Borderland-stalkers and Stalking Horses. Horse Sacrifice as Liminal Activity in the Early Iron Age.

Published in "Current Swedish Archaeology" vol.14. 2006

I've Read This

Please login or signup to add this paper to your Papers I've Read list

Added to your Read List

Readers (4)

• 36 Views
• Like Liked (1) Undo

Open Preview Close Preview Fullscreen Download (.doc)

Borderland-stalkers and stalking-horses. Horse sacrifice as liminal activity in the Early Iron Age. Anne Monikander Keywords: Borders, horse sacrifice, wetland sacrifice, bracteates, Grendel, Miðgarðr, chaos, myths, the Miðgarð Serpent. Abstract This paper investigates the fact that many Early Iron Age wetland sacrifices are deliberately placed in liminal zones. Sacrifice of horses turned them into liminal creatures by manipulating their heads, hoofs and tails. In this way they imitated the liminal creature above all, the Miðgarð Serpent, who bites its tail in the ocean surrounding the world, and the horses too became guards against dangerous powers from outside Miðgarðr. The attention to tongues can also be seen on the bracteates which may have served as amulets reminding their users of the creatures guarding the borderlands. Borders and walkers in the borderland Quite a few of the most well known Iron Age wetland sacrifices in Sweden are situated on, or close by, parish borders. For example the fen of Skedemosse on Öland is placed where the borders between the parishes of Köping, Bredsättra and Gärdslösa meet. Finnestorp in Västergötland is on the border of Larv and Trävattna parishes, Käringsjön in Halland is between Övraby and Enslövs parishes and Lillemyr on Gotland is between Barlingbo and Endre parishes. Smaller, and less known is the sacrificial bog of Frösvi in Närke, which is situated on the border between Edsbergs and Hackva parishes. Both Skedemosse and Finnestorp are also situated on borders between different hundreds (sv. härader)(Lindqvist 1910: 119, Hagberg 1967a: 10 Fig. 3;1967b: 72 ff.). There are, however, a number of wetland sacrifices to which these conditions do not apply. Gudingsåkrarna, Eke parish and Roma, Roma parish on Gotland, Hassle-Bösarp in the Scanian parish with the same name and Äversta, Glanshammars parish in Närke are exemples of this (Hagberg 1967b:72 ff.). The same seems to be the case in Denmark. Of a few randomly picked wetland sacrifices there, Illerup, Dover parish, Hjortspring, Svenstrup parish, Ejsbøl, Gammel Haderslev Landsogn parish, and Vimose, Allese parish, appear to be on, or very near, parish borders, while Valmose, Rislev parish, Nydam, Sottrup parish and Kragehul, Flemløse parish are not (Trap 1953-1972). It is uncertain whether parishes and hundreds in Sweden reflects pre-Christian conditions or not. The topic has been cause for lively debate during most of the last century (see for instance Bebyggelsehistorisk tidskrift 1982:4). It appears that the age of the parish borders vary between different Scandinavian regions and in some cases they may correspond with older boundaries, but in most cases they do not. In Denmark the situation is much the same. The construction of parishes was not necessary until the beginning of collecting thites (Brink 1990: 113 ff., 120 ff.; Blaaberg 1992:15). I do not believe that the parishes reflect an older, pagan geographic unit. They were designed by the Christian clergy to respond to the needs of a congregation but I think that in some cases already existing borders may have been used. The parishes may occasionally have been constructed for reasons such as population density or natural borders. It might for instance have been practical to place the border between parishes on marshy and useless land which was less likely to give cause to conflicts of ownership. It might also have been practical to use the presence of a natural border such as a river or, indeed, a fen. Pagan society, on the other hand, may also have used existing natural borders, or they may have thought it fit to locate their sacred sites outside the inhabited area but still on a place which was easy to access from several different settlements. However, I would like to suggest
1

that this placement of sacrificial sites on geographic, demographic and/or judicial borders is not coincidental but has a deeper meaning. The discussion of the farm and the cultivated lands inhabited and ordered by men in opposition to the wild, uninhabited and foreign lands is wellknown in archeaology as the Miðgarðr-Utgarðr-debate (Ström 1999: 96 f., Lindgren-Hertz 1997: 48 f., Andersson & Hållans 1997: 585, Brink 2004: 291 f.). For more than a decade it has been the dominant theory about how the pagan Scandinavians perceived their universe. More recently, however, it has come to be seen as being too simplistic. The pagan world view was much more complex and consisted of several dimensions in different positions around Miðgarðr (Brink 2004: 292 ff., Clunies Ross 1996: 50 f.). In fact the word Útgarðr only occurs once in Snorris Edda, as the home of Útgarða-Loki, and it is never mentioned in the skaldic verses (Clunies Ross 1996: 51, Snorres Edda Gylfaginning 45 1997: 72). ”Rather than posit a binary opposition between the central space and the world outside the fence, there seems to be more textual support for a spatial conceptualisation of a series of territories belonging to different classes of beings arranged like a series of concentric halfcircles, the perimeter of each circle being imagined as a kind of protective rampart, a garðr” (Clunies Ross 1996: 51). Thus any attempt to turn Snorris descriptions of the pagan world view into a logical spatial model or to outline it in a schematic illustration is an intellectual anachronism (Brink 2004: 195 f.). Since our sources of the myths were written down as late as the 14th century and thus are much younger than many of the archaeological finds it is always problematic to use Scandinavian pagan mythology to interpret prehistoric society. We do, however, have other sources to the mythic world from the earlier, pre-litterate epochs, such as picture stones and jewellery and when these correspond to a specific myth or to a cognitive stucture which is also represented in the written sources it would be wrong to disregard the mythology. We must, of course, compare and interpret with great caution and be aware that we look at both written and archaelogical material through the multiple filters of two millenia of changeing religions and social conditions. But if we do this, and at all time try to be aware that our interpretations always reflect our pre-judice, I think it is possible to use the myths to understand the archaeological material. As Margaret Clunies Ross puts it: ”any modern interpretation can only be partial and suggestive of social relevance and modes of thought rather than indicative of them” (Clunies Ross 1996: 15). Clunies Ross proposes a description of the pagan world view based on different classes of beings in the Old Norse myths.Gods, humans, and elves belong to the first class. They are dominant and normative and the myths are told from their point of view. Giants and dwarfs belong to the second class. Their domicide is seen as outside of Miðgarðr, not because of its actual location but because the users of Old Norse myths were willing to accept ”the world of the gods and humans as central and the interests of others as peripheral” (Clunies Ross 1996: 54). Although I agree that the binary opposition of Miðgarðr-Útgarðr is too simple, and that the point at issue is problematic and needs further clarification, I still consider it justified to speak of ”otherworldly” creatures and places. Bogs and wetlands must be categorized as out-lands in the most literal sense of the word. As seen above they are often borderlands between different inhabited areas, which in some cases later might have become parishes. They are the limit of what people in one specific settlement or area thought of as ”ours” and the beginning of ”theirs” and ”the Realm of the Others´”. One of the earliest sources we have on how Early Medieval people regarded bogs and wetlands is the Anglo-Saxon poem Beowulf (Beowulf 1958). The poem was written by a Christian poet but it describes a pre-Christian epoch and it is generally agreed that the poet did his best to capture the minds of the heathen protagonists and that he conciously uses an archaic
2

language (Tolkien 1975: 35 ff., 84 and note 20). On linguistic grounds the earliest probable date of composition is the 8th century (Beowulf 1994:5). In Beowulf the creature Grendel and his mother are referred to as belonging to the race of eotenas (l. 112) and thyrs (l. 426), giants (ON. jotnar; tursar). They are also called maercstapan (l. 103), border-stalkers. In the glossary to his 1958 edition C. L. Wrenn translates it as: ”walker in the borderland; haunter of the bordercountry” (Beowulf 1958 s. 273). In the context of the poem it is clear that it is not a question of proprietal or national borders but the border between the known world, ordered by man and exemplified by Heorot, and some dark and ominous dimension outside. The home of the monstrous creatures is situated on the bottom of a bottomless lake. It is described as a great hall but the way to the lake is bleak and its surroundings are filled with fog, frost and darkness (Beowulf 1958 l.1397-1430, 14921590). It is also noted in the text that the lake is filled with wyrm-cynnes (l. 1425), a race of serpents and sæ-draca (l.1426), sea-serpents (Beowulf 1958: 282, 306). In these circumstances it is especially interesting that the poet speaks of Grendel and his mother as eotenas and thyrs. In scandinavian pagan myth these are the creatures who live outside Miðgarðr and who are indeed hostile toward the gods and the race of men. It would seem that Beowulfs poet and his contemporary audience thought that the creatures who lived in the fens and walked the borders of the world of men were identical to the frostgiants of the old myths. The scholar of Old English literature, Edward B. Irving Jr., discusses the action in Beowulf as a combat between the good forces who create and defend the order of the world of men by fighting the evil forces of chaos and disorder ”and thus protecting the `cultivated plot´ from the `encompassing wilderness´”(Irving 1968: 86). Irving presented his theories before the concept of Miðgarðr-Útgarðr became popular in arcaeological circles, but his formulation sums it up very neatly. The pagan Scandinavians may never have used the concept of Útgarðr, in the way modern archaeologists do, but I am convinced that they did conceive of certain parts of the world as Outside or as borders to the Outside. What Irving refers to as the ”cultivated plot” is not necessarily the lands belonging to a particular farm, family or people but the idea of an ordered and controllable world. This world is called Miðgarðr and the ”encompassing wilderness” outside is not a certain place, but any place or situation which evades the normative order and culture. Wetland sacrifices of horses During the Iron Age all kinds of animals but above all horses were sacrificed in lakes, bogs and wetlands, sometimes as part of larger and more varied sacrifices. The deposits naturally differ from each other, between regions and over time but they also have a lot in common. In Valmose near Rislev in Jutland a number of animals, humans and artefacts have been sacrificed on different occasions. One of the horses, Nr. VI, was found in situ with its tail placed in its mouth with the thick end first. The presence of only a part of one of the stylohyoid bones indicates that the tongue had been cut out before the tail assumed its place. Together with the head and tail was found the complete lower extremities of the same horse. Altogether eleven horses were found in Valmose. Six of them were deposited with head, hoofs and tail in much the same way as horse Nr. VI, the five others were less complete. The remaining parts of the horses were probably eaten. In fact, horse Nr. XII, is really a conglomerate of marrow split bones from several individuals, which the excavators interpreted as the left overs of the other horses (Ferdinand J. & K. 1961: 57 ff.). When sacrifices of horses during the Iron Age are discussed, it is generally assumed that they took place in the same way as they did in historic times among some Siberian nomadic peoples, i. e. the horse is slaughtered and skinned in such a way that the head, the tail and the feet are still attached to the skin. This is hung up on a pole and the meaty parts are consumed (KlindtJensen 1957: 84 ff., Ferdinand J. & K. 1961: 79 ff.). Horse Nr. VI in Valmose clearly shows that this is not the case here. The tail lacks the caudal vertebrae 1-4, that is, the ones which are
3

situated inside the horses body. Both this and the fact that it was placed in the mouth of the horse with the thick end first contradict its being attached to the skin at the time. Since the caudal vertebrae lay in order it is plausible that the tail was fresh at the time of the deposit. The sacrifice was dated to the 4th or 5th centuries AD based on the typology of deposited ceramics (Ferdinand, J.& K. 1961: 73). Remains of four humans were also found in the Valmose sacrifice. Only two of them have been preserved and recently they have been radiocarbon dated to the Pre-Roman Iron Age. Skeleton I was dated to 380 BC cal. (K-3598) and Skeleton II was dated to 225-335 BC cal. (K-3599); (Jansen Sellevold et al. 1984: 96 f.). ”The human bones, which comprised of two complete skeletons and some bones from two other individuals, were in fact found a little deeper than the rest of the find, but, because of the general instability of the soil, this was not considered to be important” (Jansen Sellevold et al.1984, App. B, p. 248). This ”general instability of the soil” may of course put some doubt to the age of the animal bones. However, their general likeness to other animal sacrifices of the Roman Iron Age, and not least the presence of the pottery, makes me regard the excavators date as valid. The Hjortspring find is the oldest of the Iron Age weapon sacrifices and wood from the boat has been radiocarbon dated to 350-300 BC (Arkæologiska udgravninger i Danmark 1987: 230, 240). It is most famous for the find of the well preserved boat but at least one horse was also found here, the bones of which were spread in the southeast corner of the excavation area. It had been deposited on its back, partly under the boat, so that ribs and long bones had perforated the boardplanks. Its head, which was found about 1 m from the rest of the skeleton was turned the right way up. The lower jaw was in its place and the tonguebone lay in situ in the peat. In the horse´s mouth one of its splint-bones had been deliberately placed between the mandibles in the same position as the tongue (Rosenberg 1937: 33 ff.). Around the boat were also found a young sheep, a dog and a puppy, sacrificed at the same time as the horse. A tibia from another, larger horse was also found in this context (Rosenberg 1937: 34). Below the finds dated to the Pre-Roman Iron Age were deeper wells in the bog, where more remains of animals were found. These probably belong to the Bronze Age (Rosenberg 1937: 35). In Skedemosse on central Öland large quantities of weapons were sacrificed in the 2nd to 6th centuries AD (Rau, A. 2005: 632), but sacrifices of animals began already in the 4th century BC (Hagberg 1967a:103). Among these are the remains of about one hundred horses. As in Valmose they seem to have been deposited with a special attention to head, hoofs and tail whereas the rest of the bodies have been eaten. When the meaty parts were deposited they were marrowsplit and broken, a condition not made better by being washed by the waves of what was then a lake (Hagberg 1967a: 79, Boessneck et al. 1968: 16). It is thus more difficult to say anything about the original positioning of the bones than at the two previously mentioned places. The sex distribution was fairly even and no particular age of animals was preferred (Boessneck et al 1968 s. 16). Two horses have been radicarbon dated. F 307 (St-1975)1935+75 BP, calAD 73 +-91 and F 139 (St-2383)1825+-100 BP, calAD 204+-117 (Hagberg 1967a p.91,103, Hagberg 1967b p.138, my own calibration made with the help of www.calpal). In only six cases has it been possible to collect more or less complete skeletons. These are F 1, F 218, F224, F 545, F 1053 and F1055. Strangely enough only two of them have skulls and only three have tails (Boessneck et al. 1968: 19 Tab.4). Since heads and tails as mentioned above are generally over-represented compared to other parts of the skeleton this suggests that special interest was directed towards these parts and caused them to end up separated from the rest of the body. In Illerup Ådal, the most recently excavated and perhaps most well known of the large weapon sacrifices in Denmark, remains of horses have been found. They show severe head injuries as a result of blows of sharp and heavy objects. Brøndsted shows a picture of a horse from Illerup with cracked forehead and a pierced eye and rib (Brøndsted 1960: 225). Ilkjaer describes how several horses were led out in the water, cut to pieces and left were they fell. A
4

hind leg of a horse was packed together with weapons in a piece of cloth and deposited with them in the lake (von Carnap-Bornheim & Ilkjaer 1996, Ilkjaer 2002: 203, 211, www.illerup.dk). Remains of three stallions were found in the weapon offering of Nydam I, scattered among the other finds and with iron bits still in their mouths. Nydam mose was first excavated by Conrad Engelhardt in 1859-63 and the skeletal material was investigated by Professor Japetus Steenstrup (Engelhardt 1865). The oldest horse, 10-11 years old, is the most complete skelton and was probably deposited whole. The workers on the site claimed that they found it with its legs pulled up beneath it, but Engelhardt would not take it on his word. The second horse, a 6 year old, consisted of the skull and the lower extremities and of the third, a 3 year old, only the skull was found. The bones are well preserved but appear to have been weathered for a while before they were covered with peat. They have also been gnawed by wolfs or large dogs. The pelvis of the oldest horse is damaged and the youngest´s nose, its processus styloides and the vertebrae of the neck have been partly eaten away. Steenstrup says that the fact that only the parts wich are usually attacked first have been gnawed suggests that the carnivores had only a brief access to the bones. All three horses have been victims of severe violence, and have the same type of injuries as the horses from Illerup. Their skulls show traces of heavy blows of swords, in some places hard enough to cut all the way through the bone tissue. The impact angles show that the horses must have been on the ground when they were injured. Since the lower jaws are untouched by the blows, but show severe gnaw marks, Steenstrup speculates that they had already fallen from the rest of the skull when the sword blows were delivered. He also says that the sharp edges of the cut marks in the bones of the lower extremities, and the little splintering of the bone indicates that the bones were skeletized and not very fresh when they were cut (Engelhardt 1865:39 ff.). However, it is in fresh bones the cuts leave sharp edges, while old and dry ones tend to splinter. The drawings of these bones with cut marks look as if the horsees were injured in a fresh condition, as were the horses in Illerup (Molnar, verbal commentary 2005-09-13). The oldest horse has been shot in its left shoulderblade with an arrow with a square iron head and one of its ribs has a smaller, triangular hole. Traces of rust show that the projectiles were left in the wounds when the horse ended up in the bog (Engelhardt 1865: 37 ff.). In view of the Illerup horses which were similarly treated, and the fact that the bone tissue had to be fresh for the arrow head to get stuck in it, it is possible that at least the oldest horse in Nydam I was also cut down alive. In 1984 a new weapon find, the so-called Nydam III-find, was excavated in Nydam mose. The find consisted of heaps of swords and spears from the Early Migration period. Spread among them were two teeth and an elbow-bone from a medium sized horse. The elbow has been radiocarbondated to AD 280 +-50 = 265-429 calAD (K-5129) (Petersen 1987: 121). These South Scandinavian places can be contrasted to the sacrificial site of Oberdorla in Thüringen. It has been in use during approximately the same space of time as Skedemosse, but only 24 horses have been sacrificed in Oberdorla compared to 100 in Skedemosse. The same figures for bovines are 114 and 80 respectively (Behm-Blancke 2003: 98 Abb.40). This suggests an entirely different view of which objects that were proper to sacrifice. In this context it is particularly interesting that no caudal vertebrae at all were found in Oberdorla, altough heads and feet seem to have been deposited together as in the other places. Since the much smaller tonguebones have been found (Teichert 1974: 52 f.) this points to a quite different sacrificial ritual but still one which payed special attention to the tails. The presence of large numbers of poles and stakes with holes or cross-bars in them indicates that different sacrificial objects were suspended as part of the rites at Oberdorla (Behm-Blancke 2003: 54 f.). However, if this is what happened to the tails, one would think that at least some caudal vertebrae would have fallen down when the tails decayed and consequentely been found
5

in the excavations. When none is, this implies that the tails were removed from the site and deposited somewhere else. Bracteate beasts and liminal creatures The creature usually connected with the limit of Miðgarðr is the Miðgarð Serpent. This child by Loki and the giantess Angerboda lie in the ocean that surrounds the world, while he bites his own tail. In this way he is depicted on the reverse of a bracteate from Lyngby on Jutland (Ellmers 1970: 222 Fig. 23). Birgitta Johansen discusses the many liminal functions of the serpent in Old Norse traditions (Johansen 1997). She states that the serpent is the boundary between gods and giants at the same time as it encircles the world. The borderland it is guarding is situated between the known and settled region and the unknown wilderness, or between different regions, or it can be in an area which is not land nor water. She also refers to Beowulf as an example of how the dangerous powers in Old Norse cosmology were seen as inhabiting bogs and desolate fens and marshes. The Miðgarð Serpent, she concludes, maintains the border between chaos and order, between the world and its destruction (Johansen 1997: 42). Margaret Clunies Ross, on the other hand, defines the Serpent as belonging to a third of her classes of creatures, which represent natural forces of the kind Þórr meets at Útgarða-Loki, such as fire, old age and the sea (Clunies Ross 1996: 52, 258). I do not agree with this, the Serpent lives in the sea, it is not an allegory of it. The Serpent is a threat to the world of men and as such it will emerge in Ragnarøkr, but until then it is a necessary defence from things outside the world, which could be a lot worse than itself. When Þórr in Hymiskviða tries to pull the Serpent from the sea he almost starts the end of the world prematurely (Den Poetiska Eddan 1964: 107 Hymiskviða verses 23-25, Johansen 1997: 42). As I see it, the Serpent´s mixed genealogy, contributes to making it into a liminal creature, belonging to both Miðgarðr and the world outside, and to none of them. To return to the bracteates, the most frequently depicted animal on them is a big, horned, horselike beast. This animal has been interpreted as a billy-goat, a bull, a horse or as a combination of them all. Most scholars seem to think that the horse is the most plausible beast, and they also seem to agree that the bracteates depict scenes from the Old Scandinavian mythology (for instance Andrén 1991, Axboe 1997:122 f., Gaimster 1997:147 f.). Different gods as Odin, Tyr, Loki and Balder can be seen on them. One of the most well known interpretations has been put forward by Karl Hauck. In a number of publications he suggests that the animal is Balder´s stumbling horse as it is described in the second spell from Merseburg (for example Hauck 1970, 1988, 1993). The two spells from Merseburg was found in the Chapter Library in Merseburg, jotted down inside a prayerbook from the 10th century (www.fh-augsburg.de). The second spell is a magic cure for sprained or broken legs of horses. It belongs to a canon of Old High German spells aimed at the curing of both horses and people but it is unique because it is the only one which does it in the names of Old Scandinavian gods (von der Leyen 1935:205 ff., Hampp 1961:247 ff., Eis 1964). It relates how Phol (Balder) and Wodan ride through a forrest when Phol´s horse stumbles and injures its leg. A number of gods partake in its healing and Wodan finally succeeds. It is Hauck´s belief that this is a central scene from Norse mythology and that the injury and death of the horse is an omen of Balder´s death. To the pagan users of bracteates Wodan´s ability to heal and resurrect the horse was a sign of hope for Balder´s return and this explains the popularity of the motif. The divine head depicted above the horse on the bracteates Hauck interprets as Odin in the act of healing the horse. Odins recital of sacred words is shown as an”Atem-Chiffre”, lines coming out of the god´s mouth or as the ear of the horse being put in the god´s mouth. The ”Atem-Chiffre” also refers to Odin´s character as lord of the wind (Hauck 1993: 408 ff.). Most scholars agree with Hauck altough not all of them go into such detail as he does (see for instance Kolstrup 1991: 185, Axboe1997:122 ff., Gaimster 1997:148).
6

A different explanation is suggested by Detlev Ellmers. He also thinks that the animal is a horse, although not Balder´s, but the horse which was sacrificed in the rituals he believes is described in the runic inscriptions of the bracteates. The depicted horse has been raced to exhaustion and is then killed, and this is the reason for its outstreched tounge and the unnatural position of its legs (Ellmers 1970: 237 ff.). Ellmers links the animal depicted on the bracteates to the horses on picture stones, such as the Häggeby stone from Uppland and the stone from Roes on Gotland (Ellmers 1970: 205, 243 f.). Ellmers is not the first to connect horse sacrifices with horse races. Hagberg discusses this extensively with reference to the Skedemosse sacrifices, but I do not intend to go further into the subject here (Hagberg 1967b: 85 ff.). Recently Anders Kaliff and Olof Sundqvist have explored the connection between the bracteates and the continental cult of Mithras. They think that the bracteates through acculturation borrowed iconographical details from pictures used in the cult of Mithras but that the scandinavians filled the borrowed symbols with their own meaning (Kaliff & Sundqvist 2004: 89). Thus the animal and the god are depicted in the same pose as Mithras killing the bull, but when a scandinavian saw this he read it as Odin and a horse. It seems to me that Hauck and Ellmers theories are mutually contradictive. Hauck gives a plausible explanation of the object protruding from the gods mouth, but he does not explain why the horse should have been depicted with the attribute of a wind god, especially if the horse as he claims belongs to Balder, who has no connection what so ever with climatic conditions. Ellmers on the other hand explains the outstreched tounge of the horse, but does not say why the god should put his out. Of course one could argue that Odin, as the hanged god, has every reason to do so, since hanging or strangulation causes the tounge to swell and force its way out between the victim´s teeth. Thus depicting Odin this way might simply allude to his self-sacrifice and his capture of the runes.(Den poetiska Eddan 1964: 70 Havamål vers 139-141), but Ellmers does not make this connection. Nor does this explain why Odin is chewing on the horse´s ear. Another interpretation which may explain the objects protruding from the mouths of both the god and the horse would be that they symbolized prophetic abilities. Hauck touches upon this when he discusses bracteates where Odin is sucking his thumb (Hauck 1988 p.470 f.). Already Tacitus writes in Chapter 10 of Germania that sacred white horses were kept in holy groves and used to foresee the future (Tacitus 1914: 278 f.) If this is true, the ”Athem-Chiffre” with both the god and the horse could signify prophesy, as could the god speaking into the horse´s ear. Guarding against chaos The sacrifices of horses in lakes, bogs and other wetlands are generally older than the bracteates and seem to have ceased in the 6th century AD, but they continued in pits under houses and in the so called horse-burials in connection with human graves or on the limits of burial grounds (Müller-Wille 1970-71: 185 ff.). Bracteates were manufactured and used roughly between AD 450/75-525/560 (Axboe 1998: 232). Even before the earliest Abracteates, animals were depicted on various ornaments with some strange object protruding from their mouths. Ellmers gives a number of examples (Ellmers 1970: 252 ff.). Although I do not find all of his examples convincing I would like to mention some that are. An animal shaped brooch from the 4th-5th century from Gotland (Ellmers 1970:253 Fig. 54), a langobardian goldbracteate from Poysdorf, Niederösterreich, and a very similar horse on a ceramic shard from Mähren (Ellmers 1970:254 Fig 57, 58), the Tangendorf disc-shaped brooch, the belt fitting from Skedemosse, and the second ornamented disc from Thorsberg (Ellmers 1970: 256 Fig.62, 63, 65). One might also mention the bronze bull´s head from Childerics grave, which have four long bands hanging from its mouth and which was found on the forehead of a horse cranium(Ellmers 1970: 249 f. Fig. 51). All these animals are not horses, and some are, like the beast on the bracteates, difficult to identify with regard to species. The animal on the belt fitting from Skedemosse is a very good example of this ambiguity. It has the
7

body of a deer, the feet and jaws of a wolf, and its tail and general stance also have some horselike features. As Hagberg states it is ”a creature of fantasy” (Hagberg 1967b: 24). These objects originate from quite different parts of Europe and are not from the same time. What they all do have in common is the long object hanging out from the mouths of the animals. Ellmers links this to horse Nr.VI from Valmose and suggests that the lines in front of the animal´s mouth on the bracteates and the other objects may be something that in the course of the sacrificial ritual was put in the mouth of the beasts. (Ellmers 1970: 241 f.). This may have been the tail as in Valmose, or a leg bone as in Hjortspring or something else which is not preserved. The fact that the ritual is focused on hoofs and tails, and that precisely these parts are found deliberately placed in the mouths of the horses´ is significant. As mentioned above, the liminal creature usually known to bite its tail is the Miðgarð Serpent. I would like to suggest that the sacrifice of horses in wetlands with various objects placed in their mouths is a way of creating a guardian of the border to the Outside by alluding to the Serpent biting its tail in the sea around Miðgarðr. Bogs and fens were seen as potentially dangerous places, borders to the unknown and uncontrolled. The sacrificial ritual transformed an ordinary horse into a liminal creature, something which just as the Miðgarð Serpent belonged to both worlds and to none. In this way it became a sentry against the dangerous forces living in the liminal zone of the fen but its aquired liminal character also made it one of them, a borderland-stalker. One problem with this interpretation is the variation in ritual between different sacrificial sites. In Hjortspring, Rislev and Skedemosse the horses have been slaughtered either by a blow to the forehead or with a method that leaves no trace of the killing. Only in the two first mentioned places manipulations in the horses mouths have been identified. The horses in Illerup were subjected to very grave violence and hewn to pieces while still alive. In Nydam I the violence was equally brutal but the evidence is contradictory as to whether they were alive or not. Perhaps they were just freshly slaughtered and lying on the ground while being cut up? This might explain why there are no traces of cuts on the mandibles. Since the arrow wound in the shoulder hardly is the cause of death, the horse may have received this at the same time. It is beyond doubt that at some point in the ritual, both at Illerup and Nydam I, the horses were used as stalking-horses and exposed to various projectiles being shot and hurled at them. Could it be possible to discern a gradual change in the ritual towards more flagrant violence? At first horses were killed for a purpose, they had to die to become liminal creatures, part of both this world and the Outside. Later on this was not enough and the act of killing itself became more important. In Illerup and in Nydam I they had to be cruelly mowed down by the participants for the ritual to have the desired effect. This also poses a problem, since with the exeption of Hjortspring, these sites are more or less contemporary. However, the sacrifice of horses at Skedemosse started long before the sacrifice of weapons and in Rislev there are no weapons at all. This may imply a ritual difference between sites with and without weapons, but Hjortspring is a classic ”spoils of war”-sacrifice and in Skedemosse the same type of horse sacrifice continued along side the weapon deposits. All things regarded I still find it plausible that the sacrificial rituals turned more violent over time. A similar change is also, in a way to be seen in the literature and in iconography. The pagan epics of gods and heroes are filled with fierce combats against serpents and dragons. Birgitta Johansen implies that the view of the serpent changed over time. To begin with it was perceived as having both positive and negative characteristics but in the Late Iron Age and Medieval Period it became increasingly threathening, especially to male individuals, and the only way to control it was to kill it (Johansen 1997: 106). If the sacrifice of horses in wetlands is indeed a way of transforming them into liminal guardians the possible increase in violence could be a reflection of this cognitive change. Any liminal creature has to incorporate in itself something from both sides of the limit, the Miðgarð Serpent is a chaotic being but it is also an insurance against chaos.
8

In the pagan cosmology the uncontrolled and unknown may stand against the known and ordered, but both are needed to balance each other. It may be that as influences and ideas from the Christian world spread in Scandinavia during det late Iron Age there was a cognitive shift which made the pagan chaotic powers to be seen as more malign than before, more in analogy with the Christian opposition of good versus evil. With this view in mind both the big horned animal on the bracteates and the animals sacrificed in wetland sites can be seen as representations of the liminal creature keeping guard against the uncontrollable. The bracteates were used as amulets, they may have been given as gifts on important social and political occasions (Andrén 1991:250 ff., Axboe 1991:200), and in this way the picture of the liminal creature, the maerc-stapan, could have functioned to uphold and define social borders. The horses sacrificed in wetlands became liminal creatures haunting the borderlands to other dimensions, the realms of the dead, of chaos and of anything unknown. In this way they helped upholding the social order and protecting the inhabitant of Miðgarðr from the terrifying and uncontrollable. If the social occasions where the bracteates were used were sacrificial feasts, as the runic inscriptions alu, laukaR and laÞu may imply (Andrén 1991: 249 ff) this could have strengthened their protective powers by reminding the users of the actual sacrificed animals. Wearing, or using, a gold bracteate could, among other things, have been a sign of having participated in a particular sacrifical ritual and thereby being protected against chaotic powers from across the border of the unknown. Literature: Andersson, Carolina & Hållans Ann-Mari 1997 No Trespassing. Physical and Mental Boundaries in Agrarian Settlements. In: Andersson, H., Carelli, P., Ersgård, L. Visions of the Past. Trends and Traditions in Swedish Medieval Archaeology. Lund Studies in Medieval Archaeology 19 pp.583-602 Andrén, Anders 1991 Guld och makt - en tolkning av de skandinaviska guldbrakteaternas funktion. In: Fabech, C. & Ringtved, J. Samfundsorganisation og Regional Variation. Norden i Romersk Jenralder og Folkevandringstid. Aarhus pp. 245-256 Arkæologiska udgravninger i Danmark 1987. Ed: Rigsantikvarens Arkæologiske Sekretariat. Det Arkæologiska Nævn København 1988 Axboe, Morten 1991 Guld og guder i folkevandringstiden. Bracteaterne som kilde til politisk/religiøse forhold. In: Fabech, C. & Ringtved, J. Samfundsorganisation og Regional Variation. Norden i Romersk Jernalder og Folkevandringstid. Aarhus pp.187-202 1997 Amulet Pendants and a Darkened Sun. On the Function of the Gold Bracteates and a Possible Motivation for the Large Gold Hoards. In: Magnus, Bente Roman Gold and the Development of the Early Germanic Kingdoms. Symposium in Stockholm 14-16 November 1997. KVHAA Konferenser 51 pp.119-135 1998 Die innere Chronologie der A-C-Brakteaten und ihrer Inschriften. In:Düwel, K. Runeninschriften als Quellen interdiziplinärer Forschung. Abhandlungen des Vierten Internationalen Symposiums über Runen und Runeninschriften in Göttingen vom 4.-9. August 1995 Berlin pp.231-252 Bebyggelsehistorisk tidskrift 1982:4 Äldre territoriell indelning i Sverige. Andersson, T. & Göransson, S. (Eds).
9

Behm-Blancke, Günter 2003 Heiligtümer der Germanen und ihrer Vorgänger in Thüringen. Die Kultstätte Oberdorla.Teil 1: Text und Fototafeln Stuttgart Beowulf. With the Finnesburg Fragment. 1958 Ed: Wrenn, C.L. London Beowulf. A student Edition. 1994 Ed: Jack, George. Oxford Blaaberg, Claus 1992 Sognedannelsen i dansk middelalder. Hørsholm Boessneck, Joachim, von den Driesch-Karpf & Gejvall, Nils-Gustaf 1968 The Archaeology of Skedemosse III. Die Knochenfunde von Säugetieren und vom Menschen. Stockholm Brink, Stefan 1990 Sockenbildning och sockennamn. Studier i äldre territoriell indelning i Norden. Acta Academiae Regiae Gustavi Adolphi LVII Uppsala 2004 Mytologiska rum och eskatologiska föreställningar i det vikingatida Norden. In: Andrén, A., Jennbert, K., Raudvere, C. Ordning mot kaos. Studier av nordisk förkristen kosmologi. Vägar till Midgård 4. Lund pp. 291-316 Brøndsted, Johannes 1960 Danmarks Oldtid III. Jernalderen. Köpenhamn von Carnap-Bornheim, Claus & Ilkjaer, Jørgen 1996. Illerup Ådal Vol. 8 Die Prachtausrüstungen. Grabungsdokumentation und Fundliste. Aarhus Clunies Ross, Margaret 1996 Prolongued Echoes. Eis, Gerhard 1964 Altdeutsche Zaubersprüche. Berlin Ellmers, Detlev 1970 Zur Ikonographie nordischer Goldbrakteaten. Sonderdruck aus Jahrbuch des Römisch-Germanischen Zentralmuseums Mainz. 17. Jahrgang. Mainz Ferdinand, Johannes & Klaus 1961 Jernalderofferfund i Valmose ved Rislev. KUML pp.47-90 Gaimster, Märit 2001 Gold Bracteates and Necklaces. Political ideals in the sixth century. In: Magnus, Bente German Gold and the Development of the Early Germanic Kingdoms. Symposium in Stockhom 14-16 November 1997. KVHAA Konferenser 51 p.143-156 Stockholm Hagberg, Ulf Erik 1967a The Archaeology of Skedemosse I. The Excavation and the Finds of an Öland Fen, Sweden. Stockholm 1967b The Archaeology of Skedemosse II. The Votive Deposits in Skedemosse Fen and their Relation to the Iron-Age Settlement on Öland, Sweden. Stockholm Hampp, Irmgard 1961 Beschwörung, Segen, Gebet. Untersuchungen zum Zauberspruch aus dem Bereich der Volksheilkunde. Stuttgart Hauck, Karl 1970 Völkerwanderungszeitliche Bilddarstellungen des Zweiten Merseburger Spruchs als Zugang zu Heiligtum und Opfer. In: Jankuhn, H. Vorgeschichtliche Heiligtümer und Opferplätze in Mittel- und Nordeuropa. Bericht üger ein Symposium in Reinhausen bei Göttingen vom 14. bis 16. Oktober 1968. Göttingen pp.297-319

10

1988 Frümittelalterliche Bildüberlieferung und der organisierte Kult. In: Hauck, K. Der histiorische Horizont der Götterbild-Amulette aus der Übergangsepoche von der Spätantike zum Frümittelalter. Bericht über das Colloquium vom 28. 11-1.12. 1988 in der WernerReimers-Stiftung, Bad Homburg.. Göttingen pp. 433-574 1993 Das Aufkommen des erfolgreichsten Motivs der völkerwanderungszeitlichen Brakteaten. In: Arwidsson, G et al. Sources and Resources. Studies in Honour of Birgit Arrhenius. PACT 38 pp.403-434 Ilkjaer, Jørgen 2002 Den bevidste ødelæggelse i krigsbytteofringerne. In: Jennbert, K., Andrén, A., Raudvere, C. Plats och praxis. Studier av nordisk förkristen ritual. Lund pp.203-214 Irving, Edward B. Jr. 1968 A Reading of Beowulf New Haven Jansen Sellevold, Berit & Lund Hansen, Ulla & Balslev Jørgensen 1984 Iron Age Man in Denmark. Prehistoric Man in Denmark, Vol. III København Johansen, Birgitta 1997 Ormalur. Aspekter av tillvaro och landskap. Stockholm Kaliff, Anders & Sundqvist, Olof 2004 Oden och Mithraskulten. Religiös ackulturation under romersk järnålder och folkvandringstid. OPIA 35 Uppsala Klindt-Jensen, Ole 1957Bornholm i Folkevandringstiden København Kolstrup, Inger-Lise 1991 Ikonografi og religion. In: Steinsland, G., Drobin, U., Pentikäinen, J., Meulengracht Sørensen, P. Nordisk Hedendom. Et symposium. Odense pp.181-203 von der Leyen, Friedrich 1935 Das Reich deutscher Volksdichtung. Märchen, Sage, Legende, Zauberspruch, Segen, Rätsel und Volkslied. Berlin Lindgren-Hertz, Lena 1997 Farm and Landscape. Variations on a Theme in Östergötland. In: Andersson, H., Carelli, P., Ersgård, L. Visions of the Past. Trends and Traditions in Swedish Medieval Archaeology. Lund Studies in Medieval Archaeology 19. pp 43-72 Lindqvist, Sune 1910 Ett ”Frös-vi” i Nerike. Fornvännen p.119-138 Müller-Wille, Michael 1970-71 Pferdegrab und Pferdeopfer im frühen Mittelalter. Sonderdruck aus Berichten van de Rijksdienst voor het Oudheidkundig Bodemonderzoek. Jaargang 20-21. Petersen, Peter Vang 1987 Nydam III - et våbenoffer fra ældre germansk jernalder. Aarbøger for Nordisk Oldkyndighed og Historie. København pp.105-137 Den Poetiska Eddan 1964 övers. B. Collinder Stockholm Rau, Andreas 2005 Skedemosse-Krigsbeuteopfer. In: Beck, H., Geuenich, D., Steuer, H. Reallexikon der Germanischen Altertumskunde. Band 28. Berlin pp.632-634 Rosenberg, G. 1937 Hjortspringfundet. København Ström, Folke 1999 Nordisk hedendom. Tro och sed i förkristen tid. Göteborg Sturlasson, Snorre 1997 Edda övers. K.G. Johansson & M. Malm Stockholm
11

Tacitus, Publius Cornelius 1914 Dialogus de Oratoribus, Agricola, Germania. trans. Peterson, Sir William London Teichert, Manfred 1974 Tierreste aus dem germanischen Opfermoor bei Oberdorla. Text och Tafelteil i två volymer. Weimar Tolkien, J. R. R. 1975 Om Beowulfsagan övers. Åke Ohlmarks Stockholm Trap, J.P. 1953-72 Danmark. Ed: Nielsen, Niels. København Unprinted: www.calpal.de 2005-08-26 www.dkconline.dk 2005-08-05 www.fh-augsburg.de/~harsch/germanica/Chronologie/08Jh/Merseburg/mer_intr.html 200509-13 www.illerup.dk/images 2005-08-14 Verbal commentary: Molnar, Petra 2005-09-13

12



Add Comment

To add a comment, you need to login or signup (you can signup without leaving this page)