Depositing objects in the Viking Age

The first arguments for claiming the existence of a Viking Age concept of particular artefacts as being animated built upon the links between the deposition of artefacts in wetlands in Scandinavia and the role of artefacts of the same type in the Old Norse written sources. It was previously assumed that the act of depositing artefacts in Scandinavian wetlands ceased in the 6th century A.D. simultaneously with the increase of dryland depositions in prominent settlement structures (see Fabech Reference Fabech, Fabech and Ringtved1999, with references). A number of scholars have subsequently pointed out that the number of wetland depositions may have been very low in the 6th–8th centuries (as only a few finds from this period have been made in total), but that they increased significantly in the Viking Age. These finds include weapons, but also tools, keys, whetstones, coins, jewellery and other objects of precious metal, whole and fragmented, that were deposited in lakes, bogs and watercourses at places with standing water (Zachrisson Reference Zachrisson1998; Hedeager Reference Hedeager, Gustafsson and Karlsson1999; Reference Hedeager, Rolfsen and Stylegar2003; Andrén Reference Andrén, Jennbert, Andrén and Raudvere2002; Lund Reference Lund2004; Reference Lund2005; Reference Lund2008; Reference Lund2009; Pedersen Reference Pedersen and Lund2004; Ryste Reference Ryste2005). The majority of the south Scandinavian depositions consist of only one to three objects, but a few locations such as Råbelöv Sjö in Scania, Tissø on Zealand and Gudingsåkrarna on Gotland contained respectively 35, 50 and around 500 objects (Müller-Wille Reference Müller-Wille1984; Jørgensen Reference Jørgensen, Raudvere, Andrén and Jennbert2002; Lund Reference Lund2009, 77–79). Weapons and jewellery make up the largest proportions of these finds. Noticeably, far higher-quality and unique objects are found in wetlands compared to contemporary graves (Lund Reference Lund2009, 108, 272; Reference Lund, Carver, Sanmark and Semple2010). The swords from Dybäck and from Oppmanna Sjö serve as examples. The sword from Dybäck of Jan Petersen's type Z was found in a bog. The handle was made of gilded silver and decorated with bird figures in a style typical of contemporary English Anglo-Saxon material (Strömberg Reference Strömberg1961). The sword from Oppmanna Sjö of Jan Petersen's type D is made of iron and bronze with elaborate decoration. Half of the very small group of finds with this rare type of decoration are found in Britain (Petersen Reference Petersen1919; Rydbeck Reference Rydbeck1932, 32). The decoration on the hilt of the sword, the so-called véttrim, includes bearded masks and stylized animal heads (Strömberg Reference Strömberg1961, 72). These two swords are unique weapons of high-quality metalwork. The reasons why these precious objects were discarded in bogs may be found in their social biographies.

A biographical perspective

An approach focusing on the social biography of the artefacts deposited in wetlands provides insights into the changing meanings and social roles, and the social identity, that an artefact gained and acquired throughout its social life (Appadurai Reference Appadurai and Appadurai1986; Kopytoff Reference Kopytoff and Appadurai1986; Helms Reference Helms1993; Gosden and Marshall Reference Gosden and Marshall1999; Gosden Reference Gosden2005). By possessing an object, a social relationship is created between the person and the artefact. In this process, the artefact becomes more than its own materiality; it may come to possess the qualities of a social agent (Weiner Reference Weiner1985, 212; Gosden and Marshall Reference Gosden and Marshall1999, 120).

The choice of artefacts for deposition was hardly accidental. The two swords were presumably produced abroad and thus the result of plunder, gift giving and/or trade. They were owned, used as weapons or for display, and finally deposited in wetlands in Scania, providing the two artefacts with long and complex biographies. The swords are among the most exquisite of their kind in the Viking Age, and such finds hardly ever appear in Viking Age grave contexts. In Scania and Denmark, where the phenomenon of wetland depositions has been most thoroughly explored, the high number and the high quality of weapons found in wetlands are striking compared to the relatively low number and lower quality of weapons in grave contexts (Lund Reference Lund2004; Pedersen Reference Pedersen and Nicolle2002; Reference Pedersen and Lund2004; Strömberg Reference Strömberg1961; Svanberg Reference Svanberg2003 for overviews and comparison).

Using the Old Norse written sources in a study of the Viking Age

In the middle of the 20th century, a hypercritical position towards using Old Norse sources in studying Viking Age religion peaked. At present, a renewed trust in using the Eddic and Skaldic poetry in research on Old Norse paganism and ritual actions can, however, be found in philology as well as in the history of religion (Harris Reference Harris, Lindow and Clover1985, 95–96; Clunies Ross Reference Clunies Ross1994, 13–20; Schjødt Reference Schjødt2004, 109–113; Steinsland Reference Steinsland2005, 37–38). A remarkable change can be observed in the last 15 years in Viking Age archaeology towards what could be termed a historical archaeological approach. This highly text-dependent, innovative Late Iron Age and Viking Age archaeology is present in a number of research environments in Scandinavia (Price Reference Price, Mortensen and Arge2005, 379). The interplay between objects and texts is in focus in the methodological reflections. Particular written sources in Old Norse, such as the Eddic poetry, the Skaldic poetry and the Icelandic sagas, are again being used in archaeological analyses, where these written sources are used analogically with the archaeological evidence. These texts were written in the high medieval period, but the narratives most probably took place in Viking Age Scandinavia and parts of the narratives may have been preserved in oral form. The potential of these new approaches has, for instance, been demonstrated in the analyses of connections between written (although the poetry was originally oral) and material metaphors of the Late Iron Age and Viking Age (Andrén Reference Andrén2000; Herschend Reference Herschend2001; Price Reference Price2002; Domeij Reference Domeij2004; Hedeager Reference Hedeager, Andrén, Jennbert and Raudvere2004; Domeij Lundborg Reference Domeij Lundborg, Andrén, Jennbert and Raudvere2006). While maintaining a source-critical position, it is possible for an archaeologist to use written sources with content that can be argued as being older than the time it was written down. An archaeologist will by definition ask different questions of a text than a historian, a philologist, or a historian of religion. For instance, the homogeneous presentation of paganism, as presented in the written sources, may be challenged by the archaeological finds (Price Reference Price, Mortensen and Arge2005, 378–79).

What is in a name?

Janet Hoskins states that ‘things tell the story of people's lives’ (Hoskins Reference Hoskins1998), but based on the Old Norse written sources it could likewise be said that in Viking Age Scandinavia people were telling the stories of things’ lives. In these texts, we hear of swords and pieces of jewellery being acquired, passed on as gifts or as heirlooms. Weapons could bear names, and were described as having opinions (Wever Reference Wever1961; Davidson Reference Davidson1962, 171–73; Drachmann Reference Drachmann1967; Reference Drachmann1969; Idsøe Reference Idsøe2004; Lund Reference Lund, Carver, Sanmark and Semple2010, 50–51; see also Pearce Reference Pearce2013). Based on Keith Thomas's work on magic and popular beliefs, Roberta Gilchrist points out that naming objects and using them in ritual contexts included a process of transformation, as the mere utterance of the name in a ritual context could change the character of the material object (Thomas Reference Thomas1991; Gilchrist Reference Gilchrist2012, 227). In the Old Norse sources, even where swords were broken and forged into new weapons, they are described in a way that indicates that they were believed to have their own personality, which survived through this transformation (Davidson Reference Davidson1962,171–73). These objects with complex social biographies are presented as agents, as personified or animated artefacts. In this sense they are presented in manners suggesting that they were recognized as possessing personhood (Fowler Reference Fowler2004; Reference Fowler, Beaudry and Hicks2010; Reference Fowler and Insoll2011). If we return to the high-quality weapons and pieces of jewellery found as wetland depositions, these finds indicate that artefacts with complex biographies were being handled differently to other artefact types. Thus they appear to have been categorized in a different manner to artefacts in general from this period. The artefact types found in wetlands are objects of the very types which in the Old Norse written sources are presented as animated.

Reopening burial mounds

Another phenomenon that gives us insight into these perspectives on specific artefacts is the reopening of burial mounds during the Viking Age. Such actions are primarily documented during the excavation of the large mounds, for instance Gokstad, Oseberg and several of the large mounds of Borre, all in Vestfold in Norway; in Grønhaug from Karmøy and several of the large mounds of Trøndelag, also in Norway; in the large mound in Årby, Uppland, in Sweden; and in Jelling in Jutland and Ladby in Funen, in Denmark. This type of reopening can, however, also be identified in relation to smaller grave mounds, such as Gulli in Vestfold (Brøgger Reference Brøgger1945; Capelle Reference Capelle, Jankuhn, Nehlsen and Roth1978; Krogh Reference Krogh1982; Brendalsmo and Røthe Reference Brendalsmo and Røthe1992; Sørensen Reference Sørensen2001; Gjerpe Reference Gjerpe and Gjerpe2005a). Viking Age graves did not generally contain large amounts of precious metal. The vast majority of jewellery and coins in precious metal from the Viking Age have been found deposited in hoards (Skovmand Reference Skovmand1942; Hedeager Reference Hedeager, Gustafsson and Karlsson1999; Kilger Reference Kilger, Skre and Pilø2008; Myrberg Reference Myrberg, Danielsson, Gustin, Larsson, Myrberg and Thedéen2009a; Reference Myrberg, Kaenel and Kemmers2009b). Cenotaphs make up a considerable part of the mounds without traces of reopening, whereas none of the reopened burial mounds were cenotaphs, indicating that the content of the grave was known to the intruders (Brøgger Reference Brøgger1945; Brendalsmo and Røthe Reference Brendalsmo and Røthe1992, 84–85). As many valuable objects such as bronze cauldrons, glass beakers and pearls were not removed from the graves, the reopening of the graves cannot be interpreted as mere treasure hunting (Brøgger Reference Brøgger1945, 1–9; Myhre Reference Myhre, Hansen and Bjerva1994, 75).

Several factors indicate that the reopening took place during the Viking Age. In most cases the mound was opened from the side, part of the furniture was destroyed by chopping with an axe, and the deceased was removed, but some of the bones were found spread in the trench made for the reopening, which points towards the deceased being drawn out of the mound in a partly decayed condition (Brøgger Reference Brøgger1945; Brendalsmo and Røthe Reference Brendalsmo and Røthe1992; Myhre Reference Myhre, Hansen and Bjerva1994). This suggests that the intruder knew where the deceased's body lay. Neither the body nor the furniture was fully decayed. Rather, the wood in the grave was so fresh that chopping it with the axe destroyed the timber. A radiocarbon date from the stratigraphic layer from the plunder in Borre, Grave 7, shows that the reopening took place between the late 9th and the early 11th century, most likely in the earliest part of the period, and thus within the period in which the cemetery was still in use. In conjunction with estimations for the speed of deposition in the gytje layer from the ring ditch surrounding one of the mounds,Footnote ¹ pollen analyses indicate that the reopening happened less than 50 years after the burial (Brendalsmo and Røthe Reference Brendalsmo and Røthe1992, 88; Myhre Reference Myhre, Hansen and Bjerva1994, 71; Gansum and Risan Reference Gansum and Risan1999, 71). These factors indicate that reopening took place one or two generations after the construction of the burials (Brendalsmo and Røthe Reference Brendalsmo and Røthe1992, 87–88). Jan Bill and Aoife Daly have radiocarbon-dated the spades used in the reopening of the Viking Age burial mounds Oseberg and Gokstad in Vestfold. Due to the lack of sapwood, the spades can only be given a terminus post quem. The dates fall into two groups, from after 801 A.D. to after 845 A.D. and from 898 A.D. to after 953 A.D. Based on the dendrochronological dates and the fact that the result may be influenced by the different ways the individual spades were cut from the timber on production, they state that both ship graves were reopened in the late Viking Age, and most likely during the second half of the 10th century (Bill and Daly Reference Bill and Daly2012, 812–15).

If the reopenings are understood as ritualized actions, they may also be understood as the last phase in a long and complex series of burial rituals (Brendalsmo and Røthe Reference Brendalsmo and Røthe1992). In almost all of the graves that were reopened, the body of the deceased has been moved, partially or totally, from its original location. The bodies thus appear to have been one of the targets of the reopening (Brendalsmo and Røthe Reference Brendalsmo and Røthe1992, 89; Gjerpe Reference Gjerpe and Gjerpe2005a). By focusing on the social context of the reopening of Merovingian graves, Christoph Kümmel has argued that the reopenings can be characterized as representations of either friendly, hostile or indifferent feelings towards the buried (Kümmel Reference Kümmel2009). Building on Kümmel's theory of hostile reopening as an indication of intentional destruction for political reasons, Jan Bill and Aoife Daly have interpreted the reopening of these sites as the result of the Danish king Harald Bluetooth's campaign in the Viken region, where the reopenings may be understood as a means of desecrating the local elite group, performed a century after the burial mounds were built (Bill and Daly Reference Bill and Daly2012).

A possible division in Viking Age research between scholarship focusing on the social, economic and political aspects of the period (see, for instance, Skre Reference Skre1996; Jørgensen Reference Jørgensen, Pestell and Ulmschneider2003; Sindbæk Reference Sindbæk2007; Skre Reference Skre and Skre2007; Pedersen Reference Pedersen2010) and research emphasizing mentality, cognition and religion (see, for instance, Price Reference Price2002; Solli Reference Solli2002; Andrén, Jennbert and Raudvere Reference Andrén, Jennbert, Raudvere, Andrén, Jennbert and Raudvere2006; Hedeager Reference Hedeager2011; Price Reference Price2014) may also be identified in studies of the reopening of graves. Independent of whether the reopening should be understood as part of the burial rituals (an interpretation emphasizing the religious aspects of the Viking Age), or as politically motivated actions (an interpretation emphasizing social aspects of the Viking Age), the materiality of the reopenings is profound: the huge amounts of soil removed, the destruction of grave goods, and the acts of removing artefacts and body parts are striking. Whilst the inventory of the grave in terms of boats, textiles, bronze kettles, kitchen equipment, caskets and chests was kept on, and often destroyed on-site, other objects were removed. The lack of weapons in male graves and of pieces of jewellery in female graves is striking, particularly considering the otherwise high quality and quantity of the grave inventory. As the trench dug for the reopening is generally located towards the part of the body of the deceased where the weapons or jewellery were most likely to be placed, acquiring these artefacts was most likely one of the purposes of the reopening (Myhre Reference Myhre, Hansen and Bjerva1994, 79). The grave goods from the burial mound in Oseberg, with an extremely rich grave inventory, may serve as a prime example. Dress ornaments of precious metal of a character and quality that would match the remaining inventory appear to be lacking in the grave goods (Herschend Reference Herschend, Barnes and Ross2000, 145). At the cemetery at Gulli, the seven plundered graves are all disturbed around the location of the deceased's head (Gjerpe Reference Gjerpe and Gjerpe2005a, 142–44). This indicates that the intruders knew the location of the body and the grave goods within the grave prior to the intrusion. In one grave from Gulli, the traces of reopening are in the centre of the grave; this contained a burial which during excavation was found to contain many artefacts, but lacked jewellery and weapons (Gjerpe Reference Gjerpe and Gjerpe2005b, 54–56). Two graves in Gulli contained only one oval brooch, as opposed to the standard two; the other one was presumably removed at the reopening (ibid., 74–76, 80). Another grave contained an upper sword hilt placed two metres from the additional grave goods, but no blade (ibid., 89–91). Similarly, at the boat burial from Årby, most of the bones of the deceased had been removed in the reopening; the grave contained many objects, but the typical personal equipment was lacking (Capelle Reference Capelle, Jankuhn, Nehlsen and Roth1978, 204). What is lacking in these reopened graves is swords and high-quality pieces of jewellery, objects generally placed in relation to the body. This indicates that one purpose of reopening burial mounds may have been to acquire these specific types of artefact.

The Old Norse sources as an approach to studying the Viking Age

Stories of reopening mounds appear in several of the Icelandic sagas, in the fornaldersagas, and in the Skaldic poetry (Beck Reference Beck, Jankuhn, Nehlsen and Roth1978; Wellendorf Reference Wellendorf2002).Footnote ² Whether the fornaldersagas can be used as sources for analysing material culture from the Viking Age or whether they are complete fiction has been the object of debate (Olsen Reference Olsen1966, 56; Price Reference Price2002, 54). The fornaldersagas were generally written down in the 14th century, whereas the content mainly concerns Viking Age Scandinavia of the 9th–10th centuries. Several of the sagas even include parts of the Eddic poetry, clearly centuries older than the time the sagas were written down. The fornaldersagas contain fantastic or magic elements. Some scholars have focused on the parallels with, and thereby inspiration from, chivalric romances from the Continent from the High Middle Ages. One element that indicates that the fornaldersagas do carry traces of narratives from the Viking Age, presumably preserved in oral form, is the stories of the reopening of graves. The descriptions of grave interiors match Viking Age Scandinavian burial traditions, including ship burials, wooden burial chambers, and the choice of objects for grave goods, whereas they by no means match the burial customs of the time and place they were written down. It is notable that literary models for these stories of reopening mounds do not seem to exist in any other type of literature. This motif is unique and characteristic of the Old Norse texts (Beck Reference Beck, Jankuhn, Nehlsen and Roth1978). Reopening of graves also appears in the Skaldic poetry, which can be dated much more accurately than the sagas, i.e. in Haugbúi from the 10th century.

In the Old Norse sources, it is possible to identify two motives for the reopening of graves: to gain esoteric knowledge and knowledge of the family and lineage, and to get hold of specific objects (Beck Reference Beck, Jankuhn, Nehlsen and Roth1978; Wellendorf Reference Wellendorf2002; Lund Reference Lund2009, 245–55, with references). Either the acquired object is described in general terms as fé (Old Norse for treasures or valuable belongings), or in most cases the artefact taken from the grave is a specific, named sword, ring or helmet (Beck Reference Beck, Jankuhn, Nehlsen and Roth1978, 233). Weapons, rings and pieces of jewellery were buried in a mound with their owner, and acquired by new owners by reopening the mound and defeating the spirit living within it. The sword seems in particular to be associated with the dweller or the inhabitant of the mound. In several of the sagas the intruder removes the equipment of the dweller of the mound – except the sword. The intruder then uses the sword of the inhabitant of the mound to kill him – whereby this time he is more permanently dead (Beck Reference Beck, Jankuhn, Nehlsen and Roth1978; Wellendorf Reference Wellendorf2002; Lund Reference Lund2009).

Anton Brøgger suggested that a goal for the reopening of Oseberg was to acquire the magic powers of specific artefacts (Brøgger Reference Brøgger1945, 41). Though the concept of magic powers does not completely encompass the qualities and meaning that appear to have been linked to these objects, Brøgger's interpretation points towards a central element in understanding the reopening of the graves. The social biography of the objects removed from the grave also clearly influenced the biography of the intruder (Lund Reference Lund2009, 245–55; for examples see Beck Reference Beck, Jankuhn, Nehlsen and Roth1978, 224–25; Wellendorf Reference Wellendorf2002, 49–50). From the objects being removed from the mound it is clear that these objects have been ascribed new and profound layers of meaning through this type of circulation – travelling into the world of the dead and back again to the living. The mound dweller is portrayed as emotionally attached to its belongings in the narratives (Wellendorf Reference Wellendorf2002, 72). In several of the stories, reopening the mound and acquiring the possessions of the mound dweller is only possible because the intruder, be it a man or a woman, is related to the mound dweller. Qualities connected to the family, and to honour and shame, were attributed to the acquired objects (Beck Reference Beck, Jankuhn, Nehlsen and Roth1978, 224). A sword could return from the dead by being conquered or regained in or at the mound. These swords may have been made abroad, acquired through trade or plunder, or passed on as a gift or an heirloom; they travelled around different parts of Scandinavia and even travelled from the living to the dead and back again – obtained through reopening a grave.

In the stories of the reopening of graves, meeting the mound dweller is always described as a conflict, and in the narratives part of this conflict is due to the strong emotional bond that the mound dweller feels for his possessions. The sagas do indeed focus on the social biography of the objects in the graves, and especially on the concept of heirlooms. The acquired objects from graves were often named after the former owner, i.e. the ring Sóta-nautr in Harðar saga named after the mound dweller Sóti, or the sword Kárs-nautr in Grettis saga named after the mound dweller Kárr. The last part of the name, -nautr, is also used in the Eddic poetry and the Icelandic sagas for named swords, pieces of jewellery and ships that were received as gifts or acquired by force. Nautr describes the relation to the former owner, independent of how the object was obtained. They are what Weiner would describe as inalienable objects, artefacts which carried a link to an original producer or owner, an object not suited to buying and selling (Weiner Reference Weiner1985; see also Mauss Reference Mauss1954). The word nautr also means ‘companion’ or ‘fellow traveller’ (Vigfusson Reference Vigfusson1991). They are perceived and presented as personified objects, whose life stories intervene in and intertwine with the lives of their owners.

Considering that the reopening of mounds can be documented in Scandinavia during the Viking Age, the very period of time and area when and where these sources claim to take place; that the literary motif is unique; and that actions of these types did not take place in the High Medieval Period, one can argue that the sources do, on this particular point, build upon a real knowledge of actions that really did take place. Such an interpretation is strengthened by the fact that the description of the inner appearance of the graves, especially in the Skaldic poetry, matches the graves of the Viking Age as evidenced from the archaeological material. Consequently, the stories of reopening mounds in the Old Norse literature need not be complete fiction. The reference to the removal of weapons and pieces of jewellery from the graves in texts that presumably existed in oral form in Viking Age Scandinavia may be perceived as a particular mindset or attitude towards objects of these types. Furthermore, objects of the very type which were removed from Viking Age graves were also deposited in wetlands, and thus clearly treated and handled differently to other types of material culture. The motive appears to be connected to personal relationships between humans and objects and through these the relationship between living and deceased relatives in the Viking Age. As the social biographies of the former owners were inscribed in the social biographies of the objects, the new owners could connect to the history of their ancestors.

To summarize, if we move from the narratives to the archaeological material, it is clear that some objects were treated very differently than artefacts in general: high-quality weapons and pieces of jewellery – unique objects that seem to have been buried in mounds that were reopened soon after the burial (Lund Reference Lund2009; Reference Lund, Carver, Sanmark and Semple2010), and these are the very types of object which were deposited in wetlands. They appear to have been perceived as personified or animated objects.

Reopening Merovingian graves

This phenomenon of animated objects has a striking parallel. The reopening of graves in Scandinavia resembles the reopening of graves on the Continent from the Merovingian period shortly after burial (Myhre Reference Myhre, Hansen and Bjerva1994, 75). Across a vast area, from southern Germany and Austria to present-day France, the Netherlands, and southern England (though with clear regional variations), a practice of reopening graves and removing grave goods can be identified, particularly in the 7th century. These actions have traditionally been interpreted as mere looting for valuables, but have in recent years been re-evaluated by a number of scholars (Kümmel Reference Kümmel2009; Van Haperen Reference Van Haperen2010; Aspöck Reference Aspöck2011; Klevnäs Reference Klevnäs2013). Reopening of graves is also referred to, directly and indirectly, in a number of written sources from the same period (Krüger Reference Krüger, Jankuhn, Nehlsen and Roth1978). The custom is particularly common in the Frankish and eastern Langobard areas, which in this period were dominated by Reihengräber. More than a third of the Merovingian graves were apparently reopened (Roth Reference Roth, Jankuhn, Nehlsen and Roth1978). Archaeothanatological studies of disturbed graves in eastern Langobard areas show that most reopenings were performed at a time when the corpse was fully skeletonized, but was still surrounded by a hollow space from a coffin or other structure (Aspöck Reference Aspöck2011, 299–300).

Objects of specific types were removed, whilst others were left in the graves. In the male graves, the removed objects were mainly swords, seaxes and belts (Roth Reference Roth, Jankuhn, Nehlsen and Roth1978), whereas brooches, parts of girdle hangers and necklaces appear to have been removed from the female graves, as indicated by a study of the Langobard cemetery Bruun am Gebirge, Austria, by Edeltraud Aspöck (Reference Aspöck2011, 310; see also Naji Reference Naji, Rakita, Buikstra, Beck and Williams2005, 178). In other cemeteries, necklaces and one of four brooches were left behind (Roth Reference Roth, Jankuhn, Nehlsen and Roth1978; Aspöck Reference Aspöck2011). The removal of skulls or crania in the Langobard graves has been interpreted as related to relics taken from the graves as the population migrated (Aspöck Reference Aspöck2011, 313). In addition to the removal of parts of the body and specific objects, pieces of clothing were also being removed (Van Haperen Reference Van Haperen2010).

Like the reopened graves in Scandinavia, the disturbance of Merovingian graves has been interpreted as a final act in the burial rites (Brendalsmo and Røthe Reference Brendalsmo and Røthe1992; Van Haperen Reference Van Haperen2010). These actions took place in a Christian context, not long after the Conversion. In light of the worshipping of saints’ relics and contact relics, Martine C. van Haperen identifies the similarities between the disturbed Merovingian graves and the descriptions of saints’ cults in the same period. Objects and bones from graves were presumably perceived as a specific category of things. In her optic, the purpose and consequence of reopening the graves were thus to enable the ancestors to be physically present among their living descendants (Van Haperen Reference Van Haperen2010, 17–23).

Relics and furta sacra

This has led me to think it might be relevant to follow how the cult of the saints’ relics was constituted on the Continent in the time of the Viking Age; that is, in the Carolingian era. Patrick Geary points out that even though the cult of saints may be identified almost everywhere in the medieval Christian world, it was only in the West that the worship primarily centred on the physical remains of, and relics related to, the saints (Geary Reference Geary and Appadurai1986; Reference Geary1994, 41). The promotion of the cult of saints and their relics in the 8th–9th centuries by the Carolingians caused demand for relics that exceeded supply (Andrea Reference Andrea and Taylor2010). In late antiquity, the protecting figures within Christianity were, to a growing degree, human beings in the guise of saints (Brown Reference Brown1981, 58–59). In the Carolingian period, relics appear as agents in written sources, as they owned property, received donations and were acknowledged as the proprietors of the churches in which they were buried (Geary Reference Geary1990, 28–43; Reference Geary1994, 42). This means that the physical remains of the deceased, in terms of bones and artefacts of specific types (contact relics), were undoubtedly treated and perceived as primary actors.

From the 7th century, a new type of narrative turns up in the sources: rather short stories of translationes, including the opening of graves in order to move saints’ relics (Geary Reference Geary and Appadurai1986; Van Haperen Reference Van Haperen2010). In the 8th–10th centuries, the Carolingian period, and into the Ottonian period, the number of these stories increased and they became a genre of their own (Geary Reference Geary and Appadurai1986; Reference Geary1990, 3–44). One important type of translationes is the furta sacra, the stealing of holy objects. From the late 8th century and the following 100 years, men of the Carolingian church acquire a number of remains of saints from Italy and Spain by ordering the theft of bones from, presumably, holy men and women (Geary Reference Geary1990). The phenomenon of furta sacra only occurs within a short time frame. These stories of furta sacra vary, but they flourish in the 9th century. In the correspondence between different bishops, the same composition appears. We hear of wandering monks who give relics to churches in the Carolingian Empire, the Roman or Italian origin of the relics, the more than questionable means of their acquisition, and the instant excitement they caused not just within the diocese, but wider afield. This was a process of translating saints from Italy into the heart of the Carolingian Empire (Geary Reference Geary1990, 18–19, 44–107). Relics could also be bought from merchants selling the bones of saints (Geary Reference Geary and Obelkevich1979; Reference Geary and Appadurai1986), yet these relics were not as popular as those stolen from other churchly buildings or from catacombs by churchmen motivated by pure religious zeal (Geary Reference Geary1990, 7–43).

Christianity changed radically during the Early and High Middle Ages in Europe, including the cult of saints. From the 11th century, the cults of the Virgin Mary and of Christ dominated, promoted primarily by the Cistercians at the expense of local saints (Geary Reference Geary1990, 24–26). This means that the retrospective perspective that has dominated studies of early Christianity and the Conversion in Scandinavian research may be biased or unsuitable when it comes to understanding these changing phenomena. Perceptions of bodies and objects in early Christianity certainly do not resemble mentalities or mindsets regarding bodies and objects in the following centuries (Lund Reference Lund2013). Independent of the Christian context, a similar phenomenon occurred within a short time span: the reopening of graves, and the removal of bones and particular artefacts in terms of weapons and pieces of jewellery. Contemporarily, physical relics were perceived as agents.

Linking biography and animism

If we return to the archaeological material of Viking Age Scandinavia, it is clear that some objects were treated very differently from artefacts in general: high-quality weapons and pieces of jewellery – unique objects – were buried in mounds that were reopened after the burial and objects of the same categories were deposited in wetlands (Lund Reference Lund2009; Reference Lund, Carver, Sanmark and Semple2010). The act of deposition is the final stage in the social life of the artefact in its ancient context. The reason for this ending can be that the identity of the artefact was bound to the identity of a person, so that, for instance, when the owner died the deposition of his sword could be a way of dealing with this object. The sword would in this sense be an inalienable object, connected to its original owner even as it circulated between men (Weiner Reference Weiner1985; Fowler Reference Fowler2004, 57) – an heirloom that entailed and contained the stories of the previous owners. As such, inalienable artefacts cannot be destroyed (Weiner Reference Weiner1985, 210). The acts of deposition could in this sense be a way of handling objects with complex social biographies – taking them out of everyday life, but keeping them in places endowed with special meaning. Most likely, the relics in Carolingian churches, as well as the high-quality, potentially unique swords and pieces of jewellery in Scandinavia that were removed from graves or deposited, were artefacts with long and complex biographies. These were artefacts that were handled as categorically different to material objects in general. In both contexts, contemporary and later written sources indicate that they were perceived not only as inalienable, but also as animated objects. In this context, the choice of the term ‘animated’ rather than ‘personified’ points towards links between perspectives on the social biography of objects and the concept of animism that have yet to be fully explored.

The study of animism has long been toned down or avoided within social anthropology due to the primitivistic associations the concept bears (Descola Reference Descola and Kuper1992; Willerslev Reference Willerslev2007, 2–9). However, animistic perspectives may give us new insights into ontologies of the past. In many circumpolar contexts, such as the Canadian Ojibwa, the Siberian Yukaghir, or the Siberian Khanty, not only human beings, but also beings of other types, can be considered as people, including stones, trees and animals (Ingold Reference Ingold2000, 89–131; Jordan Reference Jordan2003; Willerslev Reference Willerslev2007, 2–9). Among the Siberian Khanty, humans, animals, water running in a river, falling snow or thunder are considered as animated, whereas snow that has fallen, water in a bucket or a rock that cannot be moved are considered as unanimated (Jordan Reference Jordan2003, 103). These types of categorization of what it means to be a person in a given context are also found outside the circumpolar areas, for instance in the Amazon, and show noticeable similarities with the circumpolar examples (Viveiros de Castro Reference Viveiros de Castro1998, 472; Brightman, Grotti and Ulturgasheva Reference Brightman, Grotti, Ulturgasheva, Brightman, Grotti and Ulturgasheva2012; Hugh-Jones Reference Hugh-Jones, Brightman, Grotti and Ulturgasheva2012). It must be kept in mind that animism in this context by no means implies that all animals and objects are considered as having personhood and agency. Notably, personhood emerges in a given context of a specific time and place (Ingold Reference Ingold1993, 89–131; Jordan Reference Jordan2003; Willerslev Reference Willerslev2007). A tree or an animal is not considered to simply take the role of a person, but to be a person (see also Willerslev Reference Willerslev2007, 8). This means that a Western ontology of personhood, concerning what is animated and unanimated, is not necessarily relevant in other contexts – including the mindsets and world views of past communities (for further discussion see, for instance, Descola Reference Descola and Kuper1992; Alberti and Marshall Reference Alberti and Marshall2009). In other words, the ontological duality of nature and culture might not be relevant or applicable if we are aiming to understand the mindsets, world views and ways of life of people living in Viking Age Scandinavia or under the Frankish reign.

As the social identity of an object is intimately connected to its human relationships throughout its social life, the biographical perspective partly coincides with these perspectives on animism. In animism, reciprocity is a central element in the relationship between humans and objects or humans and places. In the Andes, for instance, offerings and daily work are used to develop reciprocal social obligations with animated objects and places. In this process, humans create social relationships with the material world. Furthermore, this relationship gives social identity to the implicated artefacts and places. In essence, animism is about social relations between humans and the material world (Sillar Reference Sillar2009, 367–74; Lund Reference Lund2015). If we turn to the animated objects of the Viking Age, such as particular named swords, they are also perceived as part of a reciprocal system as they act upon and with the humans that own them, treating them well or even abandoning them (Lund Reference Lund, Carver, Sanmark and Semple2010).

As they are presented in the Old Norse sources, the unique swords or pieces of jewellery are presented as being rather than becoming animated, but the animation appears in the narratives to be the result of their biographies, and thus as a state which occurred. Objects of Viking Age Scandinavia became animated at production, in other words by ‘birth’, through gift giving, and even by theft in the sense of the reopening of graves. In the case of the Carolingian relics, their animated quality was also produced or enhanced by theft. The relics became animated either by being declared relics, in other words by ‘birth’, and through gift giving, but also by theft. Naming objects was also a way of imbuing them with personhood. Roberta Gilchrist states that in the action of naming and/or inscribing an object, agency may be transferred from a person to an object (Gilchrist Reference Gilchrist2012, 224). Yet in the cases presented here the personhood of the sword or the relic did not come directly from a human agent. The relics were not simply the representation of the saint, or representations of relations or personhood, they were material and thus physical specimens of such mindsets (see also Weismantel Reference Weismantel2015, 2–3). They operated through their materiality, and thus were not mere instantiations of an already given framework, but contained these transformational frames (see also Hastrup Reference Hastrup2013, 42–43).

The cognitive schemes of the 9th–10th centuries?

As Mary Weismantel points out, out the categorization of diverging ontologies within anthropology, including the Amazonian perspectivism of Viveiros de Castro, may be criticized for lacking a historical dimension. Thus archaeology may provide such a material and historical dimension to these perspectives (Weismantel Reference Weismantel2015, 4–5). Aiming at grasping the ontology of Viking Age Scandinavia or the Carolingian world also includes moving beyond the outsider etic perspective towards an emic perspective. When we reduce our studies into searching for one category known from our own modern perspective, such as belief or power, what we see is only ourselves. Rather we should act in order to make room for the other, and let the other, ontologically speaking, have it their own way, to use de Castro's wording (de Castro Reference de Castro2015, 12).

The presence of animated objects in Scandinavia was hardly the result of a mere diffusion of Carolingian ideas. In numerous ways, the elite of Viking Age Scandinavia used similar cultural expression as the Carolingians, but with a completely different content. For instance, the reuse of stylistic elements, manuscripts or sarcophagi from the Christian Roman Empire in what has been termed the Carolingian renaissance (see, for example, Koenigsberger Reference Koenigsberger1987, 126–35) may be seen as equivalent or correspondent to the use of multiple pasts through the reuse of burial sites in Viking Age elite centres (Pedersen Reference Pedersen, Andrén, Jennbert and Raudvere2006; Holst et al. Reference Holst, Jessen, Andersen and Pedersen2012; Lund and Arwill-Nordbladh forthcoming). Aspects of Viking Age attitudes towards material culture may thus be recognized as a counterpart to the Carolingian. The personhood of a Carolingian relic was not identical to the personhood of a unique, Scandinavian sword. What they have in common is the understanding, and handling, of specific objects as animated within the same period of time in two areas that were in contact. One could argue that the similarities were not the result of contact, but were caused by overlaps in the cognitive schemes of the Norse of the Viking Age and the Carolingians. Descola's notions on animistic ontology appear to apply to the Norse, and even to Carolingian Christianity, with regard to the Viking swords and Carolingian relics. Yet animist thinking by no means dominated mentalities or world views in either of these two contexts. On another level, in terms of the links between cosmology and ritual, Old Norse and Carolingian Christianity may rather be understood in light of Descola's notions of analogist ontology (Descola Reference Descola2013, 129–43, 201–31). These classifications or groups of basic sets of ontologies should by no means be considered a typology of world views, but as a heuristic device activated in specific contexts (Descola Reference Descola2010, 338–39). As the animation of the object is understood as occurring in specific situations, we may turn our attention towards such artefacts, which in specific cultural contexts were treated, handled and categorized differently than were object features in general. We may understand this as an expression of a specific type of connectedness with things. This does not in any way imply that these are traces of an animistic religion per se. It means that animistic perspectives may offer an insight into the mindsets relating to the ritual depositions of artefacts and the reopening of graves in the south Scandinavian Viking Age, the reopening of Reihengräber and the stealing of relics. Thus it is likely that people of the Viking Age, and perhaps more surprisingly Carolingian Christians, considered some, but certainly not all, artefacts to be animated, in certain contexts.