ANCIENT DEAD Post-burial Practices and Ontologies of Death in Southern Sweden AD 800 – 1200

137 This text discusses reuse and modifications of older graves in southern Sweden during the Late Iron Age and early medieval period (c. 9th to 12th centuries AD). Post-burial practices in the Late Iron Age have in general been interpreted as means to negotiate status, identity and rights to land, while in the later part of the period they are comprehended as expressions of religious insecurity and syncretism. In this text, the continuity of post-burial practices during the whole period is stressed and instead of general top-down interpretative models, the ontological status and material aspects of death, dead bodies and their graves is emphasized. It is argued that the post-burial actions generally constituted ways of relating to a specific type of materiality, the bones of the ancient dead, which transgress binary categorizations such as living–dead, past–present, heathen–Christian, and human–nonhuman. The argument builds on five recently excavated sites in southern Sweden: Bogla, Broby Bro, Lilla Ullevi, Valsta and Vittene.

This text discusses reuse and modifications of older graves in southern Sweden during the Late Iron Age and early medieval period (c. 9th to 12th centuries AD). Post-burial practices in the Late Iron Age have in general been interpreted as means to negotiate status, identity and rights to land, while in the later part of the period they are comprehended as expressions of religious insecurity and syncretism. In this text, the continuity of post-burial practices during the whole period is stressed and instead of general top-down interpretative models, the ontological status and material aspects of death, dead bodies and their graves is emphasized. It is argued that the post-burial actions generally constituted ways of relating to a specific type of materiality, the bones of the ancient dead, which transgress binary categorizations such as living-dead, past-present, heathen-Christian, and human-nonhuman. The argument builds on five recently excavated sites in southern Sweden: Bogla, Broby Bro, Lilla Ullevi, Valsta and Vittene. Keywords: reuse, post-burial practices, Iron Age, burials, materiality, perspectivism

INTRODUCTION: RE-(AB)USE OF THE OLD DEAD
Burials that have been reopened, reused, destroyed and manipulated in various ways are known from all periods of prehistory (e.g. Randsborg 1998;Nilsson Stutz 2003;Kümmel 2005;Brinch Petersen 2006;Andrews & Bello 2006;Olofsson 2006;Fahlander 2008bFahlander , 2010Klevnäs 2013). Such post-burial actions do not comprise a single category of practices, but span over a large array of different types interferences with older graves. Grave robbery with the aim of gathering valuables or collecting artefacts or bones of the dead for ritual purposes has traditionally been a frequently advocated reason for post-burial activities (cf. Aspöck 2011:299). Another typical explanation is accidental superimpositions and intercuts with unmarked graves (Burenhult 2002: cf. Andersson 1997. More recently, however, it has become more common to emphasize ritual and social aspects of post-burial actions. Ritual interpretations tend to revolve around the buried individual as either a dangerous or a benign entity. For example, one reason for meddling with a grave can be to prevent the dead from interfering with the living by adding or removing things from the grave as a way to amend a failed burial ritual (Fahlander 2010;Runer & Sillén 2014:33). To destroy a grave can also be a way to neutralize an enemy's reputation or memory by ritually "re-killing" an already dead individual (Fahlander 2008b;Klevnäs 2016). However, reuse of a grave can likewise work as "mnemonic citations", negating memory and history, and as spectacles to commemorate a person and his/her reputation (Wickholm 2008;Price 2010). A particularly persistent theme in post-processual archaeology concerns social interpretations that emphasize different strategic ways in which old graves are employed to negotiate ideology. For instance, the idea that burying the dead in relation to old dead may constitute ways among the living to claim genealogy in order to legitimize status or rights to territories (Zachrisson 1994;Williams 1998;Andersson 2005).
From being largely a marginal issue, the literature on reuse and postburial practices in the past has grown rapidly in the last few decadesespecially in Late Iron Age research (e.g. Williams 1998;Gansum 2004;Andersson 2005;Pedersen 2006;Artelius 2004Artelius , 2010Artelius , 2013Artelius & Lindqvist 2007;Appelgren & Renck 2007;Aspeborg 2007;Olsson 2007;Thäte 2007;Wickholm 2008;Bratt 2008;Lindqvist 2010;Aspöck 2011;Hållans Stenholm 2012;Klevnäs 2013;Lund 2013;Satalecki 2014;Wuopio 2015). However, despite the growing numbers of cases, generalizing models based on memory and ancestry are still a dominant perspective in understanding post-burial practices in Swedish archaeology. Suffice it to mention Hållans Stenholm's (2012:43) re-cent survey of post-burial research in Scandinavia, in which three main themes are distinguished: the past and the ancestors as a legitimizing principle, the mound as a representation of the past and ancestors, and reuse of a mound as a ritualized practice in order to legitimize ownership and status. This state of things is somewhat reversed when it comes to interpretation of reuse in the early medieval period (11th-12th centuries AD). Social interpretations based on collective memory and ancestor veneration are frequent here too, but, because they coincide with a change of religion to Christianity, ritual aspects tend to be emphasized more. For instance, post-burial actions are understood in terms of hostility towards the pre-Christian dead, religious insecurity, a lack of proper knowledge of the Christian doxa, or as a token of religious dualism or syncretism (Olausson 1993;Andersson 1997;Lundström & Theliander 2004:79;Pedersen 2006:351;Artelius 2010:215;Hed Jakobsson & Lindblom 2011:89;Lund 2013:51f;Runer & Sillén 2014:33;Vinberg 2015;Tesh 2015).
There is no question that burials can be important parts of a social arena, and that they can express religious and eschatological concerns. However, what is at stake here is an unfortunate polarization and the streamlined interpretations of a rather complex set of practices. When looking more closely at the post-burial actions in the Late Iron Age and early medieval period, two particular aspects emerge that tend to be overlooked in Scandinavian research. The first is the quite large timespan that often separates the original burials and the later modifications. In Sweden, post-burial actions were generally made about 500-1000 years after the first phase of burials had ceased. It is thus important to thoroughly examine the circumstances in which people during the Late Iron Age "returned" to or "recolonized" previously abandoned burial sites -and more importantly: how the remains of the dead and their graves were understood in an ontological sense. For instance, are the ancient cremated remains necessarily recognized as "ancestors" -or even as human remains? The second aspect that tends to be neglected is the apparent continuity in post-burial practices over the whole period (9th to 12th centuries) and that such actions are not queer incidents, but must be considered an integrated part of Late Iron Age and early medieval burial practices (cf. Thäte 2007:5f;Lund 2013:53).
In this paper I wish to examine some of the common assumptions about why older burials were rearranged, reused and manipulated during the Late Iron Age and early medieval period. I will employ a symmetrical approach that to a greater extent recognizes the material and practical aspects of the monuments and the mortal remains. This involves dismantling a few preconceptions concerning death and dead Fredrik Fahlander bodies in terms of ancestors and personhood, but also avoiding dichotomized concepts such as pagan-Christian, past-present, life-death, human-material, etc. Instead I wish to emphasize what is "hidden" behind the hyphen in concepts such as "pre-Christian" (cf. Latour 1999). The interesting aspects are not necessarily found in either "the old ways" or in the "Christian", but in the ways they intersect, oppose, and relate to each other. In order to achieve that I employ a microarchaeological method and approach the issue from the bottom up, working together with the archaeological material rather than approaching it from the top down (cf. Fahlander 2008a). This means keeping a necessary level of specificity to allow that material to affect the outcome of the study. Holbraad & Pedersen (in print), who have developed a similar approach in anthropology, may serve to illustrate the approach. They point out that when ethnographers study small-scale societies, their concepts (e.g. gender, religion, social structure etc.) tend to remain more or less static throughout the study, working mainly to categorize the assembled data. As an alternative, Holbraad & Pedersen suggest that we should approach the Other from an ontological point of view, allowing for alterity by fo- cusing not so much on what we expect to find, but what there actually is to see. Such an ontological inquiry is not about deconstruction of general concepts, but about reconstruction -an empirical perspective that is open for other ways of categorizing the world (Holbraad & Pedersen in print). The present study will hence begin with the material, which in this case comprises five case studies from southern Sweden: Vittene, Lilla Ullevi, Bogla, Valsta and Broby Bro (Figure 1). The examples are primarily chosen because they are recently excavated sites (with the exception of Valsta) that comprise sufficient documentation of post-burial activities during the 9th to 12th centuries AD. After a brief review of the documented post-burial actions at these sites, I will elaborate on what how we may understand the different post-burial actions taken towards the material remains of the old dead.

VITTENE -CURIOSITY AND INCLUSION
The first example concerns Vittene, a small, seemingly ordinary Iron Age burial ground in the county of Västergötland (Artelius & Lindqvist 2007). The site comprises 67 graves ranging from various stone settings, small mounds, a ship setting and five inhumations in cists ( Figure 2). The first phase of burial started in the Bronze Age with a single stone setting (A85). After some 500 years, it was accompanied by additional burials during the pre-Roman and Roman period. The site was then abandoned for about 600 years until it was reinstalled as a burial ground in the Vendel/Viking Age, continuing into the 11th century (2007:51). During the last phase there are examples of different kinds of reuse and post-burial manipulations. In Vittene there are not only burials superimposed on older ones, but also examples of old burials incorporated in new monuments. It is important to note that the site by no means is spatially limited, but that there is there plenty of room for more burials without the need of overlapping. There is thus no practical reason for the new graves to be superimposed or intersecting with so many of the previous ones (2007:145). One interesting example is the large mound (A5) constructed during the Viking Age. When this monument was built, special attention and measures was taken to find and modify the previous urn burials which the mound was placed upon (2007:123). This seemingly careful interest in the older burials is also apparent in several other Viking Age cremations that were placed directly on top of the cover slabs of the old urns. This was done in a manner that never damaged the older graves, but seemingly because of a desire to physically relate to the older dead (2007:146).
Another noteworthy example in Vittene is the Viking Age ship setting (A6) that apparently incorporated an older pre-Roman cremation (A45) at the "bow" end of the new monument (2007:88). Artelius and Lindqvist argue that this inclusion of a much older grave was not so much a ritualized appropriation as a "curious interest" and knowledge of the earlier burials (2007:122). Artelius and Lindqvist even suggest that the Viking knowledge of the location and constitution of previous burials had been retrieved from a systematic careful excavation of old graves. At Vittene, this interest in the old dead continued for the whole period of use. One late example is the construction of the rectangular stone setting (A10), which is interpreted as a Christian imitation of a pre-Christian grave because of similarities in shape and the ways in which a few flakes of ground stone were added at the foot end of the coffin (2007:152). The inclusion of stone flakes is a common feature in the pre-Roman and Roman graves at Vittene. None of the presumably Christian inhumations (A10, A37, A58, A59 and A75) interfere with the older ones, but they are placed close to the largest mound (A5) at the site (2007:91, 155). Even though Artelius and Lindqvist may exaggerate the extent of the Viking knowledge of earlier burials and the mimicking practices, Vittene nonetheless comprises subtle ways of relating to the older dead that is characterized by Artelius and Lindqvist as a "caring and inclusive" attitude (cf. Thäte 2007:277).

VALSTA -APPROPRIATION AND MANIFESTATION
The second case concerns Valsta in the province of Uppland. It is not a recently excavated site, but is included here mainly because it is one of the most frequently discussed examples of reuse in the Late Iron Age (e.g. Andersson 1997;Gräslund 2001:128;Ersgård 2006:99f;Bratt 2008:247;Mejsholm 2009:35f;Tesh 2015). The post-burial practices at Valsta are understood in terms an appropriation of an early burial ground during the Viking Age. The burials at Valsta began in the Roman Iron Age (c. 200-300 AD) with 8-10 stone settings placed on the top of a ridge. After 500 years, the place was put into use again and during the Viking Age and the Early Middle Ages when 35 cremations and 19 inhumations were added on the small hillock.
The most outstanding feature in Valsta is a large mound (A1), a cremation burial that was constructed in the early 9th century that partly covers five earlier stone settings. As such, the mound is considered by Andersson (1997:53) as a prime expression of an "Odal mentality", that is, emphasizing claims of ancestry and territory, but, of course, it may also be a case of reuse of the most prominent spot on top of the hillock. However, the mound was later reopened sometime during the early medieval period (c. 1050-1150) when three stone cists (presumably containing inhumations) were built in a "cross-like" manner into the mound's inner stone cairn ( Figure 3b). This action partly destroyed the old cremation burial which contents were randomly spread in the filling of the mound. On a later occasion the mound suffered a "brutal and rushed" opening, which is interpreted by Andersson as an attempt at grave robbery. This post-burial action destroyed the east-west oriented cist and the north-western part of the original cremation layer (1997:58). Because no bones were found in the cists, it is not possible to establish whether it was an attempt at robbery or if the purpose may have been to relocate bones of the dead. Besides the post-burial actions in mound A1, the other earlier graves at Valsta were left undisturbed. The superimposition of earlier stone settings by mound A1 can be seen as inclusive -making the old burials part of the new one -although not as expressive as in Vittene. The destructive rebuilding of mound A1 in the 11th and 12th centuries, however, shows little reverence for the previous dead. This negligent attitude to the old dead during the latter part of the period is, with the exception of Vittene, a common issue for the sites discussed here.

BOGLA -FROM CURATING TO REUSE
Bogla is another small burial site in the province of Småland in southern Sweden (Artelius & Kristensson 2005;Artelius 2010). This site started with the construction of a single stone setting in the pre-Roman period (A12158). After a long period of abandonment, the site was put to use again during the Vendel Period and continued into the late 11th century (Artelius 2010:38). The first pre-Roman grave at Bogla did not pass unnoticed when the burials resumed. About 800 years after its construction, another cremation and a mound were added and some of the original kerbstones were rearranged (2010:164, 170). A number of mounds, stone settings and a ship setting were subsequently constructed, and during the final phase of the site, ten inhumation burials were also added ( Figure 4). The inhumations are all east-west aligned and considered Christian -or at least constructed during the early Christian period. Four of them were buried under a mound (A114462a, A114462b, A114787 and A114429), while the subsequent six had no recognizable superstructure. A boundary ditch with stones and poles was also created to enclose the area in the east (2010:166).

Fredrik Fahlander
Artelius argues, as he did in the case of Vittene, that the reuse of the pre-Roman stone setting is further evidence of how older burials evoked curiosity during the Late Iron Age (2010:175). Also in the latter part of the period, he suggests that the inhumations under mounds were intentional as to allude to "the old ways" by using the mound as a superstructure (2010:208). In one particular instance (A11429), a Viking Age mound was reused for a later inhumation during the late 11th century (1060-1070). In contrast to the Viking Age reuse of the pre-Roman stone setting, the coffin was dug into the mound with no apparent consideration to the previous cremation (cf. Valsta). Artelius sees conflicting ideas here, in the first case of the reused stone setting, a careful relation to the old dead was manifested, while in the latter case, an earlier mound was "only" reused with no recognizable interest in the previous cremated remains (2010:207).

LILLA ULLEVI -MESSING WITH THE RECENT DEAD
The fourth case considers Lilla Ullevi in the province of Uppland. Here, at least 34 cremations and 30 inhumations have been excavated on a small hillock, all of which are dated to the Late Iron Age, c. 750 to 1050 (Hed Jakobsson & Lindblom 2011). At least four, possibly five, cremation burials have been affected by later inhumation burials. Judging by their positions and lack of reused elements, the superimpositions seem to have been unintended -only in one case has an older grave been emphasized when the new one was constructed. The later inhumation burials do, however cling quite tight to the old cremations, even though there is plenty of space on the hillock ( Figure 5).
There are at least two instances of deliberate post-burial actions at the site. In grave 22 a small pit had been dug above the torso of the buried person, which was filled with stones and ashy soil (2011:85). In grave 60, a hearth-like feature was found within the stone-lined square that contained fragments of bone, beads, metal fragments and an arrowhead. The excavators interpret this as more or less contemporary with the inhumation. Most interesting, however, is grave 57, which is a double inhumation burial that at a later point was superimposed by another inhumation. The superimposing grave share a kerb lining similar to that of the double grave and it is assumed that the previous grave was known when the subsequent one was added. Yet, on a later occasion, this grave too was reopened. The interior of the grave was jumbled and parts of the stone lining were disturbed. The excavators find no apparent reason for the latter intervention; an iron knife was found left in the grave, indicating that it was not a case of grave robbery. Unfortunately no non-cremated bones are preserved at the site, which makes it difficult to tell if the bones were relocated or manipulated in any way. Interestingly, the stratigraphy reveals that the two inhumations superimpose three earlier cremations of which at least two seem to have been noticed when the new, presumably, double burial was made (2011:92ff). It seems that this spot has attracted particular interest during the whole period of use. It is also interesting that inhumations and cremations have been practiced parallel during the period AD 950-1050 (cf. Olausson 1993:23). The stratigraphy is complex, but it is possible to establish that at least one inhumation is older than at least two of the cremations (2011:82ff). In contrast to the previously mentioned sites, Lilla Ullevi lacks an earlier phase of burials from the Early Iron Age. It is difficult to discern any particular hostile or benign relation to the superimposed cremation burials, but the post-burial actions at Lilla Ullevi display an indifferent and inconsistent attitude towards previous inhumations as indicated by the sequence of grave 57.

BROBY BRO -REBUILDING AND RELATING
The fifth and final example concerns the burial ground RAÄ 36 at Broby Bro in the province of Uppland. The site is only partly excavated, but consists of about 50 stone settings mainly from the Pre-Roman and Roman Iron Age, a stone cairn, one rectangular and two round mounds from the Late Iron Age, as well as at least two inhumations from the 11th century. In the vicinity (possibly related) is also a burial ground consisting of 19 inhumations from the 11th century (RAÄ 620). The first phase of burials at RAÄ 36 consist of a number of stone settings and spans from the pre-Roman to the Roman period. After a period of some 700 years the site was again put in use during the 10th century, continuing into the early medieval period.
Also at Broby Bro, the inhabitants of the Late Iron Age seem to have been well informed about the layout and contents of earlier graves. Some post-burial actions seem to have respected the older burials, while others were destroyed and rebuilt. The earliest example of reuse is the rectangular stone setting containing an urn cremation from the 10th century (A15000) which was placed slightly asymmetrically on top of a round stone setting from the Early Iron Age (Figure 6). The large kerbstones lining the superimposing grave are dug down into the stone packing of the earlier grave, but otherwise left the older grave intact. An expression of a quite different attitude towards the earlier dead is found a few metres west of the two graves in a much larger (13 × 8 m) rectangular structure (A8000). This feature too is placed upon a round stone setting dated to the Pre-Roman Iron Age (c. 400-200 BC, 2s). However, in this case a large pit was dug into the centre of the stone setting, destroying the central burial, and in its place two large stone blocks were positioned. One was buried halfway down in the old grave to provide a flat surface on its top. The second block was placed in upright position next to the buried one (Figure 7b). In conjunction with this, a rectangular kerb of stones was also erected and the interior filled with soil that created a concave mound covering all but the tip of the second stone slab. It has not been possible to date the time when the old burial was rebuilt. An indication is, however, found in another rebuilt grave in the area. A7000 was also originally a round stone setting from the first century AD that was manipulated in similar manner. The stone setting had also been dug out and a large slab with a flat surface was put in the centre (Figure 7a). This old grave was likewise expanded with a larger kerb (although retaining the round shape) and built up with soil. These interferences and manipulations in this case could be stratigraphically related to an eastwest oriented 11th-century inhumation burial placed in the mound. The many similarities in modus operandi between A7000 and A8000 sug- gest that both interventions are contemporary. It is important to note that the labour invested in the reuse in these cases is quite extensive (as in the case of mound A1 in Valsta).
In addition to these rather large-scale constructions are also traces of less substantial and less visible post-burial actions present at Broby Bro. One example is a flat and non-conspicuous Early Iron Age stone setting, which was reused for an 11th-century inhumation burial (A16000). In this case no additional kerb or fill was added. The inhumation pit was, however, dug through the main concentration of cremated bone, leaving just a crescent of the originally circular cremation layer. An interesting aspect of this reuse is that the original cremation (and the coffin) was not placed in the centre of the round stone setting. This is yet another indication that the layout of the original burial underneath the cairn apparently was well known when the inhumation was added (see Figure 6).
The intersections in Broby Bro give a mixed message. It is not possible to determine the nature of the Viking Age attitude to the previously dead from a single case (A150000). During the early medieval period, however, the original burials in all four cases were partly destroyed (cf. Bogla and Valsta). In A16000 it was evidently important to place the new burial on the exact same spot as the original. Some bones were left in situ, but a larger part was probably shovelled aside when the pit for the coffin was dug. A relation to the old dead was obviously sought, but the actions show no apparent traces of inclusion or respect -on the contrary, the destruction of the original cairns rather points in the opposite direction.

FROM CURIOSITY TO NEGLIGENCE? CHANGING ATTITUDES TOWARDS THE OLD DEAD
The five sites discussed here are all distinct in their own individual ways at the same time as they serve to illustrate the variability and complexity of post-burial actions during the period in question (Table 1). As a general phenomenon, similar post-burial actions are by no means rare instances in the grave fields of southern Sweden. As comparison it has been established that during the Late Iron Age in the Mälar valley region of south-eastern Sweden, about 9% of the larger mounds are superimpose on older graves (Bratt 2008:98). The proportion of sites in the same area and period that include one or more superimpositions is about 20% (Hållans Stenholm 2012:110ff, 131). In Denmark, Pedersen found a similar rate of at least 20% of the Viking Age burial sites that show evidence of reuse or "association with an ancient burial mound" (2006:348). The sites discussed here are thus not unique in terms of post-burial activities, but are a part of a general phenomenon in the Late Iron Age and early medieval periods in northern Europe (Pedersen 2006;Thäte 2007;Aspöck 2011).
Although diverse, the five sites all share some common aspects in the execution as well as in the context of the practices. For example, with the exception of Lilla Ullevi, there is a substantial gap (about 500-1000 years) in time between the intersected graves and the later reuse which contests a genealogic relation between the old and the new dead. This phenomenon is duly noted by the excavators, but not elaborated upon at length. Artelius and Lindqvist argue in terms of "a return" -that people were coming back to Vittene during the Late Iron Age -and others about "recolonization" or "reappropriation" of a place. This terminology implies a familiarity with the former occupants, or at least a sense of sameness (after all, they are all from the Iron Age). There is, however, little discussion about where the new occupants return from or what constitutes their possible previous relations to the area. To put things in perspective, it is worth noting that the intrusions in Early Iron Age graves at Broby Bro were made on graves that at the time were more than a thousand years old. Those graves were thus older then than the added inhumation burials are in relation to us today. This is a central issue to which I will return in the discussion. Another common trait among these cases is also the attitude towards the previously dead. In the 9th to 10th centuries the remains of the old dead seem to been handled with something resembling curiosity and respect. In general, the engagements with older graves during this period were subtle and inclusive. The cremated remains of the old graves are not disturbed even though some urns and cremation layers were superimposed or "excavated". This attitude seems also to include the burial monuments, which generally are left intact (cf. Biuw 1992:81;Thäte 2007:277).
In the later part of the period (11th and 12th centuries), when inhumation burial becomes more common, the five cases demonstrate a different attitude towards both the remains of the dead and their monuments. There are still indications of a continued interest in and knowledge of the older dead and their graves. The inhumation burials in all five cases are often situated close to, or reusing, ancient mounds. The attitude towards the remains of the older dead, however, differs. Instead of respect and inclusion, older burials were substantially rebuilt and the remains of the dead removed or ignored when new inhumations were added to the old monuments. A common reason for the manner of reuse during the later part of the period is related to the change of faith. The inhumations have all been assumed to be "Christian", with the possible exception of Valsta (Andersson 2005:93). The definition of what constitute a Christian burial varies, but generally concerns eastwest orientated inhumation burials where the dead are to be placed on their backs with the head to the west (Artelius 2010:117; but see Andersson 2000:12). Christian burials should not (but often do) include interments such as coins and a knife, but not combs and never animals or food (Gräslund 2001:47ff;Andersson 2005:144). Others have argued that in order to be considered fully Christian, the burials need to be placed within a restricted space separated from earlier pre-Christian burials (e.g. Theliander 2010:174). Judging from these criteria, with the exception of the last, the majority of the inhumations in the five cases seem to correspond to a Christian burial tradition -or at least being inspired by continental Christian burial practices (cf. Artelius 2010:109, 154). However, there are several instances that indicate a less dichotomized relation between "the old" and "the new" ways of burial. For instance, the continued use of the mound as superstructure for the inhumations, is, besides the square flat stone settings, similar the ones of the Late Iron Age (Gräslund 2001:53;Andersson & Fahlander ms). In Lilla Ullevi and Valsta a parallel practice of inhumations and cremation can be demonstrated (Andersson 1997:80). There are also several cases of burning associated with inhumation burials (e.g. Andersson 2011;Wikström 2010:103;Runer & Sillén 2014:26;Holback 2016, cf. Holloway 2008. Other anomalies include the case of a cow's tooth placed on the lid of one of the coffins in Broby Bro and the "pagan" amulet rings in the coffins at Valsta (Andersson 1997:76;Andersson 2011:23). All sites with inhumation burials discussed here display one or more such discrepancies or hybrid variations. The idea that the change of faith is the sole reason behind the changed attitude towards the remains of the old dead thus seems less likely.
But how are we to comprehend these varying attitudes towards the old dead? Following the suggestions of Holbraad & Pedersen's ontological inquiry, we may benefit from exploring alternative ways of understanding the Late Iron Age interest in older burials other than to manifest ties to ancestors in order to emphasize memory and to legitimize power and right to land. There may also be more behind the early medieval intersections of older graves than religious syncretism, ambiguity or a reluctance to abandon the old ways. One aspect that tends to be neglected in the discussion is the ontological facets regarding how the materiality of the burials and the remains of the dead were appropriated. The long periods of abandonment of the burial sites suggest that old groups of partly overgrown graves were rather encountered in the landscape to whose presence and origins the new occupants probably had to relate to in some way or another. That scenario raises the question of how the pre-existing ancient graves and their content actually were apprehended over time. It is essential here to allow for ways to look at graves and human remains other than from a Western modern perspective. A grave must not necessarily constitute a final resting place for an individual, and the mortal remains need not represent a dead individual. In addition, we must allow for other ways of experiencing a past than the modern typological one. For example, Early Iron Age graves are not necessarily more similar or related to Late Iron Age ones merely because they are situated in the same general period. When working from the bottom up one should not perceive the past from the view of the present, but to allow every past to have its particular array of perspectives on possible pasts and futures to come. Gosden and Lock (1998) have suggested that the past in the past may be perceived as either genealogical history, based on a lineage of ancestors, or as mythical history, where the past is diffuse and fluid. The first Fredrik Fahlander type is most common in post-processual archaeology, which by tradition has focused on ancestors, memory and personhood (e.g. Parker-Pearson 1999;cf. Whitley 2002:124). The latter type of mythical histories is less explored -despite its potential of opening up for alternative views on the ancient dead (but see Hållans Stenholm 2012:42). One interesting illustration of this way of understanding the past is the suggestion by Williams that the older human remains that occasionally were found by the Anglo-Saxon gravediggers may have encouraged mythical and supernatural interpretations rather than recognizing the remains as dead ancestors (1998:97). In the poem Beowulf, and for its Anglo-Saxon audience at the end of the first millennium AD, Williams argues, the past is not described as the past, but rather as a foreign, distant mythological past of an ancient race whose world was enmeshed in supernatural powers. The ancient dead thus need not be about genealogy and ancestors, but rather about something that is unrelated to the present social conditions and thus less likely to employed in social strategies The concept of mythical history is indeed interesting and actually fits quite well with the substantial time-gap between the original burials and the subsequent reuse in the case studies discussed here. In oral societies the past becomes increasingly mixed up with myth after some 150-200 years (Montell 1996:178) and it would thus not be surprising if the 300-800-year-old graves that were manipulated could have been considered as something different. From that point of view, the post-burial actions in these cases need not necessarily be directed towards dead ancestors of predecessors, but as different ways of relating to the materiality of the ancient dead.

DISCUSSION: ONTOLOGIES OF DEATH AND THE MATERIALITY OF THE ANCIENT DEAD
This line of reasoning resonates well with Holbraad and Pedersen and other advocates of the so-called "ontological turn" which view people, things and animals as relationally constituted rather than being discrete categorical entities (e.g. Alberti & Bray 2009;Salmond 2012;Watts 2013). Especially the proponents of perspectivism accentuate alterity and seek to explore other ways to understand the world that are not necessarily within the boundaries of western scientific categorization. It is argued that binary categories, such as culture-nature, living-dead, or human-non-human are not as strict in some ontologies, and that, for instance, animals and materialities can possess certain humanlike properties and vice versa (Vivieros de Castro 2004). Perspectivism is superficially similar to traditional animism, but is actually about animacy, which focuses on relations between humans, and non-humans in a similar sense as actor-network theory and other symmetrical perspectives (Fahlander 2016;cf. Gell 1998).
Although the cremated bones probably were recognized as humanlike, the cremated remains of the old dead, from a perspectivist point of view, need not necessarily represent dead persons or even death -but can constitute a special type of materiality oscillating between nature and culture. For the new settlers of the Late Iron Age, the remains of the ancient dead encountered in the landscape, could, for instance, be understood as an affordance -a resource that, if not growing, at least is known to be found, and perhaps even cultivated on small hillocks. As a materiality, the remains of the ancient dead can be attributed powers and properties that need not to be restricted to ideas of individual agency beyond the grave (i.e. the will of the ghost of a dead person). Indeed, fragments of human bones are known to been used in magic during the Late Iron Age (e.g. Stoklund, Rasmussen & Brinch Madsen 2004;cf. Bill 2016). If the ancient dead and their graves were considered as a powerful materiality rather than representations of dead ancestors, the recently dead could possibly benefit from being buried close to them, as we find in the case studies. These arrangements can be discussed in terms of providing aid along the transformation towards (or in) the afterlife or as protection against the living. Combining two burials can also be interpreted in terms of merging bodies together, creating a "duovidual", a hybrid, or even a downright new entity (Fahlander 2013). The concept of merging entities is by no means foreign in Late Iron Age mythology; suffice it to mention the floating boundary between humans and animals (Price 2002;Hedeager 2011:81-96), or in the cremation as a means of unify animals, humans and artefacts by fire (Williams 2005;Fahlander 2014). As a materiality, the remains of the ancient dead may also been comprehended as something that demanded actions. Perhaps the ancient structures and human remains had to be cultivated by the new occupants during the Viking Age only to become later on, in the early medieval period, something that had to be disarmed or controlled? Such a perspective could make better sense of the Late Iron Age inclusions and imitations of older graves e.g. at Vittene and Bogla.
In the 11th and 12th centuries the time gap is significantly smaller between the old graves and the reuse. At Bogla, Lilla Ullevi and Valsta there may even be a continuity of burials from the Late Iron Age, with the exception of Broby Bro where pre-Roman and Roman graves were reused, while the Viking Age graves were left intact. The idea that the apparent ambivalence of the early medieval inhumation burials is the result of religious syncretism or an expression of ambiguity and reluctance to abandon the old ways does not satisfactorily cover the variability of the postburial actions during this period. What really stands out is the apparent lack of reverence for the remains of the ancient dead and their graves. The cremated bones of the old graves seem to simply have been dug up and left scattered among the filling. This, despite the much shorter time-span Fredrik Fahlander between the original graves and the new additions (100-400 years). Besides Broby Bro it is actually possible that the buried people in the reused graves at Bogla, Valsta, Vittene and Lilla Ullevi were known when the inhumations were added. Why some burials still were manipulated and altered is difficult to determine. One possibility is that the ideas concerning the materiality of the cremated dead for some reason changed from being something benign/powerful to something considered unpleasant or sinister. But this would not explain why some old burials were chosen for new burials instead of simply making new monuments. One reason for the radical difference in the view of the dead and their graves may at least partly be the different materiality between the ancient and the recent dead (i.e. cremated versus non-cremated). Building cremation burials over older cremations involves a different set of relations with the previous dead than burying complete bodies in coffins (cf. Appleby 2013). Burying corpses in coffins in (or on) older cremations requires a quite different material engagement. It is also evident that the materiality of the cremated dead is very different from that of the non-cremated (also in terms of longevity). Thus, one part in the change in attitude towards the remains of the old dead may be because the burial practices differ in a material and practical sense. The case in Lilla Ullevi, where an old inhumation grave was left intact when a later inhumation was superimposed, may indicate that the material form of the ancient dead indeed mattered.
The different ways in which the old graves were apprehended are difficult to grasp, and the reasons for the post-burial activities would most probably have varied over time and between regions. However, either if one wishes to emphasize social or ontological aspects, the long time span between the old and the new graves in the five case studies must have had consequences for how the materiality of the ancient dead was appropriated during the Late Iron Age and early medieval period. To emphasize ontological and material aspects, however, does not simply mean that we should swap a social view of the past for a less structured, mythical perspective, but emphasizes that the post-burial practices of the 9th to 12th centuries are a much more varied and complex affair that needs more elaborate interpretations which also involve ontological recategorization of familiar concepts.

CONCLUSION
In this text I have discussed post-burial activities from a bottom-up perspective in order to understand the varying relations to the ancient dead from an ontological and material point of view. Working from the bottom up is a way to limit the influence of preconceived notions in the study and to be open for the actual archaeological data in all its variability and inconsistencies. Judging from the five cases discussed here, the whole period from the 9th to the 12th centuries comprises a varied and complex set of post-burial actions. A common aspect of the period as a whole is the quite detailed knowledge of the inner structure of the old graves that could only be gained by some kind of investigations. For example, new burials were placed at the exact spot of the old ones even though their actual position is not always obvious from the outline of the graves' superstructures.
The period saw a change from cremation to inhumation, which seem to have altered the way in which the remains of the older dead were appropriated. At the beginning of the period (9th to 10th centuries AD), the remains of the ancient dead were respected -perhaps even "cultivated". They were sometimes incorporated in the new burials, but their superstructures were only rarely destroyed, adjusted or modified. In general, the superimpositions and inclusions of older graves were rather subtle. This attitude towards the original human remains changed in the early medieval period (11th to 12th centuries AD). Old graves were still reused and rebuilt, but with less concern for the old cremated bones. In some instances, quite drastic means were employed. For example, the destructive modifications of the Early Iron Age graves at Broby Bro (A7000 and A8000), Valsta (A1) and Bogla (A11429). It is however, important to point out that even though the medieval reuse is destructive, the people nonetheless still manifested an interest and intimate relations to the ancient dead. The difference could, at least partly, be due to the different materiality between a cremated and an inhumated body, but perhaps also because of the smaller time span between the old and the new dead. By the 11th and 12th centuries the old graves from the Early Iron Age may not have been considered that much different from the more recent Late Iron Age cremations.
Despite the complexity and variability, the post-burial actions of the Late Iron Age and early medieval period nonetheless constitute a continuous set of practices that must be considered a recurrent part of the general burial customs of the whole period. To understand the variability and seemingly contradictory attitudes towards the ancient dead. ontological and material aspects of the ancient dead are emphasized. The large time gap (500-1000 years) that separates the old graves and the later engagements is a key to understanding the postburial actions in the Late Iron Age and the early medieval period. This aspect alone makes it less likely that the old graves were remembered and understood as ancestors, predecessors, or perhaps even as dead individuals at all (although the remains probably were recognized as human-like), but as a particular materiality unsettled between nature and culture. It is argued that post-burial actions in this period are entangled in different sets of conceptions regarding the past and the materiality of the ancient dead. Considering the way in which the old graves were discovered by the new occupants in the Late Iron Age as something that was already present, it is suggested that the encounters with the ancient dead and their monuments may have fostered ideas of the cremated bone as a kind of materiality, or affordance, that may or may not have been charged with certain properties or powers. This perspective does not exclude practical and ideological use of the old burials, but it makes better sense of the keen interest in older burials and the variability in attitudes to the ancient dead during 9th to 12th centuries AD in southern Sweden.