A Re-Evaluation of the Roles of Domestic Cats in Viking Age Scandinavia

The role of cats in Viking Age society is little investigated and has been dominated by uncritical adoptions of medieval mythology. Based on literary sources, the domestic cat is often linked to cultic spheres of female sorcery. Yet the archaeological evidence indicates an ambivalent situation. Cat bones from many trading centres show cut marks from skinning and highlight the value of cat fur. In contrast, the occurrence of cats in male burials points rather to a function as exotic and prestigious pets. The influence of Old Norse mythology on the traditional interpretation of cats as cultic companions therefore needs critical reconsideration. For this, a broad range of literary and historical sources – from Old Norse literature to Old Irish law texts – will be analysed and confronted with the archaeological evidence for domestic cats in Viking Age Scandinavia. The results will be discussed on a broader theoretical approach, involving concepts such as agency, and embedded in current research on human-animal-relations in order to achieve a more nuanced perspective on the roles and functions of cats in day-to-day reality as well as in the burial context.

We did not domesticate the cat so much as the cat became tolerant of, and adopted us. (Reed 1980:16) Introduction: Animals and agency Through the recent 'animal turn' in archaeology (see Andersson Cederholm et al. eds 2014) understanding of human-animal relations has increasingly been regarded as a research question of the highest importance for the interpretation of prehistoric societies (see especially Ryan & Crabtree eds 1995;Pluskowski ed. 2005Pluskowski ed. , 2012. Animals are much more than merely passive objects of human agents. Instead, humans and animals interact on different levels. On the one hand, these interactions take place in everyday life -the animal as vital workforce, companion or livestock -and in social discourse -the animal as an expression of social status, affiliation or descendant. They also exist on a religious level -the animal as mythological symbol, as a totem of a mythically defined identity or as a psychopomp, a being which is able to cross the borders between this world and the other world. Through their natural behavior, the particular qualities attributed to a species (for example nobility, loyalty, strength, fertility, bravery), their function in everyday life, but also through their individual character and the context of this interaction, animals have their own agency in their interaction with humans (see Steward 2009;Carter & Charles 2013). The concept of agency in archaeology is not clearly defined (see Dobres & Robb 2000a:9), but generally speaking it can be regarded as the capability or ability to influence surrounding structures (see Dobres & Robb 2000bBarrett 2012). This ranges from a broad definition including a 'nonhuman agency' of objects through their entanglement with humans (see Latour 2005:122; Knappett & Malafouris eds 2008; for the concept of entanglement see Hodder 2011) to a very restricted definition as the 'strategic carrying out of an intentional plan in accordance with a specific culturally constructed idea of personhood' (Dobres & Robb 2000a:9). Even though the definition of agency as following 'a specific culturally constructed idea of personhood' does not fit the agency of animals, it is obvious that animals are capable of intentionally modify or manipulating their surrounding structures. The best example might be the process of domestication as the deliberate establishment of close relationships with humans for better access to food and shelter. This is especially striking in the case of the domestication of cats, as the above-mentioned quotation exemplifies. Another example is animal migration that forces humans to follow the movements of other species and to adopt their habitat. Especially relevant for the current analysis is the agency of animals that is mirrored in certain attributes such as strength, fertility or loyalty that were regarded as desirable for humans. Those animals were incorporated into a cosmology and self-concept as totem, spirit or heraldic animal. Thus, they were especially suited to present, construct or manipulate descent, status or affiliation to particular religious, social or otherwise defined groups (for social identities as a dynamic and fluid process, see Toplak 2019), thereby shaping human culture and identity (see for example Sundkvist 2001 for horses). Well known examples are the totem or spirit animals of Native American tribes, animals in medieval coats of arms or mascots of sports clubs. The perception of this agency in turn reciprocally modifies the symbolic meaning of the species in the particular cultural context. The understanding of human-animal relations as a means of expression of a certain socio-cultural identity reveals much information about the self-conception of societies (see for example Williams 2001Williams , 2005Thomas 2005;Bertašius 2012).
This strong impact of animal agency on human-animal-relations and interactions, and on human behaviour and culture, illustrates the importance of a much more symmetrical approach in archaeology that shifts away from the dualism of humans : things (see Murdoch 1997;Shanks 2007;Witmore 2007;Olsen & Witmore 2015). However, despite the justified criticism of an often far too stringent definition of agency (Witmore 2007:551-552) which consolidates a simplistic differentiation between active humans and a passive world, the concept of agency as an 'active influence on the surrounding structures' seems to be suitable for the investigation of human-animal-relations.
Human-animal relations in the form of both symbolic roles and functions are quite well investigated for two of the most formative animals in early medieval history: dogs as earliest domesticated animals and important companions for hunting, herding and as guard dogs; and horses as an indispensable means of transport and iconic symbols of status for a mounted warrior elite (Jennbert 2003;Sikora 2004;Gräslund 2004;Kuczkowski & Kajkowski 2012;Leifsson 2012Leifsson , 2018Stelter 2013;Cooke 2016;Nichols 2018;Strehlau 2018). Furthermore, the consumption of horse meat was common in many cultures and was in part connected to religious ideas, as for example in Viking Age Iceland (Leifsson 2018:264-265). In contrast, the social roles of another common animal -the cat -remains much vaguer.

New perspectives on animal sacrifice
While cats occur in Viking Age in settlement contexts as element of the common livestock, their frequent appearance in burials, as sacrifices or as burial goods, implies a similarly specific cultic, mythological or ideologi-cal symbolic value as those of horses or dogs, and requires further investigations. Following Insoll (2011:151), the term 'sacrifice' is hereby defined as the ritual killing of an animal. This means that the significant aspect is the action of the public and ritualized slaughter, not the deposition of the animal as grave goods. In addition to the concept of sacrifice given by Hubert & Mauss (1898;see also Wunn 2006) as a means of communication with a divine sphere (see also Kaliff & Østigård 2013:88), it might fulfil not only religious but also social functions; as a social catharsis (Leifsson 2012:321-322) and a socially accepted substitute for violence (Girard 2010) or to establish or strengthen group identity (Bertašius 2012) through the repetition of mythological/cosmological events (Williams 2005:19;Strehlau 2018:7). Both concepts -animals as a sacrifice and animals as grave goods -might overlap. Horses can be sacrificed to re-enact a certain mythological or cosmological event as an mnemonic action to establish ideas of identity and affiliation (see for example Cooke 2016:11), at the same time being deposited in the grave as symbols of status. Therefore, the differentiation between animals being sacrificed and animals being deposited as grave goods is diffi cult (see Morris 2012:13, 17-18;Vretemark 2013:52, 58) and a primarily theoretical construction (see Wunn 2006:32).
As with dogs, cats served as nutrition only in times of famine and were not regularly consumed (Hüster Plogmann 2006:112). Therefore, it can be assumed that cats were not intended as a food offering. Instead, it seems likely that cats -just like horses or dogs -were sacrificed in the burial cere mony because of their special symbolic meaning. This is in contrast to cattle, sheep/goats or pigs, which were often killed at the optimal slaughter age (see Magnell & Iregren 2010:233-235;also Andersson 2005:85). While whole cattle carcasses in some richer burials might be regarded as symbols of wealth, sheep/goats and pigs were often dismembered. Single parts of the animal -often fleshy joints of meat which are traceable by single bones, for example femora or ribs -were deposited in the graves (Bond & Worley 2006:91-92), as symbolic nourishment for the deceased or leftovers from ritual feasts (Williams 2005:32). However, this does not rule out a preceding sacrifice of the animals. Due to the fragility of cat bones it is not always certain whether the animals were deposited in their entirety, with their bones articulated in correct anatomical order nor in which relationship to the deceased they were deposited in the grave. Yet these factors are of enormous importance for the interpretation of animal sacrifices (see Williams 2001:199;Strehlau 2018:8). However, from three graves with cat bones analyzed in the work of Strehlau (2018:36), two seem to contain only phalanges -a potential indication for cat furs -but in at least one grave a complete cat skeleton was deposited.

The Warrior and the Cat
The need for a much more differentiated and critical interpretation of animal sacrifices in burial contexts, set against the contrasting background of day-to-day reality, is illustrated by the deconstruction of traditional parad igms in the newest discussions on horse burials in Anglo-Saxon England and the Norse colonies in the north Atlantic (Ashby 2002;Williams 2001Williams , 2005Sikora 2004;Fern 2005Fern , 2007Fern , 2012Bond & Worley 2006;Leifsson 2012Leifsson , 2018Stelter 2013;Cooke 2016;Strehlau 2018). Horses in burials were long regarded as classical signs of high status and as symbols of a mounted warrior elite (see for example Hills 1999). But as new investigations have highlighted, animals can also serve as symbols of identification -as a form of totem -for social groups that are not necessarily defined by a specific high social status, but by a common ancestry, origin or mythology (see also Richards 1992;Crabtree 1995;Williams 2001Williams , 2005. Even though their function in everyday life might be mirrored in this symbolism, the decisive aspects are the specific attributes of these animals such as for example strength, loyalty, courage or fertility. It can be assumed that the function of cats in everyday life was completely divergent from the type of roles that horses and dogs had in Viking Age society, which were highly interwoven with ideas of social status, and especially with male-linked activities such as combat or hunting. Therefore, the investigation of the symbolic roles and significance of cats as sacrifices in Viking Age funerary rites allows for new points of view on the relevance of the agency of animals and human-animal relations in the presentation or construction of social identities in the burial context.
The burial of several wildcats in a Neolithic settlement in Scania, southern Sweden (During 1986:151), illustrates the high value or perhaps even the veneration of cats before their domestication in the north, while finds of bones from wildcats from Mesolithic and Neolithic find spots in Scandi navia bear evidence for the hunting of wildcats (Lepiksaar 1975:144;1982:119;During 1986:49, 151;Trolle-Lassen 1987). The oldest finds of domesticated cats in northern Europe date back to the Roman Iron Age, even if the new find of a caudal vertebra from a domesticated cat in a posthole in Uppland, Sweden, radiocarbon-dated to the Pre-Roman Iron Age, might alter this general picture (Zachrisson 2017). Bones of tame cats are known from a couple of graves in Denmark and Sweden dating to around 200 AD (Boessneck et al. 1979:176;Aaris-Sørensen 1988:223-224;Hatting 1994:94;Jennbert 2011:67), which corresponds with the increasing appearance of domesticated cats in several settlements in northern Germany from the second/third century AD onwards (Johansson & Hüster 1987:10-11;Benecke 1994:146;Reichstein et al. 2000:333). During the Migration Period, domesticated cats became a more frequent element of the Scandinavian fauna, as found mainly in settlements such as Vallhagar on Gotland or Eketorp on Öland (Gejvall 1955;Boessneck et al. 1979:176;Colling 1986:193;Lepiksaar 1986:22;Iregren 1997:14-17, tables 1-3) as well as in some early graves (Andersson 1993:13, 39-41). From the Vendel Period onwards and especially in the Viking Age, cats appear regularly in graves (Iregren 1997;Andersson 1993:13). DNA analysis of some of these early domesticated cats from northern Europe suggest that they were direct descendants of the north African/southwest Asian wildcat and not related to the European wildcat (Ottoni et al. 2017:5).

The cat in Old Norse literature and mythology
In the Old Norse sagas, except for some short sentences which refer to the cat's habit of hunting or playing, the appearance of cats is very limited and The Warrior and the Cat totally detached from being a part of the domestic fauna. 1 This is a clear contrast to for example horses, which appear in nearly every saga as a means of transport and an important part of everyday life, 2 and beyond that even, as highly symbolic 3 and partially sacred animals (Stelter 2013:17-31; see Rohrbach 2009 for a detailed survey of the human-animal relations in Old Norse literature). The absence of cats as an everyday element of the domestic fauna might be an indication that they were still unusual and exotic animals at the time of the transcription of the saga literature in the thirteenth and fourteenth centuries. This impression is supported by the low number of cat bones from the archaeological records. In the sagas they serve mostly as negative forces (Prehal 2011:8-9). Examples for this can be found, for example, in Vatnsdoela saga (chapter 28), where a man called Þórólfur sleggja owns twenty demonic cats for his protection (Sveinsson 1939:73) or in Orms þáttr Stórólfssonar, in which the protagonist Ormr is confronted with a man-eating giant called Brúsi and his demonic mother in the shape of an enormous cat (Jónsson 1904). This is also emphasized by the fact that 'cat' -'kattvr' (Jónsson 1931:195) -is used as a kenning for a giant in the Þulur of the Edda (see also Holtsmark 1962-65:148-149), while the draugr Þráinn in chapter 4 of the legendary saga Hrómundar saga Gripssonar is said to be of 'kattakyn' -'cat-kin' (Jónsson & Vilhjálmsson 1976b:412). A similar attribution lies behind the famous scene in the Gylfaginning from Snorri's Edda, in which the Old Norse thunder god Þórr is trying to lift a cat, who is later on identified as the Jǫrmungandr, the giant sea snake, who surrounds the whole earth (Jónsson 1931:57-58, 60). Beside this association with monsters and the underworld, the labelling as 'cat' might have been regarded as an insult, perhaps due to the connection of cats with negative forces, (female) sexuality (Prehal 2011:8-11) or laziness. 4 The most famous reference to cats in Old Norse saga literature is the appearance of the Greenlandic sorceress (a völva) Þorbjǫrg in Eiríks saga rauða (chapter 4), who is wearing a black hat of lambskin lined with cat 3 In Egils saga Skalla-Grímssonar, the father of the protagonist Egill is buried together with his horse (chapter 58 (Nordal 1933)), which reflects the prevalent custom of horse burials in Viking Age Iceland (Leifsson 2018 (Sveinsson 1934b:283)), in Orkneyinga saga ('at þú liggir hér sem kǫttr í hreysi' ('to lie crouching aside like a cat among stones') chapter 26 (Guðmundson 1965:67)) and in Helgakviða Hundingsbana in fyrri ('konung óneisan, sem kattar son', ('kinglike, as the son of a cat') stanza 18 (Neckel & Kuhn 1983:133)). A completely neutral reference seems to be the name 'Urðarköttr' ('Wildcat'; literally 'scree-cat') for one protagonist in Finnboga saga ramma, because he was found on a scree as an infant (Halldórsson 1959:257). skin and cat skin gloves also lined with cat fur: 'Lambskinnskofra svartan á hǫfði ok við innan kattskinn hvít.
[…] kattskinnsglófa ok váru hvítir innan ok loðnir' (Sveinsson & Þórðarson 1935:207-208). Because of this explicit description of cat skin and fur in the context of clothing -which is unique in the whole corpus of saga literature -the seeress's appearance is often regarded as an indication of the connection of cats with Old Norse sorcery (Davidson 1964:120;Price 2002:108;Mansrud 2004:32;Prehal 2011), at least based on literary sources. Less well-grounded but widely quoted is the chariot of Freyja, goddess of love, fertility and beauty, which according to Gylfaginning is pulled by two cats (Näsström 1995:18). Thus, cats are literally/mythologically linked with the female sphere of sexuality and also with female sorcery and cultic activities, as Freyja was the goddess who brought the sorcery of seiðr (for seiðr see Strömbäck 1935;Price 2002Price , 2006Solli 2002) to the AEsir (Simek 2006:113). Even if this link between cats and the female domain is often quoted and a popular subject in media, literature and esoteric circles, its reliability for Old Norse mythology during the Viking Age remains uncertain. The cat's chariot is only mentioned shortly on two occasions in Snorri Sturluson's Prose Edda (Gylfaginning, chapter 24 and 49 (Jónsson 1931:31, 66)), whilst in the eddic poem Hyndluljóð (stanza 7 (Neckel & Kuhn 1983:289) Freyja rides on a boar. While the boar's name Hildi svíni is mentioned in Hyndluljóð, the cats pulling Freyja's chariot remain nameless, in contrast to all other animals which serve as companions of the Old Norse gods (see Jennbert 2011:49). Furthermore, in a poem by the 10th/11th century Old Norse poet Hallfreðr vandraeðaskáld, Freyja is given the nickname 'sýr' ('sow'), which appears also in Snorri's Prose Edda (Gylfaginning, chapter 35 (Jónsson 1931:38)). This association of Freyja and her brother Freyr (Simek 2006:114-115) as gods of fertility with wild boars (or pigs in general) seems to be more authentic and incorporated in Viking Age beliefs of fertility rather than with the fragmentary image from Snorri Sturluson's literary work.
This might indicate that the association of cats with Freyja does not originate in Viking Age mythology but could be interpreted as an interpolation from Christian or antique traditions (Näsström 1995:23-28). The chariot pulled by cats seems to be an adaption of the antique trope of female deities of fertility or mother goddesses with a wagon pulled by big cats such as lions or panthers, for example Cybele or Artemis (Halvorsen 1959:618;Polomé 1999:587;Simek 2018:313-314;contra Bernström 1963:363). Consequently, even the rather negative associations of cats in Old Norse saga literature do not necessarily reflect Viking Age beliefs but may result from Christian influence. This might be strengthened by the absence of cats as everyday elements of the domestic fauna in Old Norse saga literature.

The Warrior and the Cat
The image of the cat in the eyes of the Christian church changed dramatically during the time when the Old Norse literature was written down. In early Christianity, until the 12th century, cats were appreciated as companions and eudemonic even in monasteries (Spahn 1986:53;Delort 1991Delort :1079Steuer 1994:661), where they were sometimes the only animals allowed (Reeves 1998:110;Walker-Meikle 2011:12). In the Old Irish poem Pangur Bán from the 9th century, an Irish monk in the monastery of St. Paul in Carinthia, Austria, praises his cat Pangur as a precious and helpful companion (Poole 2015:872) and a door in Exeter cathedral as an entrance hole for the bishop's cats (Reeves 1998:110;Walker-Meikle 2011:12). It was not until the papal bull Vox in rama, issued by Pope Gregory IX sometime between 1232-1234, that (black) cats were demonized in official Christian doctrine (Hergemöller 2007). At the very latest during the trial of the Templars (1307-1314) cats were explicitly associated with the devil and heresy (Gray 1990;Lipton 1992), as the Templars were accused of worshipping a cat (Walker-Meikle 2013:23). This perception of cats as devilish animals, as companions of witches and fetishes for black magic and witchcraft, remained constant during the Late Middle Ages and Early Modern Times (Gray 1990;Prehal 2011:22-24) and cats were an integral element in magic rituals of popular belief (Bernström 1963:366-367;Hoggard 2004:175-176;Hukantaival 2016:34-35).
The complicated picture given by Old Norse saga literature -the lack of cats as common animals, their association with negative forces and witchcraft and the interpolation of classical antique topoi -as well as the dramatically shifting perception of cats in the Middle Ages demands a more sceptical perspective on the historical value of Old Norse literature for the symbolic meaning of cats in Viking Age Scandinavia, due to their potential skewing by later medieval, Christian ideology.

The archaeological evidence for cats in Viking Age Scandinavia
In the following sections, the function and symbolism of cats in Viking Age Scandinavia will be analysed and discussed, based on the archaeological evidence and supported by several historical sources.

CATS IN VIKING AGE CULT AND SORCERY
The archaeological evidence for magical connotations for cats in Late Iron Age Scandinavia remains vague. Leather fragments from the famous Ose berg ship burial of a potential völva or priestess (see Brøgger et al. 1917Brøgger et al. , 1920Brøgger et al. , 1928Christensen et al. eds 1992;Nordström 2007:250-365) could have been cat skin (Ingstad 1992:254-255), which would ostensibly strengthen the validity of the saga account concerning the appearance of the seeress Þorbjǫrg, even if the saga remains silent as to whether the cat skin was a definite necessity for magic/cultic rituals or just an insignia or dress accessory. In addition to these leather fragments, several wooden arte facts from the Oseberg burial are decorated with the so-called gripping beast motif that is characteristic of early Viking Age art styles as Oseberg, Broa or Borre (Graham-Campbell 2013:48-81), and which show animals that might be interpreted as cats, for example, on a wagon (figure 1). Within the context of saga literature and restricted to the Oseberg complex, this seems to argue for a cultic role of cats in Old Norse sorcery and resembles the image of Freyja's cat-pulled wagon, which led Ingstad (1992:254-255) to interpret at least one of the two females buried at Oseberg as a priestess of Freyja (see Price 2002:160;Mansrud 2004:32;Prehal 2011:47-49).
Cat bones are known from some Migration and Vendel Period, as well as Viking Age, cult sites which provide evidence for the sacrifice of cats among many other species, for example from Lejre (Christensen 1991:184), Uppåkra (Nilsson 1998:92), Gamla Uppsala (Henriksson 2003:2) or the sacrificial bogs Skedemosse on Öland and Hassel Bösarp in Scania (Anders son 1993:27;Jennbert 2011:60). At the Viking Age farm of Borg in Östergötland, Sweden, a high proportion of skull bones and lower jaws from horses and dogs, as well as some cats, dated to the 9th century (Nielsen 1996:100;Einarsson 2008:155), resemble Adam of Bremen's description of the great sacrifices at Gamla Uppsala, even if human bones were lacking at Borg (Nielsen 1996:102). In the foundation pit of a Viking Age farmstead in Bunkeflo, Sweden, remains of a potential cat skull were found that might be deposited as building sacrifice (Carlie 2004:136; for building sacrifices in general see Hukantaival 2016). At the pagan cemetery from Ingirístaðir in northern Iceland, roughly dated to the Viking Age, a pit with bones from several animals (cattle, pig, caprine), fragments of two human skulls and an incomplete cat skeleton were excavated (Prehal 2011:37-47). It seems most likely that this pit contained the remains of a sacrifice, even if it might be speculated that it is the burial of a female shaman or völva (Prehal 2011:40). Remains of a cremated cat were also found in the famous burial of a völva at Klinta (grave 59:3) on Öland (Price 2002:145;Mansrud 2004Mansrud :90-91, 2005. The appearance of cat bones in some complexes connected with cult and sacrifice proves that cats were regarded as valuable sacrificial animals. But the count of cat bones on cult sites is far too small to ascribe a significant function or meaning to them in Old Norse cult, even more so, when compared to other animals, for example horses or dogs, which appear far more frequently at cult sites (Iregren 1997:22 , table 7). However, when comparing the proportion of cat bones to those of other (mostly bigger) domestic animals, the inferior preservation of (burnt) fragile cat bones to that of dog bones for instance (see Andersson 1993:7-12; Kitchener & O'Connor 2010:91) must be taken into consideration.

CATS IN BURIALS
Beside the increasing occurrence of cats in settlements from the Migration period onwards, rising appreciation of domesticated cats is also evident from the growing numbers of finds of cat bones in graves in the Vendel Period and especially in the Viking Age. As mentioned above, cats were used as grave goods in occasional burials from the late Roman Iron Age onwards, as for example in a female burial from Överbo in Västergötland, Sweden, dating to the second century AD (Jennbert 2011:67). While cat bones are missing from graves from the Migration Period -perhaps due to the prevailing burial rite of cremation that affects the representativeness of the fragile bones (Andersson 1993:9-12) -around 50 graves with cat bones from Vendel Period and Viking Age Sweden have been identified by Maria Andersson (1993) in her master thesis.
The distribution of cat bones in these burials shows a clear tendency towards male burials. More than half of the Swedish graves with cats from the Vendel Period and the Viking Age were burials of men, only one third were burials of females, even if an increasing percentage of female burials with cats in the ninth to tenth centuries leads to an almost equal distribution in the Viking Age (Andersson 1993:14). From the ninth century onwards, cat bones also appear in greater numbers in burials of infants or juveniles of unidentified sex, with a distribution of around 14 per cent of the burials with cat bones from the Swedish Viking Age. Many of the burials with cat bones from Vendel Period and early Viking Age Sweden were above-average rich graves (Andersson 1993:18-19), for instance, the so-called 'stormannagravar' ('magnate graves') from Uppland in Sweden are often equipped with larger numbers of animals such as dogs, horses, birds of prey and occasionally cats (Sten & Vretemark 1988:150;Jennbert 2011:103). An impressive example is the cremation burial of an adult male at Vibyhögen from the late ninth or early tenth century with grave goods of gold and silver and 25 cremated animals from 19 different species, among them bones of a cat (Sten 1976;Sten & Vretemark 1988;Jennbert 2011:102). Further famous examples evidencing cat bones are the late ninth-century ship burial of Gokstad in Norway (Schetelig 1904:332), and the western royal mound in Gamla Uppsala, Sweden, dating to the early Vendel Period (Major 1924:119). Surprisingly, no cat bones are known from the high-status burials at Vendel and Valsgärde in Sweden, except for the early Vendel Period grave mound known as Ottarshögen (Sten & Vretemark 1988:150). An interesting case is a potential cenotaph from Helgö, Uppland, in Sweden (grave 6 in cemetery 116). The grave was equipped with a larger number of precious beads, but human remains were missing. Instead, the bones of a cat were found in the burial pit. It seems to be unlikely that grave 6 was intended as burial of a cat, but rather as a cenotaph or a substitute for a woman who died abroad and could not be buried at home (Jonsson & Sander 1997:95;Zachrisson 2004:165). In the later Viking Age, cat bones became more common in average burials as well, apparently since cats had established themselves as a regular element of the domestic fauna in Scandinavia during the eleventh century, as their increasing numbers in settlement contexts show (Reichstein et al. 2000:333).

CAT FUR IN VIKING AGE DRESS
While cats were occasional grave goods in high-ranking burials of the Vendel Period and early Viking Age, the shift in their function in later Viking Age society, which is visible in the increasing appearance of cat bones, even in average burials, can also be detected in many emporia and early urban centres, such as Hedeby or Sigtuna. Many cat bones from (proto-)urban settlements in Scandinavia show clear cut marks from skinning -especially on the skulls around the snout, at the distal ends of the long bones and on the pelvis (figure 2). The best examples are from the detailed investigation of cat bones from Hedeby, nowadays Germany (Requate 1960;Johansson & Hüster 1987;Hüster Plogmann 2006:90-97), which prove that cats The Warrior and the Cat were selectively killed and skinned for their fur, while a regular slaughtering for consumption can be ruled out because of the missing evidence for the dissection of the long bones (Johansson & Hüster 1987:47-48;Hüster Plogmann 2006:90-91).
Most of the bones with cut marks from skinning come from juvenile or younger adults, full-grown animals that were killed between 8-12 months of age or at least in their second year of life. Since most cats were born in spring, they were slaughtered with their thick winter pelt, which was, in most cases, not yet damaged by diseases, parasites or malnutrition as is typical for older freely roaming cats (Johansson & Hüster 1987:21;Hüster Plogmann 2006:93). The development of their denture suggests that the cats from Hedeby were mostly free to roam and their nutrition was speciesappropriate based on small rodents (Johansson & Hüster 1987:40;Hüster Plogmann 2006:95) in contrast to the more human-dependent and, in part, additionally fed animals from later Schleswig (Benecke 1994:213).
A large pit dating to the late Viking Age or early Middle Ages with the remains of at least 68 cats as well as bones from several other animals -among them a fox and a raven -is known from the town of Odense in Denmark (Hatting 1989(Hatting , 1990, which most likely testifies to commercial fur production in an early urban environment (figure 3), even if a cultic context has been postulated, mainly because of the raven skeleton (Hatting 1990:192). Cat bones are almost absent from Viking Age Ribe, Denmark, but the few examples show at least faint cut marks as potential result from skinning (Hatting 1991:52). Further evidence for the regular skinning of cats can be found at several late Viking Age and medieval sites such as Sigtuna (Hårding 1990:107;Wigh 2001:123) and Gamla Lödöse (Colling 1986:196) in Sweden, medieval Schleswig in northern Germany (Spahn 1986) andYork in England (O'Connor 2003:3233-3234). A similar situation can be traced in the early medieval town of Novgorod, Russia (Maltby 2012:377), as well as on a number of sites from the late Viking and Middle Ages in England, Ireland and Scotland (Noddle 1974:378-408;Maltby 1979:12;Richards 1991:74;McCormick 1997:835;McCormick & Murray 2007:612). The high value of cat fur as trimming or lining for clothing is likewise documented in literary and legal sources from medieval Iceland and Sweden (Bernström 1963:364-366;Colling 1986:196;Hårding 1990:107).

CATS IN VIKING AGE BIRKA AS PETS AND COMPANIONS IN THE AFTERLIFE
In contrast to the examples from many early towns in late Viking Age Scandinavia such as Hedeby or Sigtuna, the skinning of cats has not so far been The Warrior and the Cat documented in Birka, despite the expectation for such an important trading place. In Sigtuna, phalanges from cats appear only in the archaeological layers from the 10th century, while the occurrence of cat skulls with cut marks in the later layers from the Middle Ages indicate a shift (and maybe a centralization) in the production of cat fur from the import of hides to the breeding of the animals in the town itself (Andersson 1993:25-26). At Birka, the distribution of skin and fur from fur-bearing animals such as fox, marten, beaver, squirrel, lynx or hare is evident from the bone material, mainly in form of phalanges, which indicates that the furs were imported to Birka as hides with the paws attached (Wigh 1995(Wigh :88-89, 1997. However, the recorded cat bones from Birka stem from all parts of the body. They show no cut marks from skinning and many cats were found as articulated skeletons (Wigh 1997(Wigh :609, 2001. Although a larger population of domesticated cats existed in Viking Age Birka, for some reason cat fur was not used as trading goods or dress accessories. So far, no evidence for the use of cat fur could be detected in the textile material from the Birka graves (Geijer 1938:185-186;Ågren 1995).The only exception is a man's grave with bones from a cat's paw (grave 886), which can be discussed as potential evidence for the use of cat fur as trimming for clothes.
One possible explanation for this situation could be a greater importance of cats in the pest control of small rodents such as the house mouse (Mus musculus), as is documented at Birka from the first phase of the settlement onwards (Wigh 2001:126-127), or the black rat (Rattus rattus), which appeared at Birka at the beginning of the 9th century (Wigh 2001:125-126). The significance of domesticated cats in pest control, especially for rural communities, is mentioned in several early sources (Edwards 1990:59;Benecke 1994:160;Kelly 1997:122;Poole 2015:865) and is already evident from their name: 'The Old English term for cat was catt or catte; whilst of uncertain etymology, it may stem from the Latin cattus, a term which Isidore of Seville says derives from catching (captura)' (Barney et al. 2006:254). In the Old Irish Laws of Hywell Dda the king's cat is praised as a valuable mouser (Richards 1954:92;Clutton-Brock 1993:41).
Also, the lack of cat fur in Birka might be explained with the high esteem in which domesticated cats as prestigious pets were held. In at least eight graves from the cemeteries of Birka 5 cat bones were found; most of these graves were richly furnished (figure 4). While the sex of the deceased in three of these burials (graves no. 1, no. 11, no. 696) remains unclear, one grave (grave no. 151) was the burial of a cremated female, buried with jewellery, horse equipment and both burnt and unburnt bones from several species (dog, bird, pig, cattle, horse and cat). According to the grave goods, three graves with cat bones were the burials of men with weapons, other prestig- The Warrior and the Cat ious artefacts and bones from several animals. 6 Due to the state of preservation (and archaeological recording) it cannot be ascertained whether the cats in these graves were deposited only in parts or as intact carcasses.
A small figurine made of amber (SHM 8252:1) was found in the settlement area of Birka (the 'svarta jorden') and seems to represent a sitting cat (Arwidsson 1989:53-54). The exact function of this c. 3cm long figurine is unclear, but as it does not have any holes for suspension, it may have been a children's toy (figure 5). As cats were becoming more common as grave goods for children during the early Viking Age (Andersson 1993:14) -for example in two eleventh-century burials of children, where a cat was the only grave good (Andersson 1993:16), it may be possible that the small figurine represents an increasingly important role of domesticated cats as pets, even for children.
6 Grave 477; spear, shield boss, whetstone, gaming piece, comb, penannular brooch, unburnt bones of cat, cock, pig and cattle. Grave 628; two shields, arrow heads, penannular brooch, horse equipment, wooden bucket with iron fittings, unburnt bones of cat, pig, beaver, cock, cattle. Grave 886; sword, shield boss, penannular brooch, weights, Arabic dirhams, gaming pieces with wooden gaming board, silver braids as trimming of some form of cap, unburnt bones of a cat's paw. This relevant aspect of the emotional bonds with cats as pets and companions becomes apparent in an Old Irish legal source, which apart from their function in pest control mentions the ability to purr as important attri bute (Kelly 1997:122). Purring, developed as a signalling mechanism of reassurance between kittens and their mother, persists as an expression of contentment and affection, mainly in physical contact with conspecifics or humans. Thus, the explicit requirement of the ability to purr indicates that cats were also appreciated as pets (Poole 2015:873). These increasingly closer relations between humans and domesticated cats can also be detected from changes in dentition and growth of many cat populations in several settlements and early towns, with increasingly less species-appropriate nutrition based on kitchen refuse (McCormick 1988:223-224;Benecke 1994:213, 229;McCormick & Murray 2007:116;Poole 2015:873-874).

Discussion
The picture of the role and function of cats in Viking Age society as depicted by archaeological as well as historical and later mythological sources is highly ambivalent. Cats appear as grave goods in high status burials of both females and males. They were kept as pest control and presumably also as pets. They were slaughtered for their fur and mythological traditions link them with a special cultic role, associated with magic and female fertility. It appears difficult, indeed meaningless, to ascribe to them a consistent and defined symbolic meaning. They inherited a broad range of functions which, in part, seem to contradict each other -on the one hand as grave goods in high status burials and on the other hand as provider of fur, whose carcasses could be dumped in waste pits. Furthermore, the investigations of horse burials, mentioned above, illustrate that even animals with an ostensibly defined function and symbolism can change their meaning in burial contexts. It can be assumed that cats had a multitude of meanings, dependent on their individual ability as agents to alter the human-animals-relationship in a particular context (Morris 2012:14;Poole 2015:862-863). Their function in life reality -as pest controller, as companion and pet, and as a provider for fur -seems to be obvious. However, the symbolic meaning associated with cats as a sacrifice in burials or as a spiritual animal in mythology and cosmology can only be speculated upon, especially as one has to be aware of the potentially changing significance of specific animals in this and the other world (see Bond & Worley 2006:90). Cats -or cat bones -as grave goods appear regularly from the Vendel Period onwards. It is not always certain whether the cats were ritually killed during the funeral ceremony as a sacrifice, or if the killing was a The Warrior and the Cat profane act that was not incorporated into the ceremony. Furthermore, due to the fragility of the bones, it often remains unclear, whether the whole carcass was deposited in the grave or if only certain parts of the animals were used as a pars pro toto (see Vretemark 2013:52). Based on the fact that cat fur was used for clothing, it cannot be ruled out, that cat phalanges result from fur trimmings and have no specific symbolic meaning as grave goods. At first, they are most common in male burials of a higher status, while they were almost equally distributed in the graves of the early Viking Age. Their increasing occurrence in infant burials thereby contradicts the classic interpretation of cats as being primarily associated with seiðr or other cultic-ritual aspects, as do the finds of cat bones in the burials of weaponcarrying males, or cats as decorative elements within Viking Age art -for example as the 'gripping beast' in Borre style (Steuer 1994:650;Reichstein et al. 2000:334-335) -on swords (figure 6), sword chapes, horse gear (figure 7) or other artefacts (Hedenstierna-Jonson 2006) that are associated with the male sphere (Steuer 1994:659-660).
The classic explanation for the appearance of such an uncommon grave good would be as a symbol of high status (see Leifsson 2018:316) due to the exclusive and exotic character of cats as a new and extraordinary element of the Old Norse fauna (see Pluskowski 2004). This interpretation is supported by the increasing frequency of cats in settlement contexts during the Viking Age and, accordingly, also in average burials. So, in the Vendel Period and (early) Viking Age the domesticated cat might -on one levelhave been a profane, prestigious symbol of status, possibly originating in the exotic appearance of the cat as a relatively new element of the Scandinavian fauna, with its strong and independent character, individual agency (see Poole 2015) and its elegance as a skilled hunter. However, this interpretation is strongly unidimensional and does not take into consideration the individuality of the cat as a living being and agent, as well as the multiplicity of potential meanings of artefacts in a burial context (see Toplak 2017:128-130) and the difference between this world and the other world.
An alternative interpretation or addition to this 'social' meaning of cats in burials is based on the investigations on horse burials in Anglo-Saxon England and Viking Age Iceland (Bond & Worley 2006;Fern 2007Fern , 2012Leifsson 2012Leifsson , 2018Cooke 2016), which refer to concepts of common ancestry, origin or mythology. Investigations of horse burials in Anglo-Saxon England have shown that horses in cremation burials -in contrast to horses in inhumations -were not limited to a male social elite and might reflect memories about descendant and mythical ancestors (Fern 2007(Fern :100-102, 2012. The horse as spirit guardian was utilized as an identity-creating symbol, strongly connected with the collective descent from mythological and elitist ancestors (see Cooke 2016:11). On Iceland, horses were used as Matthias S. Toplak The Warrior and the Cat symbols by a certain group of landowning families (Leifsson 2018:323-324), and demonstrate a commingling of social status (as landowning elite) and common descendant (relating to the first settlers from Norway, who took possession of the pasture land during the Landnám). Ritual sacrifices of animals during the funeral ceremony might have been intended as performative acts in an intentional created situation of communication with the social community to illustrate or manipulate references and links to mythical ancestors, mythical or cosmological ideas and certain identities (Williams 2006:36-46;Döhrer 2011:13-14;Cooke 2016:2;Leifsson 2018:327). It might even be speculated that the sacrifice of cats at the funeral was intended as a presentation of a certain identity, linked to a specific descendant, to a common mythology, to traditions or ancestors, even if the definite symbolism of cats remains obscure.
Even the simple individual emotional bond between the deceased and his cat in its function as a companion animal or beloved pet, without any further social or religious aspects, might be a potential explanation for cats in burials. This seems plausible in regard to children's burials with single animals as cats or dogs. It is also supported by the strong emotional bonds existing between humans and cats as mentioned in the poem about the cat Pangur, and the Old Irish law texts (Poole 2015:872-873). However, this interpretation remains unproven. Furthermore, the osteological analysis of horses from Viking Age burials has shown that many of the animals were simply too young to have been ridden by the persons with whom they were buried (Leifsson 2018:316-317) so that a strong emotional bond between rider and horse, and in general, between deceased and sacrificed animal cannot be taken for granted, even for such important animals as horses on Iceland. It was not the individual horse with its characteristics, agency and shared experiences which was important for the funeral ceremony, but merely the symbol 'horse'. The same might also be the case with cats, even though this result relativizes the significance of the animal's individual agency.
Despite the complex problems of source criticism surrounding the mythological transmissions concerning the mythical role of cats and the occurrence of cat bones in the burials of armed men, a ritualistic symbolism cannot be ruled out. Due to their nocturnal activities as predators, their elegant movements and ability to reach almost every place in total silence, it seems likely that cats may have been regarded as psychopomps, which were able to cross the border between this world and the other world and to guide the souls of the deceased into the afterlife. This speculative interpretation would be in accordance with the strong association of cats with the supernatural sphere in Old Norse saga literature and mythology. Apart from their function as psychopomp, cats -and other animals -could also have been intended as sacrifices to certain deities to ensure a good afterlife for the deceased (Kaliff & Østigård 2013:88) or as requisites (or actors) for the staging of specific cosmological myths (Williams 2006:36-46) and burials as 'mnemonic events' (Fern 2012:172; see also Bertašius 2012or Leifsson 2018. As graves are the static end result of a complex, multidimensional and multimodal burial ceremony (Staecker et al. 2018:63-69) which involves a wide and often no longer traceable array of rites as well as protagonists and paraphernalia, the original intention for the sacrifice and/or deposition of animals as grave goods might have been multicausal, contextual and dependant on the perspectives of the specific actors (Toplak 2017(Toplak , 2018; see also Jung 2008:1 for the concept of 'Überdeterminiertheit' ('overdetermination') of burials.). As regards the animals, even their individual character, their agency and the specific human-animal-relationship must be taken into consideration as potential motivations for the sacrifice or deposition and also the manipulation of their meaning in the specific ceremony (Morris 2012:14;Poole 2015). So, it must be assumed that the rite of cat sacrifices in Vendel Period and Viking Age burials cannot be explained by one single symbolic meaning, which has validity for the entire cultural area. Instead it has overlapping functions and meanings (see and Svanberg 2003 for the diversity of Viking Age burial rites). Cats might be regarded as psychopomps by one specific social group because of their symbolism as a totem or spirit animal, or it might have been the individual animal that was deposited as a pet in an infant's burial, owned in their lifetime as an exclusive status symbol.

Conclusion
A closer investigation of the distribution of cat bones in the archaeological record in Viking Age Scandinavia produced no evidence to confirm the unidimensional traditional perception of cats being especially associated with the female sphere of cult and magic. While cats appear as grave goods in both male and female burials of a social elite in the Vendel Period and early Viking Age, cat fur was regularly used as a common material for the lining or trimming of clothes in the early urban milieus of the later Viking Age. This aspect supports the general validity of the passage in Eiríks saga rauða concerning the use of cat fur. As cat fur seems to have been a part of men's dress which did not conflict with the presentation of the identity as a warrior in the burial context, it therefore illustrates Þorbjǫrg's high prestige and is not directly or exclusively linked to femininity, cult or magic.
The holistic and source-critical investigation of the role and function of cats in Viking Age Scandinavia presents a far more nuanced view of the domesticated cat, which became a symbolically, functionally and probably also emotionally highly important companion for humans beyond the borders of sex or gender shortly after its first appearance in northern Europe. It demonstrates quite clearly that the previous, repetitively postulated association of cats with a female sphere of cult and magic, which is mainly based on highly sporadic and partly inconsistent references in Old Norse literature must be critically reconsidered. In contrast to the case of horses or dogs, where their functional and symbolic roles are mostly regarded as equivalent, the ambivalent picture and nebulous symbolic meaning of cats, in contrast to their function in everyday life, illustrates that the individual agency of animals is of highest importance for the contextualization and perception of their symbolic role in the human-animal-relationship.