ‘delamination’, splitting along the laminar plates that make up cortical bone); colouration of the fracture site that is similar to the rest of the element; the occurrence of anvil and hammerstone abrasions or impact scars and related diaphyseal spiral (or helical), ‘butterfly’ fractures with no sign of healing (for an example, see Sauer 1998: fig. 4); longitudinal fractures with sharp margins associated with the exposure of the marrow cavity of long bones (for example, see Lyman 1994: figs 8.6-8.7); element frequencies that may be more akin to those of faunal assemblages than to human interments (in other words, selective preservation of some elements as opposed to others).

Whole bodies (in single or group burials) Incomplete skeletons (individual depositions) Multiple, partially articulated skeletons in charnel pits Skulls or frontal bones1 Pelvic girdles Individual bones and bone fragments

1 This term is a potentially misleading one. The ‘skull’ refers to both the cranium and the mandible (Aiello and Dean 1990: 33). If these elements are found together in the same context, then this may represent decapitation or an attempted decapitation, especially if there is evidence for cut- or chop-marks in the cervical vertebrae or in the vicinity of the gonial angle of the mandible. If the cranium is found alone without the mandible, then decapitation is not a likely scenario to explain the occurrence. The mandible is exceedingly difficult to disarticulate in a fleshed state owing to the arrangement of the muscles and ligaments that form the temporo-mandibular joint. In this case, one may be dealing with a secondary burial rite in which the cranium (but not the mandible) is retrieved after the soft tissues have decomposed, as indeed appears to be the case for Neolithic remains in the Near East (see, for example, Belfer-Cohen and Hovers 1992: Rollefson et al. 1992). Given differing contextual relationships, one may argue for decapitation for the purpose of trophy collecting (headhunting), decapitation as a mortuary rite, or as a form of corporeal punishment (see Boylston et al. 2000).

168

C. REBECCA CRAIG, CHRISTOPHER J. KNÜSEL AND GILLIAN C. CARR: FRAGMENTATION, MUTILATION…

FIGURE 3. Spiral fractures of a ‘dry’ variety of a right tibia from Danebury (authors’ photograph).

FIGURE 4. Close-up of a ‘dry’ bone spiral fracture of a long bone diaphysis from Danebury (authors’ photograph).

1984: 451). Their disarticulated and haphazard appearance within the pits could be argued to imply disturbance or secondary burial and could reasonably be expected to leave identifiable signatures of their accumulation histories. Walker (ibid.) notes that there was a more complicated mass of bone than the catalogue may imply, with a number of scattered limbs and skulls not necessarily attributable to the same bodies. A third pit (P2509) was also examined in the course of this reanalysis, as was Pit 374 which contained a young male, a crouched infant, an articulated upper limb, and other isolated skeletal elements.

It is clear that this number is not representative of the burial rites for an entire population spanning 450 years of occupation. Something set these people apart from the rest of the population, leading to them being treated in a way that made them archaeologically visible. There is no evidence for burial goods and, from the categories listed above and the evidence provided in volumes 2 (Cunliffe 1984) and 5 (Cunliffe and Poole 1991) of the Danebury report, they have not all been treated in a uniform fashion. Perhaps one of the key points to notice is the occurrence of multiple burials. With cases of multiple burials, interpretations will lean towards catastrophic events or mass disasters (see, for example, Sledzik and Rodriquez 2002), epidemics (see, for example, Margerison and Knüsel 2002), or massacres (see below).

The remains within Pits 1078, 923, and 374 provide evidence for fragmentation of bone elements. Figures 3 and 4 show ‘dry’ spiral fractures of long bones, while Figure 5 displays a peri-mortem chop-mark in another long bone specimen. As may be seen from its morphology, the fracture in Figure 5 appears to have been ‘sheared off’. In addition, there are at least two possible

The two largest pits were examined for taphonomic information (see Craig 1999). Pits 923 and 1078 of ceramic phase seven contained the mixed skeletons and skulls of ten and eleven bodies, respectively (Walker

169

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY quantifies the situation, this is not the first time there has been an identification of peri-mortem fracturing at Danebury. Walker (1984) and Hooper (1984) mention the possible occurrence of peri-mortem fractures from category F (individual bones and bone fragments). However, this evidence was not taken any further or quantified.

cut-marks near the fracture site. Spiral or helical fractures occur most often when a bone is ‘wet’ or ‘green’, in other words, when it is still fresh and with some collagen content (Outram 2002). 2 In animal bone assemblages (Fig. 6) peri-mortem fractures are produced during butchery.

Pit 2509 yielded three juvenile skulls and a group of long bones. At the time of the present study, only two of these skulls were available for analysis. Both of the mandibles were found to possess traumatic injury in the form of a small chip, forming a nick, on the posterior left ascending ramus in the vicinity of the gonial angle (Fig. 7). Such damage is consistent with decapitation at the level of the C2 or C3 vertebrae. The homogeneity of these injuries and their simultaneous deposition strongly argues the case for ritual killing. FIGURE 5. A chop-mark of a long bone diaphysis from Danebury, again with a large spall removed and incised cutmarks just to the right of the chop-mark (arrowed) (authors’ photograph).

FIGURE 7. A peri-mortem sharp force cut on the lateral epicondyle of a left humerus from Danebury. The position of this cut is consistent with a defence injury (authors’ photograph).

Examination of the photographs of this deposition in Cunliffe and Poole (1991: 420) adds weight to the possibility that these skulls were displayed for a time following removal from the body, prior to deposition in the pit. Two of the mandibles appear greatly displaced from the crania (a third one not as much, but still out of proper anatomical position), which may indicate that they were not attached to the crania at the time of deposition. It is common for there to be a degree of movement of bones within the burial environment, especially in the case of a cranium (Roksandic 2002: 103). In this case, however, the picture in the microfiche in Volume 5 indicates that these remains were packed in with animal remains, therefore creating a stable environment that would have inhibited movement.

FIGURE 6. Spiral fractures in butchered cattle remains from the Coppergate site in the city of York (from Knüsel 1986).

Overall, of the 1151 specimens studied, 91 (7.9%) displayed evidence of bone fragmentation (Table 1). From Table 1, it seems that the fragmentation concentrates in the long bones of the body - the appendicular skeleton - rather then the axial skeleton, although not exclusively so. Among these fragments are some that are of a peri-mortem variety and relate to weapon-induced trauma. Although this information 2 Since this study was undertaken, the authors have become aware of another category of fracture that may relate to breakage after deposition but when the bone still retains collagen. This type of fracture relates more to depositional history than to peri-mortem events and is referred to as a ‘dry fracture’ (see Outram 2001). The ‘dry fracture’ has also been discussed by Valentin and Le Goff (1998). Ubelaker (1986: fig. 4) describes a similar, post-mortem example that is still distinguishable from a peri-mortem fracture. Figures 3 and 4 display ‘dry’ fractures while Figure 5, like Figure 6, displays peri-mortem fractures.

This, in combination with the peri-mortem chop-mark, seems strong evidence to suggest that the head was detached and - in the absence of cut-marks relating to disarticulation of the mandibles - crania and mandibles must have become separated at some point before they 170

C. REBECCA CRAIG, CHRISTOPHER J. KNÜSEL AND GILLIAN C. CARR: FRAGMENTATION, MUTILATION…

Element

Number of Identifiable Specimens

cranium mandible cervical vertebra thoracic vertebra lumbar vertebra sacrum coccyx ribs sternum clavicle scapula humerus radius ulna carpals metacarpals manual phalanges pelvis femur patella tibia fibula tarsals metatarsals pedal phalanges unidentified phalanges unidentified long bone unidentified vertebra Total

39 4 45 88 28 15 6 315 2 17 26 28 24 18 40 41 31 84 36 10 30 24 30 4 11 49 58 48 1151

Number with Possible Perimortem Fragmentation 1 2 1 0 1 1 0 12 0 0 0 5 9 4 1 2 1 6 14 0 10 6 0 0 0 0 15 0 91

Percentage with Perimortem Fragmentation 2.6 50.0 2.2 0.0 3.6 6.7 0.0 3.8 0.0 0.0 0.0 17.9 37.5 22.2 2.5 4.9 3.2 7.1 38.9 0.0 33.3 25.0 0.0 0.0 0.0 0.0 25.9 0.0 7.9

TABLE 1. Peri-mortem fragmentation at Danebury.

group, a group that is normally under-represented in mortality profiles produced by natural attrition (Weiss 1973). There are more males than females, with a very low count of older adults. Cunliffe suggests (1991: 424) the possibility that in a war-like society male mortality is likely to have been greater than that of females in the early years. However, he comments that there is no compensatory increase in female deaths in later years at Danebury. Cunliffe believes it is, therefore, best to regard the disproportion as being the result of some socially controlled selection process. This process preferentially selected younger individuals over older ones and males over females.

became buried (i.e. after the strong temporo-mandibular ligaments had decayed). We can argue, therefore, that two of these individuals were definitely decapitated and, apparently, not immediately buried, but were exposed or displayed for a period of time. These observations may be added to those made by Hooper (1991: 429) for deposition 238. Here, a portion of the left side of an adult male (?) mandible has a swordcut 18mm in length, beginning at the posterior edge of the ramus. Hooper goes on to suggest this is consistent with execution by decapitation (ibid.: 429). This element has similar traits to the two from P2509, but while they come from ceramic phase 3, this one comes from a layer in ceramic phase 7.

Hooper records that at least 10 individuals suffered injuries made with a sword or other weapons (Hooper 1991: 428). Of these, three display more than one sword stroke or puncture. Two decapitations can now be added to this total, as well as evidence of fragmentation of skeletal elements (Table 1). The combination of perimortem weapon injuries (see, for example, Fig. 8 and Hooper 1991: 428-30), fragmentation and dismemberment, including intentional decapitation, at Danebury indicates that some individuals or parts of individuals experienced peri-mortem treatment before they were placed in pits or other features.

Danebury in context The human remains from Danebury do not represent the entirety of the population (Cunliffe 1991: 424). They seem to represent individuals who were selected from among the populace. Knüsel and Carr (1995) previously noted that the demographic profile produced from burial types A-D at Danebury do not fit a ‘normal’ pattern of attritional mortality. The Danebury population sample is represented by a cross-section of age categories, with the greatest number of individuals occurring in the adolescent 171

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY

FIGURE 8. Chop-marks (arrowed) on the ascending rami of Danebury mandibles of a decapitation variety (authors’ photograph).

these sites involve the deposition of human remains in a non-normative burial environment within or near features that include structures interpreted as temples or shrines (Brunaux 1993). Similar structures are found at Danebury towards the centre of the hillfort, dating to the later and, thus contemporary, period of occupation, 350-50 BC (Cunliffe 1995: 86). There are also similarities between the human remains and their spatial patterning at the northern French cult sites and those recovered from the Danebury hillfort.

The site of Ribemont-surAncre (Somme) lies 23km northeast of Amiens on the right bank of the River Ancre, a tributary of the Somme. Its earliest phases date to the mid-third century BC and lie beneath a later Gallo-Roman site that includes a temple, amphitheatre, and bath complex, dating to between the first century BC and the second century AD (Brunaux 1996). The earliest feature at the site is a ditch dating to about the year 250 BC which contained human remains, metal accoutrements and weapons (Fercoq du Leslay 1996). A series of post-holes and nails within the ditch attest to the presence of some sort of structure, the function of which is not clear (ibid.), although this was previously thought to represent an above-ground edifice into which human corpses were placed (Brunaux 1993: fig. 4).

Such treatment might also have occurred at other sites. At the hillfort of Bredon Hill, Gloucestershire, Hencken (1938) describes fragmented human remains found with broken metal objects at the entrance to the hillfort. Given the presence of burnt material, Hencken (ibid.: 57) surmised that these human remains ‘had come down with the burning gate, and it may be tentatively put forward as a suggestion that some severed heads had been set up on the gate, which had then been fired’. She also found limb bones that she thought had been ‘tossed down upon the roadway after a barbarous mutilation’ (ibid.: 57). No further evidence, such as confirmatory cut- and chopmarks, is provided to support the notion that these remains had been mutilated, but there is ample evidence to suggest that such treatment did occur in the Iron Age.

This ditch was followed by a square palisaded enclosure, measuring some 2,500 sq m, rebuilt many times and surrounded by a ditch, which makes it the largest of its kind thus far known (Brunaux 1993). Within the enclosure are the foundations of a small temple and a well of later, Gallo-Roman date (Fercoq du Leslay 1996). A scatter of fragmented and crushed human bones (measuring 1.70 sq m) and two square ossuaries of human bones (each 1.60 sq m) at the northeast and southeast angles of the ditched palisade date to just after 200 BC – that is, they post-date the early ditch but antedate the other, Gallo-Roman internal features (ibid.).

The site of South Cadbury (Somerset) provides another very similar example. Many of the fragments of human bones from the massacre deposit also showed signs of burning, especially the fragments of human skull, which ‘showed traces of intense burning on their exterior and interior surfaces’ (Woodward and Hill 2000: 110). Woodward and Hill, too, suggest that ‘some or all of the victims were beheaded, and the severed heads of the vanquished displayed at some conspicuous point in the vicinity of the devastated gate. At some time later they could have been taken down and burnt to destruction …’ (ibid.: 111). Other evidence for mutilation can also be found at cult sites in northern France.

These ossuaries consist of some 2,000 interlaced human long bones, 246 humeri and 501 femora, mixed with weapons and horse bones. A cylindrical post-hole measuring 30cm in diameter and some 90cm deep, with a funnel-like channel leading to it, occupied the centre of the northeastern construction with a floor or ‘pavement’ of human iliac bones surrounding it. A mass of splintered and burnt human remains filled this post-hole

Cult sites in northern France ‘Cult sites’ of the Middle La Tène period (c. 300-125 BC) in northern France have revealed evidence relating to ritual behaviour involving human and animal remains. These sites do not occur on hillforts but are located along the river courses of the region. The rites identified at 172

C. REBECCA CRAIG, CHRISTOPHER J. KNÜSEL AND GILLIAN C. CARR: FRAGMENTATION, MUTILATION… The entirely male composition of the assemblage from Ribemont and the presence of peri-mortem fractures are features also seen in a group of individuals found in a mass grave in Yorkshire dating to the battle of Towton (AD 1461). These remains were recovered from a ditch measuring 12m x 18m that was about 0.50m deep (Sutherland 2000). While the Towton individuals were apparently stripped before burial, the presence of armaments found in the Ribemont grave suggest parallels with the mass grave from the battle of Visby (AD 1361) on the island of Gotland (Sweden), where the dead were interred still wearing their armour in some cases (Thordeman 1939). The few craniofacial elements that were recovered from the Ribemont ditch deposit bear peri-mortem trauma; a second maxillary molar had had its roots removed by a blow with a sharp-edged instrument and the mandibular fragment also bears a chop-mark. Similar injuries accounted for many of those wounds identified among the Towton skeletons, whose faces had been crushed by weapon blows (Novak 2000).

(Cadoux and Lancelin 1987). There are perhaps the remains of 300 individuals and 50 horses in these structures (Fercoq du Leslay 1996). Based on the fracture patterns of these remains, this treatment occurred peri-mortally before decomposition had been completed (Brunaux 1996; Duday 1998). Running on a northeast axis within the sanctuary is a series of burials of what appear to be the remains of single individuals. Almost equidistant between these constructions are the remains of a decapitated male buried supine but in a contorted position. Lying along the superior right side of this individual were a sword in its scabbard and, near the feet, the point of a lance (pers. obs.; see also Brunaux 1998: fig.104 and discussion in Fercoq du Leslay 2000: 133). This combination of a decapitated individual with weapons is similar to that identified by Wheeler (1954) at Stanwick, with the exception that the remains of the decapitated head were not found at Ribemont.

The evidence for dismemberment that can be identified on the Ribemont remains contrasts with the absence of such evidence in the Towton mass grave, although the extreme number of cranial injuries in the Towton individuals suggests mutilation (see Knüsel and Boylston 2000). The absence of differential weathering of the Ribemont remains, in conjunction with the presence of marks of gnawing by rodents and dogs, supports the argument that these individuals had been transported and exposed as a group for a period of time prior to their final deposition in a mass grave (Duday 1998).

The ditch outside the southeast corner of the palisade, the earliest feature at the site, has been partially excavated. The excavated extent of the ditch was 20 sq m (Duday 1998) and it was 0.30-0.40m deep (Brunaux 1996). It contained the remains of 75 robust and otherwise healthy adult males, none of whom were less than 15 years of age at death and with few signs of the osteoarthropathies of advanced age. These individuals were tightly packed, as suggested by the high density of three or four individuals per square metre (Duday 1998). Among the 10,000 human remains recovered, Duday and his team noted only six skull fragments, consisting of a fragment of a mandible, a piece of temporal bone and four teeth. The cranial remains must have been removed in antiquity, along with the equally poorly represented superior-most cervical vertebrae, the atlas and axis.

The presence at Ribemont of over 200 complete or nearly complete weapons, some with bent points and evidence of corrosion, suggests equally that these armaments were exposed after their modification and prior to deposition (Brunaux 1996). Brunaux (1993; 1996; 1998) interprets these remains as those of fallen foes displayed as trophies, but Duday (1998) argues that these individuals were deposited as decapitated bodies with no regard for the uniform placement of them in the deposit. This lack of a uniform burial orientation is similar to that observed at Towton (see Sutherland 2000). It is suggested here that the remains at Ribemont once formed part of a mass grave, with the deceased not being buried immediately after death (cf Duday 1998). This mass grave is unique among the deposits of human remains at these French cult sites but the fragmented remains found within the Ribemont enclosure are similar to those found at other contemporary cult sites.

Analyses of the cervical vertebrae found in the ditch reveal that these individuals had been decapitated perimortally, probably after they were already dead because the sometimes multiple cut- and chop-marks are located on the anterior aspects of the cervical vertebrae and were made with an axe or sword (Duday 1998). The postcranial remains from the deposit exhibit weapon-related cutting and piercing peri-mortem injuries with no signs of healing. Some of these appear to relate to combat trauma, while others are indicative of dismemberment of the corpse. For example, among the remains are six articulated vertebrae from T10 to L3, with T10 having been cut transversely with a sword or axe. In addition, there is a male pelvis still in articulation with a partial vertebral column consisting of T8 to L5 and the upper third of the right femur and upper fourth of the left femur, both of which bear chop-marks suggestive of their having been hacked from the body (Cadoux and Lancelin 1987). These are similar in appearance to deposition 47 of ceramic phase 3, a cut-marked and disarticulated pelvis recovered in pit 1020 at Danebury (Walker 1984: 453, tab. 45; see also Cunliffe 1992: 73, fig. 3; see below).

The sites of Acy Romance (Ardennes), Gournay-surAronde (Picardy) and Montmartin (Oise), the latter two being only three kilometres apart, provide a picture of unusual mortuary rites and exposure of human remains, prior to their being deposited in pits or ditches with butchered animal remains (Méniel 1992). Consideration of these sites suggests something of the range of treatments at smaller cult sites. 173

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY authors and to test the veracity of their statements (Brunaux 1993; 1996), it also provides us with an opportunity to identify behaviours not indicated in these texts, as well as instances in which motivations are implicated. They allow us to glimpse circumstances that might have influenced the patterns revealed in the archaeological material. In his review of the classical sources that mention Celts or Gauls, Brunaux (ibid.) notes several that allude to sacrificial rites among these peoples. Perhaps most laconic among these are several related by the first-century AD poet Lucan, who describes a ‘hideous’ deity worshipped in ‘savage sanctuaries’, associated with the Roman Mars, whom the indigenous people apparently called Esus:

At Gournay, there is evidence for decapitation in the form of multiple, incised cut-marks and chop-marks to the cervical vertebrae, in addition to a cut-mark on a half mandibular fragment (4120), similar to those previously discussed at Danebury. In some cases, these decapitating blows were delivered from different directions using both a knife and a heavier instrument, indicating that these crania were probably removed after death (cf Duday 1998 for Ribemont and Boylston et al. 2000 for British examples). The anatomical alignment of these elements, at least of the crania and cervical vertebrae, suggests decapitation and exposure before burial. Additionally, there are cut- and chop-marks delivered to the postcranial skeleton of some individuals; among them are several cut-marks to the extremities which attest to their having been received from weapons, while still other remains possess peri-mortem fractures (Brunaux et al. 1985). The small size of some femora recovered from this site suggests that females might have contributed to the assemblage (ibid.). The demographic profile of this site is thus similar, in this regard, to that from Danebury with its inclusion of females (see Cunliffe 1991: 425; Knüsel and Carr 1995).

‘Esus Mars is worshipped in the following way: a man is suspended in a tree until his limbs detach themselves’ (Lucan, Pharsalia I: 445-6, quoted in Brunaux 1996: 195; translation by CJK). Classical authors might have been assiduously recording real events but the frequency and context are often apparently ascribed, rather than demonstrated. Lucan does not tell us whether or not such individuals were alive or dead when they were placed in trees, for example. He does not say who these men were, or indicate their age or their status in society. The connection with a war god may suggest a martial link, however, so we may have here an allusion to the means by which those killed in battle were displayed after death.

Further sites in northern France have similar evidence but with the addition of some unique peri-mortem treatment. At Montmartin, there is, again, evidence for decapitation and display of crania, as well as evidence for dismemberment of remains in the form of multiple incised marks on crania and post-cranial remains, in addition to defleshing. There is also evidence for weapon-related cut-marks and chop-marks on cranial and mandibular remains and, uniquely at this site, evidence of the removal of the tongue, indicated by a series of cutmarks on both internal and external surfaces of the mandible, directed at removing musculature.

An anonymous writer of the Bellum Hispaniense leaves one in little doubt about the desired intention of the display of human body parts in a siege of the settlement of Munda. The chronicler writes: ‘… the decapitated heads of the dead were placed on the points of swords and all were turned in the direction of the town, in a manner such that their adversaries, even though being enclosed by the rampart, could still see in the direction of the enemy the mark of their warrior prowess that inspired fear’ (quoted in Brunaux 1996: 186; translation by CJK).

At the site of Acy-Romance there is evidence in the form of scattered and burnt human remains, with an example of a possible execution of an estimated 30-year-old male. This individual was found with his right forearm placed over his left wrist and the upper limbs drawn to the right side of the body and behind the back, suggestive of his hands having been tied (Lambot 1998: 83, fig. 85). He had been despatched with a blow to the right temporal, probably delivered with an axe. At the same site, a further 16 young males were interred with their torsos so strongly flexed that the crania were found between the feet. Lambot (ibid.) argues that these individuals were placed in a box before burial. Complete and partial human remains were found in a ditch at Acy-Romance and in pits at the site (ibid.). There is no mention in the report of any weapon-related trauma, except for that of the aforementioned individual exhibiting cranial trauma.

Interpretations – combatants not ancestors A study of the Danebury remains supports fragmentation, mutilation and dismemberment, at a level of 7.9% of identified elements. Deposition 47 of ceramic phase 3 from pit 1020, the cutmarked and disarticulated pelvis, is the most striking example (Walker 1984: 453), together with deliberate fragmentation of long bones located in large deposits of co-mingled remains, and decapitation with possible display. The repetitive nature of these decapitations and their similar find contexts may suggest that this was a ritualized act. It should also be noted at this stage that, within the ‘charnel pits’ (P923 and P1078), Walker (ibid.: 448) noted the presence of sling stones, an observation that seems to have been since disregarded. She suggested there might therefore have

Reports from Classical authors Although the archaeological evidence can be said to provide a means by which to elaborate upon our understanding of rites referred to cursorily by classical 174

C. REBECCA CRAIG, CHRISTOPHER J. KNÜSEL AND GILLIAN C. CARR: FRAGMENTATION, MUTILATION… killing comes from the account of John Rodgers Jewitt, an early nineteenth century English armourer aboard the American ship Boston, who left a written account of his capture by the Nootka chief, Maquina, his life on what is today Vancouver Island (British Columbia), and his eventual forced release. Immediately after his capture Maquina made Jewitt (1961 [1815]: 224) identify the severed heads of the 25-man crew displayed on the deck of the Boston. Maquina had attacked the ship after receiving rebukes in the presence of his men from the Boston’s Captain Salter. Jewitt (ibid.: 252) comments ‘… revenge of an injury is held sacred by these people…’. In later European contexts such actions have a number of motivations which provide some perspective on later European practices of violence towards the enemy.

been a correlation between the exposure of the bodies and hostile treatment. As mentioned above, Cunliffe and Poole (1991: 425) interpret some of the human remains at Danebury in terms of ritual dismemberment and battle trophies. Brunaux (1996) interprets the pits and ditches of near contemporary northern French cult sites as relating to a warrior cult and at Gournay to the ‘ancestralisation ou héroisation’ of the deceased (Brunaux et al. 1985: 177). Although the warrior status of these individuals may be a reasonable interpretation, much of the evidence seems equally likely to relate to the denigration of the deceased: weapon-related injuries in conjunction with mutilation decapitation and dismemberment - burning, execution, defleshing and exposure of bodies and body parts for some period after death, any or all of which can also be subsumed under the heading ‘human sacrifice’, for which the importance of violence and an overkill factor has been emphasized by Green (1998: 173).

The late medieval Wars of the Roses have been characterized as a series of running feuds between powerful aristocratic families (Goodman 1981; Storey 1986), which resulted in the killing and mutilation of foes (Knüsel and Boylston 2000), especially experienced ones, perhaps in an attempt to prevent their continued participation in further hostilities (Boardman 1994). From this climate come some illustrative examples of the effects of such circumstances on the treatment meted out to the bodies of vanquished foes. For example, the body of Richard Neville, Earl of Warwick (the famed ‘Kingmaker’), killed at the battle of Barnet (AD 1471) was displayed in St. Paul’s cathedral, not so much as a trophy but to forestall attempts to claim that the powerful lord was still alive (Gillingham 1981: 201).

From this perspective, such individuals are more likely to represent vanquished foes who suffered mistreatment prior to being ritually displayed, perhaps as a warning against further hostile actions by enemies. This would also account for bodies being placed in highly visible places, such as at the entrances to hillforts (Bredon Hill, South Cadbury) and outside the large enclosure at Ribemont. This desire to expose the dead such that they might be seen may be reflected in what Fitzpatrick (1997) calls ‘superfluously’ defended hillforts which are often overlooked from nearby places (Bowden and McOmish 1987; 1989). Within sanctuary sites they might have formed a ghastly decoration (at Gournay and AcyRomance, for example).

Other historical examples abound. After being killed at the battle of Wakefield (AD 1460), Richard of York and several of his followers were decapitated as traitors and their severed heads were displayed on Micklegate Bar in York, Richard’s head adorned with a paper crown, a symbol of the unworthiness of his claim to the throne (Gillingham 1981: 120). Similarly, William de la Pole, the duke of Suffolk, was beheaded in 1450 for treason, and his detached head was displayed on a pole on the beach at Dover, next to his unburied body. The form of his execution leaves us in no doubt as to the denigration implicit in this display: he was beheaded by multiple strokes from a rusty sword wielded by one of the most ignorant persons present (Daniell 1997: 80-1).

This behaviour could have existed against a background of a complex mortuary rite that involved several phases, including display and transport of the corpse, burning or cremation, collection of remains, interment, and graveside ceremonies including animal sacrifice (see Roymans 1990: 219-20; Metzler-Zens et al. 1999). This mortuary rite could have existed within a society dominated by tribal warriors in an atmosphere of feuding and raiding (cf Wells 1995). The destabilizing effects of incursions of Mediterranean imports and people would have exacerbated and extended endemic violence, even among peoples at some remove from the Mediterranean (cf Dietler 1990).

Scott and Willey (1997), Liston and Baker (1996), and Olsen and Shipman (1994) provide physical evidence of dismemberment and mutilation of the dead as a form of denigration of defeated foes. After the battle of Little Bighorn (AD 1876), the dead of Lieutenant Colonel George Armstrong Custer’s 7th U.S. Cavalry were dismembered by Sioux and Cheyenne after the battle, leaving telltale physical evidence in cut-marked and chop-marked bones (Scott et al. 1989).

The mutilation and display of bodies and body parts of vanquished foes is well known. Rather than taking vanquished foes as captives, societies that seek revenge and use terror as a means of exacting it - often in the context of feuding - summarily kill captives taken in battle. This practice is especially common, although not exclusively so, in warrior societies (Otterbein 2000). These reprisals may come not only from a desire to avenge a fallen compatriot but also as a result of a social insult. An excellent example of this type of reprisal

A few examples from the human remains from Little Bighorn will suffice to demonstrate what physical 175

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY more recent circumstances, it is likely that these actions were dressed in and coloured by more deep-seated beliefs, but the combination of physical evidence for trauma received in armed conflict, coupled with perimortem fragmentation and decapitation and the display of human remains with weapons, would suggest that the remains of at least a proportion of these individuals might more likely have been bodily talismans of vanquished foes than revered warriors and ancestors.

evidence such activities leave on the skeletal remains. Among these are fractures of the cranium (14 examples) made at or about the time of death (i.e. peri-mortem) from blunt force trauma that may relate to the despatching of the wounded or to their mutilation. There are also series of incised marks lying close together on hand and foot bones that relate to post-mortem removal of the individual’s gauntlets and boots. Bilaterally distributed chop-marks made by heavy, edged weapons (axes or hatchets) to the femoral capiti, greater trochanters and diaphyses relate to post-mortem removal of the lower limbs. A similar chop-mark to a fourth cervical vertebra relates to the immediate post-mortem decapitation of another individual.

Acknowledgements We thank Kay Ainsworth of Hampshire County Council Museum Service for access to the Danebury remains. We thank the three anonymous reviewers for their thorough and much appreciated comments on an earlier version of this manuscript. Carol Palmer read and commented upon previous drafts of this paper. Versions of this contribution were presented at the University of Durham’s Iron Age conference (2001) and the University of Sheffield’s Warfare and Violence conference (2002). We thank the organizers of these events for allowing us to contribute.

Although these individuals are no longer identifiable as such, there are accounts that relate to the treatment meted out to named individuals recovered after the battle. The body of Captain Tom Custer, brother of George Custer, was found with his head crushed and with evidence of having been scalped, in addition to his abdomen having been slit open and his body shot full of arrows (Scott et al. 1989: 247). Accounts of individuals who participated in the battle and its aftermath attest to the multiple motivations for such treatment. The Sioux war-leader, Gall, noted that his ‘heart was bad’ because of the loss of several family members during the fighting: he fought for revenge. Rage and revenge form a common motivation for such behaviour. Non-combatants need not be excluded from such behaviours. Thus White Necklace, Wolf Chief’s wife, decapitated a fallen soldier in retaliation for her niece having received similar treatment in an earlier massacre (Powell cited in Scott et al. 1989: 86).

Bibliography Aiello, L.C. and Dean, M.C. 1990. Human Evolutionary Anatomy. London: Academic Press. Behrensmeyer, A.K. 1978. Taphonomic and ecologic information from bone weathering. Paleobiology 4: 150-62. Behrensmeyer, A.K. 1982. Time resolution in fluvial vertebrate assemblages. Paleobiology 8: 211-17. Belfer-Cohen, A. and Hovers, E. 1992. In the eye of the beholder: Mousterian and Natufian burials in the Levant. Current Anthropology 33: 463-71. Berger, L.R. and Clarke, R.J. 1995. Eagle involvement in accumulation of the Taung child fauna. Journal of Human Evolution 29: 275-99. Boardman, A.W. 1994. The Battle of Towton. Stroud: Sutton. Bowen, H.C. and Wood, P.D. 1967. Experimental storage of corn underground and its implication for Iron Age settlements. Bulletin of the Institute of Archaeology 7: 1-14. Bowden, M. and McOmish, D.S. 1987. The required barrier. Scottish Archaeological Review 4: 76-84. Bowden, M. and McOmish, D.S. 1989. Little boxes: more about hillforts. Scottish Archaeological Review 6: 12-16. Boylston, A., Knüsel, C.J., Roberts, C.A. and Dawson, M. 2000. Investigation of a Romano-British rural ritual in Bedford, England. Journal of Archaeological Science 27: 241-54. Brück, J. 1995. A place for the dead: the role of human remains in Late Bronze Age Britain. Proceedings of the Prehistoric Society 61: 245-77. Brunaux, J.-L. 1993. Les sanctuaires gaulois. La Recherche 256 (24): 820-8. Brunaux, J.-L. 1996. Les Religions Gauloises: rituels celtiques de la Gaule indépendante. Paris: Errance.

In addition to these emotions, a deeper, more spiritual meaning is also indicated. These Plains peoples conceived of a spirit land that one enters in the condition in which one has perished. An individual who has had certain muscles cut away cannot perform tasks associated with living, such as pulling a bowstring or firing a rifle (mutilation of the upper limbs), or putting a foot in a stirrup or stooping to drink (mutilation of the lower limbs) (Carrington cited in Scott et al. 1989: 86). Essentially, then, a corpse treated in this manner enters the world of the spirits in a disabled state. Post-mortem mutilation, then, may be administered in revenge, to denigrate, and for spiritual or cosmological reasons that are not mutually exclusive. Such mutilation, though, seems to have at its core a desire to disfigure or disable an individual, perhaps in an attempt to remove that individual’s identity. From the evidence reviewed here, it seems that some of the mortuary remains from Danebury (and perhaps other Iron Age sites in Britain, such as those from South Cadbury) provide ample evidence for mass killing and mass burial, peri-mortem mutilation and dismemberment, and display of bodies and body parts that is in keeping with more recent instances of revenge warfare. As in 176

C. REBECCA CRAIG, CHRISTOPHER J. KNÜSEL AND GILLIAN C. CARR: FRAGMENTATION, MUTILATION… Defleur, A., White, T., Valensi, P., Slimak, L. and Crégut-Bonnoure, E. 1999. Neanderthal cannibalism at Moula-Guercy, Ardèche, France. Science 286: 12831. Degusta, D. 2000. Fijian cannibalism: osteological evidence from Navatu. American Journal of Physical Anthropology 110: 215-41. Dietler, M. 1990. Driven by drink: the role of drinking in the political economy and the case of early Iron Age Europe. Journal of Anthropological Archaeology 9: 352-406. Duday, H. 1998 Le charnier gaulois de Ribemont-surAncre (Somme). Etudes et Documents Fouilles 4, Les Celtes: rites funéraires en Gaule du nord entre le VIe et le 1er siècle avant Jesus-Christ. Namur: Ministère de la Région Wallone. 113-19. Ellison, A. and Drewett, P. 1971. Pits and post-holes in the British Early Iron Age: some alternative explanations. Proceedings of the Prehistoric Society 37: 183-94. Fercoq du Leslay, G. 1996. Chronologie et analyse spatiale à Ribemont-sur-Ancre (Somme). Revue Archéologique de Picardie 3/4: 189-208. Fercoq du Leslay, G. 2000. L’apport des fossés de Ribemont-sur-Ancre (Somme) à la chronologie et à l’interprétation du site. Revue Archéologique de Picardie 1/2: 113-46. Fitzpatrick, A.P. 1997. Everyday life in Iron Age Wessex. In A. Gwilt and C. Haselgrove (eds) Reconstructing Iron Age Societies: new approaches to the British Iron Age. Oxford: Oxbow Books, Oxbow Monograph 71. 73-86. Frayer, D.W. 1997. Ofnet: evidence for a Mesolithic massacre. In D. Martin and D.W. Frayer (eds) Troubled Times: violence and warfare in the past. Amsterdam: Gordon and Breach. 181-216. Gifford-Gonzalez, D. 1989. Modern analogues: developing an interpretive framework. In R. Bonnichsen and M.H. Sorg (eds) Bone Modification. Orono ME: University of Maine Center for the Study of the First Americans. 43-52. Gillingham, J. 1981. The Wars of the Roses: peace and conflict in fifteenth-century England. Baton Rouge: Louisiana State University Press. Goodman, A. 1981. The Wars of the Roses: military activity and English society. London: Routledge and Kegan Paul. Green, M. 1998. Humans as ritual victims in the later prehistory of Europe. Oxford Journal of Archaeology 17: 169-89. Haglund, W.D., Reay, D.T. and Swindler, D.R. 1988. Tooth mark artifacts and survival of bones in animal scavenged human skeletons. Journal of Forensic Sciences 33: 985-97. Haglund, W.D., Reay, D.T. and Swindler, D.R. 1989. Canid scavenging/disarticulation sequence of human remains in the Pacific Northwest. Journal of Forensic Sciences 34: 587-606. Harding, A.F. 2000. European Societies in the Bronze Age. Cambridge: Cambridge University Press.

Brunaux, J.-L. 1998. Un monumental trophée celtique à Ribemont-sur-Ancre (Somme). In Etudes et Documents Fouilles 4, Les Celtes: rites funéraires en Gaule du nord entre le VIe et le 1er siècle avant JesusChrist. Namur: Ministère de la Région Wallone. 10713. Brunaux, J.-L., Méniel, P. and Poplin, F. 1985. Gournay I: les fouilles sur le sanctuaire et l’oppidum (19751984). Paris: Errance, Revue Archéologique de Picardie, numéro special, Buikstra, J.E. and Swegle, M. 1989. Bone modification due to burning: experimental evidence. In R. Bonnichsen, and M.H. Sorg (eds) Bone Modification. Orono ME: University of Maine Center for the Study of the First Americans. 247-58. Cadoux, J.-L. and Lancelin, P. 1987. L’ossuaire de Ribemont-sur-Ancre (Somme). Anthropozoologica (Premier Numéro Spécial) 91/96. Carr, G. 2002. A time to live, a time to heal and a time to die: healing and divination in later Iron Age and early Roman Britain. In P. Baker and G. Carr (eds) New Approaches to Medical Archaeology and Anthropology: practices, practitioners and patients. Oxford: Oxbow Books. 58-73. Carr, G.C. and Knüsel, C.J. 1997. The ritual framework of excarnation by exposure as the mortuary practice of the early and middle Iron Ages of central southern Britain. In A. Gwilt and C. Haselgrove (eds) Reconstructing Iron Age Societies: new approaches to the British Iron Age. Oxford: Oxbow Books, Oxbow Monograph 71. 167-73. Chapman, J. 2000. Fragmentation in Archaeology: people, places and broken objects in the prehistory of southeastern Europe. London: Taylor and Francis. Craig, C.R. 1999. A taphonomic study of a sample of the human remains from the hillfort site of Danebury, Hampshire. Unpublished MSc dissertation, University of Bradford. Cunliffe, B. 1984. Danebury: an Iron Age hillfort in Hampshire. Volume 2: The Excavations, 1969-1978: the finds. London: CBA Research Report 52. Cunliffe, B. 1991. Iron Age Communities in Britain: an account of England, Scotland, and Wales from the seventh century BC until the Roman conquest. London & New York: Routledge. Cunliffe, B. 1992. Pits, preconceptions, and propitiation in the British Iron Age. Oxford Journal of Archaeology 11: 69-83. Cunliffe, B. 1995. Danebury: an Iron Age hillfort in Hampshire. Volume 6: A Hillfort Community in Perspective. London: CBA Research Report 102. Cunliffe, B. and Poole, C. 1991. Danebury: an Iron Age hillfort in Hampshire. Volume 5: The Excavations 1979-1988: The Finds. London: CBA Research Report 73. Daniell, C. 1997. Death and Burial in Medieval England, 1066-1550. London: Routledge. Darling, J.A. 1998. Mass inhumation and the execution of witches in the American Southwest. American Anthropologist 100: 732-52. 177

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Haselgrove, C., Armit, I., Champion, T., Creighton, J., Gwilt, A., Hill, J.D., Hunter, F. and Woodward, A. 2001. Understanding the British Iron Age: an agenda for action. Salisbury: Wessex Archaeology. Hencken, T.C. 1938. The excavation of the Iron Age camp on Bredon Hill, Gloucestershire, 1935-1937. Archaeological Journal 95: 1-111. Hey, G., Bayliss, A. and Boyle, A. 1999. Iron Age inhumation burials at Yarnton, Oxfordshire. Antiquity 73: 551-62. Hill, A. 1989. Bone modification by modern spotted hyenas. In R. Bonnichsen and M.A. Sorg (eds) Bone Modification. Orono ME: University of Maine Center for the Study of the First Americans. 169-78. Hill, J.D. 1992. Can we recognise a different European past? A contrastive archaeology of later prehistoric settlements in southern England. Journal of European Archaeology 1: 57-75. Hill, J.D. 1995. Ritual and Rubbish in the Iron Age of Wessex: a study on the formation of a specific archaeological record. Oxford: BAR Publishing, BAR British Series 242. Hooper, B. 1984. Anatomical considerations. In B. Cunliffe, Danebury: an Iron Age hillfort in Hampshire. Volume 2: The Excavations, 1969-1978: the finds. London: CBA Research Report 52. 463-74. Hooper, B. 1991. Anatomical considerations. In B. Cunliffe (ed.) Danebury: an Iron Age hillfort in Hampshire. Volume 5: The Excavations, 1979-1988: The Finds. London: CBA Research Report 52. 42531. Hubert, J. (ed.) 2000. Madness, Disability and Social Exclusion: the archaeology and anthropology of difference. London: Routledge. Isserlin, R.M.J. 1997. Thinking the unthinkable: human sacrifice in Roman Britain? In K. Meadows, C. Lemke and J. Heron (eds.) TRAC 96: Proceedings of the Sixth Annual Theoretical Roman Archaeology Conference, Sheffield 1996. Oxford: Oxbow Books. 91-100. James, S.T. forthcoming. A bloodless past: the pacification of Early Iron Age Britain. In C. Haselgrove and R.E. Pope (eds) The Earlier Iron Age in Britain and the Near Continent. Oxford: Oxbow Books. Jewitt, J.R. 1961 (1815). The headhunters of Nootka. In F. Drimmer (ed.) Captured by the Indians: fifteen firsthand accounts, 1750-1870. New York: Dover. 216-55. Knüsel, C.J. 1986. An analysis of a faunal assemblage from the tenth-century levels at 16-22 Coppergate in the city of York. Unpublished MA dissertation, University of York. Knüsel, C.J. and Carr, G.C. 1995. On the significance of the crania from the River Thames (with a reply by Richard Bradley). Antiquity 63: 162-9. Knüsel, C.J. and Boylston, A.E. 2000. How has the Towton Project contributed to our knowledge of medieval and later warfare? In V. Fiorato, A. Boylston, and C.J. Knüsel (eds) Blood Red Roses: the

archaeology of a mass grave from the battle of Towton AD 1461. Oxford: Oxbow Books. 169-88. Lambot, B. 1998. Les morts d’Acy Romance (Ardennes) à La Tène finale. Etudes et Documents Fouilles 4, Les Celtes: rites funéraires en Gaule du nord entre le VIe et le 1er siècle avant Jesus-Christ. Namur: Ministère de la Région Wallone. 75-87. Liston, M.A. and Baker, B.J. 1996. Reconstructing the massacre at Fort William Henry, New York. International Journal of Osteoarchaeology 6: 28-41. Lyman, R.L. 1994. Vertebrate Taphonomy. Cambridge: Cambridge University Press. Maltby, M. 1985. Patterns in faunal assemblage variability. In G. Barker and C. Gamble (eds) Beyond Domestication in Prehistoric Europe. London: Academic Press. 33-74. Margerison, B.J. and Knüsel, C.J. 2002. A comparison of attritional and catastrophic cemeteries: the palaeodemography of the medieval plague cemetery at the Mint site, London. American Journal of Physical Anthropology 119: 134-43. Marler, R.A., Leonard, B.L., Billman, B.R., Lambert, P.M. and Marler, J.E. 2000. Biochemical evidence of cannibalism as a prehistoric Puebloan site in southwestern Colorado. Nature 407: 74-8. Marshall, L.G. 1989. Bone modification and the laws of burial. In R. Bonnichsen and M.H. Sorg (eds) Bone Modification. Orono ME: University of Maine Center for the Study of the First Americans. 7-24. Méniel, P. 1992. Les Sacrifices d’Animaux chez les Gaulois. Paris: Errance, Metzler-Zens, N., Metzler-Zens, J., Méniel, P., Bis, R., Gaeng, C. and Villemeur, I. 1999. Lamadelaine: une necropole de l’oppidum du Titelberg. Luxembourg: Musée National d’Histoire et d’Art Dossiers d’Archéologie. Morlan, R.E. 1984. Toward a definition of criteria for the recognition of artificial bone alterations. Quaternary Research 22: 160-71. Novak, S. 2000. Battle-related trauma. In V. Fiorato, A. Boylston, and C.J. Knüsel (eds) Blood Red Roses: the archaeology of a mass grave from the battle of Towton AD 1461. Oxford: Oxbow Books. 90-102. Novak, S.A. and Kollman, D.D. 2000. Peri-mortem processing of human remains among the Great Basin Fremont. International Journal of Osteoarchaeology 10: 65-75. Ogilvie, M.D. and Hilton, C.E. 2000. Ritualized violence in the prehistoric American Southwest. International Journal of Osteoarchaeology 10: 27-48. Olsen, S.L. and Shipman, P. 1988. Surface modification on bone: trampling versus butchery. Journal of Archaeological Science 15: 535-53. Olsen, S.L. and Shipman, P. 1994. Cutmarks and perimortem treatment of skeletal remains on the northern Plains. In D.W. Owsley and R.L. Jantz (eds) Skeletal Biology of the Great Plains: migration, warfare, health, and subsistence. Washington DC: Smithsonian Institution Press. 377-87.

178

C. REBECCA CRAIG, CHRISTOPHER J. KNÜSEL AND GILLIAN C. CARR: FRAGMENTATION, MUTILATION… Shipman, P. and Rose, J. 1983. Early hominid hunting, butchering and carcass processing behaviors: approaches to the fossil record. Journal of Anthropological Archaeology 10: 465-74. Sledzik, P. and Rodriquez, W.C. III 2002. Damnum fatale: the taphonomic fate of human remains in mass disasters. In W.D. Haglund and M.H. Sorg (eds) Advances in Forensic Taphonomy: method, theory, and archaeological perspectives. Boca Raton FL: CRC Press. 321-30. Storey, R.L. 1986. The End of the House of Lancaster. Stroud: Sutton Publishing. Sutherland, T. 2000. Recording the grave. In V. Fiorato, A. Boylston, and C.J. Knüsel (eds) Blood Red Roses: the archaeology of a mass grave from the battle of Towton AD 1461. Oxford: Oxbow Books. 36-44. Thordeman, B. (ed.) 1939. Armour from the Battle of Wisby 1361. Vitterhets Historie. Stockholm: Antivitets Akademien. Turner, C.G. II and Turner, J.A. 1990. Cannibalism in the prehistoric Southwest: occurrence, taphonomy, explanation and suggestions for standardized world definition. Journal of Anthropological Science 103: 122. Ubelaker, D.H. 1996. Skeletons testify: anthropology in forensic science. AAPA luncheon address, April 12, 1996. Yearbook of Physical Anthropology 39: 229-44. Ubelaker, D.H. and Adams, B.J. 1995. Differentiation of peri-mortem and postmortem trauma using taphonomic indicators. Journal of Forensic Sciences 40: 509-12. Valentin, F. and Le Goff, I. 1998. La sépulture secondaire néolithique de La Chaussee-Tirancourt: fractures sur os frais ou os secs? L’Anthropologie 102: 91-5. Villa, P., Bouville, C., Courtin, J., Helmer, D., Mahieu, E., Shipman, P., Belluomini, G. and Branca, M. 1986. Cannibalism in the Neolithic. Science 233: 431-7. Wait, G.A. 1985. Ritual and Religion in Iron Age Britain. Oxford: BAR British Series 149. Walker, L. 1984. The deposition of the human remains. In B. Cunliffe, Danebury: an Iron Age hillfort in Hampshire. Volume 2: The Excavations, 1969-1978: The Finds. London: CBA Research Report 52. 44263. Walker, P.L. 2001. A bioarchaeological perspective on the history of violence. Annual Review of Anthropology 30: 573-96. Weiss K.M. 1973. Demographic models for anthropology. Memoirs of the Society for American Archeology 27. Wells, P.S. 1995. Settlement and social systems at the end of the Iron Age. In B. Arnold and D. Blair Gibson (eds) Celtic Chiefdom, Celtic State. Cambridge: Cambridge University Press. 88-95. Wheeler. R.E.M. 1954. The Stanwick Fortifications. Oxford: Society of Antiquaries Research Report 11. Whimster, R. 1981. Burial Practices in Iron Age Britain: a discussion and gazetteer of the evidence c. 700 BC to AD 43. Oxford: BAR British Series 90.

Otterbein, K.F. 2000. Killing of captured enemies: a cross-cultural study. Current Anthropology 41: 43943. Outram, A.K. 2001. A new approach to identifying bone marrow and grease exploitation: why the ‘indeterminate’ fragments should not be ignored. Journal of Archaeological Science 28: 401-10. Outram, A.K. 2002. Bone fracture and within-bone nutrients: an experimentally based method for investigating levels of marrow extraction. In P. Miracle and N. Milner (eds) Consuming Passions and Patterns of Consumption. Cambridge: University of Cambridge McDonald Institute Monograph. 51-63 Parker Pearson, M. 1996. Food, fertility and front doors in the first millennium BC. In T.C. Champion and J.R. Collis (eds) The Iron Age in Britain and Ireland: recent trends. University of Sheffield: JR Collis Publications. 117-32. Richards, C. and Thomas, J. 1984. Ritual activity and structured deposition in Later Neolithic Wessex. In R. Bradley and J. Gardiner (eds) Neolithic Studies: a review of some current research. Oxford: BAR British Series 133, Reading Studies in Archaeology 1. 189-218. Roksandic, M. 2002. Position of skeletal remains as a key to understanding behavior. In W.D. Haglund and M.H. Sorg (eds) Advances in Forensic Taphonomy: method, theory, and archaeological perspectives. Boca Raton FL: CRC Press. 99-117. Rollefson, G.O., Simmons, A.H. and Kafafi, Z. 1992. Neolithic cultures at ‘Ain Ghazal, Jordan. Journal of Field Archaeology 19: 443-70. Roymans, N. 1990. Tribal Societies in Northern Gaul: an anthropological perspective. Amsterdam: Universiteit van Amsterdam. Sauer, N.J. 1998. The timing of injuries and manner of death: distinguishing among antemortem, perimortem and postmortem trauma. In K.J. Reichs and W.M. Bass (eds) Forensic Osteology: advances in the identification of human remains. Springfield IL: Charles C. Thomas. Second edition. 321-32. Saul, J.M. and Saul, F. 2002. Forensics, archaeology, and taphonomy: the symbiotic relationship. In W.D. Haglund and M.H. Sorg (eds) Advances in Forensic Taphonomy: method, theory, and archaeological perspectives. Boca Raton FL: CRC Press. 71-97. Scott, D.D., Fox, R.A., Connor, M.A. and Harmon, D. 1989. Archaeological Perspectives on the Battle of Little Big Horn. Norman OK & London: University of Oklahoma Press. Scott, D. and Willey, P. 1997. Little Bighorn: human remains from the Custer National Cemetery. In D.A. Poirier and N.F. Bellantoni (eds) In Remembrance. Westport CT: Bergin & Garvey. 155-71. Sharples, N.M. 1991. Warfare in the Iron Age of Wessex. Scottish Archaeological Review 8: 79-89. Shipman, P. 1981. Life History of a Fossil: an introduction to taphonomy and paleoecology. Cambridge MA: Harvard University Press.

179

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY White, T.D. 1992. Prehistoric Cannibalism at Mancos 5MTUMR-2346. Princeton NJ: Princeton University Press. Whitley, J. 2002. Too many ancestors. Antiquity 76: 11926. Wilson, C.E. 1981. Burials within settlements in southern Britain during the pre-Roman Iron Age. Bulletin of the Institute of Archaeology 18: 127-70.

Woodward, A. and Hill, J.D. 2000. The human bodies. In J.C. Barrett, P.W.M. Freeman and A. Woodward (eds) Cadbury Castle Somerset: the later prehistoric and early historic archaeology. London: English Heritage Archaeological Report 20. 109-11.

180

The origins of warfare: later prehistory in southeastern Iberia Gonzalo Aranda Jiménez and Margarita Sánchez Romero Departamento de Prehistoria y Arqueología, Universidad de Granada

warfare having been formalized or ritualized rather than real (Carman and Harding 1999b: 4-5).

In this paper we develop an approach for studying the origin of warfare from the archaeological evidence of southeastern Iberia. The process that leads to the appearance of warfare began at the end of the fourth millennium and is particularly evident during the third millennium BC alongside the development of social complexity and conflict. Warfare accompanied the emergence of permanent settlements and a significant demographic increase. Both these elements provoked on the one hand the ascription and identification of social groups with a specific territory and its resources and, on the other, the appearance of high-density population levels. In this context a new settlement pattern developed, characterized by strong hierarchization. Through the first half of the third millennium, there appeared several sites of considerable dimensions with central and strategic positions and complex defensive systems. They contrast with less sizeable settlements without defences which developed in a great variety of locations orientated to the exploitation of particular economic resources.

In recent years this situation has begun to change. The attention paid by many recent authors to conflict in prehistoric societies, mainly concerning the origin and causes of warfare (Keeley 1996; Carman 1997; Carman and Harding 1999a; for Iberia Monks 1997; 1999; Oosterbeek 1997; Kunst 2000) has opened an interesting debate. In our opinion this discussion should develop the analysis of different types of conflicts and violence and, especially, their relationship to society at large. From this perspective it should be possible to assess and understand the various causes of different conflicts. The concept of conflict is richly nuanced: it works at different scales, from individual to organized aggression (raids, revenge attacks, initiation rituals, wars etc.) to ideological mechanisms of control and subversion. Hence violence or coercion embodies material and ideological elements together in the same process. Similarly, conflict is played out at inter- and intra-group level. Investigation of conflict regulation requires a detailed analysis of social organization.

Also during this period arrowheads were manufactured as specialized weapons. Evidence for this comes from the archaeological contexts in which arrowheads have been found, and their production in workshops dedicated exclusively to their manufacture. Moreover, arrowheads as grave goods had symbolic meanings related to the exhibition of wealth and power by specific social groups. This symbolic relationship allows us to identify the emergence of warriors during the third millennium BC.

Conflict and warfare are not synonymous. A fight between two people, or one involving a small group of people, or even a revenge attack or raid, are not warfare (Brothwell 1999: 25; Mercer 1999: 144). Warfare is just one complex means of conflict regulation. Warfare can be understood within complex societies as the long-term institutionalization of violence through the consolidation of warriors into a specific social group. In societies of less social complexity, it would be more appropriate to talk about aggression or conflict being episodic and developing through specialized elements such as weapons, fortifications, alliances, and so on. For this reason we prefer to link the origin of warfare to specific forms of social organization rather than simply placing it in a cultural or chronological period such as the Neolithic (Camps 1992; Campillo 1995) or Mesolithic (Vencl 1984: 121).

Introduction Studies of the origins of violent behaviour have been a traditional issue in the development of philosophical thought. From the first philosophers the origin of human aggressiveness has been a recurrent theme; it also has a long history in anthropological research (Haas 1990; Keeley 1996). Nevertheless this concern has not been manifest in the analysis of prehistoric societies, and consequently the origin and development of conflict and its repercussions for social change have been of little interest despite some of the archaeological evidence. This has resulted in the propagation of a pacific view of prehistory in which ideological, symbolic and ritual interpretations have been emphasized when dealing with evidence such as fortifications or violent deaths (Keeley 1996: 18; Kristiansen 1999: 175). Ancient warfare has often been defined as ritual, heroic or primitive, and characterized by rules based on the control of violence in which the preservation of the enemy is as important as its destruction. As a result, a conception has developed of

Southeastern Iberia in the late fourth and third millennia Southeastern Iberia (Fig. 1) has a dynamic prehistoric sequence. By the end of the nineteenth century several sites, mainly necropolises (Siret and Siret 1890), had produced a rich body of archaeological evidence that implied an increase in social complexity in later prehistory. The region became a useful setting to promote hypotheses about processes of cultural change, from communal social groups in the Neolithic period to

181

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY

FIGURE 1. Location of archaeological sites mentioned in the text. 1, Los Millares; 2, Las Pilas; 3, Cerro de la Virgen; 4, El Malagón; 5, Almizaraque; 6, Cabezo del Plomo; 7, Campos.

stratified societies in the Bronze Age. Interpretations have developed from diffusionist perspectives, in which technological and social advancement has been related to eastern Mediterranean influences (Siret 1913; Blance 1961; Sangmeister 1975), to autochthonist positions in which self-evolution of social groups has been considered as the key in process of change (Chapman 1981; 1990; Ramos 1981; Mathers 1984; Gilman 1987).

when there was an important shift in social organization with the advent of the Argar culture at the beginning of the Bronze Age. In what follows, we analyse the origins and cause of warfare during this period, and the emergence of social groups that can be defined as warriors. We also consider the evidence associated with conflict, not only in the sense of physical coercion but also as expressed by its symbolic or ritual meanings. From this perspective we have tried to avoid the anthropological distinction between ritual and real warfare, integrating both elements within the same discussion.

Within these processes of growing social complexity, our study focuses on early stages: from the Almeria culture of the Late Neolithic (3400 to 3000/2950 cal. BC) to the Millares culture or Copper Age. The Copper Age has been subdivided into phases: x x x

Territory and population The increase of conflict in southeastern Iberia is clearly connected to two basic processes: on the one hand the move to sedentism in contrast to earlier strategies of environmental exploitation that imply mobility and, on the other hand, a significant growth in population. The importance of these factors has been pointed out by several authors who have considered, essentially, that pressure on resources – owing to increases in population led to violence and the spread of conflict (Díaz-Andreu 1995; Monks 1997; 1999; Vencl 1999; Haas 1999).

Ancient Copper Age (3000/2950 to 2800/2700 cal. BC) Middle Copper Age (2800/2700 to 2500 cal. BC) Late Copper Age or Beaker period (2500 to 2250 cal. BC) (Molina 1988: 261-2; Castro et al. 1996: 79-82).

Hence our period of study is from the later fourth millennium to the end of the third millennium cal. BC 182

GONZALO ARANDA JIMÉNEZ AND MARGARITA SÁNCHEZ ROMERO: THE ORIGINS OF WARFARE Although the relationship between this increase in population and the move to sedentism with the development of conflict should not be assumed to be a deterministic one, we consider both elements as necessary requirements for the appearance of a society dominated by offensive and defensive strategies at the end of the fourth and during the third millennia BC in southeast Iberia.

with is strongly aggregated high-density population (Molina 1988: 258-9). These processes of sedentism and demographic growth, resulting in a high density of population, imply fundamental changes relevant to increases in conflict. With the loss of mobility and an increase in territorial fixity, populations became strongly linked with specific territories and their resources (Tringham 1972: 469). Relationships of dependence between people and territory led to a new concept of property based on exclusion and competition between different social groups: in other words, rights over exclusive exploitation of local resources, together with avoidance of other social groups’ resources, formed the new basis for the reproduction of the community. In this context increases in conflict are related to pressure on economic resources, whereas the ascription of a specific territory does not necessarily result in the growth of violence. The latter seems to be the case in several Neolithic societies in Europe characterized by low-density populations and plentiful local resources, for example the first Neolithic communities of the Tripolye culture (Dolukhanov 1999: 81-2) or sedentary societies such as the Anasazi of the American Southwest (Haas 1999: 18-20).

From archaeological evidence so far recovered, it appears that the settlements of Neolithic communities in the southeast were mainly in caves. However, at least in specific areas such as Vega de Granada, this settlement pattern was articulated with seasonal open-air sites with very short occupation sequences (Sáez and Martínez 1981). From the end of the Neolithic and during the Copper Age this situation changed dramatically. Cave sites were abandoned and new permanent open-air settlements appeared. The documentation for the analysis of settlement patterns comes from surveys carried out recently in several areas of the southeast such as the Andarax basin and the Tabernas corridor (Almeria; Delibes et al. 1996; Molina 1988; Alcaraz et al. 1994), the Vera basin and Almanzora valley (Almeria; Delibes et al. 1996; Martín et al. 1998) and the Cullar-Chrivel corridor (Moreno et al. 1991/92).

Nevertheless, the situation in southeastern Iberia has different characteristics. The increasing sedentism of Neolithic and Chalcolithic communities parallels an increase in population which can be quantified by the number and sizes of settlements. In the Vera basin, for example, the number of sites increases more than five times from the Neolithic (Martín et al. 1998). In the survey carried out in the Cullar-Chirival corridor, site numbers rise from five settlements of the Late Neolithic/Early Copper Age to 16 of Middle/Late Copper Age date (Moreno et al. 1992/92). There were more severe situations in the Tabernas corridor (Alcaraz et al. 1987; 1994) and Andarax valley (Cara and Carrilero 1987; Cara and Rodríguez 1987) where evidence of Neolithic settlement is very scarce in contrast to the high number of settlements from the Copper Age.

All of these areas can be characterized by a scarcity of population until the end of the Neolithic when new openair settlements emerged in zones of high productivity, mainly in low-lying areas and fluvial valleys. These settlements had short occupations although a few of them survived into the Chalcolithic period. Although none appear to have had constructed defences, they are located in naturally defended positions, normally with good views over the surrounding area. During the Copper Age, a significant shift in settlement pattern had the following features: x

x x x x

a dramatic increase in the number of sites, for example in areas such as the Vera basin and Almanzora valley where the number of settlements rises from 40 the Neolithic to 214 in the Chalcolithic period (Martín et al. 1998); newly founded settlements, although, in some cases, the occupation of Neolithic sites continued; bigger sites than the previous Neolithic settlements; a greater degree of sedentism, inferred from long sequences of occupation without interruption on many sites; construction of complex defensive systems associated with the hierarchization of settlement pattern (see below).

In addition to this demographic growth there was, as we have noted above, a strong aggregation of population in small areas such as the Andrarax estuary. The population of Los Millares, the main settlement of this area, has been estimated as about 1000 inhabitants (Chapman 1990: 152) without taking into account the rest of the region’s settlements. Both elements – demographic increase and high-density population – could lead to considerable pressure on economic resources. This pressure can be identified from paleoecological analysis of settlements such as Los Millares where severe deforestation of its surrounding landscapes is documented for the end of the Copper Age (Rodríguez and Vernet 1991). Within this general context, the identification between society, territory and a high-density population is important for understanding the emergence and development of warfare.

The maximum degree of social complexity appears to have been attained in areas such as the Andarax estuary 183

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Hierarchization of settlement pattern and fortifications

x

Traditionally research on southeastern Iberia in later prehistory has concentrated on several sites of great size with complex defensive systems. Nevertheless, recent surveys carried out in different areas (as listed above) have documented a complex settlement pattern. Taking into account the particularities of each area, and differences in scale and intensity of settlement, it appears that Copper Age settlements were characterized by rigid hierarchization. In fact the population of certain territories was centralized in specific settlements. These central sites are characterized by their size (more than several hectares), their strategic position for the control of surrounding land and resources, and the construction of complex fortifications.

x

x

Fortified Chalcolithic settlements are rather common in the southeast and archaeological research carried out in the last two decades has increased their number substantially. Our knowledge, however, differs considerably from site to site: for some the defensive structures are documented only at a superficial level through regional survey and for others the fortifications have been thoroughly studied. We shall now present the features of the main defensive systems in order to assess their relevance in the development of warfare.

In a second category, we find smaller settlements (less than a hectare) positioned without dominating views of surrounding land, and without any use of natural or manmade defences. This second types of sites, found in a wide variety of locations, is linked to the exploitation of specific economic resources; the inhabitants may have been dependent on the central settlements. This two-tier settlement pattern can be found in a generalized way throughout the southeast although the results of research in specific areas allow us to infer different levels of complexity (Molina 1988; Moreno et al. 1991/92; Delibes et al. 1996; Martín et al. 1998; Monks 1999). In summary, several sites enjoyed a centralized position, incorporating significant defensive systems as their main features.

One of the most significant areas in the southeast, characterized by a high density of population, is the Andarax valley, with the site of Los Millares at the centre of the territory. This settlement is positioned on a terrace bounded by the confluence of the Andarax river the Huechar gully; hence it has an eminently strategic position with important natural defences which allow access only on its west side. In this western area, three lines of fortification in parallel rows close off and divide the terrace. A fortified citadel within the innermost area completes the defensive system. The outer wall is the most complex, with towers and bastions located at regular intervals. Two strongly fortified gateways have been identified; the main gateway with its barbican is the more complex (Fig. 2). The population of Los Millares could have occupied an area of approximately five hectares at the moment of its maximum size (Arribas et al. 1981; 1983; 1987; Arribas and Molina 1984a).

During the first half of the third millennium BC (Early/Middle Copper Age), new defensive strategies emerged. With varying levels of complexity, these included: x x x x x x

sites with naturally defensive locations; the construction of stone enclosure walls protecting the more accessible areas of settlements; bastions and towers constructed on walls; forts located in strategic places with wide visual control of territory; strongly defended gateways; complex systems of ditches.

The overall defensive system at Los Millares has two further lines of forts placed in strategic positions on the tops of higher hills. These forts protect the southern flank of the settlement for two kilometres and close off access along the rambla of Huechar and other adjacent areas. The thirteen known forts can be divided into three groups: a)

forts with a large circular tower of simple plan, in which only the gateway is reinforced; b) small or medium-sized stone enclosures reinforced with bastions; c) forts with complex structures.

Among the fortified sites of the southeast, the best known were probably all first-order central sites. Those highlighted here (Fig. 1) are: x x

Los Millares (Almagro and Arribas 1963; Arribas et al. 1981; 1983; 1987; Arribas and Molina 1984a) in the Andarax valley; El Cerro de la Virgen (Schüle 1980; 1986) and El Malagón (Arribas et al. 1978; de la Torre et al. 1984; de la Torre and Sáez 1986) on the high plateau of Granada; Cabezo del Plomo (Muñoz 1986) in Mazarrón (Murcia).

Almizaraque (Delibes et al 1986; 1996) and Campos (Camalich et al. 1987a; 1987b) in the Almanzora basin (Almería); Las Pilas (Alcaraz 1990) in the Aguas basin (Almería);

The best known fort (Fort 1; Fig. 3) can be included in this latter group. It is defined by two concentric enclosures of approximately circular shape, reinforced with towers and bastions. A ditch 4-5 metres deep

184

GONZALO ARANDA JIMÉNEZ AND MARGARITA SÁNCHEZ ROMERO: THE ORIGINS OF WARFARE

FIGURE 2. Main door and barbican at Los Millares (courtesy of Department of Prehistory, University of Granada).

FIGURE 3. Idealized reconstruction of Fort 1 at the site of Los Millares (courtesy of Department of Prehistory, University of Granada).

185

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY (approximately 5.5. hectares) and its long temporal sequence of occupation, one of the most complete from this region. Finally the discovery of the partial remains of a wall (Alcaraz 1990; Martín et al. 1998) makes this stand out as an important site with considerable potential for future research.

surrounds the outer wall and bastions, and further out a second ditch surrounds the fort completely (Arribas et al. 1981; 1983; 1987; Arribas and Molina 1984a). Defensive strategies at Los Millares also include loopholes located through the walls at regular intervals (Figs 2 and 3). There is high visibility of the surrounding area through these loopholes; such loopholes have also been documented in the outer wall of the settlement, especially in the barbican at the main gateway and in Forts 1 and 4 (Arribas et al. 1987). These features have been interpreted as vantage points from which archers could fire projectiles while remaining protected behind the wall. The visual angles of the loopholes towards strategic points, and their height above the ground, support this idea (Arribas and Molina 1984a).

The settlement of Cabezo del Plomo is situated in the south of Murcia, on the northeastern boundary of the Millares culture’s maximum expansion (Fig. 1). Following the pattern described for the previous sites, it has an important strategic position on a long terrace with a steep drop on its northern and eastern sides. Moreover its location provides control over an important routeway between the Guadalentín basin and the sea. An L-shaped wall defends the most accessible part of the sites, on its western and southern sides. This stone enclosing wall has eight towers and bastions at regular intervals, and delimits an area of about 3000 sq m (Muñoz 1986: 1467). With these characteristics, Cabezo del Plomo could have had a central position within the settlement organization of the surrounding territory.

The Vera basin is another area where several research projects have been carried out recently. This is a welldefined region that includes the Almanzora, Antas and Aguas rivers, all of which flow into the Mediterranean (Fig. 1). In this area several sites with major fortifications show the process of hierarchization mentioned above. Among these settlements, we would like to emphasize Almizaraque and Campos, located in the Almanzora valley, and Las Pilas in the Aguas valley. All have been studied exhaustively in recent years. In terms of their role in territorial organization, Almizaraque is the most controversial. Its current investigators argue that this site was small, less than 3000 sq m (0.3 hectares) in size, and that its wall belongs to a late period of the Copper Age (Delibes et al. 1996). Nevertheless, other researchers point to this site’s central position and its potential control of the Vera basin (Arribas and Molina 1984b: 96; Martín et al. 1998: 153).

Finally, in recent years there has been a significant amount of research on the high plateau in the eastern part of the province of Granada. One of the best-known archaeological sites currently being studied is El Malagon (Fig. 1; Arribas et al. 1978; de la Torre et al. 1984). The area of the site includes a high hill with wide visual control of the surrounding territory. Its dimensions (almost 4 hectares), its strategic situation and its complex fortifications give El Malagon the status of a central site (Moreno et al. 1991/92). As for its defensive system, several lines of stone walls have been documented in the surveys carried out in the area of the settlement; they could delimit several concentric enclosures with the same pattern as Los Millares. Excavations at the site have revealed a line of fortifications with different construction phases and a maximum thickness of 4.8 m. Some structures associated with this wall are interpreted as possible bastions or towers (de la Torre et al. 1984; de la Torre and Sáez 1986).

The site of Campos is less problematic. It is located on the edge of a spur delimited by the Almanzora river. Excavations at the end of the nineteenth century revealed a double stone-walled enclosure of trapezoidal shape with an extension of 200 sq m. The outer line of fortifications has towers and bastions in each of its three surviving corners and a ditch associated with the wall on its northeastern side (Siret and Siret 1890). More recent research has suggested that its dimensions, shape and arrangement indicate that this fortification would have been merely the inner part of a defensive complex similar to Los Millares or Zambujal (Martín and Camalich 1986).

Within the same region, the last site to mention here is Cerro de la Virgen, in the Orce valley (Fig. 1). It is located on a hilltop in a strategic position and has a long cultural sequence from the Copper Age to the Bronze Age. During the Copper Age several defensive constructions – only partially known because of limited fieldwork – formed a line of fortification in the more accessible area, and some semicircular stone structures interpreted as bastions (Schüle 1980; 1986).

Some of the most important research in the Vera basin has come out of excavations undertaken on the site of Las Pilas, located near the mouth of the Aguas river. Although we only have preliminary results (Alcaraz 1990), it seems clear that its features point to its having been a site occupying a central position and exercising control over the Vera basin (Martín et al. 1998). The settlement enjoys a strategic location on an important route of communication, positioned on a terrace separated from the surrounding landscape by distinct natural topography. It is prominent in terms of its size

These sites with their central positions in the settlement pattern can be summarized thus: x

186

each of the main regions in the area of the Millares culture has at least one of these sites. (The long tradition of research in southeastern

GONZALO ARANDA JIMÉNEZ AND MARGARITA SÁNCHEZ ROMERO: THE ORIGINS OF WARFARE

x x

labour was still very considerable in the construction of these fortified settlements, although with notable differences in scale.

Iberia makes is unlikely that any others remain undiscovered.); the sites reveal a strong preference for locations on terraces separated from the surrounding landscape and with important natural defences; despite different levels of complexity, the fortification systems appear markedly homogeneous. The more accessible areas are defended by one or more stone enclosing walls linked to forts, towers, bastions and ditches. Construction techniques are highly standardized.

Although we have highlighted so far those elements relating to increases in conflict and warfare, we are conscious that this evidence has other connotations. These defensive constructions, permanent and highly visible, were symbols of power adding to the authority of the large settlements which centralized the exploitation of the surrounding territories. The monumentality and complexity of some of these fortifications were also far more substantial than can have been strictly necessary for purposes of defence. This is the case for Los Millares, Campos, El Malagón and Cabezo del Plomo, where these constructions display the capacity to mobilize a considerable labour force. These defensive systems were important symbolically as ideological forms of coercion, directed not only at any rival populations but also at different elements within the resident community. They were physical and symbolic boundaries that divided the inhabitants’ space but also reinforced internal cohesion and a sense of identity between people, place and territory.

Overall, we conclude that the relationship between fortifications and specific sites is not only a significant factor in settlement hierarchization but also a factor in an increase in conflict and social complexity. To appreciate the development of important defensive strategies, we must take into account the effort expended in construction and maintenance. The planning and construction of walls, bastions, towers, forts and ditches must have required a significant investment of labour. Furthermore, the maintenance of these structures – attested to by the frequency of rebuilding episodes documented in the settlements discussed above – is likely to have been a major investment in the long term, in some cases over a period of hundreds of years.

The emergence of warriors Assessments of the emergence and development of weaponry point, on the one hand, to the difficulty of giving specific meanings to multifunctional tools made of stone, metal or bone (arrowheads, awls, axes, maces, knives and daggers) and, on the other hand, to the absence of weapons made of perishable materials (Vencl 1984; 1999; Keeley 1996). The clear evidence of deaths caused by violence indicates the existence of weapons in European later prehistory, but the identification of weaponry is difficult since it is both typologically diverse and also of low archaeological visibility. The standardization and specialization of weapons that developed throughout the Bronze Age was associated with the ritualization of power as legitimization and institutionalization of differences in social status (Kristiansen 1999: 176).

Taking into account the size and complexity of the defensive constructions, the volume of work could have required groups of craftsmen with different levels of specialization. The complexities of the building process required various elements: raw materials from quarries such as that documented at Los Millares; the transport of large stones; specialized knowledge of wall construction, including mortar preparation, lifting tackle and erection; rebuilding because of changes in spatial organization; and long-term maintenance of the finished constructions. In the case of Los Millares, several calculations have been made of the amount of work invested in the construction of walls, bastions and towers (Monks 1997: 20-1). Two factors have been taken into account: the volume of the defensive constructions, assuming that the walls were 4 metres high and made entirely of stone, and the labour investment. To measure this, anthropological and experimental studies were used to calculate that the building of each square metre (including supply and transport) took ten days’ work. Consequently it has been calculated that the defensive complex of Los Millares could have taken over 100,000 days of work (ibid.: 21).

With regard to the problem of multifunctional artefacts, Vencl defines three analytical units: x x x

In our opinion, this seems excessive since the walls were built on a stone platform upon which a framework of wood and clay could have been erected. The stone walls might well have been less than 4 metres high. Furthermore, the number of workers has not been considered (or, at least, is not mentioned by Monks) and the working day is deemed to have been only five hours long. Nevertheless it is clear that the investment of

occasional weapons, used to attack or repel aggression in an emergency situation; non-specialized weapons, used not only as tools but also as weapons (axes, arrowheads, daggers, bows); specialized weapons such as swords and shields (Vencl 1999: 65).

Chapman similarly identifies two fundamental categories for grouping potentially multifunctional weaponry: tools/weapons and weapons/tools, distinguished according to the main function of each artefact (Chapman 1999: 108). 187

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY An initial assessment of material culture during the third millennium BC identifies artefacts which fall into the category of non-specialized weapons or tools/weapons; axes, awls, daggers and arrowheads have clear uses as tools but their potential as weapons has also been demonstrated by archaeological and ethnographic evidence (Vencl 1984; 1999; Keeley 1996). Flint arrowheads, in our opinion, can be associated with an increase in conflict. We consider them to have been specialized weapons which were used less frequently for hunting.

lodged in the bones (Vencl 1984; 1999; Keeley 1996; for Iberia see Armendáriz and Irigaray 1995; Vegas et al. 1999). In the southeast, there is no such evidence, perhaps owing to the lack of recent excavation or anthropological study of burials. There is, however, important information on arrowheads in their contexts of production and use. Workshops specializing exclusively in arrowhead production have been found in Fort 1 at Los Millares (Molina et al. 1986; Ramos et al. 1991) and at Almizaraque (Siret 1948). In this latter settlement, Siret documented a workshop with arrowheads under manufacture, a large amount of debris, and different tools related to their production – blades with splintered and/or polished ends and animals bone use for pressure retouch (ibid.: 121-4). In the case of Fort 1 at Los Millares, we can highlight a workshop within an oval habitation with a central hearth. Inside are all the elements related to arrowhead production: flint cores, blades, flakes, debris, tools for pressure retouch, and facilities for heat treatment of flint. Concave-based arrowheads were the principal typological form manufactured here (Ramos et al. 1991: 177-81).

Arrowheads appeared in the southeast during the Copper Age but not as substitutes for microliths, previously used for hunting, since these remained in use (Leisner and Leisner 1943; Almagro and Arribas 1963). Morphologically arrowheads are characterized by their relative variety, which has been interpreted in terms of their different functions and evolution over time. For example, the arrowheads with concave bases and plain in shape have been related to the emergence of warfare and warriors (Ramos et al. 1991: 62). In addition to these qualitative aspects, arrowhead production was considerable in settlements such as Las Pilas or Campos, where they form 17% and 15% respectively of the entire lithic assemblage (Martínez and Afonso 1998a: 247). This high proportion is even more marked when we consider all the specialized tools employed in their production, such as blades with splintered and/or polished ends used in pressure retouch.

The general context of this workshop has an essential implication, namely the relationship between the arrowheads’ production and their probable context of use: the fort. This association is confirmed by a find in Fort 7 of a pot with flint flakes for arrowhead production. We have other contextual evidence for the use of arrowheads: in the outer defences at Los Millares, several arrowheads have been found broken as a result of their impact against the wall (F. Molina pers. comm.).

Regarding the analysis of technology, important studies have been carried out on the site of Las Peñas de Los Gitanos (Montefrío, Granada; Martínez 1985; Afonso 1993; Sánchez 2001). This site is not strictly located in the southeast but it has a very long cultural sequence that illustrates the general trajectory of lithic production through time. We also have recent lithic analyses from Chalcolithic sites such as Zájara and Campos in the Vera basin (Martínez and Afonso 1998b) that confirm the homogeneity of this trajectory throughout the region. In general, the appearance of arrowheads is associated with important changes in production techniques, giving rise to a better quality of finished tool. The principal innovation was the heat treatment of cores and flakes before secondary modification. This increases fragility to allow finer knapping of arrowheads with thinner edges. Their great lightness would have enabled these arrows to have been fired over greater distances whilst the sharpness of the blade gave them greater penetrative power. These arrowheads would have broken easily, thus remaining within the wounded target’s body (Martínez 1985; Afonso 1993).

The increasing significance of arrowheads as weapons can be characterized by x x x x x x

significant production of debris; continuity of microliths as tools related to hunting; technological changes to improve the quality of arrowheads; specialized workshops for arrowhead production; a context of use, documented mainly at Los Millares; low proportions of hunted animals among the faunal assemblages.

Although arrowheads could be used for hunting, their appearance, development and specialization is best understood in terms of their relation to the increase in conflict and warfare during the third millennium BC. Arrowheads have a symbolic and ritual dimension which is reflected in their appearance as grave goods within funerary contexts. Another artefact recorded almost exclusively in burials is the flint dagger. These daggers are characterized by their typological variety and their

Systemic contexts provide information for analyzing the use of arrowheads. Again, low archaeological visibility presents an obstacle. Important evidence in this respect is the identification of collective burials or human bones associated with arrowheads, especially when they have 188

GONZALO ARANDA JIMÉNEZ AND MARGARITA SÁNCHEZ ROMERO: THE ORIGINS OF WARFARE been used to identify differential access to wealth and, consequently, the emergence of unequal social groupings. This hypothesis is supported by the location of burials with prestige grave goods within the inner half of the necropolis – all but one are situated less than 150 metres from the outer line of fortification.

scarcity; they could have had a pre-eminently ritual use similar to that of the arrowheads (Ramos et al. 1991: 63). But what were the symbolic meanings of daggers and arrowheads? Were they used in displays of power and ideological coercion? Or, on the contrary, did they symbolize the reproduction of community, as we may expect from a collective burial ritual?

Chapman’s definition of prestige grave goods at Los Millares did not include flint arrowheads or daggers; presumably he considered that these two items were either not used in exchange networks or were used only marginally. Current research, however, allows us to include arrowheads and daggers in this group because they meet two basic criteria defined by Chapman (1981).

The answer to these questions is complicated by the rapidity of social changes during this period. From the end of the fourth and especially during the third millennium BC, the process of growing complexity can be presumed to have caused important changes to symbolism and ideology. In this way, what were at one time symbolic representations of community might have been transformed into symbols of the status and wealth of specific social groups. Equally, the variation in complexity between social groups in the southeast could have also had a significant influence on the evolution of symbolic meanings. For example, the best archaeological evidence for the symbolic connotations of arrowheads comes from the region in which greatest complexity was reached – the Andarax valley and the funerary evidence from Los Millares.

Firstly, the grave goods recovered by the Siret brothers do seem to be representative of those present in the tombs before excavation. From the 21 sepultures re-excavated and published by Almagro and Arribas (1963), 14 could be correlated with those published by the Leisners (1943). With regard to arrowheads, during these more recent reexcavations six tombs yielded no new finds, three only a simple additional item whilst those with more finds were the sepultures in which the quantity of arrowheads recovered by the Sirets has also been high (Chapman 1981: 392, tab. 14.2). Daggers were better represented in the Sirets’ excavations than arrowheads; no daggers were found during Almagro and Arribas’s re-excavations.

At the end of the nineteenth century the Siret brothers – two Belgian mining engineers – excavated the Los Millares necropolis almost completely. From the 75 tombs and their grave goods published by G. and V. Leisner (1943), it is apparent that funerary ritual consisted principally of collective burial or of tholoi, with a central circular chamber, a false vault, an access corridor and a mound. Arising from these basic features are variations with different degrees of complexity. In some cases the corridor is divided into sections by perforated orthostats. In other cases there are sidechambers which were used for burial, located not only in the central chamber but also in the corridors. There are also important differences in tomb size.

Secondly, recent research in the southeast has provided important new data about production processes and exchange of flint on a regional scale. The importance of flint production is confirmed by the systematic exploitation of secondary deposits and, especially, by the development of flint mines. This is case at Sierra de Orce and Maria on the Grenadine high plateau, where one of these flint mines at La Venta has been thoroughly researched. The appearance of these mines implies a considerable intensification of production which has been related to increasingly complex exchange networks (Ramos et al. 1991: 79-92).

After the Leisners’ studies Almagro and Arribas (1963) developed an important research programme focused on the re-excavation of several sepultures and the identification in the field of the tombs excavated by the Siret brothers. The 57 tombs nearest to the settlement were analyzed and of these 31 (41%) can be correlated with those published by G. and V. Leisner (Chapman 1990).

The exchange of flint resources can be recognized throughout the southeast by the identification of non-local raw material on several sites. This is the case for the workshops specializing in arrowhead production at Almizaraque (Siret 1948: 123-4) and Fort 1 at Los Millares (Ramos et al. 1991: 181) where flint comes from inter-regional exchange. Lithic assemblages at the sites of Campos and Zájara were derived not only from secondary deposits in the surrounding area but also from several non-local sources of raw materials (Martínez and Afonso 1998b: 248). Consequently, intensification in flint production, its participation in exchange networks and its use in arrowhead manufacture provides the added value that Chapman uses to define prestige grave goods.

Robert Chapman (1977; 1981; 1990) has carried out more recent research on the Los Millares necropolis, analyzing differences principally among grave goods in order to establish social inferences. He interprets as prestige goods items made from ivory, ostrich egg shells, amber, jet and copper as well as some of the decorated ceramics. The value of these objects as symbols of status and wealth is defined by their exotic character, their elaborate manufacture and their participation in complex and longdistance exchange networks. The differential distribution and frequency of these prestige goods in the burials has

The distribution of arrowheads and daggers among the sepultures at Los Millares can be studied in terms of their associations with the prestige tombs defined by Chapman 189

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY

TABLE 1. Distribution of prestige grave goods in sepultures at Los Millares (after Chapman 1990). Arrowheads and daggers have been added by the authors. In calculating the total number of arrowheads, we have considered the grave goods published by Leisner (1943) and the re-excavation of tombs undertaken by Almagro and Arribas (1963). F=sherd, N=unspecified number.

(1981; 1990) in order to establish their symbolic meanings. Arrowheads appear as grave goods in 32 (42.6%) of the 75 sepultures published by the Leisners (1943), although with dramatic differences in their distribution. In 62.5% of these 32 sepultures with arrowheads, there are fewer than five arrowheads present. In 12.5% (four) of the sepultures, there are between five and 20 arrowheads present. In just two tombs (6.2%) are there between 20 and 30 arrowheads; a further two tombs have between 40 and 50 and three burials (9.3%) each has

more than 70 arrowheads (Table 1). These are significant patterns because more than 50% of sepultures have no arrowheads and 79.1% of the entire arrowhead assemblage derives from only seven burials. How does this distribution and concentration of arrowheads relate to other prestige grave goods defined by Chapman? Of the 32 sepultures with arrowheads, 26 (81.2%) have prestige goods. Only in six burials were there arrowheads without any other status item; all six of 190

GONZALO ARANDA JIMÉNEZ AND MARGARITA SÁNCHEZ ROMERO: THE ORIGINS OF WARFARE warfare and warriors. Our starting point was the significant social and economic changes which took place in the Late Neolithic and Early Copper Age, leading to the colonization of plains and fluvial valleys with high agricultural productivity, the sedentism of the population, and a considerable demographic growth with consequent high population densities.

these burials have fewer than five arrowheads. In the burials with both arrowheads and other prestige goods, there is a clear pattern of association between grave goods that include the greatest quantity and variety of prestige items and a high concentration of arrowheads. As Chapman has established, there are two categories of tomb defined as prestige sepultures and sepultures with prestige grave goods. The number and variety of status elements are the key to this differentiation (Chapman 1981: 402-3; 1990: 191). The first category includes burials 5, 7, 8, 9, 12, 16, 40, 57 and 63. Within this group are the seven sepultures in which 79.1% of the arrowheads were found.1 Consequently all the graves with the highest concentrations of arrowheads are among the prestige sepultures. With regard to daggers, the situation is similar. Of the six daggers, three were found in prestige sepultures and the other three in tombs with prestige grave goods if we consider arrowheads as such.

All these processes provoked the ascription and identification of each group with a territory and its resources, and the development of a concept of landed property characterized by exclusion and competition. In this context, settlement patterns with strong hierarchization appeared. Large settlements were built in strategic locations, with complex defensive systems that combined walls, bastions, towers, forts, strongly fortified gateways and ditches. These sites centralized the exploitation of broad territories that included other lower level, undefended settlements. At the same time as these defensive systems were built, arrowheads emerged as specialized weapons. Although they can be connected to hunting, their appearance and development is only comprehensible in terms of their relationship to an increase in conflict. This can be measured by the considerable volume of production of arrowheads, the development of specialized workshops, changes in techniques to improve weapon efficiency, a context of use clearly related to the defended sites, and the relative unimportance of hunting as represented among the faunal remains.

These data show a clear association of arrowheads and daggers with specific social groups with differentiated access to prestige goods. Burial 40, for example, with the highest number of arrowheads (85), also has the highest number of artefacts in ivory and copper, along with pottery with painted decoration. The ritual and symbolic connection of arrowheads and daggers with specific social groups allows us to confirm their link with warfare. In terms of their symbolic meaning, arrowheads and daggers could indicate warrior status, the display of wealth and the embodiment of power. Despite clear social differences, however, collective burial rituals continued throughout the Copper Age. In conclusion, the appearance of arrowheads as specialized weapons, and their relationship with specific social groups, allows us to talk about the emergence of warriors during the third millennium BC.

All of these elements indicate that the appearance of arrowheads at that time was not a casual development but a necessary innovation of specialized weaponry. At the same time, arrowheads as objects included in burials, had a ritual dimension. From the funerary evidence at Los Millares, the relationship between arrowheads and those specific social groups with the richest grave goods allows us to confirm the appearance of warriors during the third millennium BC. Arrowheads demonstrate warrior status, emerging hand in hand with wealth and power.

Conclusion In this paper we have tried to investigate the origins of warfare in prehistoric societies through the archaeological evidence of southeastern Iberia. In our opinion the concept of warfare is possible in complex societies in which violence has been institutionalized through the creation of a specific social group: the warriors. Hence, amongst the wide variety of possible violent behaviour, warfare should be considered to be a complex form of conflict regulation. We have attempted to address the differences between ritualized and real warfare. It seems likely that all conflicts have ritual and symbolic aspects which regulate the exercise of violence, although this need not imply less efficiency or pursuit of goals other than the destruction of the enemy.

Bibliography Afonso, J. 1993. Aspectos técnicos de la producción lítica de la prehistoria reciente de la alta Andalucía y sureste. Unpublished PhD thesis, Universidad de Granada. Alcaraz, F. 1990. Excavación arqueológica de emergencia en Las Pilas-Huerta Seca (Mojácar, Almería). Anuario Arqueológico de Andalucía (1989):18-24. Alcaraz, F., Castilla, J., Hitos, M.A., Maldonado, M.G., Mérida, V. and Rodríguez, F.J. 1987. Proyecto de prospección arqueológica superficial llevado a cabo en el pasillo de Tabernas (Almería). Anuario Arqueológico de Andalucía (1986) 2: 62-5. Alcaraz, F., Castilla, J., Hitos, M.A., Maldonado, M.G., Mérida, V., Rodríguez, F.J. and Ruiz, M.V. 1994.

From these premises out attention has been focused on the analysis of processes that lead to the appearance of 1 Burial 5: 75 arrowheads; burial 7: 44; burial 9: 28; burial 16: 27; burial 40: 85; burial 57: 48; burial 63: 76.

191

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Arribas, A., Molina, F., Sáez, L., de la Torre, F., Aguayo, P., Bravo, A. and Suárez, A. 1983. Excavaciones en los Millares (Santa Fe de Mondújar, Almería). Campaña de 1982 y 1983. Cuadernos de Prehistoria de la Universidad de Granada 8: 123-48. Arribas, A., Molina, F., Sáez, L., de la Torre, F., Aguayo, P. and Nájera, T. 1981. Excavaciones en los Millares (Santa Fe de Mondújar, Almería). Campaña de 1981. Cuadernos de Prehistoria de la Universidad de Granada 6: 91-122. Arribas, A., Molina, F., de la Torre, F., Nájera, T. and Sáez, L. 1978. El poblado de la edad del cobre de ‘El Malagón’ (Cullar-Baza, Granada). Cuadernos de Prehistoria de la Universidad de Granada 3: 67-116. Blance, B. 1961. Early Bronze Age colonists in Iberia. Antiquity 35: 192-202. Brothwell, D. 1999. Biosocial and bio-archaeological aspect of conflict and warfare. In J. Carman and A Harding (eds) Ancient Warfare. Stroud: Sutton. 25-38. Camalich, M.D., Martín, D. and Acosta, C. 1987a. Excavaciones en el yacimiento de Campos (Cuevas de Almanzora, Almería). Campaña de 1985. Anuario Arqueológico de Andalucía (1985) 2: 134-40. Camalich, M.D., Martín, D., Acosta, C. and Meneses, M.D. 1987b. Excavaciones arqueológicas en el yacimiento de Campos (Cuevas de Almanzora, Almería). Anuario Arqueológico de Andalucía (1986) 2: 288-95. Campillo, D. 1995. Agressivitat i violència a les societats prehistòriques i primitives. Limes 4/5: 5-17. Camps, G. 1992. Guerre ou paix? Origines des conflits intraspécifiques humains. Prehistoire et Anthropologie Mediterranéennes 1: 9-16. Cara, L. and Carrilero, M. 1987. Prospección arqueológica superficial del estuario del Andarax y piedemonte de la Sierra de Gador (Almería). Anuario Arqueológico de Andalucía (1985) 2: 63-66. Cara, L. and Rodríguez, J.M. 1987. Prospección arqueológica superficial del valle medio del río Andarax. Anuario Arqueológico de Andalucía (1986) 2: 58-61. Carman, J. (ed.). 1997. Material Harm: archaeological studies of war and violence. Glasgow: Cruithne Press. Carman, J. and Harding, A. (eds) 1999a. Ancient Warfare. Stroud: Sutton. Carman, J. and Harding. A. 1999b. Introduction. In J. Carman and A. Harding (eds) Ancient Warfare. Stroud: Sutton. 1-10. Castro, P., Lull, V. and Micó, R. 1996. Cronología de la Prehistoria Reciente de la Península Ibérica y Baleares (c. 2800-900 cal ANE). Oxford: BAR International Series 652. Chapman, J. 1999. The origins of warfare in the prehistory of central and eastern Europe. In J. Carman and A. Harding (eds) Ancient Warfare. Stroud: Sutton. 101-42. Chapman, R.W. 1977 Burial practices: an area of mutual interest. In M. Spriggs (ed.) Archaeology and Anthropology: areas of mutual interest. Oxford: BAR Supplementary Series 19. 19-31.

Prospección arqueológica superficial en el pasillo de Tabernas: primeros resultados y perspectivas metodológicas. In special issue ‘Origens, estruturas e relacoes das culturas calcolíticas da península ibérica (Actas das primeras jornadas arqueológicas de Torres Vedras 3-5 Abril de 1987)’. Trabalhos de Arqueología 7: 217-23.Almagro, M. and Arribas, A. 1963. El Poblado y la Necrópolis Megalítica de Los Millares (Santa Fe de Mondújar, Almería). Madrid: Biblioteca Praehistorica Hispánica 8. Armendariz, J. and Irigaray, S. 1995. Violencia y muerte en la prehistoria: el hipogeo de longar. Revista de Arqueología 168: 16-29. Arribas, A. and Molina, F. 1984a. The latest excavations of the Copper Age settlement of Los Millares, Almería, Spain. In W. Waldren, J. Chapman, R. Lewthwaite and C. Kennard (eds) The Deya Conference on Prehistory. Early Settlement in the Western Mediterranean Islands and Their Peripheral Areas, 3. Oxford: BAR International Series 229 (4). 1029-50. Arribas, A. and Molina, F. 1984b. Estado actual de la investigación del megalitismo en la península ibérica. In J. Fortea (ed.) Scripta Praehistorica: homenaje a Francisco Jordá. Salamanca: Acta Salmanticensia. 63-112. Arribas, A., Molina, F., Carrión, F., Contreras, F., Martínez, G., Ramos, A., Sáez, L., de la Torre, F., Blanco, I. and Martínez, J. 1987. Informe preliminar de los resultados obtenidos durante la VI campaña de excavaciones en el poblado de los Millares (Santa Fe de Mondújar, Almería, 1985). Anuario Arqueológico de Andalucía (1985) 2: 2345-262. Almagro, M. and Arribas, A. 1963. El Poblado y la Necrópolis Megalítica de Los Millares (Santa Fe de Mondújar, Almería). Madrid: Biblioteca Praehistorica Hispánica 8. Armendariz, J. and Irigaray, S. 1995. Violencia y muerte en la prehistoria: el hipogeo de longar. Revista de Arqueología 168: 16-29. Arribas, A. and Molina, F. 1984a. The latest excavations of the Copper Age settlement of Los Millares, Almería, Spain. In W. Waldren, J. Chapman, R. Lewthwaite and C. Kennard (eds) The Deya Conference on Prehistory. Early Settlement in the Western Mediterranean Islands and Their Peripheral Areas, 3. Oxford: BAR International Series 229 (4). 1029-50. Arribas, A. and Molina, F. 1984b. Estado actual de la investigación del megalitismo en la península ibérica. In J. Fortea (ed.) Scripta Praehistorica: homenaje a Francisco Jordá. Salamanca: Acta Salmanticensia. 63-112. Arribas, A., Molina, F., Carrión, F., Contreras, F., Martínez, G., Ramos, A., Sáez, L., de la Torre, F., Blanco, I. and Martínez, J. 1987. Informe preliminar de los resultados obtenidos durante la VI campaña de excavaciones en el poblado de los Millares (Santa Fe de Mondújar, Almería, 1985). Anuario Arqueológico de Andalucía (1985) 2: 2345-262. 192

GONZALO ARANDA JIMÉNEZ AND MARGARITA SÁNCHEZ ROMERO: THE ORIGINS OF WARFARE el sudeste. Unpublished PhD thesis, Universidad de Granada. Martínez, G. and Afonso, J.A. 1998a. La producción lítica: un modelo para el análisis histórico de los conjuntos arqueológicos de piedra tallada. In J. Bernabeu, T. Orozco and X. Terradas (eds) Los Recursos Abióticos en la Prehistoria: caracterización, aprovisionamiento e intercambio. Valencia: Universidad de Valencia. Martínez, G. and Afonso, J.A. 1998b. Producción lítica tallada de Campos y Zájara. In M.D. Camalich and D. Martín (eds) El Territorio Almeriense desde los Inicios de la Producción hasta Fines de la Antigüedad. Un modelo: la depresión de Vera y cuenca del río Almanzora. Sevilla: Conserjería de Cultura de la Junta de Andalucía. 235-49. Mathers, C. 1984. Beyond the graves: the context and wider implications of mortuary practices in south-east Spain. In T.F.C. Blagg, R.F.J. Joves and S.J. Keay (eds) Papers in Iberian Archaeology, 1. Oxford: BAR International Series 193 (I). 13-46. Mercer, R.J. 1999. The origins of warfare in the British Isles. In J. Carman and A. Harding (eds) Ancient Warfare. Stroud: Sutton. 143-56. Molina, F. 1988. El sudeste. In special issue ‘El Calcolítico de la Península Ibérica. Congresso Internazionale L’Età del Rame in Europa (Viarenggio, 15-18 Ottobre 1987)’ (eds G. Delibes, M. FernándezMiranda, A. Martín and F. Molina). Rassegna di Archaeologia 7: 256-62. Molina, F., Contreras, F., Ramos, A., Mérida, V., Ortíz, F. and Ruiz, V. 1986. Programa de recuperación del registro arqueológico del Fortin 1 de Los Millares. Análisis preliminar de la organización del espacio. Arqueología Espacial 8: 175-201. Monks, S. 1997. Conflict and competition in Spanish prehistory: the role of warfare in social development from the late fourth to third millennium BC. Journal of Mediterranean Archaeology 10: 3-32. Monks, S. 1999. Patterns of warfare and settlement in southeast Spain. Journal of Iberian Archaeology 1: 127-71. Moreno, A., Contreras, F. and Cámara, J.A. 1991-92. Patrones de asentamiento, poblamiento y dinámica cultural en las tierras altas del sureste peninsular: el pasillo de Cúllar-Chirivel durante la prehistoria reciente. Cuadernos de Prehistoria de la Universidad de Granada 16-17: 191-245. Muñoz, A.M. 1986. El eneolítico en el sudeste. Historia de Cartagena 2: 143-62 Oosterbeek, L 1997. War in the Chalcolithic? The meaning of western Mediterranean Chalcolithic hillforts. In J. Carman (ed.) Material Harm: archaeological studies of war and violence. Glasgow: Cruithne Press. 116-32. Ramos, A. 1981. Interpretaciones secuenciales y culturales de la edad del cobre en la zona meridional del península ibérica. Cuadernos de Prehistoria de la Universidad de Granada 6: 203-56.

Chapman, R.W. 1981. Archaeological theory and communal burial in prehistoric Europe. In I. Hodder, G. Isaac and N. Hammond (eds) Patterns of the Past: studies in honour of David Clarke. Cambridge: Cambridge University Press. 387-411. Chapman, R.W. 1990. Emerging Complexity: the later prehistory of the south-east of Spain, Iberia and the west Mediterranean. Cambridge: Cambridge University Press. Delibes, G., Díaz-Andreu, M., Fernández-Posse, C.M., Montero, I., Muñoz, I.K. and Ruiz, A. 1996. Poblamiento y desarrollo cultural en la cuenca de Vera durante la prehistoria reciente. Complutum Extra 6: 153-70. Delibes, G., Fernández-Miranda, M., Fernández-Posse, M.D. and Martín, C. 1986. El poblado de Almizaraque. In Homenaje a Luis Siret (1943-1984). Sevilla: Conserjería de Cultura. 167-77. Díaz-Andreu, M. 1995. Complex societies in Copper and Bronze Age Iberia: a reappraisal. Oxford Journal of Archaeology 14: 23-33. Dolukhanov, P.M. 1999. War and peace in the prehistory of eastern Europe. In J. Carman and A. Harding (eds) Ancient Warfare. Stroud: Sutton. 73-88. Gilman, A. 1987. Unequal development in Copper Age Iberia. In T. Earle and E.M. Brumfiel (eds) Specialization, Exchange and Complex Societies. Cambridge: Cambridge University Press. Haas, J. (ed.) 1990. The Anthropology of War. New York & Cambridge: Cambridge University Press. Haas, J. 1999. The origins of war and ethnic violence. In J. Carman and A. Harding (eds) Ancient Warfare. Stroud: Sutton. 11-24. Keeley, L.H. 1996. War before Civilization: the myth of the peaceful savage. Oxford & New York: Oxford University Press. Kristiansen, K. 1999. The emergence of warrior aristocracies in later European prehistory and their long-term history. In J. Carman and A. Harding (eds) Ancient Warfare. Stroud: Sutton. 175-89. Kunst, M. 2000. A guerra no calcolítico na peninsula ibérica. Era-arqueología 2: 128-42. Leisner, G. and Leisner, V. 1943. Die Megalithgraber der Iberischen Halbinsel: der Suden. Berlin: Walter de Gruyter. Martín, D. and Camalich, M.D. 1986. Las excavaciones en el poblado de Campos (Cuevas de Almanzora, Almería) y su problemática. In Homenaje a Luis Siret (1934-1984). Sevilla: Conserjería de Cultura. 267-306 Martín, D., Mederos, A., Chávez, E., Díaz, A., Aramburu, E. and López, J. 1998. Estudio del territorio. In M.D. Camalich and D. Martín (eds) El Territorio Almeriense desde los Inicios de la Producción hasta Fines de la Antigüedad. Un modelo: la depresión de Vera y cuenca del río Almanzora. Sevilla: Conserjería de Cultura de la Junta de Andalucía. 137-70. Martínez, G. 1985. Análisis tecnológico y tipológico de las industrias de piedra tallada del neolítico, la edad del cobre y la edad del bronce de la alta Andalucía y

193

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Siret, E. and Siret, L. 1890. Las Primeras Edades del Metal en el Sudeste de la España. Barcelona. Siret, L. 1913. Questions de Chronologie et d’Ethnographie Ibériques. Paris: P. Geuthner. Siret, L. 1948. El tell de Almizaraque y sus problemas. Cuadernos de Historia Primitiva 3: 117-24. Torre de la, F., Molina, F., Carrión, F., Contreras, F., Blanco, I., Moreno, Mª. A. and Ramos, A. 1984. Segunda campaña de excavaciones en el poblado de la edad del cobre de ‘El Malagón’ (Cullar-Baza, Granada). Cuadernos de Prehistoria de la Universidad de Granada 9: 131-46. Torre de la, F. and Sáez, L. 1986. Nuevas excavaciones en el yacimiento de la edad del cobre de El Malagón (Cullar Baza, Granada). In Homenaje a Luis Siret (1934-1984). Sevilla: Conserjería de Cultura. 221-26. Tringham, R. 1972. Territorial demarcation of prehistory settlements. In P. Ucko, R. Tringham and G.W. Dimbleby (eds) Man, Settlement and Urbanism. Hertfordshire: Garden City Press. 463-76. Vegas, J.I., Armendariz, A., Etxeberria, F., Fernández, M.Sª., Herrasti, L. and Zumalabe, F. 1999. La sepultura colectiva de San Juan ante Portam Latinam (Laguardia, Álava). Saguntum Extra 2: 439-46. Vencl, S. 1984. War and warfare in archaeology. Journal of Anthropological Archaeology 3: 116-32. Vencl, S. 1999. Stone age warfare. In J. Carman and A. Harding (eds) Ancient Warfare. Stroud: Sutton. 57-72.

Ramos, A., Martínez, G., Ríos, G. and Afonso, J.A. 1991. Flint Production and Exchange in the Iberian Southeast, III millennium B.C. Granada: Universidad de Granada. Rodríguez, M.O. and Vernet, J.L. 1991. Premiers resultats paléoecologiques de l’établissement chalcolithique de Los Millares, Almería, d’après l’analyse anthropologique de l’établissement. In W.H. Waldren, J.A. Ensenyat and R.C. Kennard (eds) 2nd Deya International Conference on Prehistory. Recent Developments in Western Mediterranean Prehistory: archaeological techniques, technology and theory. . Oxford: BAR International Series 573. 1-16. Sáez, L. and Martínez, G. 1981. El yacimiento neolítico al aire libre de la Molaina. Cuadernos de Prehistoria de la Universidad de Granada 6: 17-34. Sánchez, M. 2001. Espacios de Producción y Uso de los Útiles de Piedra Tallada del Neolítico. Oxford: BAR International Series 874. Sangmeister, E. 1975. Spätes Neolithikum und Kupferzeit der iberischen Halbinsel (ed. K.J. Napp). Handbuch der Urgeschichte 2: 545-54. Schüle, W. 1980. Orce und Galera. In Zwei Siedlungen aus dem 3. bis 1. Jahrtausend v. Chr. im Südostem der iberischen Halbinsel. I, Übersicht über die Ausgrabungen 1962-1970. Mainz am Rhein. Schüle, W. 1986. El cerro de la Virgen de la Cabeza, Orce (Granada): consideraciones sobre su marco ecológico y cultural. In Homenaje a Luis Siret (19341984). Sevilla: Conserjería de Cultura. 208-20.

194

Weaponry, statues and petroglyphs: the ideology of war in Atlantic Iron Age Iberia José Freire MA student, Department of Archaeology, University of Sheffield

Habitats as hillforts As external pressure from the Roman Empire increased, the regional groups inhabiting many of these settlements were forced to change their habits and preferences. Many of the great hilltop oppida were abandoned, for they did not fit the Graeco-Roman model of an acceptable form of civilization.

The presence of a warlike environment, or at least the possibility of armed peoples or cultures, is normally implicit within descriptions of the regions of Iberia during the Iron Age. The existence of several types of fortresses and equipment associated with warfare is commonly accepted as providing clear evidence that war was a reality for societies of this area during this period. Although a state of constant war could not have been the prevailing pattern, we have to understand the presence of defensive structures as an indicator of violence and consequent survival tactics. Many of the regions of Iberia can be defined as having been chiefdoms since at least the final moments of the Atlantic Bronze Age. Several regions have examples of a distinctive type of defensive site - the walled castrum – situated at the tops of hills (numbering over 1,000 in the northern area; Fig. 1; Silva 2000); these are interpreted as centres of power and residences of the leading elites. After the decline of some of these hillforts during the period known as Iron Age II (sixth-first centuries BC), the settlements that were built subsequently in the vicinity of rivers and on plains were also heavily defended, by impressive stone walls around their perimeters. The presence of monumental architecture need not necessarily imply ‘defensive structure’ – there may always be more than one explanation for such monumentality - yet the impressive sets of ramparts of great thickness seen at such sites, together with the building of towers around their main entrances, may well be interpreted as a clear expression of warlike behaviour or purposes. The layout of the settlements and their increasing complexity over time demonstrate that buildings with such potential as these must in one way or another be linked to a society experiencing institutionalized warfare. Many of the so-called ethnic migrations or incursions (Ligurian, ‘Celtic’, Phoenician, Greek, Celtiberian, Carthaginian and Roman) that occurred alongside the development of iron metallurgy during this period were linked either definitely (or, in some instances, possibly) to a social structure involving dominant elites. As these types of defended settlements evolved over time, so too did the technology and structural complexity. The sheer effort of maintaining these types of settlements through the centuries is a clear indication of their use and necessity.

Figure 1. Distribution of oppida in northwestern Portugal, showing major hillforts (squares) and secondary hillforts and settlements (dots). A: the port of trade (Cale hillfort, Porto); B: mining facilities (Castro de Vandoma, Baltar); C: political and agricultural capital (Citânia de Sanfins, Paços de Ferreira) (by permission of A.C. Silva [2000])

195

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY representation, mainly on artefacts such as statues or in petroglyphs and symbolic decoration.

Artefacts as weapons Turning to the artefacts which have been recovered from Iron Age sites in the region, it is clear that, alongside tools and other very different types of implement, weaponry was indeed an important component of the material culture of these past societies. Weapons are usually found mingled with many kinds of prestige artefact, such as torcs, belts, arm-rings, fibulae, Greek vases, horse-gear and chariot fittings, etc. Women’s graves are defined by the ‘normal’ feminine items, whereas male graves also contain warfare-related artefacts and prestige goods that were in fashion at the time: swords, daggers, shields and spears (Table 1). Such a grave good assemblage is almost a ‘classic’ indicator of a renowned European Iron Age chieftain. Artefact type

Composition

Period

Sword: antennae-hilted

iron + silver

V – IV BC

Sword: falcata (curved blade)

iron + silver

V – IV BC

Sword: La Tène Dagger: antennae-hilted

iron iron

IV – III BC V – II BC

Spear: long-headed

iron

V – IV BC

Shield: Celtiberian

bronze + wood

V – II BC

Shield: La Tène

bronze + wood

IV – III BC

Helmets: Montefortino + Lanhoso

bronze

I BC

Figure 2. Symbolically bent weapons from Nécropole do Olival do Senhor dos Mártires, Santa Maria do Castelo, Alcácer do Sal, Portugal: (L to R) a spear-head, a dagger and an antennaehilted sword (from Museu Nacional de Arqueologia 1996).

Table 1. Iberian Iron Age weaponry

Common weapons such as antennae-hilted swords, daggers and falcatas (curved swords medium to long in size) were sometimes turned into prestigious items during the manufacturing process. Such is the case for weapons from the well-known group of Iron Age II elite graves of the Necrópole do Olival do Senhor dos Mártires (Fig. 2; Álcacer do Sal; Silva and Gomes 1994). These particular weapons were made of iron and silver, with an additional layer worked with beautiful La Tène motifs. These graves are quite typical of the high-status graves thought to belong to the wealthiest and most powerful elites of the late Iron Age period (Vasconcelos 1913; Silva and Gomes 1994). Such artefacts are fairly similar to those found in more northern and central areas of Europe, for example those from the early La Tène tomb of SommeBionne (La Marne, France; Stead in Moscati et al. 1991: 179; Furger-Gunti in Moscati et al. 1991: 356-9).

Figure 3. Iron Age II Galaic warrior stature (1.73m x 0.54m) (from Museu Nacional de Arqueologia 1996).

Statues as symbols Dating mainly to the period of Iron Age II (sixth-first centuries BC), statues depicting warriors (Fig. 3) can be seen as part of a carefully composed symbolic system which clearly implies the idea of a community’s warlike

A common pattern of symbolism linked to war or warlike components is a clear expression of an ideology of warfare. Such components of warlike behaviour can be understood through the study of many kinds of 196

JOSÉ FREIRE: WEAPONRY, STATUES AND PETROGLYPHS: THE IDEOLOGY OF WAR IN ATLANTIC IRON AGE IBERIA identity or an individual’s status as a chieftain. Once thought to have been a component of funerary ritual, these intricate statues were mostly placed at the main entrances of very important hillforts (Silva and Gomes 1994; Branco Freire 1998).

of philology. A true melting pot of cultures of alien origin, the eastern regions of Iberia had early contacts with the civilizations of the Mediterranean. Alongside such links to the Classical world, Atlantic and continental influences were always present.

The statues’ symbolism displays how the elements of warrior or chiefly identity were envisaged. Usually representing a powerful leader standing in a commanding posture, the statues also show some of the essential items typical of warfare, such as swords, daggers, shields and helmets (Fig. 4). Also present on the statues are carvings of symbolic or pan-religious motifs, including features such as triskeles, solar-wheels, etc.

Through the use of toponymy and anthroponymy in conjunction with the archaeological evidence, we can uncover the existence of an overarching ideology of warfare within these Iron Age societies. Particularly significant are instances of Celtic names attached to settlements and chieftains; these are usually composed of a mixture of suffixes and prefixes that are clearly indicators of a society addicted to power, made up of communities which, in one way or another, are commonly inter-reacting within a warlike environment. These philological elements, including signs of Celtic grammar and vocabulary, are found well inside Iberia’s Indo-European regions (Untermann 1987). The most common element to be found in names of hillforts is an old Celtic word briga (‘elevated fortress’). This name-element usually meant that the settlement’s buildings were walled or in some way heavily protected (Untermann 1987). Examples of this can be seen in the names of many of the region’s most important castri: Cetóbriga, Lacobriga, Miróbriga, Conimbriga, Calambriga, Nemetobriga, Segóbriga, Brigantia, etc. Other Celtic elements also occur in the names of such hillforts as Budens (old Celtic ‘victory’), Catraleuco (old Irish Cat–, ‘battle’), Caladunum (derived from –dunos, ‘castle’) and Uxbonna (from –bonna, ‘fortified palace’). The existence of these Celtic forms does not exclude the presence of other ethnic influences; they are merely more representative of the subject in question. Nevertheless, the Celtic-speaking peoples are well known across Europe as having been a martial people, an identity which is attested to both by the archaeological evidence and by sources in mythology and later texts. A simple example of the relationship between war and chiefdoms can be found in a traditional old Irish saying ‘Ni gebtar cad sin rig!’ (Can’t win a battle without a King!).

Figure 4. Iron Age II Lusitanian warrior statues (height 1.70m) (Museo Martins Sarmento, Guimarães, Portugal).

Such a combination of elements shows the clear outline of a society which to some extent tended to institutionalize war. It is unlikely, however, that the system of dispute settlement created an environment of unremitting war: it would surely have been more normal to live in conditions of ‘armed peace’ (Senna-Martinez 1998), given the deleterious effects of war on the economy of any constantly belligerent area. Such a state of ‘armed peace’ goes some way towards explaining the motivations for building particular types of settlement, each inhabited by a concentration of people identified closely with their settlement.

Petroglyphs versus sacred acts as warfare imagery In exactly the same area of the Vila Nova de Foz Côa valley where Palaeolithic rock carvings were found, a new set of petroglyphs was discovered which surely belong to the Late Iron Age, mainly depicting a great ensemble of warriors, horses, weapons and battles (Zilhão 1997). These petroglyphs are of great importance and very enlightening. Their warlike atmosphere is quite similar to that of other rock art from the Celtiberian nucleus. The full paraphernalia of warfare can be perceived within the battle sketches: swords, spears, shields, helmets and armour are all present.

Language as ideology Further evidence for the existence of warfare, settlement identity and chieftain-led society can be found in the field 197

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY populations had been settled in these areas since at least the later periods of the era of bronze metallurgy.

Symbolism within the imagery can be found in particular segments that show compositions of ravens, snakes and other such features which can be interpreted as a pictorial form of past beliefs. This rock art contains no peaceful representations of, for example, farming or the redistribution of goods, as can be seen in Neolithic or, more commonly, medieval art. The representation of power relations and their justification was then more important than in earlier or later periods.

Interest in the acquisition of land and raw materials, particularly the special motivations of regional control and commerce, developed early. During the trading conditions of the Atlantic Bronze Age, land had already started to rise in importance accordingly to its particular geographic situation (see papers by Gomes; Soares and Tavares Silva; Parreira; Vilaça; Cardoso; Kunst; SennaMartinez and Bettencourt in Secretario de Estado da Cultura et al. 1995). Moving forward in time, during the second stage of the European Iron Age, under conditions of greater social complexity, a larger number of controlled territories appeared with a clear connection to silver and gold mines or other mineral resources (Silva 1986; Collis 1997; Cunliffe 1997; 1998; Arnold and Gibson 1999; Kristiansen 2000).

Another source of cultural information is the bending or breaking of artefacts (particularly weapons; Fig. 2) which is usually considered to be a symbolic or even sacred act, varying from region to region or from culture to culture. Throughout Europe, people of different cultures bent or broke particular types of objects before placing them in graves or depositing them in significant watery environments. Artefacts were even manufactured specifically for the occasion, hence the general opinion that this was a symbolic practice (Bruneaux and Almagro-Gorbea in Moscati et al. 1991: 364-5, 386-405).

For this era of increasing complexity in societies all over Europe, it is impossible to separate the conditions for the creation of wealth (see Brumfield and Earle 1987) from local settlement structure and social organization. During the Roman domination of the subsequent period, all these conditions tended to merge within the Empire’s priorities, but the terms had been established long before under the potent influence of the commercial institutions of the Mediterranean.

This kind of annihilation of artefacts tends to be understood as a pan-religious activity, through which the warlike cultures that practised it could make contact with the gods or the sacred forces of nature. The destruction of weapons was accepted as an offering or even a trade between heaven and earth. This practice does not contradict the evidence drawn from rock art; it appears to be a complementary activity.

Warfare and violence was then institutionalized, not only for the obvious reasons of conquest of land but also for upholding all the powers of the local elites. Chieftainship was a difficult but privileged way of maintaining a recent or a hereditary monarchic status. After the early collapse of chiefdoms, several peoples instituted an important system of alliances throughout the land, thus intensifying all the social pressures and conditions likely to lead to warfare, which in turn transformed oppida and their hinterlands into synonyms of elite glory and preservation.

Institutionalized warfare: conclusions Given the presence of oppida, weaponry, petroglyphs and the traces of ancient languages, western Iberia seems to present a convincing example of the existence of a warlike environment. Although not a constant issue, battles and other kinds of violent confrontation are sure to have taken place. In a simple anthropological view, if the inhabitants of the region had the means with which to fight, they surely made use of them.

Table 2 sets out the possible evolutionary paths for, and differences between, chiefdoms and chieftainships in order to clarify the sheer complexity of these past societies. Several important societies of the region were constructed according to a pattern of institutionalized warfare. Turning to the number of oppida, I infer that the base of society was made up of a certain number of chieftainships which controlled a smaller number of hillforts. A larger number of hillforts (and also families and warriors) was then created, beyond the number indicated by the ‘standard’ rules. A superior number of oppida (three or four or more) was a necessary factor in the growth of chiefdoms. This brief analysis of social structure helps us to understand some of the reasons why warfare was institutionalized by those past societies.

As reported clearly by Classical authors, many of the Iberian peoples managed to fight back against formidable alien invaders, at least for some time. From the onset of the Punic Wars, both Carthaginians and Romans fought throughout the entire peninsula and, during this period of war, the indigenous peoples and confederations started to take sides to pursue their own interests and motivations. The existence of an ideology of war within some of the elite groups can be demonstrated but this situation does not characterize all the peoples of Iberia. Generally speaking, accordingly to the Classical texts, there were only two possible means by which large number of individuals could move to another region (as happened, for example, around the fifth century BC when the Turduli Veteres or the Celtici migrated from the south to the northern territories). These enterprises always relied either on force or on an alliance, since other indigenous

This was the period that saw the creation of major groups in social terms, a time of the appearance of confederations which sometimes existed only for a brief phase of hostilities. Their names were often to be 198

JOSÉ FREIRE: WEAPONRY, STATUES AND PETROGLYPHS: THE IDEOLOGY OF WAR IN ATLANTIC IRON AGE IBERIA

Confederation Chiefdom

Territory

vs.

Chieftain

Warfare

Territory

People’s organization

War conditions

vs.

Redistribution

Group organization

Multi-ethnic

Standard society

Institutionalized war

1-2 ethnicities

Institutionalized behaviour

(3-4 oppida >)

(1-2-3 oppida) Table 2. Motivations for social behaviour

Cunliffe, B. 1998. Prehistoric Europe. Oxford: Oxford University Press. Kristiansen, K. 2000 [1998]. Europe Before History. Cambridge: Cambridge University Press. Moscati, S., Frey, O.H., Kruta, V., Raftery, B. and Szabó, M. (eds) Les Celtes (I Celti). New York & Milan: Grupo Editoriale Fabbri/Bompiani. Museu Nacional de Arqueologia. 1996. De Ulises a Viriato: o primerio milénio a.C. Lisboa: Museu Nacional de Arqueologia. Secretaria de Estado da Cultura, Instituto Português de Museus and Museu Nacional de Arqueologia. 1995. A Idade do Bronze em Portugal: discursos de poder. Lisboa: Secretaria de Estado da Cultura et al. Senna-Martinez, J.C. 1998. The Central Portugal Late Bronze Age: contribution to a study on regional ethnogenesis. Lisboa: APEQ, Estudios do Quaternário 2. Silva, A.C.F. and Gomes, M.V. 1994. Proto-História de Portugal. Universidade Aberta. Untermann, J. 1987. Lusitanisch, Keltiberisch, Keltisch. In J. Gorrochategui et al. (eds) Studia Palaeohispanica: actas del IV coloquio sobre lenguas y culturas paleohispánicas. Vitoria: Veleia 2/3. 57-76.

repeated by their future enemies: Lusitanians, Gallaecci, Celtiberians, etc. The stronger elites succeeded in pursuing all possible means of conserving or increasing their social importance. An alliance or a war was simply a means to an end. That end was mainly a desire for regional power and control. Bibliography Arnold, B. and Gibson, D.B. 1999. Celtic Chiefdom, Celtic State. Cambridge: Cambridge University Press. Branco Freire, J. 1998. La toponymie celtique de la proto-histoire du PortvGal – un rapport interdisciplinaire: archaéologique, épigraphique, géographique, historique et philologique. Unpublished DEA thesis, Université de Rennes 2. Brumfield, E.A. and Earle, T.K. 1987. Specialization, Exchange and Complex Societies. Cambridge: Cambridge University Press. Collis, J.R. 1997. The European Iron Age. London: Routledge. Cunliffe, B. 1997. The Ancient Celts. Oxford: Oxford University Press.

199

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Vasconcelos, J.L. 1897–1913. Religiões da Lusitânia. Vols I - III. Imprensa Nacional.

Zilhão, J. (ed.) 1997. Arte Rupestre e Pré-História do Vale do Côa. Lisboa: Ministerio da Cultura.

200

A palaeodemographic investigation of warfare in prehistory Neil A. Bishop and Christopher J. Knüsel Department of Archaeological Sciences, University of Bradford

Limited and ambiguous archaeological evidence makes it difficult to investigate the occurrence and scale of prehistoric warfare. As an alternative approach, this study uses palaeodemographic estimators in order to examine prehistoric warfare. Though no palaeodemographic study can provide definitive results, this study offers an interesting perspective on early warfare. It seems that, in agreement with some anthropological hypotheses (Otterbein 2000; Keeley 1996), early warfare was extremely dangerous and hardly ‘ritualized’, if this term is taken to mean non-fatal. Although casualties would appear to have been few in number, the dead and wounded would have made up a large proportion of the affected community. There is strong evidence to suggest that the distinctions drawn in historical warfare between combatants and noncombatants were absent in prehistory. This may be due to differences in individuals’ roles in society, differing moral standards, or deliberate policies of genocide. It is also interesting to note that the raiding strategies characteristic of early warfare, rather than showing a lack of sophistication in military matters, appear to have been an optimal form of attack, given the combatants and the terrain involved.

warfare may be necessary as part of a rite of passage (Van Gennep 1960). Given the importance attached to warfare, it would be expected that its study would be high on the archaeological agenda. However, this is not necessarily the case. Most of the literature on prehistoric warfare is based on socio-cultural studies. The investigation of warfare through the archaeological record is difficult, given the limited amount of archaeological evidence left by warfare (Vandkilde 2003; Vencl 1984a). A further complication is that perceived evidence of warfare can be interpreted in a number of ways. For example, artefacts interpreted as weapons can have alternative functions, such as for use in hunting (Childe 1941), or as indicators of social status (see, for example, Härke 1992). Similar criticisms can be levelled at other traditional archaeological indicators of warfare (for further discussion of this subject see Vencl 1983; 1984a; Childe 1941; and Knüsel this volume). Possibly the clearest evidence of violence in past populations is offered by peri-mortem trauma on skeletal remains. The problem here, however, is that evidence of violence does not necessarily equal evidence of warfare. There are many time periods and geographical locations for which such evidence is absent. In such cases, many authors writing on prehistoric warfare have cited less obvious osteological evidence for warfare. Many of these accounts have been published for their palaeopathological (Schutkowski et al. 1996; Bloom and Smith 1991) or artefactual (Siiriäinen 1977) interest. Such cases may represent rituals (Vencl 1984b), punishments (Vencl 1984b; Dent 1983), murder (Dent 1983) or accidents (Cybulski 1999). Even sites containing the remains of a large number of individuals may not provide clear evidence. A number of large mass burials have been interpreted as sacrificial sites (Verano 1986; Jamieson 1983; Lothrop 1954), mass executions of social outcasts (Darling 1998) or some unknown form of ritual behaviour (Browne et al. 1993). Even where studies of skeletal trauma are more wide-ranging in their sample choice (e.g. Standen and Arriaza 2000) and conclude that violence is endemic in a population, it is again difficult to determine the nature of the violence owing to the loss of much of the contextual information.

The results of this study support the arguments of Vandkilde (2003) and others who argue that warfare was an important form of social interaction in prehistory and, as such, requires greater consideration in archaeological explanations of culture change. Warfare in prehistory It is clear that warfare1 has formed a central social interaction in history. Given its more recent apparent importance, it may be expected that warfare should exhibit similar importance in prehistory (Vandkilde 2003). To a large extent this statement appears to be true. Warfare has been associated with the formation of early chiefdoms (Earle 1997) and state-level societies (Kosse 1994; Carneiro 1970). It has been suggested that successful participation in warfare can enhance the material conditions of an individual’s life, so improving their inclusive fitness and benefiting both the individual and the group (Durham 1976). Warfare may play an important role in the regulation of population levels (Engelbrecht 1987; Cowgill 1975). Success in war can be a way for leaders to increase their status and power (Monks 1997), and a certain level of proficiency in

Another problem is that it may be expected that prehistoric societies were more violent than those of the modern Western world, given such factors as a lack of an organized and independent judicial system. Indeed, Earle (1997: 105) suggests that intimidation and violence were essential factors in the maintenance of hierarchies in chiefdoms and other early stratified societies. Also, the

1

Warfare is defined for the purposes of this work as ‘any prolonged conflict between two [or more] rival political groups by force of arms’ (Montgomery 1968: 11). The term ‘political groups’ in this sense is used only generally, and could be replaced by ‘ethnic’, or other types of group.

201

WARFARE, VIOLENCE AND SLAVERY IN PREHISTORY Deviations from the historical patterns may suggest fundamental differences in the nature of warfare in prehistoric times.

prehistoric way of life may be expected to have produced an increased number of trauma-related deaths (owing to a higher number of accidents and poor health care), which may mimic violent death.

Methods Work by Angel (1974) suggests that there is indeed a relationship between social complexity and the prevalence of trauma. While some types of trauma (e.g. blade wounds or parry fractures) can be attributed to violence, it is not always possible to distinguish between accidental trauma and trauma resulting from deliberate violence (Ortner 2003; Larsen 1997). For example, a depressed fracture can result from violence or another form of heavy impact to the head. Furthermore, violence can leave little in the way of osteological evidence (Kennedy 1994). One question not easily addressed by such analyses relates to the presence of warfare (intergroup conflict) as opposed to that of domestic and similar social violence (Etxeberria and Vegas 1988). Does a survey of traumatic injuries manifested in skeletal material provide evidence for warfare, or for a violent society? Ultimately, this is a question of scale and intensity. One method of addressing this question is to undertake a palaeodemographic study of populations bearing evidence of weapon-related trauma within their archaeological context.

Sample validation It would be unreasonable to expect this study to be completely free of the problems that are so common in palaeodemography (Margerison 1997). It is necessary to be aware of the limitations of palaeodemography (Guy et al. 1997), to have conservative expectations (Margerison 1997) and to consider the results as tendencies rather than as certainties (Fóthi and Fóthi 1996). With regard to this study, the main problems are accurate ageing of adults, sample representivity and juvenile sex assessment. The accurate ageing of adult skeletal material has been a central problem in palaeodemography for many years (see Bocquet-Appel and Bacro 1997; Bocquet-Appel and Masset 1982; 1996). Despite many attempts to rectify the situation, this problem still persists, although it can be limited through the use of appropriate palaeodemographic methods, such as broader age categories or Bayesian statistical approaches. While the sex determination of adults is usually possible, the sex determination of sub-adults is much more difficult, despite attempts by a number of authors to rectify this problem (e.g. Schutkowski 1993; see also Piontek 2001). As a consequence, sub-adults are not divided by sex in this study. Where data for this paper have been derived from a study in which such sex determination had been attempted, the number of male and female subadults have been treated here as a single category.

The palaeodemography of war Ferguson (1997) notes that demographic patterns of warfare are rarely reported, although their increased use in investigations of warfare has been advocated (Lambert 2002; Kennedy 1994). The analysis of demographic patterns allows an insight into the ways in which war was waged and the ways that it affected human populations. Recent work (Margerison and Knüsel 2002; Paine 2000; Keckler 1997) suggests that it is possible to detect catastrophic mortality events from palaeodemographic data. Other authors (Bocquet-Appel and Arsuaga 1999; Bouville 1980) have used palaeodemographic evidence to suggest that some form of catastrophic event occurred in prehistoric times. Warfare is one of the phenomena that can produce a catastrophic palaeodemographic profile.

As it was not possible for the authors to examine the skeletal material directly, only well-analysed sites were included in the present study. Whilst it is acknowledged that a number of older sites have been assessed to standards which, though suitable for the time, are not on a par with the standards in use today, this problem is largely unavoidable in synthetic treatments such as this. Nevertheless efforts were made to include only sites which had been subjected to a multi-factorial analysis, something deemed essential for palaeodemographic studies (Piontek 2001; Acsádi and Nemeskéri 1970).

A small number of studies have provided direct demographic evidence of warfare (Bouville 1980; Owsley et al. 1977) but these studies are rare. It is noticeable that there is little in the way of comparisons drawn between sites and surprisingly little historical data has been used to make comparisons with prehistoric sites, a method advocated by Kennedy (1994).

Sample representivity presents another large problem. For this work many sites are not expected to be fully representative, for reasons which may be taphonomic (Piontek 2001), cultural (Meindl and Russell 1998), from excavator bias (Ubelaker 1999) or due to later disturbance of graves (Sellevold in Ammerman 1989). However, it is essential that the control sites are representative. The most commonly noted discrepancy is the under-representation of children (Piontek 2001; Meindl and Russell 1998; Bietti Sestieri et al. 1997; Guy

It seems clear that greater use must be made of historical graves that are known to contain victims of warfare. These sites can allow the construction of demographic profiles, which can then be compared with those from prehistoric and protohistoric sites. If warfare was present in prehistory, it may be expected that the demographic patterns of the historic and prehistoric sites would be similar if the same social events created these contexts. 202

NEIL A. BISHOP AND CHRISTOPHER J. KNÜSEL: A PALAEODEMOGRAPHIC INVESTIGATION OF WARFARE IN PREHISTORY suggested that this will lower the value of the data, Ubelaker (1999) suggests that more error is introduced by the exclusion of such individuals than their inclusion, and this method is deemed to be of more analytical value than including an ‘unknown’ age category.

et al. 1997; Acsádi and Nemeskèri 1970). Data from historical demography (Masset 1973; Hollingsworth 1969) and anthropological studies (Ray and Roth 1984) suggest that, in the past, high infant mortality was normal in pre-industrial societies, but this is not always apparent from skeletal samples. There is debate as to whether this is due to taphonomic (Guy et al. 1997) or cultural reasons (Meindl and Russell 1998), although it is probably a combination of both (Bietti Sestieri et al. 1997).

Given the problems of assigning ages to adult individuals and of infant under-representation in skeletal samples, it is desirable to use a method of investigation which avoids these problems as far as possible. One such method is the Juvenility Index (Bocquet-Appel and Masset 1982; Bocquet and Masset 1977). This is represented by the formula D5-14/D20-Z where D equals the number of people dying at age x, and Z is the maximum age achieved before death. This formula is unaffected by the underrepresentation of young infants, and it also has the advantage of being unaffected by the erroneous classification of older individuals.

The sites The sites are divided into three groups: 1. 2. 3.

the attritional cemeteries; warfare-related historical sites (which form the control groups); and the prehistoric sites.

The attritional cemeteries are used for comparative purposes to show the expected mortality in pre-Jennerian populations (prior to the advent of vaccines and public health measures associated with modern times). They are also used to estimate the age distributions of living populations. The warfare-related sites are all taken from the historical period and are historically documented as deriving from warfare. Most are mass graves, but two shipwrecks and some military cemeteries are included. These provide insights into the expected mortality pattern in warfare-related situations.

This formula cannot be easily accommodated by the age intervals described above. Rather than introduce inaccuracies into the data by further division of the figures, the formula has been adjusted to D7-16.99/D17-Z. However, there are no reference data for such a formula. When developing the Juvenility Index, Bocquet and Masset (1977) used a number of cemetery sites to create their own reference data. As such the reference data for this study is created via application of the formula to 20 pre-Jennerian cemetery sites. An indiscriminate catastrophe (such as a massacre) should produce a skeletal sample that accurately reflects the living population structure (Bocquet-Appel and Arsuaga 1999). Consequently, it was necessary to produce values of D7-16.99/D17-Z which represent a living population for comparative purposes. The living population structure (termed Cx) can be produced from a life-table analysis (Bocquet-Appel and Arsuaga 1999) although, to do this, it is necessary to assume a stable and stationary population (Mensforth 1990).

The prehistoric sites selected for analysis are characterized by a high prevalence of peri-mortem trauma, and the vast majority have been cited as evidence for some form of violence or warfare. They (and some of the protohistoric sites) also tend to be mass graves, but some of these sites are represented by assemblages of scattered, unburied human remains (Table 1). Palaeodemographic methods The source data for this study come from a wide variety of published reports. Unfortunately, different authors divide data into different age classes depending upon a wide range of factors. This has led to the suggestion that it is ‘useless’ to make demographic comparisons of populations studied by different authors (Masset 1976). Most palaeodemographic studies use 10-year intervals at least for adults (e.g. Margerison and Knüsel 2002; Blakely 1971) - although others (e.g. Alesan et al. 1999; Fóthi and Fóthi 1996) favour 5-year intervals, which palaeodemography ideally requires (Jackes 1992). Given the wide variety of age classes in use, it is impossible to classify individuals into narrow age categories. As a result individuals were classified as aged >17, 7 to 16.99 and