The Sandby Borg Massacre: Interpersonal Violence and the Demography of the Dead

European Journal of Archaeology 22 (2) 2019, 210–231 This is an Open Access article, distributed under the terms of the Creative Commons Attribution- NonCommercial-ShareAlike licence (http://creativecommons.org/licenses/by-nc-sa/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the same Creative Commons licence is included and the original work is properly cited. The written permission of Cambridge University Press must be obtained for commercial re-use.

1 2 CLARA ALFSDOTTER AND ANNA KJELLSTRÖM 1School of Cultural Studies, Linnaeus University, Kalmar, Sweden 2Department of Archaeology and Classical Studies, University of Stockholm, Sweden

During excavations of the Iron Age ringfort of Sandby borg (AD 400–550), the remains of twenty-six unburied bodies were encountered inside and outside the buildings. The skeletons and the archaeological record indicate that after the individuals had died the ringfort was deserted. An osteological investigation and trauma analysis were conducted according to standard anthropological protocols. The osteological analysis identified only men, but individuals of all ages were represented. Eight individuals (31 per cent) showed evidence of perimortem trauma that was sharp, blunt, and penetrating, consistent with interpersonal violence. The location of the bodies and the trauma pattern appear to indicate a massacre rather than a battle. The ‘efficient trauma’ distribution (i.e. minimal but effective violence), the fact that the bodies were not manipulated, combined with the archaeological context, suggest that the perpetrators were numerous and that the assault was carried out effectively. The contemporary sociopolitical situation was seemingly turbulent and the suggested motive behind the massacre was to gain power and control.

Keywords: violence, trauma, Migration period, Sandby borg, Iron Age, unburied

INTRODUCTION mainland in the Baltic Sea. The remains of fifteen Iron Age ringforts (i.e. fortified vil- The timespan between AD 400 and 550, lages) are still visible on Öland. These ring- the so-called Migration period when the forts might be the result of a general Roman empire was declining, is considered change in settlement, as many settlements a period of social instability and political were abandoned during the Migration turmoil in Europe. The regions north of period (Stenberger, 1933; Näsman, 1988). the Rhine and Danube were populated by Initial studies have shown that a massacre various tribes, the tribes of Scandinavia was carried out in the ringfort of Sandby each being ruled by a few elites as a result borg, and that the victims were not buried of growing social differentiation. While the (Victor, 2015; Alfsdotter et al., 2018) written sources are few, archaeological (Figure 1). The word massacre is used to records document widespread contact with describe a collective act of intentionally the continent (Näsman, 1984; Lund killing a group of people unprepared for Hansen, 1987; Fischer, 2005). This era is battle. Here, we examine the violent event characterized by finds of Roman solidi and in general, and its victims in particular. hoards of gold, not least on the island of Though it is clear that many people died Öland, located east of the Swedish during a single event at Sandby borg, the

© European Association of Archaeologists 2018 doi:10.1017/eaa.2018.55 Manuscript received 6 March 2018, accepted 3 September 2018, revised 5 July 2018 Downloaded from https://www.cambridge.org/core. IP address: 170.106.33.42, on 27 Sep 2021 at 02:04:47, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2018.55 Alfsdotter and Kjellström – The Sandby Borg Massacre 211

Figure 1. Map showing the location of Sandby borg, Öland, Sweden (basemap © OpenStreetMap contributors, modified by Helena Victor).

history behind the massacre is not known. culture-specific, the results are further con- We present the demography of the dead, textualized in an attempt to discuss the the trauma patterns, the type of trauma, and event, the perpetrators, and the possible the positions of the dead. Since violence is motives that prompted the violence.

Downloaded from https://www.cambridge.org/core. IP address: 170.106.33.42, on 27 Sep 2021 at 02:04:47, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2018.55 212 European Journal of Archaeology 22 (2) 2019

The Sandby borg ringfort to have housed 200–300 inhabitants (Victor, 2015). A mix of prestigious jewellery and The vast majority of artefacts found within everyday items was discovered in the the ringfort indicate that Sandby borg was dwellings (Victor, 2015; Gunnarsson et al., occupied between AD 400 and 550 (Victor, 2016). The large quantity of valuable arte- 2015; Alfsdotter et al., 2018). Since 2011, facts, in association with the discovery of small-scale annual excavations have been articulated human and animal remains, gives carried out by Kalmar County Museum. the impression of a ‘moment frozen in time’ Three houses have been fully excavated, and (Alfsdotter et al., 2018). Articulated and six partially. Additionally, parts of the street partially scattered skeletons of humans and surrounding the central block have been animals (sheep, dogs, and pigs) were found subject to excavation (Figure 2). The exca- within the houses. Macro-botanical and vated area amounts to a modest 9 per cent taphonomic analyses indicate that the ring- of the ringfort interior, but the results are fort was not revisited after the violent attack nevertheless informative in terms of the (Heimdahl, 2014, 2016;seediscussionin history of the ringfort. Future excavations Alfsdotter et al., 2018;forataphonomic will clarify this further. The fort is believed discussion, see Alfsdotter & Kjellström, in

Figure 2. The internal structures of Sandby borg in accordance with geophysical prospection (Viberg 2012). Identified plausible house structures have been numbered 1–53. Marked areas have been subject to excavation. (Figure by Helena Victor.)

preparation). The commingled skeletal and necrosis, or an abnormal elemental remains are predominant in an outdoor shape (i.e. inadequate fusion or pseudoar- environment. The archaeozoological analysis throsis). When a fracture does not show has shown that the event is likely to have any signs of healing in fresh bone, it is clas- taken place between late spring and early sified as a perimortem trauma, which indi- autumn (Alfsdotter in Gunnarsson et al., cates that it occurred around the time of 2016). Recent isotopic analyses indicate that death (Knüsel, 2005; Ubelaker, 2015). at least two of the individuals (ID1 and However, since the perimortem interval is ID2) were probably locals (Wilhelmson, related to the moisture retention of the 2017: 143). For a detailed description of bone, the period can be prolonged in comparative materials, contemporary certain environments (Kjellström & Ham- weapons, and tactics, see the online supple- ilton, 2014). Characteristics of perimortem ment (Supplementary materials 1). trauma include smooth, obtuse fracture angles with the same coloration as the rest of the bone, sometimes in combination MATERIAL AND METHODS with a plastic response and distinctive fracture types described in the medical litera- Minimum number of individuals ture (Ubelaker, 2015). Perimortem trauma (MNI), sex, and age determinations was further categorized according to type: sharp force trauma (SFT), puncture trauma Standard osteological techniques were (PT), and blunt force trauma (BFT). SFT applied for sex and age determinations of are injuries (incisions, stab wounds, or the individuals. The methods are listed in chop wounds) sustained from a sharp the online supplement (Supplementary object which leaves linear marks, V-shaped materials 2). All individuals were assigned in cross section, and a smooth kerf wall to the following age groups: fetal=

Table 1. Preservation, demographic distribution, and trauma in relation to context (AT = antemortem trauma; PT = perimortem trauma; F = find number, used for elements without clear association to a unique individual). ID Majority of skeleton Age group (years) Sex Trauma Context articulated (in brackets, per cent of skeleton preserved)

1 Yes (100) Adolescent (17–19) Male PT + AT House 40 2 Yes (100) Young adult (19–25) Male PT House 40 3 No (20) Middle adult PT House 40 4 Yes (75) Young adult Male? PT + AT House 40 5 Yes (60) Young(?) adult Male House 40 6 Yes (90) Adolescent (12–15) PT House 40 7 Yes (100) Adolescent (12–15) House 40 8 No (5) Infant/Child (2–5) House 40 9 Yes (85) Old adult Male House 52 10 No (5) Infant House 52 11 No (5) Middle adult House 52 (pit 4) 12 Yes (80) Child (6–8) House 4 13 No (40) Adolescent (10–13) PT House 4 14 No (20) Old(?) adult PT House 4 15 No (75) Old adult Male House 4 16 No (20) Middle adult Street 17 No (5) Child (5–9) Street/alley 18 No (30) Young adult (20–25) Alley 19 No (20) Old adult Male Alley 20 No (5) Child (3.5–6.5) Street/alley 21 No (25) Adolescent (12–15) Street 22 No (5) Child (3.5–6.5) Street 23 No (20) Young adult (20–25) Male? PT Street 24 No (20) Middle adult Male House 3 25 No (5) Infant (1.5–3 months) House 40 26 No (5) Infant (2–4) House 3 F6671/F6141 No (1) Adult AT House 40

the preservation of different elements). three; Street: eight) (see Figures 5A and This means that it is likely that the identified B). The age distribution is varied: the injuries are only a minimum representation youngest individual (ID25) was an infant of the full extent of the original trauma. 1.5–3 months old and the two oldest individuals (ID9 and ID15) were over 50 years old. The nine adults with sufficiently RESULTS preserved morphological traits to deter- mine biological sex exhibit male character- A minimum of twenty-six individuals was istics. The sex of the children is unknown. identified (House 3: two individuals; The sample size was too limited to iden- House 4: four; House 40: nine; House 52: tify any demographic patterning on the

site; but, in the house with the most 7 × 8 mm and is located on the distal- bodies (House 40), all age groups were lateral part of the shaft. represented, from infants to old adults. ID2: the skeleton of a young adult Similarly, children, teenagers, and adults male, was lying in a prone extended pos- were also discovered in the street. ition in House 40. The cranium presented Perimortem trauma was identified in its left posterior-lateral side and was lying eight individuals (Table 2)(Figure 3 and 4): adjacent to the left hip and lower thorax ID1: the skeleton of a male adolescent, of ID1. The left upper limb was abducted, was found in a supine position in House rotated medially, and slightly flexed at the 40. The humeri were abducted (c. 30°) and elbow, with the forearm pronated. The the lower limbs flexed at the knees. SFT right upper limb was slightly abducted, was identified on the right scapula and the rotated medially, and projected posteriorly. cranium together with a circular defect, The PT identified consists of a rhombic likely to be a PT, on the right femur. The puncture wound of the central occipital, linear scapula lesion is located diagonally just above the external occipital protuber- over the acromion and extends into the ance. The injury, 7.3 mm in length, pene- corpus in posterior view, left superior to trated the outer table and diploë but not right inferior (c. 110° to the sagittal plane). the tabula interna. (In a preliminary ana- Fragmentation prevents us from determin- lysis, SFT on a rib and a tooth of ID 2 ing the maximum length. The SFT on the was suggested by Wilhelmson (2017) but left parietal runs from just below the sagit- this could not be confirmed in the present tal suture towards the squamous suture, analysis.) aligned to the frontal plane (Figure 4A). ID3: the skeleton of a middle adult The cut measures 102 mm and penetrates individual, was found disarticulated in the tabula interna. The kerf wall is primar- House 40. Adjacent to a fragmented ily visible on the posterior part in the cranium, presenting its lateral right aspect, superior section of the wound and on the a radius and a humerus were found. An anterior part in the inferior section. The articulated tibia, fibula, and a foot were smooth femoral circular defect measures encountered 1.2 metres to the north.

Table 2. Perimortem trauma identified (R = right, L = left. SFT = sharp force trauma; PT = punctur- ing trauma; BFT = blunt force trauma). ID Element Side Trauma Description

1 parietal L SFT Linear lesion (102 mm) on central section 1 scapula R SFT Linear lesion (30 mm) diagonally in acromion and corpus 1 femur R PT Circular lesion (7 × 8 mm) on lateral-distal shaft 2 occipital central PT Rhombic lesion (4.3 × 7.3 mm) in central section 3 parietal L SFT Linear lesion (50 mm kerf wall) on central section 3 frontal R SFT Linear lesion (>30 mm) on medial section 4 occipital R SFT Linear lesions (both fragments >16 mm) lateral occipital 6 parietal R BFT Circular lesion (36 × 29 mm in diameter) on central section 13 cervical vertebra corpus SFT Linear horizontal lesion (right: 10 mm + left: 22 mm) through lamina (C3–5?) 14 parietal L BFT? Horizontal convex lesion (>24 mm) on central section 23 ilium L SFT Linear lesion (19 × >19 mm) through arc composé

Figure 3. Distribution of perimortem trauma of all the victims. One lesion to the occipital bone of ID4 is not included, as the exact location could not be established (original illustration by Buikstra & Ubelaker, 1994, modifications by Alfsdotter and Kjellström).

However, it cannot be established whether perimortal secondary fracture radiates pos- the latter unit of elements belong to the teriorly. A second SFT of the medial right same individual as the cranium. The sub- part of the frontal bone was identified. In section of the house (House 40) where the frontal view, the cut runs from left super- individual was discovered contained ior to right inferior (c. 45°). The maximum several skeletons with partial postmortem length of the injury is unknown as frag- disarticulation, likely to have resulted from ments are missing bilaterally. The cut has the partial destruction of the roof follow- penetrated tabula interna. ing a fire associated with the assault ID4: the highly fragmented skeleton of (Alfsdotter & Kjellström, in preparation). a young adult, was encountered in House SFT to the centre of the left parietal was 40 in a prone extended position with the identified, in lateral view the injury radi- right upper limb extended and the left ates right superior to left inferior. The kerf upper limb flexed, with the hand under wall is primarily on the anterior side, the the left shoulder. The positioning of the posterior fracture surface is rough. A cranium could not be established. The

Figure 4. A: Sharp force trauma on the left parietal bone of individual ID2. B: Blunt force trauma on the right parietal bone of a young teenager (ID6) (photograph by Daniel Lindskog). C–D: The skeleton of a 10–13 years old adolescent (ID13) shows evidence of decapitation. Sharp force trauma is visible on a cervical vertebra.

skeleton was slightly rotated towards the blow aimed from the lower right. The right, judging from the position of the other lesion could be part of the same thorax and the pelvis. The femora were traumatic incident, but it could also be the extended. All bones inferior of femora result of a different blow. It cannot be were missing (this might be the result of a established whether the cuts penetrated disturbance in form of an earlier archaeo- the tabula interna, as the diploë is logical trench). The calvarium was frag- fragmented. mented and scattered. Two sharp force ID6: the skeleton of an adolescent, was injuries (both at least 16 mm in length) found in House 40 in a supine extended were identified on the right side of the position with the left upper limb and the occipital. Due to fragmentation, the exact lower limbs extended. The right upper location of only one of the lesions could limb was partially missing. The cranium be established. The injury radiates hori- was fragmented but presented its lateral zontally in posterior view and has probably right side. The vertebral column shows a crossed the lambdoid suture on the right postmortem dislocation between T12 and side. The angle of the cut in relation to L1. The feet were partly on top of the hip the horizontal plane is c. 20° and striae are of ID5, whose hand bones were discovered visible radiating diagonally, indicating a under the left fibula of ID6, implying that

ID6 fell after ID5. The cranium of ID6 The vertebra(e) affected is between C3 displays BFT of the right parietal bone and C5. The mandibula of the individual (Figure 4B). Just medially of the parietal was not affected. tuber, a circular injury (36 × 29 mm) with ID14: the remains of another predomin- internal bevelling was observed. At least antly disarticulated skeleton of an older indi- five secondary fractures radiate from the vidual, were found in front of the entrance point of impact. All showed perimortem of House 4. A fragment of the left parietal characteristics but only three of these ends displayed perimortal BFT. Centrally on the at the lambdoid suture. Two of the frac- element, above the squamous suture, a hori- tures continue across the lambdoid suture, zontal convex fracture radiates. The injury is indicating a postmortem origin (Berryman funnel-shaped, internal bevelling is visible, & Symes, 1998). and the fracture surface of the tabula externa ID13: consists of disarticulated bones is smooth. The preserved part of the fracture that are seemingly from the same young measures 24 mm medio-laterally. However, adolescent. They were found in front of the anterior extension of the fracture is the entrance of House 4 (Alfsdotter, in missing. From the posterior end of the frac- press). Two bones display gnawing from a ture, a postmortal fracture radiates towards rodent. SFT was noted on two vertebral the lambda. arch fragments from the cervical region of ID23: On the street, a fragment of a the vertebral column. One of the frag- left ilium from a young adult displaying ments, the right superior articular process SFT was found. The injury is located next and part of the lamina, exhibited a linear to the auricular surface, through the com- horizontal cut angled inferiorly and radiat- posite arch. The recovered cut surface ing laterally-medially (Figure 4C, 4D). measures 19 × 19 mm. In lateral view, the The inferior part of the right side of the injury radiates from right superior to left vertebrae is missing. The second arcus inferior, of the ilium (at an angle of c. 40° fragment (inferior articular process and to the sagittal plane). The bone is frac- lamina) displaying SFT is from the left tured anteriorly, hindering analysis of the side. The lesion is similar to the first frag- full extent of the trauma. No antagonist is ment, a horizontal cut, but the kerf wall present. The weapon seems to have pene- has another angle (superior rather than trated the lateral aspect of the acetabulum inferior). Since both fragments lack and cut the entire left side gluteus antagonists and striation of the cut sur- muscles. faces, it cannot be ruled out that the There are few antemortem injuries traumas are the result of the same blow. (Table 3). Healed fractures were

Table 3. Antemortem trauma identified (F = find number, used for elements without clear association to a unique individual; R = right, L = left). ID Element Side Trauma Description

1 scapula R Fracture? Separation of coracoid 1 ribs R + L Fractures Posterior corpus, >2 fractures right side and >3 left 2 metacarpal V R + L Fractures? Bilateral swelling of shafts 4 humerus R Fracture Spiral fracture of proximal shaft 4 clavicle L Pseudoarthrosis On costal tuberosity, inferior-posterior on sternal end F6671/F6141 neurocranium Fracture Circular depression (15 mm in diameter).

documented in three, or possibly four, indi- 2010). The investigation showed that eight viduals. Two individuals (ID1 and ID4, of twenty-six individuals exhibit perimor- both found in House 40) show traces of tem trauma. This means that at least 31 two injuries each. The first (ID1), an ado- per cent were subject to violent acts near lescent around 17–19 years old, showed a the time of death. In comparison, if we separation of the right coracoid of the look at only cranial blade wounds, 96 per scapula and callus formation on the poster- cent of the men from the Battle of Towton ior section of the ribs from both sides (a (England, War of the Roses, AD 1461; minimum of 2 right and 3 left), most likely Novak, 2000), 60 per cent from the Battle caused by direct trauma (i.e. through force of Good Friday at Uppsala in AD 1520 transmitted at a right angle). As for the (Kjellström, 2005), 90 per cent from medi- second individual (ID4), the right humerus eval Sandbjerget (Næstved, Denmark; exhibited a malalignment of the proximal Bennike, 1998), and 40 per cent at the humeral diaphysis and a pseudoarthrosis in Battle of Wisby in AD 1361 on Gotland the sternal end of the left clavicular diaph- (Ingelmark, 1939) were affected. However, ysis, indicating oblique or spiral fractures the comparison is impaired by a number of and indirect trauma. In addition, a neuro- factors: these comparative data are not only cranial fragment still not allocated to a of a later date, but they represent a com- skeleton (F6671/F6141, House 40) had a pletely different composition of people, healed, circular depressed fracture measur- who were often trained for battle, died in ing 15 mm in diameter. ID2, again from battle, and were subsequently buried. The House 40, shows possible trauma-induced human remains excavated at the Iron Age anomalies bilaterally on the lateral aspects hillfort of Fin Cop (Derbyshire, England) of the shafts of the fifth metacarpals. The are a better comparison in terms of date (c. projecting bone growth takes the same 300 BC), type of violence encountered, and form on both sides but is located slightly deposition of bodies. At Fin Cop, the superior on the left side. The appearance of remains of fifteen individuals (women, the bone anomalies bears some resem- adolescents, and newborn/unborn chil- blance to boxer’s fractures (i.e. a fracture of dren) were found in a ditch together with the fifth metacarpal due to the direct debris from the demolition of the hillfort impact of a clenched fist with a person or (Waddington et al., 2012). Some skeletons object). The bilateral presence makes a were articulated and others commingled. traumatic cause less likely, though not Despite the low amount of perimortem impossible. An alternative explanation injury (only one confirmed), the cause of could be a congenital malformation. death of the individuals is believed to have been interpersonal violence. This interpretation is based on the manner of interment, DISCUSSION with bodies thrown into the ditch at the time of the hillfort’s destruction. The The unhealed sharp force lesions are con- researchers suggest that the manner of sistent with interpersonal violence. death was probably soft-tissue wounds that Furthermore, the location of the two cases left no skeletal marks (Waddington et al., of blunt trauma, above the hat brim line, 2012: 226). As at Sandby borg, only a suggests that the individuals were subjected small part of the whole site has been exca- to aggressive acts rather than accidental vated and the number of individuals is falls or similar mishaps (Kremer & likely to have been much greater at both Sauvageau, 2009; Guyomarc’h et al., sites. The unburied corpses in Sandby borg

were exposed to the weather, animals, objects varied in location and direction. Of insects, and house debris (such as that the eleven perimortem lesions identified, from stone walls, wooden beams and turf five were aimed at the back of the body roofs), leaving the skeletal material frag- (scapula, ilium, femur, and two on the mented. It should be noted that most of occipital) of the victims, implying that the individuals found in Sandby borg only they were attacked from behind. The comprise a few bone units, sometimes only trauma identified in the cervical region of teeth, as the human remains recovered ID13 could perhaps also be included in from the street are commingled and have this group. The lesions on top of the been subject to vast taphonomic degrad- cranium of at least ID1 may indicate that ation (Table 1). The nature of the skeletal the individual was in a lower position in remains gives us reason to believe that relation to the perpetrator, or that the traumatic lesions inflicted on several bodies latter was on horseback (that scenario is have not been identified. Although the however very unlikely given that the skel- number of perimortem traumatic injuries eton was found indoors1). Only two of the encountered in Sandby borg is compara- lesions (on ID 3 and ID14) are consistent tively low, it is their connection with the with battle wounds that could be expected sample’s demographic aspects and body in face-to-face combat with a right- positions that enable us to attempt a recon- handed assailant (Larsen, 1999: 157). struction of the course of events and the However, there might be an under- underlying motive. representation of facial trauma identified since the preservation of the delicate facial bones was poor. Characteristic defence How? the perimortem trauma pattern injuries, such as parry fractures on the forearms or hands, or sharp incisions on By studying the positions of the unburied the palmar aspect of the finger bones, are skeletons at Sandby borg in tandem with missing (Dolinak et al., 2005: 149–50). the perimortem trauma, it is possible to Moreover, the bodies of the remaining obtain additional information about the individuals who do not have observable modus operandi of the perpetrators. perimortem lesions were also found in the In total, eleven perimortem lesions were dwellings and in the street (Figure 5). identified in eight individuals. Sharp force They were discovered in a flexed position trauma was documented in seven cases, (ID7, figure 5A) and lying extended, both blunt force was used in two cases, and prone (ID9) and supine (ID 5 and ID12, puncture wounds were caused in two cases the latter in figure 5B), and possibly (Table 2). Two individuals (ID1 and ID3) sitting (ID15). In one case, individuals suffered at least two sharp force blows. had fallen over each other (Alfsdotter Bones of the cranium were the most affected, suggesting that this was the 1The body seems not to have been arranged in a formal primary target, but postcranial wounds position but was likely left behind after the assault. The remains of nine individuals in house 40 follow no were noted in the neck (cervical vertebra pattern in terms of skeletal position or postmortem (e)), shoulder (scapula), and hip region treatment. In House 52, a skeleton (ID9) was lying (the ilium) of the individuals, with no prone across a hearth. Parts of his skeleton were charred, a testimony to the hearth having been lit as the preference as to which side of the body. man fell(?) over it never to rise again. The Sandby borg The blunt injuries were located on the top context so far indicates a rapid killing with no or limited handling of the dead bodies (the taphonomy is of the head, near the sagittal suture, further discussed in Alfsdotter & Kjellström, in whereas the wounds inflicted by sharp preparation).

Figure 5. A: The articulated skeleton of ID7, an adolescent, is one of nine individuals whose remains have been found inside House 40. B: The partially articulated skeleton of a child (ID12) excavated in House 4.

et al., 2018). An older man (ID9) was caused by an axe (Reichs, 1998), swords lying face down diagonally over the central were most likely to have been the weapon hearth of House 52. The anterior pelvic used. The penetrating lesion did not have region was partially charred, suggesting enough characteristic traits to identify the that the fireplace was still burning, or at weapon type, and could have been caused least smouldering, when the man fell by any sharp, pointed weapon (such as a (Papmehl-Dufay & Alfsdotter, 2016). spear or lance) or other object. The same Although the number of identified goes for the blunt force traumatic injuries. injuries is low, the varied distribution lacks By combining the perimortem lesions patterning, which implies that the victims with body positions, it may be possible to did not carry any protective gear (this is establish the circumstances surrounding hardly surprising given the demographic the deaths of certain individuals. The spread). This is in line with the general supine skeleton of individual ID1 had impression that the inhabitants were taken three injuries where the sharp force trauma off guard. If the trauma analysis is repre- on the scapula must have severed the sentative of a massacre, the preferred muscles of the right rotator cuff (teres weapon of the assailants was the sword. It minor, supraspinatus and infraspinatus). is not possible to rule out axes; but, as This impaired the stability and movement none of the sharp force lesions exhibit the of the shoulder and upper limb. The blow fracture pattern expected from a combin- does not seem to have penetrated the pos- ation of sharp and blunt force trauma terior thorax. The penetrating wound on

ID1’s femur could, in a similar way, have group is not yet known, but it is likely damaged muscles and tendons (musculus that the youngest victims of the massacre biceps femoris and tractus iliotibialis). Both were born on the island. Moreover, it is lesions must have led to blood loss but probable that all the victims were local need not have been lethal. The cranium, inhabitants. Only male remains have been however, shows a wound which penetrated identified among the adults for whom sex the inner table of the parietal bone, expos- could be determined. The presence of an ing the brain tissue. The sequence of the infant and young children, and the finds lesions is hypothetical, but it is likely that of female attire (Magnus, 1997, 2002), the blow to the back and thigh preceded imply that women lived within the walls. that on the head since the latter caused Within the bioarchaeological field, there severe cranial damage which could, in are examples where a biased sex distribu- itself, have resulted in the death of the tion (with no or few women) in a skeletal man. Consequently, it is possible that the collection has been interpreted as raiding man managed to survive an initial blow, for women and girls. This practice is not before being struck on the head and uncommon in attacks on small settlements falling on his back. In another house, the (e.g. Kohler & Kramer Turner, 2006; skeleton of the 10–13 year old adolescent Cameron, 2013; Martin & Harrod, 2015). (ID13) had SFT in the neck region. Although the main purpose of the attack Though the cervical vertebra(e) are was not necessarily to take prisoners, the severely fragmented, it is clear that at least kidnap of women may have been an added one horizontal blow cut through the soft benefit. They could be taken as wives, tissue and penetrated the medulla, which used as extra labour, or sold as slaves (e.g. would have damaged major arteries (pos- Martin et al., 2010). However, since only terior inferior cerebellar artery and possibly a small part (c. 9 per cent) of the ringfort anterior spinal artery). The cut(s) are likely has been excavated, the absence of female to have led to the instant death of the remains could be a coincidence, and we individual at the site where the human must be wary of over-interpreting this remains were discovered (if they were not aspect. It should also be stressed that the moved secondarily). These two case sex distribution among the children is yet studies suggest that, at one point, the to be established through DNA analysis. wounded shared space with the dead and Nevertheless, it is not uncommon that dying, illustrating a chaotic situation with women are treated differently in massacres, people in a state of shock. Perhaps ID1, and perhaps the female remains are to be who died just inside the entrance of discovered in another part of the ringfort. House 40, was wounded in the street, If we consider the taking of slaves as a after which he entered the house (for possible reason behind the attack, this is shelter?) where he then received the final complicated by the presence of children blow to his head. and teenagers, who we may assume were easy prey. Could the children have taken part in the battle? Little is known about Who? trauma and demography the socialization processes of children during the period, and it is not clear when The isotopic data (Wilhelmson, 2017) a child was considered to have become an indicate that at least two of the Sandby adult. The transition from childhood into borg victims grew up in the region. adulthood was likely to have been based Whether this is true for the rest of the on physical and mental maturity, but

could also be connected to the social status individuals in this group (ID1 and F6671) of the individual (Mejsholm, 2009: 24). show signs of healed direct fractures, and For boys, it could be associated with a the number of antemortem traumatic time when the youth was physically fit for injuries was generally modest in the weapon training. The youngest individuals Sandby borg skeletons. The prevalence is at Sandby borg with documented perimor- similar to that found in contemporary tem injuries were the decapitated 10–13 cemetery populations. For instance, c.8 year old adolescent (ID13), and a 12–15 per cent of the individuals (both men and year old (ID6) with a blunt force lesion to women) in the Migration-period cemetery the right parietal (Tables 1 and 2). The of Prague-Zliín, Czech Republic, showed sexes are not known, and it is not clear traces of healed traumatic lesions (Vargová whether these individuals were passive et al., 2016). A high frequency of ante- victims or if they put up armed resistance. mortem trauma may indicate previous Attempts to identify active engagement of battle experience. In the study of the skel- non-adults in violent encounters have been etal remains from the Battle of Towton made in prehistoric skeletal assemblages (England), for instance, at least 32 per (Fibiger, 2014). Adolescents of similar age cent of the crania exhibited healed, to those found in Sandby borg were weapon related lesions, implying experi- assessed as having probably participated in ence of previous encounters (Novak, armed conflict in early medieval Ireland 2000). Overall, the demographic compos- (Geber, 2015). Even if this were the case ition, along with the trauma pattern, with the two young individuals from shows that most individuals exhumed in Sandby borg, the younger children were Sandby borg had little chance to defend either deliberately killed (but without themselves, and that only a few are likely lesions identified) or left to die. This indi- to have had battle experience. Although cates that the killings were indiscriminate grown men were probably the primary with regard to age. Furthermore, the target group, all members in the Sandby young individuals (twelve of the twenty-six borg community seem to have been sub- individuals were younger than 15 years jected to violence, and non-combatants old) could not, if we disregard the signs of could at least be considered collateral visible trauma, have had much, or any, damage (see Tegtmeyer & Martin, 2017). battle experience. At the other end of the The deliberate killing or abandonment of age spectrum, there were three men older young children further suggests that the than 50, none of whom had identifiable intention of the attack was to annihilate antemortem or perimortem wounds. Both the inhabitants. Though it is possible that groups could be seen as ‘atypical warriors’. some of the fallen adult men belonged to The remaining group (n = 12) of young the attackers, it appears unlikely that the or adults in the category ‘middle adult’ assailants would have left any of their own may be considered able-bodied and fit for behind, as the victorious group in a battle fighting. Indeed, it is this group that exhi- normally has the opportunity to tend to bits the highest number of perimortem their dead. lesions, and it is tempting to interpret Does the trauma pattern at Sandby borg these visible weapon wounds as indicative and the identification of weapon injuries of a more active resistance to the attack. reveal anything about the assailants? It has However, the biased distribution of been suggested that Scandinavian men wounds could be due to preservation could have received military training in the factors. At any rate, only one or two Roman Empire (Nicklasson, 1997: 173;

see online supplement: Supplementary Ember, 1997). Studies of different geo- Material 1). At least in theory, formal graphical regions, dating from 10,000 BC training with weapons could lead to ‘stan- to today, have shown that there is a strong dardized’ violence, a recognizable pattern relationship between climate change and with regard to injuries sustained by the intergroup conflict (Hsiang et al., 2013). victims. However, no such injury pattern This is interesting, since tree ring data has was discernible, since the small size of the shown that the period AD 450–550 in sample and the poor preservation of Europe was a time of climatic changes the remains has affected the outcome of (Büntgen et al., 2011). It is difficult to the analysis. Given the few perimortem establish to what degree a small-scale traumatic injuries identified, the same can society like that at Sandby borg would be said about the primary choice of have been affected by fluctuations in tem- weapon, thought to be the sword. In close perature and precipitation and what spe- combat, this is the most effective weapon; cific social stress factors could have led to but, as discussed, other types of weapons, the Sandby borg massacre. A combination pointed, edged, or blunt, could also have of several factors was probably at play. been used. A change in settlement pattern can be seen in Scandinavia during the Migration period, where several settlements were Why? contextualization of trauma and abandoned or destroyed. This has been the inhabitants of Sandby borg interpreted as a socioeconomic crisis, possibly related to changing political struc- Martin and Harrod write that massacres tures in the aftermath of the fall of the (i.e. multiple killings of all age groups and Western Roman Empire (e.g. Stenberger, of both sexes) can be seen as a political 1933; Näsman, 1988; Fischer, 2005; instrument that is ‘[…] often used by per- Heather, 2006; Hedeager, 2011; see also petrators to either subjugate or eradicate a discussion in Alfsdotter et al., 2018). The group defined as having cultural character- presence of dead children at Sandby borg istics that separate them in some way from may be indicative of extraordinary circum- the aggressors’ (Martin & Harrod, 2015: stances. Non-accidental perimortem injur- 129). Since interpersonal violence is ies documented among young children seldom random, it allows the perpetrators from the fourth–fifth-century AD cemetery to gain control over land or other groups. of Lisieux in Normandy in France have The competition for women can be a con- previously been interpreted as the result of tributing factor behind conflicts in both social stress caused by political distur- prehistoric and present-day small-scale bances and perhaps raiding groups societies (Sanderson, 2014: 291–96); (Timmins et al., 2017). In another study, although groups demonstrate a great vari- in the context of the fifteenth-century ̌ ability regarding the causes of violence massacre of Cepin in Croatia, bioarchaeo- (McCall & Shields, 2008). Furthermore, logical results were used to identify the massacres are part of complex processes, causative agent behind a specific raid, and not only because they imply planning and to differentiate attempts to take prisoners the involvement of many people, but also of war from attempts to strategically because they operate at a sociopolitical and spread fear and panic in a population cultural level. Environmental stress factors (Šlaus et al., 2010). Judging from the and risk of losing resources could also be a current analysis of the Sandby borg motivating factor for violence (Ember & assault, it is unlikely that the massacre was

the result of an attempt to obtain wealth, very interested in staying and searching for since potential prisoners (e.g. children), loot, considering the quantity of jewellery luxury items, and animals were left and other luxury items found during the behind. Theoretically, the attack on excavations. The injury pattern and the Sandby borg could have been some form positions of the bodies is another indica- of retribution. In discussing violent behav- tion of a brief attack. The victims do not iour and its aftermath during the Iron Age have the appearance of having been moved in Scandinavia, Lindbom (2006: xviii) after they died, and the injuries are almost suggests that an attack on a farm (i.e. ‘efficiently’ distributed, in the sense that attacking people in their home) was an only the minimum amount of violence extreme act with dire consequences. The necessary to kill was used (for a tapho- attackers were consequently forced to kill nomic analysis, see Alfsdotter & everyone to eliminate witnesses and, thus, Kjellström, in preparation). The compara- avoid being subjected to vengeance. So far, tively few instances of perimortem frac- there is no archaeological or osteological tures (even in the rather complete evidence of feuding between Sandby borg skeletons, e.g. ID7 and ID12) suggest that and other local groups (e.g. evidence of each victim was overcome with a burning in cultural layers caused by previ- minimum of effort by the attacker. The ous fires, large collections of weapons, or injury pattern and reconstructed body healed weapon-related trauma). positions also indicate that the victims’ In skeletal series associated with con- bodies were not further brutalized or texts similar to those of Sandby borg, manipulated after death. Furthermore, the bioarchaeological analyses have revealed repeated pattern, with dead humans inside information about the extent of the vio- and outside several houses, suggests that lence. In well-known cases of attacks on the perpetrators were numerous and villagers in other periods and geographical coordinated. regions, documented for example on pre- Dutton et al. (2005) explored the historic and historic sites in the US, varying motives behind massacres and dreadful acts of violence, such as mutila- genocides during the twentieth century. tion and trophy-taking, have been docu- Although the social systems of modern mented (e.g. Owsley et al., 1977; Willey, states and small Iron Age chiefdoms differ 1990; Andrushko et al., 2005). Acts of greatly, some arguments put forward in body processing in a ritual or sacrificial this study may be pertinent to the discus- manner have also been observed in Iron sion of the motives behind the Sandby Age wetland contexts, showing that post- borg massacre. A common denominator of mortem manipulation of human bodies mass killings is that the perpetrating group was not a completely uncommon or perceives the target group as having extreme behaviour (see Supplementary obtained some form of unfair advantage. Material 1). However, the scene at Sandby Historical and political circumstances borg suggests a different situation. The affect the identification of the target attack appears to have been an expedient group, and the mass killing is legitimized and brutal affair. The villagers were seem- among the perpetrators as revenge ingly taken by surprise rather than (Dutton et al., 2005: 470). Often, mass engaged in a conventional battle. killings are the result of feared future con- Although it is impossible to say anything flict, and rapid social change is frequently about how much has been plundered, the behind the perception of a threat that perpetrators do not appear to have been must be eliminated. When indiscriminate

killing takes place (e.g. of children), the wiped out. The finds at Sandby borg actions are often justified on the basis that strengthen the perception of the the target group poses a future threat. A Migration period on Öland as a socially plausible alternative (or additional) motive turbulent time, probably involving strife can be to cause terror through the ruthless with conflicting groups of inhabitants. killing of vulnerable individuals. The The proposed mass killing at Sandby excessive violence commonly witnessed in borg can be contrasted with the image of massacres (mutilations, rape, torture, and warfare associated with finds such as the killing of defenceless victims) differs from weapon deposits in Hjortspring and the violence often seen in genocides, Nydam in Denmark (see Supplementary where the violence is ‘dispassionate’ and Material 1). The weapon offerings are ‘instrumental’ (since the violence is carried deposited in an organized and orchestrated out by those who only execute orders) form, demonstrating control and power in (Wolfgang, 1958; Dutton & Kerry, 1999). relation to the contemporary narrative and Dutton et al. (2005) argue that socio- ideology. At Sandby borg, the grim result political circumstances may be enough to and reality of violence is evident. explain instrumental violence, whereas the ‘passionate’ killing seen in many massacres is not yet well understood. If we combine SUMMING UP theories of modern mass killing with the efficient slaughter so far evidenced in Archaeological excavations have revealed Sandby borg, it is plausible to interpret the evidence of a Migration-period massacre dispassionate killing as having been caused in the ringfort of Sandby borg on Öland. by sociopolitical instability. The lack of Thus far, the remains of twenty-six indivi- traces of overkill combined with the effect- duals have been found in several houses ive killing pattern could indicate that the and on the surrounding street outside individuals who carried out the slaughter them. Perimortem sharp, blunt, and pene- were not personally involved with the trating trauma consistent with interper- victims. Possibly, the killing was ordered. sonal violence was identified in eight Credible motives behind the massacre individuals (31 per cent of the assem- were likely to have been connected to blage). In addition, the locations and skel- experienced past injustice, the perception etal positions of the remaining individuals of the Sandby borg inhabitants as a threat, imply that they were killed in the same and the will to spread fear to achieve event. The victims belonged to all age control. This hypothesis fits with the groups, from infants to individuals over 50 interpretation that the massacre in Sandby years of age. However, only men have borg was not a raid carried out for the been identified so far. The perpetrators purposes of plundering. The motives sug- were seemingly numerous and coordinated. gested point to the attackers having been a The assault was carried out expediently, neighbouring group (also discussed in judging from the efficient distribution of Victor, 2015; Alfsdotter et al., 2018). trauma, the absence of body modification, Though a ‘personal’ scenario cannot be and the fact that items of value still remain. verified based on the current analysis, and Though most injuries were located on the remains speculative, the unburied bodies skulls, a clear trauma pattern could not be appear to signal a provoking motive. They established. The identified weapon types could be seen as a final act of violence, are consistent with contemporary weapons. and even that the entire community was The prevailing sociopolitical situation is

likely to have led to social stress. It is Evidence of a Late Fifth-Century Massacre hypothesized that the decision to kill the at Sandby Borg. Antiquity, 92: 421–36. Sandby borg inhabitants was based on a https://doi.org/10.15184/aqy.2018.21 Andrushko,V.A.,Latham,K.A.S.,Grady,D.B., feeling of past injustices and a perception Pastron, A.G. & Walker, P.L. 2005. of the Sandby borg group as a threat. The Bioarchaeological Evidence for Trophy killing of children is indicative of a fear of Taking in Prehistoric Central California. retribution among the attackers. Ultim- American Journal of Physical Anthropology, – ately, the motive behind the massacre was 127: 375 84. https://doi.org/10.1002/ajpa. 20044 probably to attain regional power and Bennike, P. 1998. De faldne fra Krigergraven- control. analys af skeletterne fra Sandbjerget. Liv og Levn, 12: 14–21. Berryman, H.E. & Symes, S.A. 1998. Recognizing Gunshot and Blunt Cranial ACKNOWLEDGEMENTS Trauma Through Fracture Interpretation. In: K. Reichs, ed. Forensic Osteology. This work was supported by the Riksban- Springfield (IL): Charles C. Thomas, pp. kens jubileumsfond under Grant P15- 333–52. 0138:1, the Knowledge Foundation, Bohus- Boylston, A. 2000. Evidence for Weapon- Related Trauma in British Archaeological läns Museum, and Linnaeus University Samples. In: M. Cox & S. Mays, eds. through the Graduate School in Contract Human Osteology in Archaeology and Archaeology (GRASCA). The present Forensic Science. London: Greenwich article builds on osteological analysis con- Medical Media, pp. 357–80. ducted by Alfsdotter and on excavation Buikstra, J.E. & Ubelaker, D.H. 1994. Standards for Data Collection from Human records by Victor, Papmehl-Dufay, Gun- Skeletal Remains: Proceedings of a Seminar narson, Alfsdotter, and colleagues. We at the Field Museum of Natural History. wouldliketothanktheanonymous Fayetteville (AR): Arkansas Archaeological reviewers for helpful comments and sugges- Survey. tions. Many thanks to Tim Sandusky and Büntgen, U., Tegel, W., Nicolussi, K., McCormick, M., Frank, D., Trouet, V. Dejan Gajic for language review, and to et al. 2011. 2500 Years of European Anders Högberg for comments on an Climate Variability and Human earlier version of the manuscript. We also Susceptibility. Science, 331: 578–82. thank Daniel Lindskog and Helena Victor https://doi.org/10.1126/science.1197175 for providing photographs and illustrations. Cameron, C.M. 2013. How People Moved Among Ancient Societies: Broadening the View. American Anthropologist,115: 218–31. https://doi.org/10.1111/aman. REFERENCES 12005 Dolinak, D., Matshes, E. & Lew, E. 2005. Alfsdotter, C. in press. Humanosteologi i Forensic Pathology. Principles and Practice. Sandby borg. In: L. Papmehl-Dufay & H. Burlington (MA): Elsevier. Victor, eds. Sandby borg IX. Undersökningar Dutton, D.G. & Kerry, G. 1999. Modus 2016, Sandby sn, Mörbylånga kommun, Operandi and Personality Disorder in Öland. Report prepared for Kalmar läns Incarcerated Spousal Killers. International museum. Journal of Law and Psychiatry, 22: 287– Alfsdotter, C. & Kjellström, A. in preparation. 300. https://doi.org/10.1016/S0160-2527 A Taphonomic Interpretation of the (99)00010-2 Postmortem Fate of the Victims Following Dutton, D.G., Boyanowsky, E.O. & the Mass Killing at Sandby Borg. Bond, M.H. 2005. Extreme Mass Alfsdotter, C., Papmehl-Dufay, L. & Victor, H. Homicide: From Military Massacre to 2018. A Moment Frozen in Time: Genocide. Aggression and Violent Behavior,

10: 437–73. https://doi.org/10.1016/j.avb. Hsiang, S.M., Burke, M. & Miguel, E. 2013. 2004.06.002 Quantifying the Influence of Climate on Ember, C.R. & Ember, M. 1997. Violence in Human Conflict. Science, 341: 6151. the Ethnographic Record: Results of https://doi.org/10.1126/science.1235367 Cross-Cultural Research on War and Ingelmark, B.E. 1939. The Skeletons. In: B. Aggression. In: D.L. Martin & D.W. Thordeman, ed. Armour from the Battle of Frayer, eds. Troubled Times: Violence and Wisby 1361. Stockholm: Kungliga Vitterhets Warfare in the Past. Amsterdam: Gordon Historie och Antikvitets Akademien, pp. and Breach, pp. 1–20. 149–209. Fibiger, L. 2014. Misplaced Childhood? Kjellström, A. 2005. A Sixteenth-Century Interpersonal Violence and Children in Warrior Grave from Uppsala, Sweden: Neolithic Europe. In: C. Knüsel & M. The Battle of Good Friday. International Smith, eds. The Routledge Handbook of the Journal of Osteoarchaeology, 15: 23–50. Bioarchaeology of Human Conflict. https://doi.org/10.1002/oa.746 Abingdon: Routledge, pp.127–45. Kjellström, A. & Hamilton, M. 2014. The Fischer, S. 2005. Roman Imperialism and Runic Taphonomy of Maritime Warfare: A Literacy: The Westernization of Northern Forensic Reinterpretation of Sharp Force Europe (150–800 AD) (Aun 33). Uppsala: Trauma from the 1676 Wreck of the Royal Department of Archaeology and Ancient Swedish Warship Kronan. In: D.L. Martin History, Uppsala University. & C.P. Anderson, eds. Bioarchaeological and Geber, J. 2015. Comparative Study of Forensic Perspectives on Violence. How Violent Perimortem Weapon Trauma in Two Death is Interpreted from Skeletal Remains. Early Medieval Skeletal Populations (AD Cambridge: Cambridge University Press, pp. 400–1200) from Ireland. International 34–50. Journal of Osteoarchaeology, 25: 253–64. Knüsel, C.J. 2005. The Physical Evidence of https://doi.org/10.1002/oa.2281 Warfare: Subtle Stigmata? In: M. Parker Gunnarsson, F., Victor, H. & Alfsdotter, C. Pearson & I.J.N. Thorpe, eds. Warfare, 2016. Sandby borg VII. Undersökningar Violence, and Slavery (British Archaeo- 2015, Sandby sn, Mörbylånga kommun, logical Reports International Series 1374). Öland. Kalmar: Kalmar läns museum. Oxford: Archaeopress, pp. 49–65. Guyomarc’h, P., Campagna-Vaillancourt, M., Kohler, T.A. & Kramer Turner, K. 2006. Kremer, C. & Sauvageau, A. 2010. Raiding for Women in the Pre-Hispanic Discrimination of Falls and Blows in Northern Pueblo Southwest? Current Blunt Head Trauma: A Multi-Criteria Anthropology, 47: 1035–45. https://doi.org/ Approach. Journal of Forensic Sciences, 55: 10.1086/508697 423–27. https://doi.org/10.1111/j.1556- Kremer, C. & Sauvageau, A. 2009. Discrimin- 4029.2009.01310.x ation of Falls and Blows in Blunt Head Heather, P.J. 2006. The Fall of the Roman Trauma: Assessment of Predictability Empire: A New History of Rome and the Through Combined Criteria. Journal of Barbarians. Oxford & New York: Oxford Forensic Science, 54: 923–26. https://doi. University Press. org/10.1111/j.1556-4029.2009.01072.x Hedeager, L. 2011. Iron Age Myth and Larsen, C.S. 1999. Bioarchaeology: Interpreting Materiality: An Archaeology of Scandinavia Behavior from the Human Skeleton (2nd AD 400–1000. London: Routledge. ed). New York: Cambridge University Heimdahl, J. 2014. Makroskopisk analys 2014. Press. In: H. Victor, ed. Sandby borg IV. Lindbom, P. 2006. Vapnen under Wreccornas Undersökningar 2014. Sandby socken, tid, 150–500 e.Kr. Uppsala: Department Mörbylånga kommun, Öland. Kalmar: of Archaeology and Ancient History, Kalmar läns museum, pp. 399–401. Uppsala University. Heimdahl, J. 2016. Växtmakrofossilanalyser. Lund Hansen, U. 1987. Römischer Import im In: L. Papmehl-Dufay & C. Alfsdotter, Norden. Warenaustausch zwischen dem eds. Sandby borg V. Ögonblicket under mar- Römischen Reich und dem freien Germanien kytan. Seminarieundersökning i Sandby borg während der Kaiserzeit unter besonderer 2014. Sandby sn, Öland. Kalmar: Kalmar Berücksichtigung Nordeuropas. Kopenhagen: läns museum, pp. 50–51. Kongelige Nordiske Oldskriftselskab.

Magnus, B. 1997. The Firebed of the Serpent: Owsley, D.W., Berryman, H.E. & Bass, W.M. Myth and Religion in the Migration 1977. Demographic and Osteological Period Mirrored Through Some Golden Evidence for Warfare at the Larson Site, Objects. In: L. Webster & M. Brown, South Dakota. Plains Anthropologist,22: eds. The Transformation of the Roman 119–31. https://doi.org/10.1080/2052546. World AD 400—900. London: British 1977.11908859 Museum Press, pp. 194–207. Papmehl-Dufay & C. Alfsdotter, eds. Sandby Magnus, B. 2002. Ørnen flyr – om stil I i borg V. Ögonblicket under markytan. Seminar- Norden. In: K. Høilund Nielsen, ed. ieundersökning i Sandby borg 2014. Sandby Nordeuropæisk dyrestil 400–1100 e.kr sn, Öland. Kalmar: Kalmar läns museum. (Hikuin 29). Højbjerg: Hikuin, pp. 105–19. Reichs, K.J. 1998. Postmortem Dismember- Martin, D.L. & Harrod, R.P. 2015. ment: Recovery, Analysis and Interpret- Bioarchaeological Contributions to the ation. In: K.J. Reichs, ed. Forensic Study of Violence. American Journal of Osteology: Advances in the Identification of Physical Anthropology, 59: 116–45. https:// Human Remains. Springfield (IL): Charles doi.org/10.1002/ajpa.22662 C. Thomas, pp. 353–88. Martin,D.L.,Harrod,R.P.&Fields,M. Sanderson, S.K. 2014. Human Nature and the 2010. Beaten Down and Worked to the Evolution of Society. Boulder (CO): Bone: Bioarchaeological Investigations of Westview Press. Women and Violence in the Ancient Šlaus, M., Novak, M., Vyroubal, V. & Bedic,́Ž. Southwest. Landscapes of Violence,1:1– 2010. The Harsh Life on the 15th Century 19. Croatia-Ottoman Empire Military Border: McCall, G. & Shields, N. 2008. Examining Analyzing and Identifying the Reasons for ̌ the Evidence from Small-Scale Societies the Massacre in Cepin. American Journal of and Early Prehistory and Implications for Physical Anthropology, 141: 358–72. https:// Modern Theories of Aggression and doi.org/10.1002/ajpa.21152 Violence. Aggression and Violent Behavior, Stenberger, M. 1933. Öland under äldre järnål- 13: 1–9. https://doi.org/10.1016/j.avb. dern: en bebyggelsehistorisk undersökning. 2007.04.001 Uppsala: Royal Academy of Letters, Mejsholm, L. 2009. Gränsland: konstruktion av History and Antiquities. begravningsritual och tidig barndom vid Tegtmeyer, C. & Martin, D.L. 2017. The tiden för kristnandet i Skandinavien Bioarchaeology of Women, Children, and (Occasional Papers in Archaeology 44). Other Vulnerable Groups in Times of Uppsala: Uppsala University. War. In: D.L. Martin & C. Tegtmeyer, Näsman, U. 1984. Glas och handel i senromersk eds. Bioarchaeology of Women and Children tid och folkvandringstid: en studie kring glas in Times of War. Cham: Springer Inter- från Eketorp-II, Öland, Sverige. Uppsala: national, pp. 1–14. Department of Archaeology and Ancient Timmins, S., Seréville-Niel, C. & Brickley, M. History, Uppsala University. 2017. Childhood Cranial Trauma from a Näsman, U. 1988. Den folkvandringstida krisen Late Roman and Merovingian Context i Sydskandinavien, inklusive Öland och from Michelet, Lisieux, France. Inter- Gotland. In: U. Näsman & J. Lund, eds. national Journal of Osteoarchaeology, 27: Folkevandringstiden i Norden: en krisetid 715–22. https://doi.org/10.1002/oa.2581 mellem ældre og yngre jernalder:rapport fra et Ubelaker, T.H. 2015. The Concept of bebyggelsearkæologisk forskersymposium i Perimortem in Forensic Science. In: K. Degerhamn, Öland, d. 2–4 oktober 1985. Gerdau-Radonvić& K. McSweeney, eds. Aarhus: Universitetsforlag, pp. 227–55. Trends in Biological Anthropology, 1. Nicklasson, P. 1997. Svärdet ljuger inte (Acta Oxford: Oxbow Books, pp. 93–99. Archaeologica Lundensia 4. 22). Stock- Vargová, L., Horácková,̌ L., Horáková, M., holm: Almqvist & Wiksell International. Eliášová, H., Myšková, E. & Ditrich, O. Novak, S. 2000. Battle-Related Trauma. In: 2016. Paleopathological, Trichological and V. Fiorato, A. Boylston & C. Knüsel, eds. Paleoparasitological Analysis of Human Blood Red Roses: The Archaeology of a Mass Skeletal Remains from the Migration Grave from the Battle of Towton AD 1461. Period Cemetery Prague-Zlicín.̌ Interdisci- Oxford: Oxbow Books, pp. 90–102. plinaria Archaeologica,7:13–32.

Viberg, A. 2012. Remnant Echoes of the Past: is currently working on her PhD at Archaeological Geophysical Prospection in Linnaeus University’s Graduate School in Sweden (Theses and Papers in Scientific Contract Archaeology (GRASCA). Her Archaeology 13). Stockholm: Stockholm University. research focuses on the treatment of dead Victor, H. 2015. Sandby borg: ett fruset bodies. She has a special interest in the ögonblick under folkvandringstiden. In: social implications of violence, death, and K-H. Arnell & L. Pamphell-Dufay, eds. taphonomy. Grävda minnen: Från skedemosse till Sandby borg (Meddelanden från Kalmar läns hembygdsförbund och Stiftelsen Address: Clara Alfsdotter, Linnaeus Uni- Kalmar läns museum årgång 95). Kalmar: versity, School of Cultural Studies, Faculty Kalmar läns museum, pp. 97–115. of Arts and Humanities, Linnégatan 5, Waddington,C.,Beswick,P.,Brightman,J., SE-391 82 Kalmar, Sweden. [email: Bronk Ramsey, C., Burn, A., Cook, G., [email protected]]. et al. 2012. Excavations at Fin Cop, Derby- shire: An Iron Age Hillfort in Conflict? ORCID: https://orcid.org/0000-0002- Archaeological Journal,169:159–236. https:// 0181-4458. doi.org/10.1080/00665983.2012.11020914 Wilhelmson, H. 2017. Perspectives from a Human-Centred Archaeology: Iron Age Anna Kjellström is Associate Professor in People and Society on Öland (Acta Osteology at the Department of Archae- Archaeologica Lundensia 68). Lund: Department of Archaeology and Ancient ology, Stockholm University. She is cur- History, Lund University. rently a member of the Atlas Project, Willey, P. 1990. Prehistoric Warfare on the dealing with ancient human genomes. She Great Plains: Skeletal Analysis of the Crow has specialized in human health issues and Creek Massacre Victims. New York: trauma in skeletal populations from the Garland Publishing. Wolfgang, M. 1958. Patterns of Criminal Mesolithic to modern times, with a par- Homicide. Philadelphia (PA): University of ticular interest in the Viking period and Pennsylvania. Early Middle Ages. Her research focuses on bioarchaeological response associated with sociocultural concepts such as urban- SUPPLEMENTARY MATERIAL ization, interpersonal violence, and subju- gated groups. To view supplementary material for this article, please visit https://doi.org/10.1017/ Address: Archaeological Research Labora- eaa.2018.55. tory, Department of Archaeology and Classical Studies, University of Stockholm, Lilla Frescativägen 7, SE-106 91 Stock- BIOGRAPHICAL NOTES holm, Sweden. [email: Anna.Kjellstrom@ ofl.su.se]. ORCID: https://orcid.org/ Clara Alfsdotter is an osteologist and 0000-0001-8964-3771. archaeologist at Bohusläns Museum. She

Le massacre de Sandby borg : violence interpersonnelle et profil démographique des morts

Les restes de 26 personnes dépourvues de sépultures ont été découverts dans les fouilles de l’enclos fortifié de Sandby Borg (datant de l’âge du Fer, 400–550 apr. J.-C.), aussi bien à l’extérieur qu’àl’intérieur

de leurs maisons. Les squelettes ainsi que les données archéologiques indiquent que le site a été abandonné après la mort de ces individus. L’étude des ossements humains et une analyse des traumatismes subis a suivi les protocoles établis en anthropologie physique. L’examen des ossements n’a identifié que des sujets mâles ; en revanche, toutes les classes d’âge étaient représentées. Huit individus (31 pour cent de l’ensemble) présentaient des signes de traumatisme périmortem de divers types (lésions causées par des instruments tranchants, contondants ou pénétrants) que l’on peut interpréter comme indices de violence interpersonnelle. La position des corps et la configuration des traumatismes laissent penser à un massacre plutôt qu’à une bataille. La manière « efficace » d’infliger une blessure (c’est-à-dire en usant de violence de manière économique mais fatale), le fait que les cadavres n’ont pas été manipulés et le con- texte archéologique suggèrent que les responsables étaient nombreux et que leur assaut a été mené avec efficacité. La situation sociopolitique de l’époque était instable, ce qui nous amène à proposer que le motif de l’attaque fût l’obtention du pouvoir et du contrôle. (Translation by Madeleine Hummler)

Mots-clés: violence, traumatisme, époque de grandes migrations, Sandby borg, âge du Fer, morts sans sépultures

Das Massaker von Sandby borg: interpersonale Gewalt und demografisches Profil der Toten

Während der Ausgrabungen der eisenzeitlichen Ringwallanlage von Sandby borg (400–550 n. Chr.) sind die Überreste von 26 unbestatteten Menschen innerhalb und außerhalb der Häuser zutage gekom- men. Die Skelette und die archäologischen Angaben deuten darauf hin, dass die Ringwallsiedlung nach dem Tod dieser Menschen aufgegeben worden ist. Die Untersuchung der Knochen und die Analyse der Traumata folgten die anerkannten Protokolle der biologischen Anthropologie. Die osteologische Auswertung hat nur Männer identifiziert aber alle Altersstufen waren vertreten. Acht Individuen (31 Prozent der Sammlung) hatten tödliche Verletzungen (perimortem Trauma), die von verschiede- nen, scharfen, stumpfen und penetrierenden Gegenständen verursacht worden sind. Die Lage der Leichen und die Struktur der Traumata lassen eher auf ein Massaker als auf eine Schlacht schließen. Die Verteilung von „effizienten” Verletzungen (wo Gewalt spärlich aber effektiv angewendet wurde), die Tatsache, dass man die Leichen nicht nachträglich manipuliert hatte, und die archäologischen Angaben deuten darauf hin, dass die Täter zahlreich waren und dass der Angriff effektiv war. Die damalige soziopolitische Situation war offensichtlich unbeständig, was uns darauf schließen lässt, dass die Übernahme der Macht und Kontrolle der Grund für das Massaker war. (Translation by Madeleine Hummler)

Stichworte: Gewalt, Trauma, Völkerwanderungszeit, Sandby borg, Eisenzeit, unbestattet