The Dead and their Possessions: The Declining Agency of the Cadaver in Early Medieval Europe

Between the sixth and eighth centuries ad, the practice of furnished burial was widely abandoned in favour of a much more standardized, unfurnished rite. This article examines that transition by considering the personhood and agency of the corpse, the different ways bonds of possession can form between people and objects, and what happens to those bonds at death. By analysing changing grave good use across western Europe, combined with an in-depth analysis of the Alamannic cemetery of Pleidelsheim, and historical evidence for perceptions of the corpse, the author argues that the change in grave good use marks a fundamental change in the perception of corpses.


INTRODUCTION
Funerary archaeology has long been dominated by the now somewhat clichéd idea that 'the dead do not bury themselves' (Parker-Pearson, 1999: 3). Our interpretations of burials are usually from the perspective of the mourners; we recognize, for example, that the grave goods in a burial do not represent the identity of the deceased as they would have perceived it, but the ideal identity of the deceased as envisaged by the people who did the burying (Lucy, 1997: 155). From this perspective, the corpse itself is passive; it has no influence over the funerary rites at all. However, this arises from a relatively modern, Western perspective of instantaneous death, after which there is very little of the person left in the cadaver (Hertz, 1960: 28). While not an invalid approach, the idea that 'the dead do not bury themselves' has become overused (Williams, 2004: 264;Sayer, 2010: 62) and denies the presence of the deceased, along with the influence they can exert (Sofaer, 2006: 43). It implies that the corpse is an inanimate object, and that funerary rituals exist solely for the benefit of the living. Yet, the cadaver is the reason a grave is constructed in the first place (Sofaer, 2006: 19). The corpse can be active within the funeral, and its relations with the objects that accompany it have an influence on the nature of grave good deposition. Although there are many reasons for objects to be deposited in a grave, many of these rely on the concept of possession; it follows that, if a cadaver is capable of possession, it has retained some of its personhood and agency, and is capable of playing an active social role on death.
Although control over the funeral does indeed lie with the living, the way they behave is highly dependent on how they perceive the nature of the corpse; in other cultures, the soul can remain present in the corpse, and 'alive' until after certain funerary rites have been competed, which can take place over many years (Hertz, 1960: 34). We therefore need to consider the agency of the cadaver, and the influence it can have on funerary rituals. This article examines the extent to which a dead body can remain a locus of personhood and the impact this might have on grave good deposition, in order to understand a particular change in burial practices across early medieval western Europe.
A gradual decrease in grave good deposition in the seventh century AD was followed by an almost complete abandonment of furnished inhumations at the end of the seventh and beginning of the eighth century. Explanations for this change range from Christianization to increasingly stratified social hierarchies, but most scholars consider the funeral from the perspective of the mourner. Considering the role of the corpse in the funeral may provide an alternative explanation for the abandonment of grave goods. The results of a large-scale analysis of changing grave good use across 246 cemeteries in western Europe is presented here, and one site, Pleidelsheim (Baden-Württemberg) in the Alamannic region of Germany (Koch, 2001), is analysed in greater detail. The way in which grave good use changes at these sites can illustrate changing perceptions of the corpse.

THE NATURE OF THE CADAVER
In almost all societies, a corpse lies somewhere between a person and an object, and can fulfil both roles at once (Williams, 2004: 264;Nilsson Stutz, 2015: 3-4). In anthropology, there is a long history of viewing the corpse as occupying a liminal space, requiring the intervention of funerary rituals to help navigate this space (e.g. Hertz, 1960;van Gennep, 1960). Within modern archaeological studies, there is more of a tendency to classify skeletal remains as an object, a product of nature, and the subject of scientific studies (Sofaer, 2006: 40). This neglects the fact that, in most circumstances, the excavated skeleton is not the body that was originally buried, and the recently-dead corpse is likely to have held quite different connotations (Graham, 2015: 4). The corpse could be envisaged on a spectrum from living person to inanimate object. At death, the body begins to move from the former to the latter, but at a variable rate. The position of the newly dead body on this spectrum will influence the structure of the funeral ritual, the grave, and the objects within it.
There are important differences between a recently-dead body and a soon-to-bedead body; the lack of heat, muscle tone, expression, and speech all serve to create an object which, while similar to the living person, is fundamentally different (Sofaer, 2006: 68;Robb, 2013: 446). These differences may only be obvious on closer inspection, particularly in bodies which are not far either side of the point of death (Sofaer, 2006: 44;Nilsson Stutz & Tarlow, 2013: 6). The deceased body is 'simultaneously familiar and unfamiliar' (Mui, 2018: 34). A corpse cannot, on its own, conform to cultural norms as a living body could, although these can still be imposed on it through mortuary treatments (Nilsson Stutz, 2008: 23). But just because a body becomes more like an object at death, this does not mean it has been deprived of its agency (Devlin, 2015: 68). The concept of material agency has gained significant traction within archaeology (Robb, 2013: 448): the idea that objects can act as people is a point which has been made many times, but the notion that people can become objects has been less commonly aired, perhaps because of ethical considerations. Yet there is plenty of evidence for people becoming objects in contexts such as relic collecting, war trophies, or medical dissections (Sofaer, 2006: 63-64). Moreover, an object that was once a part of a living person has a special kind of power by virtue of having once been human. Although it has undergone a fundamental transformation, it is still invested with memories and emotional attachments (Hallam et al., 1999: 131), making it a 'mnemonic tool' of the person during life (Mui, 2018: 191). Rather than considering person and object to be two binary opposites, we should envisage them as fluid in nature, able to shift from one to another.
The extent to which the cadaver remains an active social person varies between societies, but it is rare for bodies to lose all personhood and there are many societies in which the deceased are believed to continue to interact with the living after death (Williams, 2004: 266). It is this transformation from one state to another which requires a funeral in the first place, a ritual that mediates the transition from social being to dead body (Nilsson Stutz, 2008: 22;Nilsson Stutz & Tarlow, 2013: 5). In many cultures, death is not recognized as the point at which metabolism ceases, but is a drawn-out process involving funerary rituals and the decay of the corpse. An example of this can be cited among the Indonesian Sa'dan Toraja people. On an individual's death, their body is kept in the family home, sometimes for many several years, and described as ill rather than dead. Death does not occur until the funerary ritual; no one can die unaided, and it is only through ritual practice that death is created (Tsintiilonis, 2000: 5).
Reconstructing such beliefs from the archaeological record is harder, but burial practices can indicate different responses to the body. For example, at the Mesolithic cemetery of Skateholm in southern Sweden, bodies were apparently buried rapidly, before decomposition could occur, and placed in positions that mimicked those of life. New graves rarely cut older ones, although in the few instances where this did occur, no effort was made to avoid disturbing skeletonized remains (Nilsson Stutz, 2008: 24). This might indicate that the body that was buried was still seen as belonging to the realm of the living, but, once decomposed, the individual in the grave had become separated from that realm, and hence bodily integrity no longer needed to be respected.
Contrasting examples come from the curation of human remains in many past societies, such as the retention of some cremated bone in Bronze Age Britain (Brück, 2017: 141), or the custom of retrieving and plastering skulls in Neolithic Turkey (Croucher, 2017: 201). Among the Chacapoya Inca, bodies were intentionally mummified, maintained with new wrappings, and presented with offerings of food and drink as part of an ancestor cult (Guillén, 2004: 153-54). These practices kept the deceased present in society through the medium of their physical remains, the cadaver providing the 'focal point … at which mortuary theatre could be anchored and enacted' (Mui, 2018: 208). When considering archaeological remains, we must therefore remember that a biologically dead person can still be socially very much alive. In the examples given above, the dead body, or part of it, acted as the focal point for a person, but this need not always be the case. With the biologically dead, it is easier for personhood to become detached from the body (Hallam et al., 1999: 2). The personhood of the deceased can be lodged where the final disposal of the remains took place, but it can also become located in inanimate objects related to the deceased, or in the consciousness of the mourners (Hallam et al., 1999: 149;Rugg, 2017: 211). We should, therefore, consider not just the extent to which the dead can remain socially alive, but also the extent to which that social persona resides in the corpse.
In an early medieval context, situations similar to those at the Mesolithic site of Skateholm, with rapid burial in life-like positions and little intercutting of graves, may be envisaged. The phenomenon of bed burials, though not a widespread rite, is perhaps a metaphor for sleep, suggesting a denial of death (Tarlow, 1999: 134-35;2013: 620). More commonly, positions where the body lay on its side with one or both hands raised to the face may suggest sleep, and children in particular tended to be buried in this way. In other instances, the placement of the body in the grave mimicked a living, upright position, for example holding a spear in a realistic posture (Mui, 2018: 139, 212, 206). This is not to suggest that the body was still considered alive; preparing the corpse for burial included stripping, washing, and redressing it, and it would have been impossible, in such close contact, to overlook the stark differences between the dead and living body (Mui, 2018: 203). Early medieval cremation burials provide another means of understanding how the corpse may have been viewed: cremations took place alongside inhumations, often in the same cemetery; the natural movements which occurred, and the sounds the corpse emitted as it burned, could have led to a belief that the corpse was still capable of movement and speech (Williams, 2004: 275).
There are hints, however, that attitudes changed from the eighth century onwards, and that, rather than being the locus of a socially active person, the body became more of an empty vessel. By the ninth century, graves were no longer carefully spaced to avoid each other; and new graves frequently cut older ones, suggesting that the preservation of earlier bodies was of little importance (Thompson, 2004: 102). This change coincides with the transition from furnished to unfurnished burial. The inclusion of objects in the grave, a practice current before the eighth century, could be seen as representing continued personhood after death. In order to explore this further, we must first consider the motivations for depositing objects in a grave, and the different relationships between a person, a body, and an object.

WHY DO PEOPLE USE GRAVE GOODS?
Grave good deposition is a common practice in many cultures but by no means a universal part of the funerary rite. The idea that grave goods were the personal possessions of the deceased, deposited for use in afterlife, was a corner-stone of antiquarian understandings, but has fallen out of favour as being overly simplistic since Ucko's (1969) ethnographic work, which revealed many other reasons for depositing (or not depositing) objects in graves. One of the earliest theories relating to the deposition of grave goods in an early medieval context related to the notions of Heergewäte and Gerade, i.e. male and female possessions, respectively, which by law could not be passed on to descendants but had to remain the property of the deceased, and so be deposited in the grave (Reinecke, 1925: 104). While this was an influential theory for some time, especially with regards to Frankish burials, a more accurate reading of the law codes in question has now largely discredited it (Effros, 2002: 25). Härke (2014) lists eleven possible motivations for the deposition of grave goods in early medieval funerary rites, such as inalienable property and equipment for the afterlife, but also as indicators of identity, conspicuous consumption, and disposal of polluted items. It has long been argued that the objects placed in the grave were not direct displays of an individual's identity during life, but that the grave goods were specifically selected to display an idealized social identity, and hence to reinforce such social identities within societies (e.g. Williams & Sayer, 2009: 3). While status cannot directly be inferred from the quantity and quality of objects in a grave (Pader, 1982: 53;Parker-Pearson, 1999: 94), conspicuous consumption in a funeral can be a way for the heirs of the deceased to display and cement their role (Fahlander & Oestigaard, 2008: 10).
The concept of conspicuous consumption focuses on the role grave goods played for the people carrying out the burial, rather than the deceased themselves. It is undoubtedly over-simplistic to view every grave good as a possession, but this does not mean the idea of personal possession should be abandoned altogether. The influence of the dead can be re-emphasized by considering the relationship between the deceased and their objects, mediated through the concept of possession. Of course, not all grave goods were the possessions of the deceased during life; some may have been gifts from the living to the dead. Gift-giving was an important aspect of the funeral, as it helped mark the deceased's new role and change in social status. King (2004) argues that early medieval grave goods were largely the product of this gift-giving network. Objects are more likely to have been gifts if deposited outside a coffin, higher in the grave fill, or in unusual positions, such as dress accessories not on the body (King, 2004: 220-21). Objects outside the coffin may also have been related to funerary ceremonies, rather than being possessions per se (James, 1988: 139). Even objects which are directly associated with the body could have been given while dressing the corpse (King, 2004: 219). In many ways, therefore, the distinction between gift and possession is meaningless, because once a gift has been made, it becomes the possession of the recipient and, thus, requires a similar level of personhood to reside with the corpse.
We also need to probe deeper into what it means to 'possess' an object, especially after death. The likelihood of possessions being deposited in graves depends very much on the nature of the connection between an object and its owner. Some objects were scarcely owned; with little meaningful connection between the object and its owner, such an object was quite easy to give away, in life or on death. Such negligible possessions were rarely deposited in graves (Klevnäs, 2015: 179). They were unlikely to have had much worth as a gift. Other objects were more meaningfully owned. The concept of inalienable possession was developed by Mauss (1990Mauss ( [1925) in the context of Pacific giftexchange networks; it refers to an inalienable connection between owner and an object, which added prestige to the object. It was possible to break these connections, but only with difficulty, and it was more common for these objects to remain within an 'owning' group, often based around kinship (Klevnäs, 2015: 170). Inalienable objects were, therefore, highly likely to be deposited in graves or, if not, then passed on to descendants. Several Icelandic sagas emphasize the deceased person's ownership of their grave goods and describe special rituals designed to break those bonds of ownership (Klevnäs, 2016: 470).
Finally, inalienable possessions should be distinguished from inseparable possessions, which were items so closely entwined with the body of its owner that there was no choice but to bury the objects with it (Harper, 2012: 56). If we accept that the boundaries of the body are not determined by the skin, then objects can be considered a part of the person, following Strathern's concept of the partible person (i.e. a person composed of relations) in Melanesian society (Strathern, 1988: 178). In a modern context, eyeglasses can be considered such an object (Harper, 2012: 49-50). Retaining these objects in society may be a way of keeping part of the deceased person present in the world of the living (Fowler, 2013); but, more often, they are not separated from the body, even in death (Klevnäs, 2015: 175). With this type of object, the connection is not with the person, but with the body; and this connection was maintained when the living body became a cadaver, regardless of how much personhood the corpse retained.
The significance of different types of possessions in an early medieval context can be approached through examining grave re-opening practices. The interpretation of this activity varies. Grave reopening may be interpreted as an act of violence against the deceased and their kin (Klevnäs, 2013: 90), but it could also have been an acceptable part of mortuary rituals to retrieve objects that had served their purpose in the initial funerary rite (Aspöck, 2011: 313) or that could act as something like relics, as a way of incorporating the deceased into the realm of the living (van Haperen, 2013: 91).
Whatever the motivations, the nature of the relationship between the deceased and the objects placed in their grave can be determined by examining the types of objects which were commonly removed from, and left in, re-opened graves. The objects which were most frequently removed were swords and brooches, particularly in graves in Kent (Klevnäs, 2013: 68-71), though they also form part of a wider repertoire of objects commonly removed from Frankish graves (Noterman, 2016: 416). Swords in the early medieval world could be described as almost like people, with names, memories, and associations with their previous owners, all of which made their transfer significant (Klevnäs, 2015: 173-74;Brunning, 2017: 409;Sayer et al., 2019). Brooches may have fulfilled a similar role to swords in the feminine sphere of exchange (Klevnäs, 2015: 174-75). Other possessions, such as knives, were, however, left in place in reopened graves (Klevnäs, 2013: 89-90;Noterman, 2016: 416). Knives were personal objects, which would have been carried and used almost every day, their blades and handles worn down by their owners' habits of use, and thus difficult for another person to use (Klevnäs, 2015: 176). They were a part of the body, not just the person.
The objects found in graves are, therefore, most likely to have fallen into the category of inalienable or inseparable possessions. Negligible possessions, those which lacked the meaningful connections of inalienable or inseparable objects, had little value as gifts, and their relative lack of connection with the deceased suggests that they may not have been the first choice within a limited repertoire of objects chosen to be deposited in graves. The use of inalienable possessions in graves implies some sort of continued possessive agency on behalf of the corpse. The same is not true of inseparable possessions; they remained with the body because they were a part of it, not because they were owned.

EARLY MEDIEVAL BURIAL RITES
The almost complete cessation of grave good deposition during the seventh and eighth centuries is a well-documented phenomenon across western Europe (e.g. Young, 1975;Halsall, 1995;Geake, 1997;Theune, 1999;Effros, 2003). Although some objects continued to be deposited in graves after the eighth century, this was minimal: only 2 per cent of later medieval British burials contained objects which can be classed as grave goods (Gilchrist, 2008: 124). Corrochano and Soulat (2017) have shown that, while some grave good deposition continued in France after the eighth century, this was no more than one or two objects per cemetery, far less than the widespread use of grave goods of the fifth to seventh centuries.
Many reasons have been put forward for the abandonment of grave goods. A simple model of Christianization has now been largely rejected as an explanation; it coincides with the abandonment of grave goods in England, but not in large parts of the Frankish world, where burial with grave goods continued far later, in some cases within churches (James, 1988: 139). Other explanations emphasize the cultural rather than religious impact of the Church (e.g. Geake, 1997), the potential for changing interpretations of the afterlife (Effros, 2003: 88), the focus on investment in the Christian Mass or above-ground monuments rather than the grave itself (Effros, 2002: 205-06), a shift towards the curation of objects rather than burial as part of changing strategies of commemoration of the dead (Williams, 2010: 33-34), or finally the religious reforms of Archbishop Theodore of Tarsus in Canterbury in the AD 680s (Hines & Bayliss, 2013: 553). The latter, at least, is too regionally specific to explain the transition as a whole. Other theories reject Christianity altogether, and instead discuss changes in craft production which could have affected the range and quantity of goods available for burial (Hines, 2017: 19), and the establishment of trading sites providing an alternative mechanism for the recycling of artefacts that would have otherwise gone into the grave (Boddington, 1990: 189).
Perhaps the most popular theory links grave good use to social stratification and conspicuous consumption. With the deposition of elaborate grave goods being limited to a select few in the seventh century, furnished burial became a method by which status was signalled and reinforced (Scull, 2011: 852;2015: 79), though changes in the levels of investment may reflect ideology more than actual social hierarchies (Carver, 2011: 846). The position of these elites was unstable, making lavish funerary displays necessary. As social hierarchies became increasingly stable, and status more entrenched in family lineages, the need to display this status through burial diminished, and artefact use eventually ceased altogether (Halsall, 2010: 175). This theory has its merits, but considering funerary practices purely through the lens of status display is insufficient, as maintaining or overthrowing the existing power structures is rarely the primary aim of mourners, even if that is an unintended consequence; rather, the immediate concerns are more emotional and less rational (Tarlow, 1999: 20-23). It also views funerary change as imposed on the majority by the elite, whose increasing wealth meant that there were fewer resources to bury with the majority of the population. This denies the agency of those lower down the social hierarchy to make their own decisions about how to bury their dead in a way that is meaningful to them.

CHANGING GRAVE GOOD USE ACROSS EUROPE
In order to investigate changing patterns of grave good use in the sixth to eighth centuries, over 33,000 graves from 246 cemeteries in western Europe were analysed using relative kernel density plots (Figures 1 & 2,   good deposition relative to the density of cemeteries per region, which prevents areas of intense archaeological investigation from being over-represented. Cemeteries included here were those with a full catalogue of graves, as well as chronological information about the cemeteries' period of use. One of the most notable trends is the variation in rates of decline of different object types, particularly the persistence of personal accessories far longer than other types of artefacts (see Table S1 for grave good classification). Figure 1 illustrates the levels of grave good use in early medieval Europe from AD 500 to 800. The key periods in England were between AD 550 and 600, when there was a marked contraction in the intensity of grave good use, and the end of the seventh century, when furnished burial almost entirely ceased. At the same time, there was a contraction in the high-density areas of grave good use in continental Europe, not as marked as in England but still more noticeable than anything before that. After around AD 700, the use of unfurnished burial rapidly accelerated, so that by AD 750, only a few furnished cemeteries along the Lower Rhine remained in use, and by the end of the eighth century, grave good use had become negligible.
In contrast, the deposition of personal accessories, the most common being knives, does not follow this pattern, remaining far more consistently used compared to other types (Figure 2). There was no decrease in personal accessories in England in the fifth and sixth centuries, nor was there any notable decrease before the late seventh century. It was only after that point that a decrease is discernible in England, when personal accessories vanished from Northumbria and became less common in Wessex. There was still a distinct concentration of personal accessories present in graves in East Anglia and Kent at the start of the eighth century, which is not evident for any other artefact type (Brownlee, 2019: 79). Although personal accessories began to decrease in continental Europe after this, it was at a slower rate compared to the overall numbers. In other words, personal accessories remained consistently used throughout the period of furnished inhumation, and it was only at the very end that they, too, began to decline.

CHANGING GRAVE GOOD USE AT PLEIDELSHEIM
The kernel density maps illustrate change at the broadest of scales, documenting primarily change in cemeteries going in and out of use. In order to illustrate these changes in greater detail, Pleidelsheim (Baden-Württemberg), an Alamannic cemetery containing 264 graves of the mid-fifth to late seventh century (Figure 3), was chosen for analysis. Its excavation report contained an in-depth chronological study (Koch, 2001), used as the standard for the region (Hines & Bayliss, 2013: 480). This dated 165 graves to one of ten phases through seriation ( Figure 4). Analysis was carried out on all 165 dated graves, as well as the subset of 123 graves not affected by re-opening.
Although grave re-opening should be seen as part of early medieval mortuary rituals rather than an impediment to their study (Aspöck, 2011: 299-300), I have focused primarily on the undisturbed graves because of the risk that disturbance will have affected the numbers and positions of objects in the grave, obscuring the intent of the people who originally created the burial. In any case, similar trends were observed in the analysis of both disturbed and undisturbed graves, although trends were stronger when all graves were included.
Following an increase from the start of the cemetery's use to the mid-sixth century, the number of grave goods decreased steadily from an average of almost five objects per grave from c. AD 530-600 to an average of 2.5 objects per grave in the last phase of the cemetery's use in AD 650-670 ( Figure 5). Many types of objects placed in the grave followed this pattern, declining in the later phases ( Figure 6). There were, however, a few notable exceptions. Jewellery, weapons, and personal accessories remained consistently present (Figure 7). Some of the rarer categories of object, such as amulets, animal remains, and coins, also continued to be deposited at the same low levels throughout the cemetery's use.
Following the work of Klevnäs on different types of possessions, I take jewellery and weapons to have been inalienable possessions, while most personal accessories were inseparable possessions. The fact that these types of objects continued to be placed consistently in graves in a period where grave good use was otherwise   Table S2 for results of statistical analysis.
declining could suggest a change in the way the corpse was perceived. A gradual shift may have occurred from perceiving the corpse as a person, albeit one in a different state, to perceiving the corpse as an object, an empty shell from which the majority of personhood has vanished. This change would have reduced the ability of Figure 6. Categories of grave goods at Pleidelsheim that show a statistically significant decrease over time. Polynomial trendlines order 3, R 2 = 0.672 (dress accessories), 0.458 (fittings), 0.801 (toilet accessories), 0.915 (vessels), 0.381 (tools). See Table S2 for results of statistical analysis.  Table S2 for results of statistical analysis.

418
European Journal of Archaeology 23 (3) 2020 the corpse to maintain its relationships with its possessions, and hence negligible possessions, such as toilet accessories, vessels, or tools, were less frequently placed in graves. These are objects in the purest sense, with very little connection to the individual, thus the corpse as an object has no use for them. However, the relationships between person and inalienable possessions could have persisted for far longer. The longest of these relationships would have been that with inseparable possessions, because they were associated not with the person, but with the body itself. The level of personhood attributed to a corpse does not affect the role played by this type of object, which continued to be placed in graves at almost the same levels as before.
A change in the physical relationships between the objects placed in the grave and the corpse was also observed at Pleidelsheim. The relationship between the body and the objects accompanying it is often an aspect of burial to which greater attention could be paid. Ekengren (2013: 183-88), Williams (2006: 74-75), and Sayer and colleagues (2019) have shown how this approach enhances our understanding of the role different objects played in mortuary rituals, and how the cadaver was perceived. Figure 6 shows the locations of objects in the graves of Pleidelsheim: either 'on the body' (e.g. objects 1, 9, and 10 on Figure 3), 'next to the body' (e.g. objects 3, 4, and 5 on Figure 3), or 'in the grave' (e.g. objects 6, 17, and 18 on Figure 3). There was a clear shift in emphasis away from the body. At the start of Pleidelsheim's use, over 80 per cent of objects were placed directly on the corpse. By the end of the seventh century, this had fallen to 40 per cent, and objects were most likely to be placed next to the body, instead of on it (Figure 8). This is also true for each individual type of object, the only exception being jewellery. Even objects such as dress accessories ( Figure 9) and personal accessories ( Figure 10) were more frequently placed next to the body, Figure 8. Locations of objects in the grave relative to the body. Polynomial trendlines order 3 R 2 = 0.839 (on the body), 0.673 (next to the body), 0.693 (in the grave). See Table S3 for results of statistical analysis.
or elsewhere in the grave in the later periods of the cemetery's use, and less frequently placed in direct contact with the body. This was particularly true of personal accessories: in the earliest graves at Pleidelsheim, they were overwhelmingly placed on the body, but this changed around the mid-sixth century, when personal accessories were more likely to be placed next to the body. While the change in dress accessories was not as extreme, objects still decreased in the frequency Figure 9. Locations of dress accessories in the grave relative to the body. Polynomial trendlines order 3, R 2 = 0.877 (on the body), 0.685 (next to the body), 0.593 (in the grave). See Table S3 for results of statistical analysis. Figure 10. Locations of personal accessories in the grave relative to the body. Polynomial trendlines order 3, R 2 = 0.637 (on the body), 0.714 (next to the body), 0.1032 (in the grave). See Table S3 for results of statistical analysis.
with which they were placed on the body from the mid-to late-sixth century onwards.
This is a change that has, in other circumstances, been interpreted as a change in funerary rituals, related to a change in the mnemonic role of grave goods (Williams, 2006: 77). However, the change could also provide further indications of a reduction in the personhood of the corpse: although it was still important to keep these inseparable objects with the body, it became less necessary, over the course of the seventh century, to lay out the grave as resembling the person's appearance during life, with dress and personal objects related to the physical body in the same way a living person would have carried them. Although the links between body and inseparable object were still present enough in the mind of the mourners for them to include then in the burial, they were not strong enough for the links between the physical body and the object to be replicated as they would have been during life.

THE INFLUENCE OF CHRISTIANITY
One potential explanation for the abandonment of furnished burial suggested by Effros (2003: 88), but which has received little further attention, is the idea of changing interpretations of the afterlife. Effros argues that the abandonment of grave goods occurred not because Christianity bans grave goods, but because grave goods are unnecessary in a Christian understanding of the afterlife. The chief teaching of the Church was that life persisted after death, but the exact way in which life persisted was heavily debated, and the seventh century was a crucial turning point in that debate (Caciola, 2016: 7). The few documentary sources regarding perceptions of the corpse were written from a Christian perspective, making the influence of Christianity worth considering.
Early Christian beliefs about the body at death have sometimes been characterized as a simple dichotomy between body and soul. But the reality is more complex and it was debated extensively throughout Christianity's history (Harris & Robb, 2013: 133). The body was clearly important in Christian theology, since it was believed to be resurrected and reunited with the soul on Judgement Day. The exact form this resurrection would take, however, and thus the importance of the body, was much discussed (Thompson, 2004, 196-97;Harris & Robb, 2013: 147).
The separation of body and soul on death was emphasized as early as the third century AD, in the writings of theologians such as Tertullian, who stated that 'not even a little bit of the soul can possibly remain inside a [dead] body, which is itself destined to disappear' (Caciola, 2016: 39-40). That such beliefs persisted can be seen in Visigothic laws: any objects stolen from a grave were to be returned to the heirs, not the grave itself, indicating that the ability of the dead to own objects did not persist beyond the burial itself (Effros, 2002: 51). An anonymous fifthcentury sermon berated the custom of offering food and drink to the dead: 'they confer drink and wine over the tombs of the dead, as if the carnal souls, having exited from the bodies, required drinks' (Effros, 2002: 75). The dead body in a Christian perspective was not the person it had been during life; the locus of personhood, the soul, had become disembodied. The corpse did not require sustenance in the way a person would, it could not own objects and had no social capacity. The body was still important for its potential to be resurrected, but it was not the person it had been during life, and since God had the ability to reconstitute the body, its preservation was not important.
Beyond this official narrative, however, there are hints that among the general population, the body continued to hold importance beyond its potential for resurrection. Tertullian's writings were not designed to voice a widely held consensus but were a polemic against contemporary belief that something of the soul remained with the cadaver after death (Caciola, 2016: 39). The need for sermons condemning food offerings suggests something similar. The seventh century was a pivotal point in Christian perceptions of the afterlife, when much more emphasis began to be placed on how a soul was to reach heaven (Brown, 2006: 257). Late sixth-and seventh-century writers, such as Gregory the Great and Isidore of Seville, gave more weight to the transformation of the corpse through funerary rituals. Isidore suggested that an unburied corpse was fundamentally different from a buried one, and that the funeral was necessary to fully separate the soul from the world of the living (Caciola, 2016: 60-61). This emphasis on the funeral was a new development, drawing greater attention than previously to the way the corpse was buried, and the importance of the correct rites for achieving afterlife. A belief in the intermediary stage of Purgatory was not officially approved doctrine until the late twelfth century, but became more prominent from the seventh century onwards (Thompson, 2004: 6). Gregory the Great's writings emphasized judgement immediately following death, not just at the final resurrection (Effros, 2002: 162), and the whole of Book IV of his Dialogues was dedicated to discussing the fate of the immortal soul. He restates forcefully the separation of the soul from the body on death. Thus, at the same time as grave good use was declining, Church writers were becoming increasingly vocal about the fate of the soul, and how best to provide for its afterlife.
The extent to which abstract theological debates about the nature of resurrection and the fate of the soul would have affected the majority of the population is debatable, but the fact that this change occurred slowly suggests that only a small section of the population was aware of such ideas. The way people understand death is often full of inconsistencies, with people able to hold several competing beliefs about the fate of the soul at once. This can combine official theology with superstition and folklore, creating a mix of beliefs which are not necessarily internally consistent (Tarlow, 1999: 47, 103). However, Brown (2006, 265) suggests that 'By the year A.D. 700, Western Christianity had … a highly individualised notion of the soul and a lively concern for its fate in the afterlife'. This increasing awareness of the Christian afterlife, and the means by which it could be achieved may well have influenced perceptions of the corpse, and thus the use of grave goods.

CONCLUSIONS
The abandonment of grave goods in the seventh century is a phenomenon seen across most of western Europe, in areas of varying initial burial practices. One thing that remains consistent across areas of furnished burial is the frequency with which personal accessories, such as knives, continued to be placed in graves, even as overall numbers of grave goods declined. The changes observed in grave good use at Pleidelsheim confirm this, and indicate an increasing tendency to disassociate grave goods from the body. This suggests a change in the way the corpse was perceived, from an embodied, active social agent, to an empty vessel which had ceased to be the locus of personhood. This later body was not capable of maintaining possessive relationships with objects that would previously have been placed in the grave, nor was it necessary to arrange it in the grave in a manner resembling its appearance during life. This transition was gradual, with the objects most intimately associated with the body continuing to be deposited after possessive connections with other objects had faded.
This interpretation would not be possible without a reconsideration of the role played by the corpse in the funeral. Even if the 'dead do not bury themselves', the actions of the mourners are shaped by what they thought of the dead body and its relationship with the material world. In societies where social death does not occur at the same time as biological death, the dead remain active participants in their own funerals, and this will undoubtedly have influenced the decisions about which, and how many, objects to include alongside the corpse, and how to relate them to the body.
In seeking to understand funerary rites, it is necessary to break down some of the traditional dichotomies between dead and living body, between objects and people. Instead, we should view a spectrum, with the living person at one end and inanimate objects at the other. Cadavers can exist anywhere along this spectrum, as can objects, which can become part of a person through association. The funerary practices initiated by a living body in the process of becoming an inanimate object depend at least in part on where on that spectrum a culture perceives a recentlydead body to exist. It is time to put the deceased back at the centre of their burials.