Jens Peter schjødt
it has often been maintained, and rightly so, no doubt, that the religion and mythology of the late (and perhaps also the Early) iron age to some extent reflected a warrior ideology, and that religion and war were strongly intertwined aspects.1 Thus, it has been said about the military band that it was ‘as much religious as it was martial’ (Kershaw 2000: 18). Therefore, the character of the military bands, the Männerbünde, as they are often called, the retinue or the warrior bands, which is the term we will use here, is of great interest in dealing with the social consequences of the religious world-view (è1).2
Military power has always been a prerequisite for any society to survive, whether it be religious or not. But particularly in relation to war and battle, the outcome is as uncertain as it is important, which is why the relation to the other World is deemed crucial in most archaic cultures. This relation surfaces in connection with several phenomena that have to do with war. from archae-1 see especially Price (2002; particularly 329–96).
2 The german term Gefolgschaft has also played an important role in research into warrior bands. for example Christoph landolt (1998: 533–37) discusses some of the terminological and semantic problems connected to various terms. Gefolgschaft, however, covers more than what is usually meant by ‘warrior band’, including large armies (cf. Timpe 1998a: 538; D. H.
green 1998: 75–77, 106; Todd 2004: 30; Heather 2009: 126–31). although the terminology and the relation between, for example, *comitatus, Gefolgschaft * and retinue has been much discussed, it is hardly possible to distinguish precisely between them (cf. green 1998). The terminology is not sufficiently clear and consistent in our sources in order for us to depict what the individual term signified in the various societies (see also J. Harris 2008: 294–96).
Jens Peter Schjødt, Professor of the study of Religion, aarhus University The Pre-Christian Religions of the North: History and Structures, Volume ii, ed. by Jens Peter schjødt, John lindow, and anders andrén, PCRN-HS 2 (Turn hout: Brepols, 2020) pp. 559–587
BREPols
PUBlisHERs
10.1484/M.PCRN-EB.5.116951
560
- Jens Peter Schjødt*
ology we know of war booty sacrifices among the early germanic tribes, particularly from Denmark and northern germany, indicating that victory was celebrated with a feast during which weapons and other items belonging to the enemy were destroyed, probably as a sacrifice to the gods (cf. orosius’s Historiae adversum paganos 5.16.5–6). at the same time, a number of literary sources tell us that a spear was thrown over the enemy army in order to dedicate it to Óðinn ( Styrbjarnar þáttr Svíakappa in Flateyjarbók; Eyrbyggja saga ch. * * 44; cf. Vǫluspá st. 24), a dedication indicating that the enemy was to be killed, and we are told by snorri that the prerequisite for being able to go to Valhǫll was that one was killed in battle ( Gylfaginning p. 32).3 all in all, we have much evidence that throughout the iron age, war and religion were linked. However, one of the most fundamental institutions was the warrior band. institutions like it can, when defined broadly, be found all over the world:4 bands of warriors, young or old, who constitute a fraternity and who are particularly devoted to the chieftain, the king, or the military leader.5 The structure and the religious content of such institutions, of course, vary a lot, but there are also many parallels.6 Most often, the members of the band are initiated through some secret rituals,7 during which they change their status in such a way that their solidarity is shifted from their original family to their new ‘brothers’, including their leader; often, an oath is involved as an indication of this change, and often it is, moreover, symbolically expressed through a transformation in the initiates’ physical appearance. as long as they are members of the band, they are not allowed to 3 This information should probably not be taken too literally, since we know that people were also able to go to Valhǫll without being killed in battle. This applies, for instance, to Njǫrðr ( Ynglinga saga ch. 9) and to sinfjǫtli ( Vǫlsunga saga ch. 10 and Frá dauða Sinfjǫtla). We shall return to these and other figures below.
4 That is, not in every single society but in societies from all parts of the world.
5 Cf. the discussion on reges and duces in (è23). as is suggested by Harris (2008: 295), there might also have been a leader among the retainers, but it is still the lord, whom the band as a whole serve, who is their ultimate leader.
6 it is important to bear in mind that talking about parallels does not mean talking about the same thing. as has been shown by many scholars, the germanic comitatus and its development probably began just before the beginning of our era, and its coming into existence undoubtedly had significant consequences both socially and politically (see Enright 1996a: 195–214, with many references), but not necessarily in relation to religion. The idea that war-like young men constituted bands connected to a certain god is clearly to be detected already in indo-European times (cf. Kershaw 2000: 201–39; West 2007: 448–51).
7 Cf. the german notion Geheimbünde used by, among others, otto Höfler (1934) and criticized by Jan de Vries (1956–57a: i, 496).
24 – Warrior Bands
561
live a ‘normal’ life with family and daily work. in the following, we shall deal both with elements we encounter in nearly all warrior bands and with elements specific to the germanic/scandinavian area.
The Historical Framework
it seems that we are able, with the help of archaeology,8 texts from antiquity, as well as sagas and other textual evidence from the North, to follow this institution in the germanic area over a very long period. also from the point of view of historical linguistics, it appears beyond doubt that the comitatus of Tacitus was an institution among the germanic peoples that survived in the North, where it was known as drótt (old High german truth), right up until the Christianization towards the end of the Viking age (cf. green 1998: 106–12; lindow 1976: 26–42).9 and again, we must reckon with a combination of continuity and development. Thus, it seems beyond doubt that parts of the late information we have on the berserkir and úlfheðnar have roots going back to Roman times (e.g., Höfler 1934) and, to a certain extent, to a European (Näsström 2006: 179–94) or even an indo-European stratum (Wikander 1938; Kershaw 2000; West 2007). Not everybody, however, has accepted this idea of continuity, and scholars taking this view are represented by Hans Kuhn, who insists that warrior bands existed only during the time of Tacitus and in the Viking age, but not in between, and that they should be seen primarily as responses to similar social and military situations (Kuhn 1956).10 However, as we shall see below, the continuity hypothesis, which is based on linguistic, archaeological, and textual evidence, appears to be far better at explaining the data. The question of how far back the idea can be traced is worth considering. Michael Enright makes a good case for proposing that the germanic *com-8 There are, as we will return to, many pictorial depictions that may be of great value to our understanding of this issue. But apart from these images, it is difficult for archaeology to * * ‘prove’ a connection between artefacts and warrior bands (steuer 1998: 546). for accounts and interesting discussions of archaeological and epigraphical evidence in connection with the warrior bands, see Näsström (2006: 107–32) and samson (2011: 287–336).
9 lindow’s book is extremely useful, because it traces the terms relevant to the warrior band institution, right from the time of Tacitus and up to post-Viking age. He considers such notions as hirð, lið and many more in terms of their semantics in both prose and poetry. He does not deal with the religious aspects of the warrior band, but simply proves that this institution lived on in a variety of similar forms right up until the Christianization period.
10 for other discussions of the history of the comitatus, see Wenskus (1977) and steuer (1982).
562
Jens Peter Schjødt
figure 24.1. The geometrically planned ringfort at fyrkat in central Jylland, dated to the 980s. The legendary Jómsvikingar and their strict rules can be compared with the four geometrically planned Danish ringforts at aggersborg, fyrkat, Nonnebakken, and Trelleborg from the 980s. Photo: anders andrén.
itatus, the war band, was taken over from the Celts during the first century bce (Enright 1996a: 169–282), which does not correspond very well with the indo-European hypothesis. a problem with Enright’s theory is that sources for germanic societal organization are totally absent before that time, the use of argumenta ex silentio thus being of no value. a possible solution to the problem could be to hypothesize that some prototype of the war band was in existence long before any group (tribes, languages) that might be called Celts or germani. Whatever the case, there is no doubt that such an entity would have changed substantially through contact with cultures such as the Roman and the Celtic. Even the very notion of war must have changed as societies became much larger, and thus also the scale of the wars fought. anders andrén, among others, has argued that we should distinguish between ritualized warfare and large-scale wars, two different forms of war that doubtless existed side by side, but which also reflect a development in society and political circumstances (andrén 2014: 98–102; cf. green 1998: 75–77). it is important here to bear in mind that, even if the social structures changed, as they no doubt did from the first century bce onwards because of the encounters with the Romans, we
24 – Warrior Bands
563
cannot automatically assume that the religious structures changed, too. This means that when Enright speaks about comitatus warfare as opposed to tribal warfare (1996a: 195–214), corresponding more or less to the distinction made by andrén, and even if it can be hard to make a clear distinction between these two forms of war,11 we do not necessarily have to assume changes in the relation between warrior bands and the other World, or for that matter in the ideals concerning loyalty to the leader. at any rate, taking Enright’s observations into consideration, it seems beyond doubt that, at least from the first century bce, we can be certain that warrior bands actually existed in the germanic area, although they changed dramatically in size and probably organization during the first centuries of our era (cf. Heather 2009: 46).
one of the best candidates for a historical, as opposed to a legendary, warrior band in the North is probably the so-called Jómsvikingar, who lived at Jomsborg, the exact location of which is uncertain but which was probably located near the mouth of the river oder on the Baltic coast. The most detailed source is Jómsvikinga saga which, although of dubious source value, tells the history of this confraternity and their final defeat in the battle of Hjǫrungavágr in the 980s or perhaps a little later.12 in this saga, we learn that there were rules that must be obeyed by all the members; for instance, women were not allowed into the fortification, war booty must be divided among the men, nobody was allowed to refuse to fight against another man of equal strength and ability, they all had to obey their leader, Palnatoke, and so forth ( Jómsvikinga saga ch.
16; cf. Hálfs saga ok Hálfs rekka ch. 10). However, we do not hear anything about the religious notions that may or may not have formed part of the ideology of the Jómsvikingar. in order to obtain an insight into that area, we have to rely on the evidence of the fornaldarsögur, which are much more informative when it comes to the religious roots of the warrior band.
11 Enright says (1996a: 195): ‘While it is not always easy to distinguish comitatus warfare from tribal warfare in the historical sources, it is obvious that the latter was much more exclusive with loose knit armies being gathered and organized according to family and clan and with little or no room, either political or economical, for the leader with an extra-tribal group of followers supported at his expense and living more or less constantly in his hall’. This is true, but still it is possible to imagine a fixed group of warriors following the leader, who are joined together through some religious rituals. for a good account of the various forms of ‘war bands’
in the Roman Early iron age, see Heather (2009: 123–34). as a sort of parallel, one could also think of the hirð of much later times; this probably developed from the drótt, but displayed rather strong hierarchies and was much larger in size (cf. lindow 1976: 26–41 and 52–63).
12 a thorough discussion of almost all aspects related to the Jómsvikingar, including the sources, can be found in Morawiec (2009).
564
Jens Peter Schjødt
figure 24.2. Destroyed weapons from the large weapon deposit at illerup in central Jylland, from about 200 ce. Photo: Moesgaard Museum, Højbjerg.
The warrior band fundamentally consists of a group of young men who vol-untarily fight for and protect the leader in wartime as well as in peace, and who are willing to die for him in battle. They are far from the only warriors in the tribe or the society, but they constitute the military elite. and alongside the social and political functions, also the religious aspects of these warrior bands seem quite clearly portrayed in the sources, since these men’s abilities were seen not only as a consequence of their strength and skills, but even more so as a consequence of a transformation of their identity into a kind of superhuman beings, strongly related to, if not directly identified with, certain animals or, as has been suggested by Höfler (1934), as an army of the dead. some of
24 – Warrior Bands
565
these ideas can be questioned, but the supernatural relation in general cannot be called into doubt. Below, we shall go through the most important sources, including the archaeological ones, and discuss the details that characterized the germanic/scandinavian warrior bands.
The Sources
Archaeology
archaeology can contribute to the issue of warfare and elite warrior bands in ancient scandinavia in different ways. large-scale warfare, with large levies of armed men,13 can be detected primarily from large weapon deposits consisting of weapons, riding equipment, personal objects, clothes, and slaughtered animals deposited in shallow lakes (fabech 1991; Randsborg 1995; ilkjær 2000; Jørgensen and others 2003; Nørgaard Jørgensen 2009). around thirty such sites of weapon deposits from the iron age have been located, most of them in present-day Denmark and southern sweden. a recurrent pattern is that weapons were deposited in the same place on repeated occasions, but often with long intervals between. some fifty deposits made at the thirty sites are known for the period 200–600 ce, and only three large deposits from the earlier period of 350 bce to 150 ce are known (fabech 1991; Kaul 2003a; ilkjær 2003).
it has proved possible to reconstruct the organization of these groups of warriors from the selection of weapons. at the early site of Hjortspring, from about 350 bce, a boat for twenty-two to twenty-three men and weapons for at least sixty-five to seventy warriors have been found. The different types of weapons indicate that the defeated army whose equipment was deposited consisted of units of ten men, with nine ordinary soldiers armed with spears and shields and an ‘officer’ armed with a chainmail, a sword, and a shield. Each boat consisted of two such units as well as one or two ‘mates’ and one higher ‘officer’ who was probably commanding the ship (Randsborg 1995; Kaul 2003a). at illerup, roughly ten thousand weapons were deposited on just one occasion at the start of the third century ce. The find may represent a defeated army of an estimated two or three thousand men (ilkjær 2000). it is possible to discern a similar military hierarchy as at Hjortspring, consisting of three different levels of warriors: namely, ‘private’ foot soldiers, mounted ‘officers’, and mounted ‘commanders’
(ilkjær 1997, 2000). The armed warriors (89 per cent) were equipped with 13 The large-scale army which could probably also include the warrior bands was equivalent to the Roman exercitus (old High german heri, old Norse herr).
566
Jens Peter Schjødt
figure 24.3. Reconstruction of a warrior dressed in a
bear skin, based on a grave from Hjärterum in Kuddy
in Östergötland, dated to about 100 bce. The bear skin
warrior invites comparison with the later known concept of
berserkir. Drawing: Jonas Wikborg in Nicklasson 1997: 241.
mass-produced lances, spears, and
shields, while the officers (9 per
cent) more often carried individu-
ally designed swords, shields, and
belts with bronze mountings. The
commanders (2 per cent) were
equipped with artfully designed
swords, shields, belts, and eques-
trian gear with gilded silver
mountings, and some of them
moreover carried chainmail and
decorative Roman helmets of sil-
ver. it is primarily to this elite that
objects with animal ornamenta-
tion and runes can be linked.
apart from large weapon
deposits, large boathouses along
the Norwegian coast as well as
large hillforts and ringforts in
sweden and Norway indicate sim-
ilar large-scale warfare with many
armed men. about ninety Norwegian boathouses for large warships are known from around 200 to 550 ce, representing a huge navy of between 1800 to 2400
warriors, probably an early form of leiðangr (Myhre 1997). Many hillforts and ringforts consist of only stone walls, but on the island of Öland from about 200
to 650 ce fifteen ringforts were constructed with houses inside the walls. it has been estimated that about 1200 houses were built in these ringforts, which means that most of the inhabitants of the island could have been housed in the ringforts during periods of warfare. Besides, it seems as if the organization of warfare in these ringforts was linked to cosmological ideas, because the layout of the forts had cosmological associations (andrén 2014: 69–115; figure è6.5
and è38.1). in later periods, the Viking attacks in Western Europe from the late eighth century onwards again involve large levies of armed men.
24 – Warrior Bands
567
figure 24.4. Plan of the fortified
hilltop settlement at Runsa in Ed in
Uppland, dated to the fifth and sixth
centuries. Map: Michael olausson.
Traces of elite warrior bands are much more difficult to detect archaeologically, but there are some indications in different parts of scandinavia. on gotland, graves from the period 200–1 bce located in different places contain weapons for several warriors; sometimes as many as fifteen warriors (arnberg 2007: 192–95). This burial custom shows that warriors were buried together and in graves separate from other ‘ordinary’ dead persons, which suggests that warriors were grouped together and were in certain ways segregated from the rest of society. in Östergötland, a grave from Hjärterum in Kuddby, dated to about 100 bce, contained not only weapons but also claws from a bear (Nicklasson 1997: 198), indicating that the warrior was dressed in a bear skin, which conjures associations to the later concept of berserkir.
along the Norwegian coast, about twenty-five so-called ringtun and tun sites are known from about 200 to 850 ce. They are shaped as clusters of houses placed around an open space in a circle or oval with one or two openings. The
568
Jens Peter Schjødt
Shield fragment
Arrow head
Spear head
figure 24.5.
Plan of a hall in Birka,
dated to the tenth century. The
building, with weapons along
the walls, has been interpreted as
a hall for professional warriors.
after Holmquist olausson and
Kitzler Åhfeldt 2002: 14.
tun sites are usually placed in the outland between settlements, although they are often surrounded by large cooking pits, fireplaces, and burial mounds.
several tun sites are linked to the placename element -leik, which means ‘play’
but also ‘training for war’. The finds are few and modest, but they show that the sites were often used on different occasions and through an extended period of time. The sites have been interpreted as places connected to eating and drinking ceremonies, which were necessary for creating a sense of community in warrior bands since these were not held together by bonds of kinship (storli 2000, 2001; stylegar and grimm 2003).
in the Mälar region in central sweden, about twenty so-called hilltop sites from the fifth, sixth, and seventh centuries ce are known. They consist of small, fortified settlements on hilltops. The best-known site is Runsa, placed on a rock overlooking one of the narrow passages between the Baltic sea and gamla Uppsala. Excavations have revealed a large hall and at least five other, smaller buildings, one probably with a ritual function (olausson 2011, 2014). The
24 – Warrior Bands
569
secluded locations, away from ordinary settlements, indicate that these hilltop sites were not ordinary aristocratic residences but instead places where various warrior bands had their bases.
an even more overtly martial environment from the tenth century has recently been discovered in Birka. on a terrace, between the hillfort and the urban settlement, a hall and a smithy have been uncovered. along the walls of the hall, remains of a wooden chest with spears and shields have been found. in addition, some of the spears appear to have been placed upright, leaning against the walls. This site has been interpreted as a hall for professional warriors, who were in all likelihood associated with Óðinn (Holmquist olausson and Kitzler Åhfeldt 2002; Hedenstierna-Jonson 2006; cf. green 1998: 79; è42).
apart from geographical sites, the pictorial world of the late iron age likewise points to warfare involving elite warriors. Warriors and different animals of prey, such as wolves, bears, snakes, eagles, and boars, are depicted on many exclusive weapons, for instance, swords, shields, and helmets. This is especially true of weapons in the rich boat graves from, for example, Vendel and Valsgärde in Uppland (Kristoffersen 1997; Hedeager 1997a, 2011; Price 2002).14
Textual Sources
for the earlier period, Tacitus is, if not the only, then definitely our main textual source with regard to the warrior bands. Chapters 13, 14, 24, 31, and 43
of Germania present important information about institutions and traditions which are probably linked to warrior bands, although, as was stated in (è12) (and several other chapters) above, the source value of Tacitus has been much debated.
We learn from Chapters 13 and 14 that the society is strongly involved in the decisions about when to carry weapons. at the people’s gathering ( concilium), for example, the young man’s father or one of the chieftains ( principum aliquis) will give him a shield and a framea (a lance or a spear) as a sign that he is now an adult and no longer a member of the household of which he was a part as a child.15 This is obviously a kind of initiation ritual, although we are not given any detail on how it is carried out. it appears to be a so-called ‘puberty 14 for an interesting analysis of one the images on the gundestrup cauldron as depicting a scene of initiation into the warrior band, we can refer to Enright (2007).
15 We may note here that there seems to be a sort of parallel between the ‘family’ of the child and the new ‘family’ of the adult, with new ‘brothers’, as is typical of the warrior band, as has been suggested by Enright (1996a: 76–77).
570
Jens Peter Schjødt
ritual’ of a type well known all across the world, but Tacitus immediately goes on to talk about the retinue ( comitatus) of the chieftain ( princeps), stating that there are various positions (sing. gradus) and that the chieftain decides who is given which position.16 The chieftain’s reputation is dependent on, among other things, the size and the courage of his retinue, and the members compete for the highest position. in Chapter 14, Tacitus continues with the notion of honour: it is disgraceful for the chieftain if the members of his retinue appear more courageous than himself, while it is disgraceful for them to be less courageous than their chieftain — but the most disgraceful thing is if someone survives a battle in which his chieftain has been killed (è21). Tacitus concludes (ch. 14): ‘principes pro victoria pugnant, comites pro principe’ (the chieftains fight for victory, the retainers for the chieftain).
it seems that Tacitus is speaking here about two different kinds of warriors: on the one hand, those who undergo a transition from boyhood to manhood, which is symbolized by their acquisition of arms;17 and then, on the other hand, those who become part of the retinue. Even if he also says that the germanic warriors are lazy and do not want to work the soil ( Germania ch. 14 and 15), we must assume that ordinary adult men would have to do some agricultural labour for a living. This is also confirmed by Tacitus when he states that the most courageous men do not carry out any form of ordinary work (which must imply that others did). These most courageous men are most likely those of the retinue. from this we can deduce, then, that all men receive weapons as part of their initiation into adulthood, whereas the most courageous become part of the retinue (professional warriors), who receive their horses and weapons18 as well as their sustenance from the chieftain they serve ( Germania ch. 14) and are therefore not supposed to do any other work than fighting for him. Tacitus thus makes reference to two kinds of warriors who probably underwent two kinds of initiation: namely, the ordinary adults and the elite warriors.19 it is noteworthy 16 The war bands were most likely intertribal (cf. green 1998: 107).
17 according to Tacitus, we are dealing with societies in which men are always armed ( Germania 13.1) except at certain, mostly ritual, occasions.
18 This statement must mean that the weaponry of the retinue is of another kind (probably of a higher quality) than those an ordinary man receives during his initiation into adulthood (see above on archaeological evidence).
19 This would also explain the strange assertion in Chapter 13 that noble birth may give the very young men rank of chieftain and the continuation that: ‘ceteris robustioribus ac iam pri-dem probatis adgregantur’ (they then join the more experienced who have already shown their abilities). if a man is a chieftain, he can hardly be seen as an ordinary member of the retinue,
24 – Warrior Bands
571
that we clearly get an impression of a hierarchical or a vertical structure, the gradus, which is not in accordance with the ideal sometimes presented in other sources, that is, that the structure was horizontal and all members of the retinue were equal (Enright 1996a: 18–20). However, the most important information here is that there existed a retinue around the chieftain. Nothing is said at this place about any religious connection, but this occurs in Germania ch. 43, which is discussed below.
in Chapter 24, we are told of a kind of ‘weapon dance’, performed by naked youths who jump between swords ( gladii) and spears ( frameae). Tacitus sees this as pure entertainment, but it is probably more than that. We know sword dances from later folklore all over the germanic area (Müllenhoff 1871), which may very well be a continuation of what Tacitus describes.20 More interesting in relation to religion, however, are some pictorial depictions, such as the helmet from Torslunda ( c. 600). Here, we see two figures, both holding spears, one wearing an animal mask (probably a wolf mask) and the other a horned helmet; the latter is one-eyed and obviously dancing.21 Whenever we encounter a one-eyed figure, the associating to Óðinn is inescapable, and although we obviously cannot be sure, this possible identification has tremendous consequences for the whole idea of the warrior band, as we shall return to below. The information given by Tacitus may not be very informative in itself, but when it is compared to later material, it suggests that on special occasions the warriors had to perform a dance,22 which displayed their skills in the use of weapons. We cannot know whether this was part of an initiation ritual, or whether it was perhaps too. What is probably meant is that the best of the young initiates may join the retinue immediately and thus become part of the circle around the chieftain. Tacitus’s statements about young men obtaining the rank of chieftains because of their noble birth or the deeds of their fathers could, however, also be seen in connection with the problem concerning reges/ duces ( Germania ch. 7), which was dealt with above in (è23). Even so, the confusion about the various categories of young warriors remains.
20 Cf. de Vries (1956–57a, 1: 443), with many references. Kershaw treats the dancing warriors extensively from an indo-European perspective (2000: 83–102). it is interesting that sword dances seem to be unknown in Norway and iceland, which may indicate that the late evidence from Denmark and sweden was, in fact, imported from germany during the Middle ages (gunnell 1995: 132). However, a continuation from the pre-Viking era can by no means be ruled out.
21 Cf. Beck (1968) and arent (1969). for further references to pictorial sources, see Mitchell (2012: 12) and Hedeager (2011: 61–98).
22 Unfortunately, we do not know which occasions.
572
Jens Peter Schjødt
part of some fertility ritual. But what is almost certain is that groups of warriors existed and that dance formed part of some of their specific rituals.
in Chapter 31 of Germania, Tacitus informs us about the Chatti,23 and with this we move outside the more general part of his work. Nevertheless, some of the information related here seems to supplement what he presents earlier, and it also appears to fit in very well with information from later sources. We are told that, on becoming adults, the young men let their hair and beards grow until they have killed an enemy. The bravest of them furthermore wear an iron ring, something which is said to be disgraceful, and only when they have killed an enemy (one more?) do they remove it. But some of them keep the ring until they grow old, and they are the elite warriors. They do not have any property, and others have to supply them with food.24
Here it seems that we have three groups: 1) the young men who let their hair and beard grow until they have killed an enemy; 2) some of them who wear an iron ring, which they remove when they have killed an (other?) enemy; and 3) some of these men who continue to wear the ring until they become old. Exactly how the status of these three groups should be understood will not be discussed here, but again we notice a distinction between different kinds of warriors and the fact that a certain group seems to constitute a special brotherhood. furthermore, we see that some kind of initiation trial is involved: being a warrior is something you are initiated into with tests and symbolic outfit (uncut and cut hair, rings). it is tempting to compare this to Tacitus’s general statement in Chapter 13 concerning the various positions of the warriors in the chieftain’s retinue.
The last chapter of Germania to be treated here is Chapter 43. This chapter describes, among other things, the Harii, who are considered, by Tacitus at least, a tribe but should perhaps rather be seen as a special group among the lugii, a federation of tribes in the eastern part of germania. Judging by Tacitus’s description, there can hardly be any doubt that this group was constituted by 23 They may actually have constituted a confraternity rather than a ‘tribe’ (cf. Harris 2008: 301).
24 There seem to be some inconsistencies in the description. since those who never kill an enemy will maintain this unkempt appearance, i.e., the cowards ( ignavus and imbellis), it would be hard to distinguish them from the bravest of the warriors (cf. fehrle and Hünnerkopf 1959: 118; see also Perl 1990: 214–15; schjødt 2008: 331–33). it has been argued that the very fact that they were fed by others is evidence that they belonged to a god (Kershaw 2000: 46), which they probably did. However, we have many examples all over the world that retinues of chieftains and kings had to be supported by ordinary people simply because the leaders had the power to demand so.
24 – Warrior Bands
573
a band of warriors,25 harii being connected to oN herr, ‘army’.26 We learn that the Harii were the strongest of the lugii, and that, although formidable as they were, they would use artificial means to look more so: their shields are black, their bodies are painted, and they choose dark nights for their attacks, so that Tacitus directly speaks of a feralis exercitus, an ‘army of the dead’. This passage has led many scholars to believe that the warrior bands were part of a cult of the dead and that the members saw themselves as dead (e.g., Weiser 1927: 39–42; Höfler 934: 163–72; de Vries 1956–57a: ii, 379; Meuli 1975: i, 275; Kershaw 2000: 41). We shall discuss this issue in more detail below; for now, it suffices to note that also among the lugii we find a certain group of elite warriors who, like the warriors among the Chatti, are distinguished from the rest of the people by a fierce appearance.
To sum up the implication of this account of the early period that Tacitus presents, it can be stated that all young men carried arms, that there were special groups with a more or less uncivilized appearance who had to be tested in particular and who, if they passed these tests, joined a group of elite warriors, a retinue or some other kind of warrior band. further, these warriors would on special occasions perform weapon dances and they were, during battle, compared to the dead.
Tacitus, as previously mentioned, is by far our best (and almost only) textual source for the existence of warrior bands in the Roman era, and the question is whether there is evidence of such an institution in the Viking age also: in other words, whether we can assume some degree of continuity. it can be stated at once that, apart from Jómsvikinga saga, we are not told directly in any of the sources that may be labelled ‘historic’ ( Íslendingasögur, other konungasögur, skaldic poems, etc.) that such bands, connected to the religious sphere, actually existed,27 although various types of warriors are mentioned. The sources we 25 The etymology of the word harii is not certain, but see de Vries (1962a: 224–25).
26 it is also worth noticing that some of Óðinn’s many cognomen are herjann, hertýr, and several others with her- as prefix, and he is also the lord of the einherjar, the dead warriors in Valhǫll.
27 a phenomenon we will return to in (è32) is the establishment of blood-brotherhood, which is described in several sources. The differences notwithstanding, these all speak of young men who, through a ritual which bears all the characteristics of an initiation ritual, transform their status in such a way that they become ‘family related’. These blood-brotherhoods seem to be related to the warrior bands in some way and the initiation into them may perhaps help us in reconstructing part of the symbolism surrounding the warrior bands (cf. schjødt 2008: 355–73).
574
Jens Peter Schjødt
must turn to in order to reconstruct warrior bands rooted in pagan religious ideas are above all the legendary sagas ( fornaldarsögur).
one stanza from a skaldic poem, however, is often taken as a starting point: namely, stanza 8 of the so-called Haraldskvæði or Hrafnsmál, known primarily from Haralds saga hárfagra and allegedly composed by Þórbjǫrn hornk-lofi shortly after the battle of Hafrsfjǫrðr.28 The relevant stanza (st. 8) goes:
‘grenjuðu berserkir | guðr vas þeim á sinnum | emjuðu úlfheðnar | ok ísǫrn dúðu’ (berserks bellowed; battle was under way for them; wolf-skins howled and brandished iron spears). later in the same poem, the two groups are mentioned and clearly with approval: they are portrayed as great warriors and highly valuable for the king. Berserkr means either ‘bare shirt’ or ‘bear shirt’,29 signifying a warrior who is either fighting without armour or one who is fighting wearing a bear skin.30 Úlfheðinn means ‘wolf skin’, thus signifying a warrior connected to wolves. it seems, therefore, that this poem — probably composed in the early tenth century and thus rather early compared to most other skaldic poems, and no doubt composed by a pagan — clearly indicates that these groups of ‘animal warriors’ constituted a kind of elite troops in Haraldr’s army.31 since there is much more information about the berserkir than about the ulfheðnar, we shall start our discussion with them.32
in the Íslendingasögur we often hear about berserkir, and they are almost always portrayed as rowdies, very strong, but without any respect for other people’s property. The way in which berserkir are introduced in these sagas is part of a literary topos where they come to a farmstead demanding food and drink, and frequently also some of the women on the farm. finally, a brave young hero stands up to them and eventually kills them, even if they are invulner-28 The date of this battle is not certain, but sometime between 870 and 900. Traditionally 872 has been assumed (Jones 1984: 89).
29 We cannot from the word itself determine which is the correct meaning (a full discussion is found in samson 2011: 66–85), but as will become clear below, the ‘bear’ etymology seems to cover the ideas about berserks as they are expressed in the medieval literature, in the more adequate way (see, however, West 2007: 449 and speidel 2002: 253–54).
30 although this should doubtless not be taken literally, since it would be impossible to fight wearing a heavy bear skin. Rather, we should envisage a warrior who is somehow seen as a
‘bear fighter’ (fighting as a bear). see also below.
31 for different views of the poem, including st. 8, see samson (2011: 93–138) and fulk (2012b: 91–117), both with references.
32 The two groups cannot be distinguished sharply from one another and it seems that, in later texts, they are more or less identified (i.e., Vatnsdœla saga ch. 9, Grettis saga ch. 2). see further below.
24 – Warrior Bands
575
able to fire and iron (cf. Blaney 1982). in many cases these berserkir number only one or two, but at other times we hear about whole gangs, consisting of twelve (for example, Egils saga, Víga-Glums saga, Grettis saga, and Hrólfs saga Gautrekssonar, as well as others). Judging from these sagas alone, the impression is that they are pure villains,33 and there is no doubt that the general view on berserkir in the thirteenth century was negative, as is evident from the prohibition against berserksgangr in Grágás (i, 23).34 But there is another view which, as stated above, is most often presented in the fornaldarsögur,35 and this is probably much more in accordance with the pagan view of the berserkir, as described in Haraldskvæði. Here, we almost always meet them in larger groups, sometimes twelve (e.g.. the sons of arngrímr in Hervarar saga ok Heiðreks ch: 1, Hrólfs saga kraka ch. 22) and sometimes constituting the retinue (or part of the retinue) of a king.
it must be conceded that we have only one source that links the berserkir directly to the sphere of religion; the rest is conjecture, albeit a conjecture which definitely makes good sense. The source is the famous Chapter 6 of Ynglinga saga in which snorri relates the following about Óðinn’s abilities: Óðinn kunni svá gera, at í orrostu urðu óvinir hans blindir eða daufir eða óttafullir, en vápn þeira bitu eigi heldr en vendir, en hans menn fóru brynjulausir ok váru galnir sem hundar eða vargar, bitu í skjǫldu sína, váru sterkir sem birnir eða griðungar. Þeir drápu mannfólkit, en hvártki eldr né járn orti á þá. Þat er kallaðr berserksgangr.
(Óthin was able to cause his enemies to be blind or deaf or fearful in battle, and he could cause their swords to cut no better than wands. His own men went to battle without coats of mail and acted like mad dogs or wolves. They bit their shields and were as strong as bears or bulls. They killed people, and neither fire nor iron affected them. This is called berserker rage.) (p. 10)
first, we notice that snorri apparently combines the two possible etymologies of the word: the berserkir fight without coats of mail, and they are strong as bears. Moreover, they go, as is an often-mentioned and defining characteristic of 33 for a general account of the literary treatment, see Beard (1978) and Blaney (1972).
34 We even see them as opponents of Þórr in the shape of women ( Hárbarðslióð st. 37). We do not know what this myth fragment is hinting at, but considered in relation to the rest of the poem it may point to the opposition between Þórr and Óðinn, the berserkir being related to Óðinn. However, we do not hear about female berserkir. for a discussion, see Näsström (2006: 133–58).
35 sometimes, however, these sagas contain the same topos as the Íslendingasögur.
576
Jens Peter Schjødt
figure 24.6. Bronze matrices for producing helmet plates, dated to c. 600, found at Björnhovda in Torslunda on Öland (sHM 4325:618351 and 4325:108869). The two matrices illustrate a warrior between two bears and a ‘wolf warrior’ together with a one-eyed warrior with a helmet. The images call to mind the concepts of berserkir and úlfheðnar. Photos: sören Hallgren, statens Historiska Museum, stockholm.
berserkir, into a kind of ecstatic rage. But, more importantly, they are regarded as warriors belonging to Óðinn and are compared to wolves and bears, exactly the two animals that are present in the terms berserkir and úlfheðnar. although this cannot be taken as proof that they were imagined in the shape of these animals, it certainly strengthens the hypothesis that there existed some sort of relationship between warriors and these particular animals, which were known for their ferocity and strength. as is the case with almost everything snorri says about religion and mythology, the value of this information to a reconstruction of pagan traditions has been heavily debated, and some have regarded it as mere folklore belonging to snorri’s own time (e.g., liberman 2003, 2005). it is true that, were this the only evidence for ‘animal warriors’, there would not be much to build an argument on. But fortunately, this is not the case, since we have in at least one fornaldarsaga a legendary, but quite detailed and apparently very old, account of how these animal warriors were part of the warrior band surrounding the king.
This saga is the late (fourteenth century or later) Hrólfs saga kraka. Here, we have on the one hand a typical medieval attitude towards the berserkir, who are portrayed, particularly in the þættir of svipdagr and Bǫðvarr bjarki, as aggressive and self-promoting, although they are members of a king’s retinue.36
36 first, we hear about King aðils’s berserkir in Chapters 18–21, then King Hrólfr’s in
24 – Warrior Bands
577
What is interesting, however, is that in two instances, first in connection with svipdagr, then with Bǫðvarr and Hjalti, we hear that the three heroes are challenged by the berserkir. it is said that it is their habit to ask everybody in the hall whether they think they are as strong as they, the berserkir, are; of course, only our heroes have the courage to accept the challenge. The point here is that, in order to become a member of the retinue, one must first prove one’s strength in what could very well be a ritual fight against a berserk (or ritually speaking: a bear). That this is part of an initiation sequence is further supported if we analyse the sequences that lead up to the acceptance into the band of Bǫðvarr and Hjalti: regarding Bǫðvarr, we get a long story in which it is said that he is the son of a man called Bjǫrn (bear) and a woman called Bera (she-bear).
His own name, Bjarki, means ‘little bear’, and he is, in the final battle in which Hrólfr and all his men are killed, fighting in the shape of a bear (ch. 50). Hjalti, however, is portrayed as a frightened little boy called Hǫttr, who is afraid of everybody in the king’s hall until, with the help of Bǫðvarr, he is forced to drink some blood and eat part of the heart of a beast,37 which has been killed by Bǫðvarr but which is subsequently raised, as if it were still alive, and then killed a second time, so to speak, by Hǫttr. after that, the king gives him a sword and a new name: Hjalti; from then on he is accepted as one of the great warriors in King Hrólfr’s retinue, among his kappar.
also the relation to Óðinn is clearly apparent in the saga (ch. 39 and 46).
Here, the king and his men come to a farmer called Hrani who turns out to be Óðinn. He tests their endurance and advises the king to send home those who are not able to stand the test, which is everyone except the kappar/ berserkir. on their second visit, Hrani offers them weapons, but the offer is rejected by the king.38 it seems that it is Óðinn himself who tests the warriors and that only
‘his’ men are able to pass the test. in this folktale disguise, we get the definite impression that behind the king and his warriors there is a deity who protects them and chooses them.
Chapters 22 and 37. in both cases, the berserkir are portrayed as elite warriors, although in the case of King Hrólfr they are to some extent distinguished from the kappar and hirðmenn (for a more detailed analyses, see schjødt 2011: 279–87). But this distinction is likely due to the negative view of the berserkir in the Middle ages. it is also worth mentioning that exactly those men who are portrayed in the saga as kappar are said by snorri to be berserkir ( Skáldskaparmál p. 58).
37 in saxo’s version of the story ( Gesta Danorum 2.6.11), it is explicitly said that the beast was a bear ( ursus).
38 it should be noted here that in the medieval tradition, Hrólfr is portrayed as a ‘noble heathen’, i.e., although he was heathen, he did not really venerate the pagan gods.
578
Jens Peter Schjødt
What we have here, then, are two stories about two heroes, intertwined yet still clearly separated, who become members of one of the most famous warrior bands of early times; and they are both, in different ways, strongly connected to bears (son of bears, killing a bear, fighting men who appear in bear-shape).39
There is no doubt that fornaldarsögur in general, including this one, are influenced by medieval folklore and popular Christianity, but also traditions reaching back into pagan times have likewise left a mark on them (cf. Mitchell 1991: 60), and the sequences analysed here would be hard to explain if we did not accept a preliterate tradition connecting warriors (of a certain kind) with bears.
and if that connection is accepted, it becomes difficult not to link these bear-warriors to berserkir. 40 We do not learn much from the saga about the wider social setting of these warriors, but if we take all the evidence into account, the following scenario can be reconstructed. in pagan times in scandinavia, there existed warrior bands who surrounded the king and who were connected to those aspects of Óðinn that have to do with war and battle.41 in order to become members of these bands, which were symbolically associated with bears, the neophyte had to go through an initiation ritual involving a fight with someone who was already a member, or perhaps a sham fight with a figure representing a bear, as may be indicated in the Hǫttr story. from then on, he would be regarded as ‘a bear’ (cf. Hasenfratz 2011: 65). This was not understood literally: he was ‘a bear’ only in regard to strength, ferociousness, and invulnerabil-ity. in other words: during his initiation, he was transformed symbolically into a bear. although saxo does not mention berserkir directly, we do hear about individuals who can hardly be distinguished from these figures; especially the negatively conceived berserkir of the saga- topos mentioned above seem to have parallels in saxo’s work. of particular interest in saxo is the reference to twelve Norwegian brothers ( Gesta Danorum 6.2.1–10), seven of whom he mentions by names and all the names have - biorn ( bjǫrn ‘bear’) as their last element. These twelve brothers have been seen as a confraternity, not as kin-related brothers, 39 a much more thorough analysis of these sequences can be found in schjødt (2008: 312–26). see also arent (1969).
40 important parallels to these episodes in Hrólfs saga kraka can be found in Grettis saga and in Beowulf. There is no doubt that we are dealing with folktale themes that are widespread all over Europe (cf. stitt 1992). The question is how such folktale themes originate, and the proposition that a memory of heroic patterns, ultimately going back to warrior initiations, is involved, seems to be valid. and there is no reason to think exclusively in terms of ‘literary loans’.
41 in Ketils saga hængs ch. 5, we also see a berserkr who is closely related to Óðinn.
24 – Warrior Bands
579
by stephen Mitchell (2012: 11), and, as we shall see in connection with the úlfheðnar, there are good reasons for reckoning with such confraternities, also in connection with the wolf warriors.
Before we leave the berserkir, it must also be mentioned that bearskins have been found in graves,42 and this has convincingly been linked to the berserkir complex by Åke V. ström (1980). Thus, we can assume that warriors who were initiated into the berserkir bands could be buried with this identity, perhaps because they were about to join their ultimate leader, their host in the other World, Óðinn.
We will now turn to the úlfheðnar. as mentioned, these are not dealt with nearly as often as the berserkir in the saga literature. Even so, there is one saga episode that gives an extensive account relating to these ‘wolf warriors’: namely the one in Vǫlsunga saga concerning sigmundr and sinfjǫtli, and this is, moreover, probably the most frequently debated episode relating to warrior bands and, not least, initiations into them. furthermore, there is onomastic material which suggests the existence of such ‘wolf ’ bands in the Viking age.
Vǫlsunga saga is probably the most famous of the fornaldarsögur and is commonly dated to the last half of the thirteenth century or later. it is a heroic saga whose content is known from the European tradition ( Nibelungenlied, Þiðreks saga and much pictorial evidence) as well as the Nordic material.43 The main figure of the saga is sigurðr — sigfried in the continental tradition — together with some of his relatives. as a kind of prolegomena, however, Chapters 1–10
deal with sigurðr’s father, sigmundr, and his half-brother, sinfjǫtli. a brief summary of the episode will be given here.44
sigmundr is a son of King Vǫlsungr and great-great-grandson of Óðinn, who intervenes in the saga on several occasions. He has nine younger brothers and a twin sister, signý. Vǫlsungr owns a great hall in the middle of which a large tree grows. The swedish king siggeirr wishes to marry signý, and although she is reluctant, her father decides that the marriage should be celebrated. The wedding takes place in Vǫlsungr’s hall, and during the feast an old one-eyed man comes into the hall and drives a sword into the tree, saying that the one who can draw it out shall own it. Everybody tries, but only sigmundr succeeds.
as he refuses disdainfully to sell the sword to siggeirr, siggeirr plans revenge 42 an inventory of these burials is found in Petré (1980).
43 Hermann schneider (1962: 125–70) has a good overview of the sigurðr tradition in scandinavian and European literature. for more recent views on Vǫlsunga saga, see, e.g., several relevant articles in Fornaldarsagornas struktur och ideologi (Ármann Jakobsen and others 2003).
44 for a more detailed summary together with an analysis, see schjødt (2008: 299–312).
580
Jens Peter Schjødt
and invites Vǫlsungr, his sons, and his followers to visit him at his home three months later. although they are warned by signý about siggeirr’s treachery, they enter the hall where they are attacked and all except Vǫlsungr’s sons are killed. The ten sons are put in stocks in the woods, but every night a large she-wolf, which is actually siggeirr’s shape-shifting mother, arrives and eats one of the brothers. This goes on for nine nights till only sigmundr is still alive. With the help of signý he succeeds in killing the wolf, and he is liberated, but stays in the woods, in a dwelling in the ground ( jarðhús), while he plans revenge.
apparently, he needs some help, and signý sends her two sons by siggeirr to him in the forest. But they do not pass the tests that sigmundr imposes on them, and signý advises him to kill them which he does. later, signý exchanges shape with a sorceress ( seiðkona) and goes to her brother in his underground dwelling. They sleep together for three nights and as she regains her own shape it turns out that she is pregnant; in due course she gives birth to sinfjǫtli. When he is ten years old (the same age as her two sons by siggeirr) he is sent into the woods and he passes sigmundr’s tests without problems. Before they seek vengeance they travel through the forests, and one day they come to a house in which there are two wolf-skins, which they put on. But they cannot get them off and they run wild, ravaging the land of siggeirr and howling like wolves.
They agree that each of them alone can take on up to seven men, but if they attack a larger group they must call for the other. as sinfjǫtli does not adhere to this rule and kills eleven men all by himself, he is killed by sigmundr, but is eventually brought back to life by the help of Óðinn. finally, they rid themselves of the wolf-skins and take their revenge on siggeirr. Then they return to the land of Vǫlsungr; sigmundr becomes a famous king and sinfjǫtli a famous warrior. in the end, both are ‘taken home’ by Óðinn himself — that is, they die and go to Valhǫll.45
a lot of details are not included in this brief summary, but it is obvious that sigmundr and sinfjǫtli can appropriately be seen as ‘wolfskins’ or wolf warriors.
it also seems quite clear that the sequence is a virtual initiation sequence,46 as was noted already by lily Weiser (1927: 70–71): the boys are tested before they are allowed to undergo the initiation sequence, and during the liminal 45 Óðinn carries sinfjǫtli with him in a boat ( Vǫlsunga saga ch. 10 and Frá dauða Sinfjǫtla), and he breaks sigmundr’s famous sword in his final battle ( Vǫlsunga saga ch. 11). Besides, we know from Eiríksmál st. 5 that they are both in Valhǫll as einherjar.
46 The sequence is somewhat obscured by the fact that we have two protagonists who have different objectives: one is to be king, the other to be warrior. in the following, we shall deal primarily with sinfjǫtli.
24 – Warrior Bands
581
phase of the ritual sigmundr and sinfjǫtli are virtually wolves, wearing wolfskins and howling like wolves.47 sinfjǫtli even dies and is resurrected, which is one of the extremely common themes of initiation rituals all across the world (Eliade 1975). Moreover, it is obvious that one of the abilities sinfjǫtli acquires during his stay in the forest is that, while he must show courage, he must not expose himself to unnecessary risks by attacking too many men. This balance between courage and realism is crucial for a warrior in order for him to perform his task in an optimal way. initiation will be treated further in (è32), but we can state immediately that, even if no text speaks directly of pagan initiation rituals in connection with admittance into the warrior band, it seems possible to reconstruct part of such a ritual. for instance, it has been proposed that drinking rites form an important part of it just as a feminine entity seems to play a significant role.48 Here, we should perhaps think of signý who becomes the mother of the future warrior.
at first glance, there is no sign of any warrior band in this sequence, but we must be aware that this is above all a piece of literature composed by an author who is unlikely to have had the slightest idea about such things as warrior bands. Even so, we may hypothesize, although it is an argument hard to support on the basis of the saga itself, that sigmundr and his nine brothers constitute such a warrior confraternity: they are all related to a wolf, which eventually kills the brothers, while sigmundr himself manage to kill the wolf.
a consequence of this is that he is able to function as the initiator of sinfjǫtli.
so, if nothing else, the two wolf episodes, one of which deals with a group of brothers, suggest some relation between bands of warriors and wolves; in short: the individual member of the band can be seen as an úlfheðinn — a member of a confraternity who is symbolically acting like a wolf. at the same time, it is also especially interesting that both sigmundr and sinfjǫtli are clearly attached to Óðinn: he is said to be their ancestor, he helps them in various situations, and he takes them ‘home’ when they die.49
47 The criteria for talking about initiations are discussed in schjødt (2008: 72–84), and the application to the sigmundr/sinfjǫtli episode is expounded in short form in schjødt (2011: 290–92).
48 for proposals as to which elements are involved in these rituals, see Enright (1996a), who suggests among many other things that an important relation between a woman and the retainers was established during the initiation, and schjødt (2008), who reaches the same conclusion from another perspective.
49 This is one of the most important results in otto Höfler’s famous book on the warrior bands (Höfler 1934: 188–226), and it is confirmed in the anonymous poem Eiríksmál st. 5 that
582
Jens Peter Schjødt
as mentioned above, there is also onomastic material that may be of help.50
among other things, we have a group of runic inscriptions from Blekinge in sweden which include wolf names, such as Haþuwulfar (battle wolf ), Heruwulfar (sword wolf ), and Hariwulfar (army wolf ). These have been thoroughly analysed by olof sundqvist and anders Hultgård (sundqvist and Hultgård 2004; è43), who conclude that they should be seen in relation to, among other things, initiatory rituals of young warriors (sundqvist and Hultgård 2004: 597).51 as with the bear names, it seems highly plausible to associate these wolf names with bands of warriors.52
so, apart from the sagas, we also have contemporary evidence which points to the existence of animal warriors. The pictorial material gives us strong indications that warriors sometimes wore animal masks, probably in order to identify them with those animals that the masks portray.53 Besides, masks have been found in the archaeological record, and although it is often impossible to determine what they depict, it is clear from some that we are dealing with animal masks. We also have depictions of masks on other objects, such as rune stones and metal items (Price 2002: 171–74).54 on what occasions such masks may have been worn we cannot say, but it seems likely that it happened in connection with some sort of rituals, and according to what has been stated so far, presumably rituals of an initiatory kind.55
The problem of whether berserkir and úlfheðnar should be regarded more or less as the same thing or be seen as two different kinds of warrior bands cannot be determined with certainty.56 There is no doubt that, in some sagas, both sigmundr and sinfjǫtli are thought to be among the warriors in Valhǫll.
50 Both sagas and runic inscriptions provide many names containing the words ‘wolf ’ and bear’ (Peterson 2007).
51 see also this work for many relevant references.
52 for these examples of wolf names and many more, among others the famous Rök stone, see Mitchell (2012).
53 for a good exposition of various kinds of dramatic activities, including rituals, and with analyses of the textual as well as the pictorial evidence, see gunnell (1995: 23–92).
54 Most of the pictorial sources that may be connected to berserkir are depicted in samson (2011: 343–58).
55 although it is also likely that they may have been worn in connection with seasonal rituals, which would bring the evidence in accordance with the later folklore, including processions of young men wearing masks (Höfler 1934: 25–31) at certain times of the year.
56 it has been argued by lotte Hedeager (2011: 89–90, 95–96) that there also existed a third kind of warrior band, connected to yet another animal: namely, the wild boar. she sup-
24 – Warrior Bands
583
the author sees them as alternative terms for the same phenomenon, whereas in other sources they appear to be distinguished. We must of course be aware that the saga writers lived centuries after the ‘animal warrior bands’ had ceased to exist as a retinue of the king or chieftain, and that their knowledge relied on oral traditions, which during this long time span had become somewhat confused and unclear. Because of this, some scholars have argued that there was no difference, that the terms were interchangeable and simply referred to wild and terrifying animals (e.g., Näsström 2006: 158–61), as is perhaps suggested in Haraldskvæði, whereas others argue that the terms actually represented two kinds of warriors (e.g., Price 2002: 374; Hedeager 2011: 89–98). Kris Kershaw even suggests that they represent two stages of the ‘education’ of the warriors, the young initiates being úlfheðnar, whereas only some of them become berserkir (Kershaw 2000: 61), a division apparently relating to the distinction made by Tacitus concerning the Chatti. We will never know for sure, but what seems clear is that both groups were connected to Óðinn.
since Óðinn, judging from the mythological and semi-mythological texts, was the god of the mental — strategic and ecstatic — and the collective side of the phenomenon of ‘war’ (cf. schjødt 2012b; è42), he is also the god we should expect to be the ‘leader’ of the warrior bands. and this is, indeed, what we see in the sources that can be linked to the warrior bands, insofar as they give us any clue at all about a divine relation (cf. green 1998: 79–80). Whereas Óðinn is mentioned several times, both in connection with ‘bear warriors’ and with ‘wolf warriors’, none of the other gods can feasibly be seen as associated with these bands. This is not to say that other deities were not connected to war and fighting, because most of the gods actually were, but there are no hints that they were ever seen as patrons or leaders of special groups of warriors. so, just as we have seen that the berserkir and the úlfheðnar were the elite among the warriors, characterized by their ecstatic fighting and by being destined to go to Valhǫll after their death, we also see that some of Óðinn’s most important characteristics are his role in connection with war, his ability to incite his chosen warriors before and during the battle,57 and being the leader in Valhǫll, the home of the einherjar. all these elements are intertwined in our sources, but whereas each piece of evidence and its individual interpretation can obviously be doubted and questioned, it would seem unreasonable to reject the sizeable accumulation of details that connect warrior bands, animal shape, and Óðinn, ports this idea with some pictorial evidence, whereas the textual evidence is extremely thin (cf. also Beck 1965).
57 Cf. his name which is etymologically related to óðr, ‘furious’ (de Vries 1962a: 416).
584
Jens Peter Schjødt
figure 24.7. Detail of the top panel of the picture stone from Hunninge in Klinte on gotland, dated to the ninth or tenth century. The panel shows a rider, accompanied by a dog/wolf, and being welcomed by a woman with a drinking horn. over the rider, a man with a ring is ‘flying’, and above him are two fighting warriors with swords and shields. similar compositions are known from the picture stones from ardre and Tjängvide in alskog, which are interpreted as representations of Valhǫll, due to accompanying images of halls. although the picture stone from Hunninge lacks a hall, the parallels indicate that the fighting warriors can be interpreted as einherjar. gotlands Museum, Visby. Photo: anders andrén.
not least because, as we shall return to in (è42), the figure of Óðinn is so closely related to that of the political leader: not only in his capacity of warlord but also at a much more general level associated with numinous knowledge and initiation.
as the last element to be treated in this chapter, therefore, we shall briefly deal with the relation among the members of the warrior bands, Óðinn, and the dead. Óðinn’s hall is the place where warriors go when they are killed in battle and where they become einherjar,58 which means ‘those who fight alone’
or ‘the only (outstanding) warriors’ or perhaps more likely ‘those who fight like 58 The locus classicus for the notions about the dead in Valhǫll is Gylfaginning pp. 32–34.
24 – Warrior Bands
585
one’, which accords extremely well with the idea of the warrior band as a unit.59
lily Weiser (1927), otto Höfler (1934), as well as Kris Kershaw (2000) have argued that the odinic warrior bands formed a counterpart to the einherjar. The feralis exercitus of Tacitus not only looked like the dead, they were the dead in a symbolic sense, and were thus identified with the ancestors. Especially Höfler used contemporary folklore from the germanic area to support this parallel by postulating a connection between the processions of young masked men, mentioned above, the ideas of ‘the wild hunt’, which is the tradition about the dead rushing through the air with a ghostly hunter as their leader, and the fact that this leader was sometimes seen as Óðinn, in order to argue a continuity from Tacitus up until the present (Höfler 1934: 77–84).60 This approach has been severely criticized, not least because of Höfler’s ideological framework, which is treated briefly in the next section.61 Undoubtedly, there are good reasons to be careful with Höfler’s exposition, but this being said, a certain link between the warrior bands and the einherjar is nonetheless relevant, although probably not exactly in the way that Weiser, Höfler, and Kershaw have seen it. again, we can never obtain an unequivocal answer, but the idea that warrior bands were in some way regarded as ‘the dead’ appears to be entirely unsupported by the extant sources, except perhaps for the Harii. There are many indications that warrior bands were in some ways regarded as particular species of animals, and thus as belonging to a non-human world, but there are no indications that the einherjar were ever seen as or even compared to animals. Being dead, they were certainly also of another world, but a direct identification is problematic. What seems much more likely is that the relation was metonymic instead of metaphoric: the warriors of the warrior bands were not regarded as dead, but when they died they became part of the group of einherjar and as such would come to belong to the ultimate retinue: namely, that of Óðinn himself. Just as the war-59 see finnur Jónsson (1931: 102): ‘de som hører til, udgør én hær’ (those who belong to, constitute one army). Kershaw (2000: 19) says: ‘the paragon of the herr’.
60 in the folklore, it seems most likely, however, that we are dealing with a combination of, on the one hand, popular Christian ideas that see Óðinn as the devil — a tradition reaching far back, at least to the thirteenth century — and, on the other hand, possibly more or less vague memories of Óðinn as the leader of the dead in Valhǫll.
61 among the most severe critics is Klaus von see, who in several publications has discussed and criticized the ‘continuity’ theory, accusing Höfler of ideologically based theories (e.g., von see 1972). a critique from another point of view is found in Näsström (2006: 232–
42).
586
Jens Peter Schjødt
riors would fight till death for their leader, so the einherjar will fight in the final battle of Ragnarǫk in which Óðinn and all his men will fall.
Scholarship
We have already looked at some of the major controversies within the study of the warrior bands: is it possible to discern a degree of continuity from Tacitus until the Viking age, and is it possible to maintain that these bands were founded in religion, that is, that they were strongly related to the god Óðinn?
lily Weiser and otto Höfler were the first to analyse the phenomenon of Männerbünde within a religious context. They were both inspired by Heinrich schurtz’s book from 1902, Altersklassen und Männerbünde: Eine Darstellung der Grundformen der Gesellschaft, which showed that the institution of secret male societies existed in all parts of the world. schurtz has been criticized for his use of the ethnographic material as well as for the theories he developed on the basis of this material. Nevertheless, subsequent research has proven that the phenomenon of certain closed societies, consisting of elite warriors, existed and continues to exist in many cultures, although in very different guises. so, in a way it is surprising that there has been so much resistance to the idea that such a phenomenon also existed among the germanic tribes and that it continued into the Viking age (Kuhn 1956). The reason for that is, no doubt, as was suggested above, because Weiser’s as well as Höfler’s results were so warmly welcomed by the National socialists of the Third Reich, which in the post-war era has led to them being seen as pure propaganda and Nazi ideology.62 it was in a certain sense understandable, and perhaps even necessary, to adopt this stance at that time, although from today’s point of view, at the beginning of the third millennium, this reaction may seem just as ideological. What seems unreasonable now is not the rejection and doubting of many details in both Weiser’s and Höfler’s arguments, because doing so is both sensible and necessary, but the rejection of the basic discoveries that they made and which we have attempted to show above from textual, linguistic, archaeological, and folkloristic evidence. as can only be expected, the growing distance in time from the excesses of the National socialists and the second World War has allowed for a more balanced evaluation of the material, one result of which is that recent scholars accept without hesitation the idea that warrior bands existed all through the 62 a brief, but very informed exposition of the post-war evaluation of the ideas of the Männerbünde concept, Höfler’s research in particular, is found in Harris (2008: 288–96). The whole article can be recommended for various aspects of the warrior band problem.
24 – Warrior Bands
587
pre-Christian era (and even into the Middle ages).63 This does not mean that there is not much left to discuss — there most definitely is — but it means that, even if the sources taken one by one could all be rejected for various reasons, the overwhelming amount of all kinds of sources taken together cannot and should not be rejected or overlooked.
Conclusion
To sum up the principal characteristics of the warrior band in its relation to religion, we can be confident that warrior bands existed among the germanic tribes, at least since the time of Tacitus (lindow 1976; Enright 1996a) and probably much earlier (Kershaw 2000; West 2007). This does not mean that they did not change sociologically during this long period, because we can be absolutely confident they did; besides their sociological/political functions, they were rooted in religion and were especially related to the cult of Óðinn or his equivalents further back in time. it is also certain that a prerequisite for becoming included in the warrior band was that the young warrior went through an initiation ritual, the symbolism of which we are quite well informed about through a number of semi-mythic texts. The fixed relation between a band and its leader — and some woman with prophetic gifts (Enright 1996a: e.g. 283–87) —, consisting of an unconditional loyalty and willingness to fight until death for him, also seems to be part of the religious core of the whole complex.
63 Just to give a few examples, which differ substantially in many ways, but all take for granted that the bands did exist, we can mention: John lindow (1976), Michael Enright (1996a), Britt-Mari Näsström (2006), Joseph Harris (2008), Jens Peter schjødt (2008), Vincent samson (2011), and stephen Mitchell (2012).
25 – Various Ways of Communicating
Jens Peter schjødt
following some introductory remarks, this chapter will fall into two parts. The first part will consist of a brief history and a discussion of some of the relevant positions within the general history of research on rituals within the history of religions, followed by an account of the terminology and the classification of rituals which are used in this and subsequent chapters. after that, we shall consider some basic structures that can be universally applied to rituals. in the second part, we shall treat certain ritual phenomena in PCRN which may be part of any one of the main ritual categories, such as sacrifice, procession, and divination.
among most modern Westerners, religion is seen as something that has primarily to do with ‘belief ’. in ‘primary’ or ‘archaic’ religions, however, religion is more often connected with what you ‘do’ than what you ‘believe’. in PCRN
there was no word corresponding to ‘religion’ as it was seen by the Christians.
Trúa as the word for religion was only introduced with the church, the more common word before that being siðr, meaning ‘custom’ or ‘habit’, perhaps ‘tradition’. Thus, when Christianity came to the North and was accepted by some, the distinction arose between the ‘old’ and the ‘new’ siðr, designating paganism and Christianity, respectively (cf. Clunies Ross 2002a). This indicates that religion, as we see it, as something that can be differentiated from everyday life, was viewed differently among the pagans. Here, religion was part of the general Jens Peter Schjødt, Professor of the study of Religion, aarhus University The Pre-Christian Religions of the North: History and Structures, Volume ii, ed. by Jens Peter schjødt, John lindow, and anders andrén, PCRN-HS 2 (Turn hout: Brepols, 2020) pp. 589–642
BREPols
PUBlisHERs
10.1484/M.PCRN-EB.5.116952
590
Jens Peter Schjødt
custom, the way people lived their lives.1 This is so because, as we saw in (è1), the interference with the other World would have consequences for almost everything that was acted out in this world. That does not mean, of course, that there was no such thing as a distinction between ‘sacred’ and ‘profane’: at certain occasions people should not and could not behave as in everyday situations.2 already we hear from Tacitus ( Germania ch. 40) that during the time when Nerthus is carried around in her wagon, humans do not begin war and do not carry weapons, as is otherwise the custom. This period, therefore, is sacred: people behaved differently from what they usually did, and this can be seen as a kind of taboo (you are not allowed to do this and that here and now), which is a universal characteristic of sacred moments. also in the old Norse sources there are many examples of periods or moments during which humans are closer to the gods than at other times and therefore have to behave differently from the way they usually do; in short: the communication between worlds is facilitated by being at the right spot at the right moment. This time and place and what goes on there is what we usually call the ‘ritual sphere’ (è27–28). This means that, although that kind of communication may take place simultaneously in various situations (a god or the valkyries show up on the battlefield or a hostile being attacks you while you are asleep), the way to discover the will of the gods, the way to please the gods, and, in general, to manipulate the gods or other beings of the other World is most often seen as more successful when a ritual setting is established.
Therefore, since the notion of belief did not play any significant role (cf. Clunies Ross 2003: 284), we must assume that there were no dogmatic controversies in pagan times because this ‘manipulation’ of the other World, carried out to the benefit of the individual or society, should be seen as the key point in the religion (è1). so, although the world-view and with that the myths would support the rationale behind the rituals, it would be fair to say that the communication with the other World constituted the more important part of PCRN. This causes some problems, however, since we do not know as much about the rituals as could be wished for because of the source situa-1 see, for instance, Max Weber’s important statement (1965: 1): ‘religious or magical behavior or thinking must not be set apart from the range of everyday purposive conduct, particularly since even the ends of religious and magical acts are predominantly economic’. The religious aspects of this ‘everyday purposive conduct’ belong to what was called ‘social and psychological consequences’ of the notions of and the communication with the other World in (è1).
2 This distinction, as we saw in (è1), is a prerequisite for speaking meaningfully about religion at all.
25 – Various Ways of Communicating
591
tion (Clunies Ross 2002a): on the one hand, we do not have very many ritual descriptions, and there are whole categories of rituals that we have no knowledge about at all, but which we must assume by way of analogy were part of the religion; on the other hand, in the descriptions that we do have, very few are detailed enough for us to carry out classical ritual analyses. in most expositions of PCRN, therefore, the myths and mythical sphere often constitute the main part. This is reasonable, insofar as we are far better informed about these things, but in all probability it does not reflect the religion as it was viewed by those who performed the rituals: namely, the pre-Christian scandinavians.
This, of course, raises many problems, because the way the sources allow us to observe PCRN will inevitably be distorted. Moreover, it creates some methodological problems for our possibilities of reconstruction. for, on the one hand, it is always problematic to attempt to reconstruct phenomena that are not treated in the extant sources, but on the other hand, it would be naïve to reject the existence of such phenomena that are known from most comparable cultures. an example could be rituals of initiation into various kinds of office. since we are not informed about such things in any sources that are traditionally regarded as reliable, some would deny their existence, although we would a priori expect them to be there; and when it is even possible to analyse sequences, mainly found in the fornaldarsögur, as relicts of initiation ideology (schjødt 2008), it would seem appropriate to deduce that various kinds of initiation rituals actually existed in PCRN. it is obvious that there are details we will never obtain access to, but in order to attain a realistic view on the ritual and thus the religious world of the scandinavians, we have to take into consideration many phenomena that are not related in the sources, but which must have been there, in order to explain various other phenomena of which we do have some information.3 What should be avoided, however, is the attempt to reconstruct individual rites in any detail within a ritual sequence, unless such are related in the sources.
archaeology is another source that provides us with material remains of the actions connected to the ritual sphere, but seldom informs us about the precise intentions behind the actions. archaeology, therefore, answers questions such as ‘what happened’, rather than why it happened.
3 an example, to which we shall return and which has been touched upon already (è23), could be the relations between rulers and the god Óðinn. This relation would be almost impossible to understand if we do not accept that the ruler and the god are tied together through some kind of initiation ritual (è32).
592
Jens Peter Schjødt
Ritual Structures and Classification
Research History
in her interesting book, Ritual: Perspectives and Dimensions from 1997, the historian of religion Catherine Bell wrote: ‘To anyone interested in ritual in general, it becomes quickly evident that there is no clear and widely shared explanation of what constitutes ritual or how to understand it. There are only various theories, opinions, or customary notions, all of which reflect the time and place in which they are formulated’ (Bell 1997: x). These are wise words that should be borne in mind when we attempt to approach ritual as a category within the history of religions. although the same words could be used about a host, if not all, of religious phenomena, they still emphasize the complexity and the many perspectives that have been used within ritual studies. This is not the place to go through the various ‘schools’ of interpretation,4 since ritual, as a research object, has already been treated by numerous historians of religion and anthropologists, and to just mention the main positions would constitute a work in itself. as was also stated in (è1), the term ‘ritual’ has been used to designate many different kinds of behaviour, and not only human. according to some definitions, it makes sense to speak of ‘animal rituals’, acts performed by animals which have no direct influence on the goal that is to be achieved, such as, for instance, the signaling of social position within a group.5 in daily language we often speak about rituals as series of actions which are repeated with certain intervals. some elements focused upon in such a definition are also relevant when we speak about religious rituals. Nevertheless, it is important to bear in mind that, within the frames of this work, talking about rituals means talking about religious rituals and thus, in discussing and analysing ritual actions in the following, the religious components will be emphasized, that is, actions which in some way or another refer to the other World. it can of course be equally meaningful to focus on other issues than those of the religious discourse; that is just not what we shall do here.6
4 Research histories can be found in many works by anthropologists and historians of religion. good overviews are Bell (1997: 3–90) and Doty (1986).
5 ‘Direct’ is important here. as opposed to, for instance, fighting or building nests where the actions are directly connected to the result, the subjection to the ‘alpha male’ or the collective, however, is just signalled through a certain behaviour that has to be ‘interpreted’ by this male or the group in order for it to work (cf. Turner and others 2018).
6 for a brief overview on definitional issues, see Zuesse (1987).
25 – Various Ways of Communicating
593
Rituals have been dealt with by most historians of religion as well as by many anthropologists, psychologists, and sociologists, because it is obvious to anybody who studies especially preliterate cultures, but also to a very large extent modern societies, that rituals play a huge role at several levels. in the book just mentioned, by Catherine Bell, the author attempts in the first three chapters to outline the history of research from the end of the nineteenth century up to the end of the twentieth, and the titles and subtitles of the chapters can thus be taken as indications of the various aspects that have attracted scholarly interest in ritual studies within the last century. We shall not deal with all these perspectives, but in order to give an impression of the variety of interests it may be worthwhile to mention these titles. although Bell’s chapters are structured in mainly chronological order, it is obvious, as she is also aware of herself (1997: 88), that most of the positions she discusses, including the older ones, are still part of the scholarly discourse. The three main chapters in this section are: ‘Myth and Ritual: Questions of origin and Essence’, ‘Ritual and society: Questions of social function and structure’, and ‘Ritual symbols, syntax and Praxis: Questions of Cultural Meaning and interpretation’, respectively.
in the first chapter, some of the subtitles are: ‘The Myth and Ritual schools’,
‘The Phenomenology of Religions’, and ‘Psychoanalytic approaches to Ritual’; in the second we find ‘functionalism’, ‘Neofunctional systems analyses’,
‘structuralism’, and ‘Magic, Religion, and science’ as subtitles; and in the third chapter, the subtitles are: ‘symbolic systems and symbolic action’, ‘linguistics’,
‘Performance’, and ‘Practice’. although the expediency of this classification can in some respects be discussed, the titles give a fairly complete view of the issues and perspectives that have been discussed within the general History of Religions. They are certainly not all of equal relevance to the study of old Norse religion, not least because, as we have seen, the sources do not allow us to reconstruct every aspect of that religion and its rituals. However, they are all relevant for the forming of general theories about rituals. it will be too much of a digression away from PCRN to go into all these points, but a few words about some of them will be necessary in order to help establish a comprehensive view on rituals within PCRN with regard to such elements as function, symbolic meaning, classification, and structure.
at first sight it may seem that the various positions mentioned by Bell are heterogenous — and it is true that they focus on very different aspects of ritual.
But it is important to acknowledge that many of these aspects can actually be seen as more or less complementary. all rituals are, for example, performed; many rituals have political alongside religious functions; rituals often have a strong impact on the psychology of the participants; they follow a certain
594
Jens Peter Schjødt
structure; and they frequently create social solidarity. it therefore makes good sense to take into consideration sociological as well as psychological factors, just as semiotics and performance theory, and so forth, can all be very useful to us if we want to understand all the various aspects of a given ritual. There are also many kinds and categories of rituals, as we shall see in a moment, and their position and function within a certain culture may vary substantially. all this makes it problematic to find elements which are relevant for all rituals, whether they have to do with function, meaning, performance, or any other aspect.
one important problem is the one concerning emic and etic interpretations and functions. During most of the 20th century, scholarship was primarily occupied with etic aspects of ritual. in discussing what rituals were good for at all, there has been a strong focus on psychological and, in particular, sociological perspectives. Taking the most influential scholars from the beginning of the twentieth century, we should of course mention sigmund freud (1913) and émile Durkheim (1912), both of whom were primarily interested in the ritual side of religion.7 To freud, sacrifice, which he saw as the origin of religion, was a celebration of the killing of the totem forefather with the purpose of obtaining access to the females within the group, and to Durkheim, as we have seen above, rituals were carried out in order to strengthen the solidarity within the group. Whereas freud has no impact on ritual studies nowadays, Durkheim certainly has, and the idea that ritual should first and foremost be considered a sociological phenomenon, organizing both the solidarity and the hierarchical structure of the society, is widely acknowledged by modern scholars, although there may also be other, compatible functions involved, not least that of communication: ritual is a form of communication through which roles are defined.8 furthermore, we can mention Jan assmann, who has suggested that ritual is a means of remembering in illiterate societies (2006: 39–40); the notion of memory has also been taken up by Harvey Whitehouse (2000), although from a different perspective wherein different kinds of memory are classified and analysed in relation to various rituals (cf. Nygaard and schjødt 7 in opposition to the enterprise of Comparative Religion in most of the nineteenth century, where myth was the main focus, not least in connection with the so called nature mythology of friedrich Max Müller (for example, 1889, 1909) and his ‘school’. The ‘change’ of focus from myth to ritual came with William Robertson smith, particularly his book The Religion of the Semites from 1889 in which he maintained that sacrifice was the fundamental religious phenomenon (see also Bell 1997: 4–8) from which the phenomenon of religion as such could be derived. freud (1913) was heavily influenced by Robertson smith.
8 Especially within social anthropology, a host of famous scholars have focused on this
‘function’ of ritual. see, for instance, Turner (1969), leach (1976), and geertz (1973).
25 – Various Ways of Communicating
595
2018). at any rate, many of these theories are by no means mutually incompatible, and these as well as many others have all contributed to a better understanding of the phenomenon of ritual, both in a religious and a non-religious sense.
Compared to the amount of these etic theories, the amount of theories dealing with ritual from an emic point of view is relatively small and have mostly been proposed by phenomenologists of religion. again, we notice that the two viewpoints are not at all incompatible: they simply reflect various interests and different levels within the rituals themselves. Thus, there can be no doubt that rituals always serve social and/or psychological functions which are, however, not the reason for their existence according to the participants. for them, it could be a question of serving god or promoting fertility,9
whereas the display of power relations that exist in many rituals is seldom consciously acknowledged by those who participate in them. Whereas questions of
‘function’, whether psychological or sociological, are usually considered from an etic perspective, questions of ‘meaning’, which are often asked by the phenomenologists, normally adopt an emic perspective.10 Perhaps we might say that the communicative function between the various participants in a ritual, that is, the ‘horizontal’ communication, is of an etic kind, whereas the ‘vertical’ communication, that between this and the other World, could be said to belong to the emic aspects of the ritual. However, the dichotomy emic/etic is often not very clear-cut and, for, instance, within semiotics and what we could broadly term structuralist studies it may be hard to decide whether the decod-ing of the symbols involved in a certain ritual should be attributed to one or the other perspective. Nevertheless, also when aiming to reconstruct pre-Christian scandinavian rituals, it is of importance to bear these different levels in mind.
another problem touched upon above is that concerning myth and ritual.
We shall not deal with the so-called ‘Myth and Ritual school’ which flourished around the end of the nineteenth and in the beginning of the twentieth centu-9 it has been discussed (e.g., Bell 1997: 108–09) whether it is possible to distinguish between exchanges made for pure devotion and those that rather have the character of bribes.
although such a distinction may have some importance, it is, however, important to be aware that whether we talk about ‘devotion’ or ‘bribe’ it is always a matter of manipulating the other World to look kindly upon our world, or at least upon the performer of the ritual. We shall return to this below in relation to sacrifice.
10 in reality, ‘meaning’ and ‘function’ are often heavily intertwined. for instance, if the purpose of a ritual, from an emic point of view, is to promote the fertility of beast and soil, it is not easy to decide whether this purpose should be seen as an emic ‘function’ or an emic ‘meaning’. for some aspects of this problem, see strenski (1993: 35–36).
596
Jens Peter Schjødt
ries (cf. fontenrose 1966; Bell 1997: 5–8; Doty 1986: 73–78; segal 1998; segal 2017) with such prolific authors as James george frazer, Jane Harrison, and arthur M. Hocart. The basic idea that myth and ritual related to one another in the same stable way as speech does to action, the myth saying in words what the ritual says in actions, has long been rejected, not least because many rituals are not linked to any specific myth and vice versa, many myths are not part of any ritual. The fact, however, remains that some rituals and some myths are strongly connected to one another.11 But often this relation is much less direct than what was proposed by the proponents of the ‘Myth and Ritual school’, so that most often we cannot expect a fixed narrative to have accompanied the performance of a ritual, interpreting the gestures so to speak for the audience.
That kind of relation was analysed within PCRN by, among others, Vilhelm grønbech (1932) and Bertha Phillpotts (1920) in the beginning of the twentieth century and, although Phillpotts in particular seems to be convincing in some of her analyses, it is obvious that such relations between myth and ritual, even in cases where they clearly do exist, are usually much more complex than this simple formula suggests. The relation between the two categories should perhaps rather be seen as that of a mythic construction of an other World vis-à-vis an attempt at ritual manipulation of this other World. for rituals do something; they are there in order to change or maintain elements in our world, whether it be the making of a king or the rhythms of nature. But in order to work properly, there must be some idea of the other World, formed via mythic narratives, so that the communicative activity can be as efficient as possible in order to achieve the wished-for goals. Therefore, a ritual performance may at times take the form of a repetition of certain deeds performed by the gods in illo tempore — for instance, the cosmogonic myth may be reenacted
— whereas in other instances it is simply a question of approaching the other World, knowing what it takes to ‘bribe’ a specific god; and such instances will often not involve the reenactment of individual myths.
Thus, whatever etic functions a ritual may have, it is, seen from the performer’s perspective, a more or less directly expressed attempt at communication with and manipulation of the other World. Prayers might be said to constitute another kind of communication. However, since most prayers are accompanied by bodily acts (bending down, folding the hands, etc.), there is hardly any reason to classify prayers as a category different from rituals in general; rather, 11 in a recent article frog (2017) has once again taken up the issue, but this time within a semiotic paradigm and using the tool of parallelism to illuminate the relationship between ritual and accompanying myths.
25 – Various Ways of Communicating
597
they should be seen as a particular category of rituals in which words play a more dominant role than is often the case in ritual performances.
Most rituals are composed of longer or shorter sequences of actions. in the following we shall call these actions ‘rites’. Thus, in the terminology used in the following, ‘rituals’ connote a sequence of ‘rites’ and are characterized by a certain goal, or perhaps several goals; whereas the rite, although it may very well have a ‘meaning’ in itself, basically contributes to the goal of the ritual.12 a
‘ritual’, then, may have as its ultimate goal the promotion of the fertility of the fields. This ‘ritual’ will normally consist of a series of ‘rites’, which may involve such elements as sacrifices, dramatic performances of fights between gods of chaos and gods defending the cosmic order, ‘imitative rites’ such as pouring water from a vat, and so forth. These all require their own interpretation; the sacrifice may be viewed as a gift to the gods in order to sway them to let it rain; fights between gods and demons enacted as certain dances may be a ‘magic’ rite, aimed at repeating what happened in illo tempore, and the pouring of water may indicate another kind of sacrificial gift to the gods of the underworld. yet first and foremost they all contribute to the overarching goal: namely, to allow the fields to prosper. apart from all the etic functions that we may be able to detect in such a ritual, the main goal seems quite clear from the point of view of the participants. it is of importance to realize that often the ‘rites’ will acquire their significance only from their position within the ‘ritual’, and therefore it is quite likely that the ‘same rite’ can be part of various rituals and therefore take on different ‘meanings’, which seems to be the case with for instance the so-called jarðarmen rite, as we shall return to below (è32). Within PCRN, especially when it comes to the archaeological record, we are often in a position where we can reconstruct (at least to some extent) some of the rites that have taken place, whereas we can seldom reconstruct whole rituals without a great deal of conjecture. as is the case with all taxonomies, there will be cases where it is not easy to make a clear-cut distinction between the categories, that is, rituals and rites.
for instance, a sacrificial rite or a divination rite may be so dominant within a ritual that the rites themselves will consist of a large amount of individual actions (perhaps even structured in the same way as a full ritual sequence), such 12 it must be noticed that this distinction and thus the terminology is not common among historians of religion and anthropologists. Usually, there is no real distinction between the two words (this is true not only in the English-speaking world), and they are mostly used synonymously. However, since it makes sense to distinguish between the individual units within a ritual sequence on the one hand and the ritual sequence itself on the other hand, the two words, rite and ritual, may be meaningfully used to denote these two entities respectively.
598
Jens Peter Schjødt
as consecrations of the place and participants involved, purifications, and invocations of the gods.
a word that will not be used very often in the following is ‘ceremony’. The reason why this term is avoided is that it is frequently used simply as a synonym for ‘ritual’ (but not for the individual ‘rite’), although a ‘ceremony’ is commonly conceived of as more spectacular than a ‘ritual’, at least the rituals of a private kind. However, most people probably associate the word ‘ceremony’
with all kinds of public celebrations, and, to a much higher degree than is the case for ‘ritual’, it is far from necessarily connected to the religious sphere. We do, of course, speak of ‘religious ceremonies’, but without the adjective ‘ceremony’ is not so closely associated with religion. Therefore, ‘ceremony’ should be regarded as a facet of certain public rituals (religious or not).
Classification of Rituals
There are different kinds of rituals, and therefore we cannot escape the problem of how to classify them. in order to organize the exposition, it is necessary to divide the vast amount of information into some ritual categories. as an example of how this may be done, we can refer once again to Catherine Bell.
in Chapter 4 of her book on ritual (1997: 91–137), she classifies various rituals into the following subcategories: ‘Rites of Passage’;13 ‘Calendrical Rites’;
‘Rites of Exchange and Communion’; ‘Rites of affliction’; ‘feasting, fasting, and festivals’; and finally ‘Political Rites’. These categories are on the one hand quite familiar within the history of religions, but on the other they show some of the difficulties that we face when we attempt to classify. The main problem seems to be that dissimilar criteria are used in the differentiation between the categories. for instance, we notice that the category ‘Rites of Exchange and Communion’ comprises offerings and sacrifices, phenomena which have been classified above as ‘rites’ and not ‘rituals’. so, whereas ‘Rites of Passage’, for instance, is a category defined through the social occasions which they accompany, ‘Rites of affliction’ and ‘feasting, fasting, and festivals’ seem rather to be defined through the moods of the participants. The classification issue, however, has been addressed by many scholars, and this is not the place to go into any deep discussion.14
13 Notice the use here of ‘rites’. To Bell and, as mentioned, to many others, the term often covers what is here called ‘rituals’.
14 a brief discussion can be found in schjødt (2008: 46–48), and a bit more detailed in schjødt (1986a).
25 – Various Ways of Communicating
599
Taking the above-mentioned distinction between ‘rituals’ and ‘rites’ into consideration, a useful classification could be the one proposed many years ago by the finnish folklorist and historian of religion lauri Honko.15 Honko divides the rituals into the following three categories: ‘Rites of Passage’,
‘Calendrical Rites’ and ‘Crisis Rites’, a categorization both similar to and different from that proposed by Bell. But Honko is much more consistent than Bell in the criteria he uses in order to make the distinctions. The three categories can be distinguished in various ways, and Honko proposes criteria such as ‘social orientation’, whether the rites are recurring and whether they are predictable (Honko 1975: 75). Thus, according to Honko a ‘rite of passage’ is characterized by being individual in its social orientation, by being non-recurring — each person will in principle only experience a certain initiation once in his or her lifetime — and by being predictable: although we may not know the exact date in advance, it is certainly predictable that a person has to go through a puberty-ritual or a marriage ritual.16 although these criteria are, as just mentioned, quite systematically applied, they are certainly not without problems when it comes to the details; for instance: is a passage ritual always individual in its social orientation? Nevertheless, the division into these three categories seems appropriate, since most religious rituals appear to belong to one or the other.
another criterion of a structural kind has been suggested by Jens Peter schjødt (2008: 47–48). in arguing that rituals in general serve to maintain or improve the initial condition, we can distinguish three qualitative levels: the ‘normal’ as well as what is ‘below’ and ‘above’ the ‘normal’. Thus, whereas a passage ritual aims at bringing the initiates to a ‘higher’ level than the ‘normal’, calendrical rituals aim at maintaining the world as it is, while crisis rituals aim at returning to a normal level from ‘below’, which is what comprises the ‘crisis’. Therefore, the relation between the initial and the final phases of the ritual constitutes a taxonomic criterion based on structural relations.
The three categories proposed by Honko will, therefore, be used in the presentations included in the following chapters, although the term ‘cyclical’ will to a great extent replace ‘calendrical’. as we saw above, many other proposals for 15 Honko did not himself distinguish between ‘rituals’ and ‘rites’, as is indicated by the title of one of his two articles that are of importance in this connection (Honko 1975 and 1979), which is called ‘Zur Klassifikation der Riten’, where the term ‘Riten’ covers what is here called
‘rituals’. That is, however, not relevant here, although it prevented Honko from relating individual rites, such as sacrifices or purification rites, to his overall categories.
16 This may be true of the ‘biologically’ oriented occasions (birth, puberty, marriage, and death) whereas it is less obvious that the entrance into some secret society or mystical organization should be predictable.
600
Jens Peter Schjødt
classifications of rituals have been suggested within the general history of religions and, apart from Honko’s three main categories, it is useful also to take yet another criterion into consideration: namely, the one of public versus private rituals. This division is applicable to all of the three above-mentioned categories. Rituals of passage may be performed as small-scale ceremonies within the family, as is often the case with rituals in connection with births, or they may be spectacular rituals for the whole people, as in coronation rituals; cyclical rituals may take the form of modest offerings to the spirits of the farm, and they may be celebrated by the entire village or even the whole nation participating.
The same is true of crisis rituals, which may be performed to cure an individual or to avert a war that would involve the entire nation. This, of course, has as a consequence that, within the three main categories, there are huge differences, and the taxonomy suggested is only a means of structuring the vast amount of ritual manifestations in religious worlds. as we shall see in this and subsequent chapters, examples of all these ‘types’ of rituals are also found within PCRN: private as well as public versions of the three main categories.
Some Basic Characteristics of Ritual Performance
one of the most conspicuous characteristics of rituals is their diversity; huge differences from one culture to another, vast differences between the various categories, and great differences within each category in each culture. function, purpose, the level of expenditure, number of participants, and so forth vary immensely. But at the same time there seem to be some elements of a symbolic and structural kind that are common to all religious rituals. one of these is the distinction between this world and the other World, often related to the distinction between profane and sacred. This basic dichotomy, which lies at the core of all religious life, generates structures and symbolisms that are more or less visible in all religious rituals,17 which is what we shall discuss briefly here.
one of the most useful general theories of ritual structure was proposed by the anthropologist Victor W. Turner back in the 1960s on the basis of the analyses carried out by arnold van gennep in his book on passage rituals (van gennep 1960).18 Turner argues that normal life in any society was 17 it is obvious that in rituals performed in private, perhaps by a single individual and lasting only a few minutes, as is the case with many prayers, it will be very hard to recognize these structures, at least in their entire complexity. Nevertheless, although some elements may be invisible, it is most often possible to apply the fundamental structures.
18 for a good, albeit brief, discussion of Turner’s contribution to both ritual studies
25 – Various Ways of Communicating
601
at certain times interrupted by liminal periods, liminality being defined as a period ‘where there is a certain freedom to juggle with the factors of existence’
(Turner 1967: 106), and that this occurs primarily during certain phases of ritual performance. But not only then since the liminal may also in some cultures become a way of life characterized as being an opposition to the ordinary life of ordinary people, a social and psycho-sociological phenomenon which is known particularly from religions with universal claims, that is, those Robert Bellah termed ‘axial religions’ (Bellah 2011).19 Concerning rituals, Turner concurs with van gennep in distinguishing three phases of ritual: namely, a separation phase during which the participants of the ritual are separated from the everyday world in order to enter the liminal space; this is followed by the liminal period, to which we shall return shortly, and following that comes a phase in which the participants return to their everyday mode of existence. The liminal period is clearly the most important since the other two can be seen as merely transitional phases between the two modes of existence: the normal, everyday existence and the liminal. This liminal phase is in some way inverted in comparison to everyday existence, and it characterizes not only religious outlooks, wherein the two modes of existence are often regarded as ‘sacred’ and ‘profane’, but is also a phenomenon of significance to psychology and sociology all over the world (è1). liminality in general, and thus also ritual liminality, does not and other aspects in his work, see Morris (1987: 235–63). Turner was an extremely productive author, and, besides many analyses of aspects of the Ndembu culture in Zambia, he wrote mostly on ritual in general, but also on pilgrimage and theatre. The works most relevant to the present topic are his books The Forest of Symbols from 1967 and The Ritual Process from 1969.
19 a kind of definition of Turner’s understanding of liminality is found in the following quotation: ‘a limen is a threshold, but at least in the case of protracted initiation rites or major seasonal festivals, it is a very long threshold, a corridor almost, or a tunnel which may, indeed, become a pilgrim’s road or passing from dynamics to statics, may cease to be a mere transition and become a set way of life, a state, that of the anchorite, or monk. let us refer to the state and process of midtransition as ‘liminality’ and consider a few of its very odd properties.
Those undergoing it — call them ‘liminaries’ — are betwixt-and-between established states of politico-jural structure. They evade ordinary cognitive classification, too, for they are neither-this-nor-that, here-nor-there, one-thing-nor-the-other. out of their mundane structural context, they are in a sense ‘dead’ to the world, and liminality has many symbols of death — novices may be classed with spirits or ancestors or painted black […] the most characteristic midliminal symbolism is that of paradox, or being both this and that. Novices are portrayed and act as androgynous, or as both living and dead, at once ghost and babes, both cultural and natural creatures, human and animal.’ (Turner 1977: 37). it is obvious from this that Turner is talking mainly about ‘novices’, that is, participants in initiation rituals, but the characteristics may easily be extended to other kinds of rituals.
602
Jens Peter Schjødt
have any inherent general properties except that it is always defined as being in some way in opposition to the non-liminal, the world of rationality, of technological knowledge, where the achieved goals are a function of the mundane procedures that are used, without interference from the other World. We might say that liminal symbolism is ‘borrowed’ from the semantic universe that constitutes the culture in question — only in inverted form. Turner gives a long list of examples of what may count as liminal symbolism (Turner 1969: 106), such as, for instance, sacredness as opposed to secularity, absence of rank as opposed to distinctions of rank, totality in opposition to partiality. The liminal is thus what everyday life is not. from a ritual perspective, then, the liminal phase is the centre of the ritual, the pivotal point so to speak. The two other phases are, of course, necessary because the two worlds are not directly compatible: in order for beings in this world to approach the other World, they must to some extent become a bit ‘otherworldly’ themselves, which is done by putting their non-liminal way of being aside. This is often symbolically expressed in separation rites: the participants, for instance, dress differently from their everyday clothing or they must be purified so as to be cleansed from any sort of profane ‘dirt’. When this has been achieved, they are ready to step into the liminal space, which is a space somewhat similar to that of the other World, so that the two worlds can meet on relatively equal terms. and that is what makes communication between the two worlds possible. in the same way, the rites of reintegration have as their main purpose to strip the participants of their sacra, the sacred attributes that have been attached to the participants during the separation rites. Thus, the degree of sacredness is gradually diminished during this phase. The foundations of these two phases are therefore obvious: because this and the other World are so profoundly different,20 it is necessary to make them compatible through mediating rites.
in spite of the variety of ritual liminality, it normally involves a liminal space, a liminal time, and a number of liminal actors. These phenomena within PCRN will be treated in subsequent chapters (è27–29), and only a couple of general remarks shall be presented here. The ritual liminal space may be constructed in many different ways (cf. Murphy 2017c). There are permanent cult sites which may take the form of buildings such as churches or they can be parts of the natural environment, such as lakes or groves, but there are also places 20 This is why we often see that the mixing of the two spheres is regarded as unclean or taboo if no (ritual) efforts have been made to bring them together, efforts which are often in the hands of ritual specialists who are more knowledgeable about the other World as well as about communication with it than are ordinary people.
25 – Various Ways of Communicating
603
that are only liminal or sacred at certain ritual occasions,21 for instance, when sacrifices are about to be carried out. However, there are some general features of sacred spaces: namely, that they are formed either as spaces with a built-in opposition between centre and periphery, the space becoming gradually more and more sacred the closer you come to the centre, or there is some sort of path that the participants move along, coming closer to the sacred as they go along.
This latter type is seen in, for example, Christian churches, but also processions roads are often structured as a path towards the sacred. liminal periods also differ greatly from each other, some being very long, even lasting for several years (although in such cases with varying intensity), whereas others only last a few minutes, which is what we often see in prayers where the hands are folded or the head bent down in a certain direction — toward the sacred — as a signal that the everyday world has been left for a short period. and also the amount and character of the liminal actors involved will naturally vary from one ritual to the next. Thus, we also see an enormous difference in dramatic expression from rituals carried out in solitude, establishing a liminal space for a brief prayer, to those carried out collectively with the whole population attending, perhaps involving large-scale sacrifices, huge processions, and feasts with eating and drinking, often led by ritual specialists who may be identical with the king or chieftain, or they may be priests or shamans, and so forth. Many of these sacred figures are sacred only during the rituals, as is typically the case with shamans, whereas others are attributed a more permanent sacredness, although, again, it usually differs in degree from liminal to non-liminal periods. The ritual actors, both religious specialists and ordinary participants, will typically dress and behave in ways which are somehow different from how they usually go about things.
so what we basically face here during the liminal phase of rituals is the turning upside down of (some aspects) of everyday life, which is in accordance with the basic religious figure of thought: namely, that there is an other World that is in important aspects opposite to the world in which we live our daily lives. going into a liminal space and a liminal period therefore means that we approach this other World and establish some sort of analogy with it in order to facilitate communication with it. This is also why in so many rituals all over the world we meet the phenomenon that the gods are present, either in the form of statues and other idols, or in the form of human beings who are dressed 21 and even the permanent ‘sacred’ places are naturally more sacred on ritual occasions than during everyday life, although they often inspire a certain degree of awe at all times (even when there is no sermon, we tend to lower our voices when we walk into a church).
604
Jens Peter Schjødt
up like beings from the other World. in such cases, these figures emphasize the fact that right now, on this spot, we are either in an other World or somewhere in between this world and the other. The main point is that we have moved into a sphere which is different (to varying degrees, according to the type of ritual we are dealing with; see above) from the one in which we live our daily life and thus have approached the other World: we have constructed a space which is compatible with that of the gods. from here, we are in a position to communicate with the other World and to manipulate it for the benefit of society and individual, whether in this daily life or in the life hereafter, in the beyond.
in (è30–32) we shall deal with the main categories of rituals, as defined above, whereas in the rest of this chapter we shall attempt to reconstruct some of the more prominent rite categories, that is, rites that may be part of any one of the three above-mentioned main ritual categories. from a phenomenological perspective, the rites dealt with here do not constitute an exhaustive review since it may very well be that in PCRN many more rites were carried out in connection with various rituals; however, even if this is likely to be the case, it is almost impossible to detect them in the sources. This is the situation that faces us with regard to such otherwise important rites as purifications, various apotropaic rites, and many more that are common in other religions.
Some Important Rite Categories
Sacrifice as a General Category
sacrifice is often seen as the most conspicuous way of communicating with the gods.22 and to many, sacrifice is defined as such because some sacrificial object
— be it a thing, some vegetables, an animal, a human, or something else — is destroyed during the ritual act, whether burned, drowned, or killed in some other way. in spite of this basic commonality, very different ideas can be linked to such a destruction, although they may take similar ritual forms. Thus, there seem to be at least two ideas which both necessitate the destruction of a victim: that involved in the so-called ‘gift sacrifice’ and that which is implied in the so-22 The category is without any doubt universal, although it may take on very different shapes from culture to culture. arguments that ‘sacrifice’, as the term is used in scholarly literature, is not universal (e.g., Berggren 2006) naturally depend on the definition. if it is defined in a culture-specific way, it is obviously not universal, whereas defining it as an object being transformed in order to mediate between this world and the other, certainly makes it universal. in that sense, the problems in connection with sacrifice are no different from those involved in any other categorization.
25 – Various Ways of Communicating
605
called ‘communion sacrifice’.23 in both types, the sacrificial object, or parts of it, is usually destroyed and sometimes eaten by the participants, although this is not essential in gift sacrifices. as the terms indicate, the idea behind the gift sacrifice is that the sacrificial object is regarded as a gift to the gods, and during the rite it is more or less directly stated that something is wanted from the gods in return. Therefore, this kind of sacrifice is commonly characterized as do-ut-des (‘i give in order that you will give’). This is also part of the simple logic connected to the very notion of the gift as this was analysed by Marcel Mauss many years ago in his famous essay about gift-giving (Mauss 1974).24 in this case, when the gift is given to the gods, the return gift is always of higher value than the gift presented by the humans: if we sacrifice this ox, we expect that you give us plenty of fertility for beast and soil; or if we sacrifice this prisoner of war, we expect that you give us victory in the battle we are about to fight, and so forth. The idea behind the communion sacrifice is quite different. Here, the victim is in some mystical sense identified with the god, and the communion therefore consists in a bond between a god and his worshippers: by eating the god, for instance the totemic ancestor, the worshippers will acquire part of the divine, thus strengthening the relation between humans and deity. again, we notice the idea that, during the ritual, a certain compatibility, or in this case perhaps rather a kind of identity, is aimed at. on an abstract level, the two kinds of sacrifice have the same goal: to strengthen the good relation between the gods and the humans performing the sacrifice. But apart from that, the two modes of thought involved are quite different, because in the gift sacrifice there is a clear distinction among the gift, the victim, and the god,25 whereas in the 23 This term is often used to characterize sacrifices in which, after the killing of the sacrificial animal, the participants eat together with the god. Here, however, it denotes the kind of sacrifices in which the sacrificial victim is in some sense seen as an incarnation of the god. sacrificial meals following the killing are, from the perspective we adopt here, still ‘gift sacrifices’ insofar as part of the victim is seen as a gift to the gods. The classical example is the main greek sacrifice, known for example from Homer, in which part of the animal is burnt so that the gods will enjoy the smell of the sacrificial victim, whereas the participants will eat the rest at a ritual banquet.
24 in introductory works on religion as well as on PCRN, we often meet many more categories of sacrifices, such as, for instance, propitiatory sacrifices, thanks offering, etc. But from the perspective we adopt here, these are simply variations of the gift sacrifice: there may be various reasons for giving the gift to the gods, but this, in our view, does not alter in any significant way the structural and semantic unity behind all these sacrifices: humans attempt to manipulate beings from the other World with some sort of ‘bribe’.
25 This said, it should also be remembered that gifts often involve the notion that part of the giver is somehow included in the gift.
606
Jens Peter Schjødt
communion sacrifice this distinction is somewhat blurred. Even if ‘eating’ plays a dominant role in both kinds of sacrifices — it is essential in the communion sacrifices and often a part of gift sacrifices in which only part of the victim is burned or otherwise destroyed whereas the rest is eaten by the participants26
— the symbolic meaning of eating is different: in communion sacrifices, the participants ‘eat’ the god, whereas in gift sacrifices they eat with the gods. seen from the outside, however, it may be hard to distinguish between these two kinds of eating in actual rituals.27
further, as was mentioned above, there is clearly a difference in the direct-ness of manipulation aimed at. in particular in many communion sacrifices, but also in some gift sacrifices, the manipulative aspect is rather downplayed, and we sometimes get the impression that we are dealing with merely devotional actions. in other instances, the direct exchange is very clear, sometimes even taken to the extreme, so that the sacrificial gift will only be given when the gods have given the ‘return gift’.28 in this end of the spectrum, we have the so-called
‘thank offerings’ in which sacrifices are performed because the crisis has been solved. But the manipulative aspect should not be overlooked: pleasing the gods is surely a means of making them look kindly upon the devotees.29 Both 26 However, this is certainly not always the case. often, the victim is completely destroyed in order to reach the realm of the gods and nothing is left for a ritual meal.
27 in actual rituals performed, the two kinds of sacrifices may also be combined. a good example is found among the ainu of Hokaido in their bear sacrifices (Batchelor 1971). Here, we learn that the sacrificial victim, the bear, is in some sense a god itself, but however the bear is also presented as a gift to its ‘parents’, the bears of the other World, so that they will send other bears to be hunted by humans. Even so, the sacrifice ends with a festive meal during which the participants eat the meat of the bear. in this case as well as in others, it is somewhat difficult to decide whether the meal is believed to be a symbolic ‘eating the god’ or whether it is simply a meal, which is supposed to strengthen the bonds between the other World bears and the humans. for instance, Marshall sahlins (1978: 46) writes: ‘offered as foods to the gods, the victim takes on the nature of the god. Consumed then by man, the offering transmits this divine power to man’. Thus, the distinction between the two categories of sacrifice is more of an analytical kind than a description of actual rites carried out in various cultures.
28 The most famous example of this is probably the Roman ver sacrum (sacred spring) sacrifice where promises are made to carry out sacrifices once the gods have fulfilled their part of the ‘contract’.
29 Bell writes (1997: 110): ‘Devotional offerings to the deity are not meant to result in direct or immediate concrete benefits, although they are understood to nurture a positive human-divine relationship that will benefit the devotee spiritually and substantively’. This is important to bear in mind because, no matter how ‘spiritual’ the emic explanation will be, it has the same characteristic as many mundane social situations: we communicate in order to
25 – Various Ways of Communicating
607
types of sacrifices are widely known and can be found in all kinds of cultures whether they are part of the ‘official’ religion or not.
Communion sacrifice was seen by William Robertson smith (1972 [1889]) as the ‘original’ form of sacrifice, providing the foundation for a host of theo-rists during the twentieth century, in a kind of opposition to the earlier dominant view, represented by, for instance, Edward Burnett Tylor (1871), that sacrifice should be seen as a gift to the beings of the other World. Whether there is such an origin that can explain all sacrifices in all cultures is, however, a question which shall not be pursued here; we shall just note that the sacrificial victim or object always serves to mediate the relation between this and the other World. in scandinavia, the sources tell us next to nothing about communion sacrifices, and the present discussion will thus concentrate on the various forms of gift sacrifices.
However, when we attempt to reconstruct rituals in PCRN, the question of structure is more important than the origin of sacrifice.30 This issue was the theme of one of the most influential books on sacrifice of all times: namely, Sacrifice: Its Nature and Function, originally published in french in 1898, by Henri Hubert and Marcel Mauss (1964). Their main point was that all sacrifices follow a certain structure, which consists of three phases: that of sacralization — the entry; that of the ritual killing — the sacrifice proper; and that of desacralization — the concluding rites. This idea recalls the structure that was described in connection with ‘Rites of Passage’ by van gennep some ten years later.31 The sacralization concerns both the victim and the participants of the ritual, sometimes even the place where the sacrifice is carried out (if it is not in a permanent sacred room — and sometimes also then) and perhaps the instruments used for the rite, because, as we saw above, the two worlds are incompatible unless the ‘profane’ is somehow transformed into a sacred entity. The desa-somehow manipulate. see also the definition of ritual proposed by Jørgen Podemann sørensen (1993: 19–20): ‘[ritual is] representative acts designed to change or maintain their object’. This definition may not be a precise description of religious rituals, but it nonetheless focuses on the main purpose of such rituals, not least sacrifices, in emphasizing the manipulative aspect.
30 for speculations about the origin, we can refer to such distinguished scholars as sigmund freud (1913), who argues that sacrifice was based on the actual killing of the father by the sons; Walter Burkert (1972), who argues that the hunters’ killing of an animal lies behind sacrificial killings; and René girard (1982), who believes that the ultimate reason for sacrifice was found in the figure of the scapegoat. Many other theories have been proposed, for which we refer to Carter (2003).
31 a brief discussion of the relation between the two ideas is taken up in van Baal and van Beek (1985: 131–32).
608
Jens Peter Schjødt
cralization, however, concerns normally only the human actors of the sacrificial act. in this way, sacrifices work as communicative acts between the two worlds with the sacrificial object, most often in the form of a victim, as the medium.
Therefore, the victim can be said to be transformed in two steps: first it is consecrated; and second, it is killed in order to be transferred to the other World.
in attempting to reconstruct sacrifices within PCRN from archaeological as well as written sources, and having almost no written records that describe in any detail the sequences of ritual actions, it may be of great importance to bear this general structure in mind in order to place the bits and pieces that we do have correctly. if we do not use this or other models, it is simply not possible to place the details meaningfully within the sacrificial ritual.
Nordic Sacrifices
Turning now to the Nordic material, we must make clear at once that the textual sources provide no such thing as a full description of a sacrifice understood as the above-described sequence of first hallowing the sacrificial object and the participants, then killing or destroying the object and then a phase of desacralization in order to return to the normal non-liminal situation. However, we do know a good deal about various elements pertaining to sacrifices, such as the seasonal timing of the three official main sacrifices (è28) and quite a bit about the various actors and participants in the sacrifices (è29) as well as about the sacrificial victims, not least through the help of archaeology (è27). in fact, there are hints at sacrifices scattered all over the written records, the most detailed being the descriptions in Hákonar saga góða ch. 14 of sigurðr jarl’s feast at Hlaðir, in Eyrbyggja saga ch. 4 of Þórólfr’s in iceland, and in Kjalnesinga saga ch. 2 of Þorgrímr goði’s, also in iceland (è31).32 although the problems relating to the archaeological sources are quite different, the information we get from them is also only partial: the situation to which we have direct access is only the last scene of the ritual (cf. Price 2008b), and we have to reconstruct 32 The source value of the sagas has been much debated, not least when it comes to the descriptions of sacrifices, as mentioned above. it has, for instance, been argued that the accounts in the three texts ( Hákonar saga góða, Eyrbyggja saga, and Kjalnesinga saga) are partly dependent on each other (Beck 1967: 21) and therefore not reliable when it comes to pagan cult. That may be so, but however the saga authors, even if they actually did know the pagan tradition, would have had great difficulties in convincing the modern source critics that this was the case.
anyway, as has been stated by anders Hultgård (1993: 235–36), when we take comparative material into consideration, most of what is related in these sagas seems to be quite in accordance with what we should expect from these sorts of sacrifices in pagan times.
25 – Various Ways of Communicating
609
from this scene all that went before, a task that we will never be able to accomplish in any detail, particularly not when it comes to the concepts accompanying the ritual performance.
The main word for sacrifice was blót, with the verbal form blóta. Usually, but not always, the grammatical construction places the receiver of the blót in the accusative and the sacrificial object in the dative, which should probably be seen as an instrumental. The purpose or the goal for which the sacrifice is carried out (the return object) is normally constructed with the preposition til (for instance til árs ok friðar, literally ‘for year and peace’). The man who performs the blót is usually called a blótmaðr,33 and there are other compounds with blót-, such as blótfé, blótbolli, and others.34 The etymology is not quite clear, but its gothic cognate blotan means ‘to honour’, and thus blóta may more specifically mean ‘to honour with sacrifice’ (cf. Maier 2003a: 108; Maier 2003b: 72; de Vries 1962a: 45). To simply ‘sacrifice’ seems, however, to be the most common semantic content of the word.
another term which denotes sacrifice is senda, meaning ‘to send’, and thus emphasizing the aspect of transportation of the object from this to the other World, but the word also may mean ‘to kill’.35 a third word which is partly synonymous is sóa; it, too, may mean ‘to kill’ and is connected to blood sacrifices in particular (cf. de Vries 1962a: 528; Näsström 2001: 25; Beck 1967: 117–19).36
These three words, blóta, senda, and sóa, are likely to have focused on different aspects of the sacrificial act, as is indicated in stanza 144 of Hávamál,37 but in the extant sources they often seem to be used more or less synonymously. it 33 or the hofgoði, the person who administrated the hof, the cult house, and who appears to be the chieftain, also functioning as a priest.
34 for further linguistic discussion, see Beck (1967: 88–95) and Näsström (2001: 23–24).
Beck’s exposition is probably still the most thorough concerning the terminology of sacrifice in PCRN and is highly recommended.
35 The semantics of this word has been vividly discussed, but, as has been convincingly shown by anatoly liberman, there can hardly be any doubt that it was used within the sacrificial sphere, perhaps particularly in connection with human sacrifices (liberman 1978: 487).
36 The word is only known from four passages in the poems (in Hávamál and Ynglingatal), cf. also Cleasby and Vigfusson (1957: 578) for possible etymologies.
37 lines 5 to 8 of this stanza say: ‘veiztu, hvé biðia scal | veiztu, hvé blóta scal | veiztu, hvé senda scal | veiztu hvé sóa scal’. since the semantic content of these words is not particular clear, it is hard to give a precise translation. seen as different elements in the sacrificial ritual, however, the following would certainly make sense: ‘Do you know how to pray? Do you know how to make the sacrificial ritual? Do you know how to send (the victim to the gods)? Do you know how to slaughter (kill the victim)?’ see also Düwel (1970: 234).
610
Jens Peter Schjødt
should perhaps also be mentioned that the word gefa ‘to give’ is sometimes used synonymously with ‘to sacrifice’, indicating that the victim is thought of as a gift to beings of the other World.
in the following, the account of Nordic sacrifices will be structured according to the constants that we find in gift sacrifices all over the world, that is: a) the sender or sacrificer, b) the sacrificial object, c) the receiver, and d) the return object. furthermore, we shall deal briefly with e) the reason for carrying out sacrifices, f ) the performance of sacrifice, g) the places of sacrifice, and h) the time for sacrifice. The two last mentioned will be dealt with more extensively in (è27–28).
a) The Sender
since the sources, as expected, are primarily occupied with the higher social strata, it is not surprising that by far the largest part of the descriptions we have of sacrifices belong to the category ‘public rituals’; therefore the sender is most often a king or chieftain, predominantly acting on behalf of the people.
But sometimes the ‘people’ act against the king, taking him as the sacrificial victim. This is what we hear in the famous story about Dómaldi, related by snorri in Ynglinga saga ch. 15 and in other texts. Dómaldi, it is said, was sacrificed because there was famine during his reign, and therefore the people made huge sacrifices at Uppsala, first of oxen, then of ordinary people, and finally of the king. Whether the reason for killing the king is that he is seen as the most valuable sacrificial victim, as seems to be snorri’s idea, or the reason is that the people want to get rid of the king, because he cannot fulfill his duty as a king —
namely, to make the land prosper — it is obvious from the sources what is is at stake in these sacrifices is the well-being of the people and the land collectively (è23).
little is known about who actually carried out the sacrifice (the killing, the cutting up, and other ritual acts; è29), but, surprisingly enough, in those examples where we are told about this, women often seem to be the main actors, at least when it comes to human sacrifices. already strabo (64 bce–21
ce) wrote in his Geography (7.2.3) that among the Cimbri, priestesses who were also soothsayers cut the throats of war prisoners over a huge cauldron and prophesied by looking at the blood as well as at the entrails of the victims.
also Tacitus ( Germania ch. 9) mentions that women were held in high esteem among the germani and were seen as having prophetic abilities. another example is the so-called angel of Death, mentioned by ibn fadlan (Montgomery 2000: 18–19), who killed the slave girl who was to follow her dead master to
25 – Various Ways of Communicating
611
figure 25.1. Panel from the gundestrup cauldron, showing an oversized woman putting a man in a huge cauldron, probably depicting a human sacrifice. The gundestrup cauldron, which is dated to the second or first centuries bce, was produced in south-east Europe and depicts Celtic mythology. it was deposited in Himmerland in northern Jylland, however, and can be used to illustrate strabo’s information about priestesses among the Cimbri killing prisoners of war over a huge cauldron. Photo: Roberto fortuna and Kira Ursem, Nationalmuseet, Copenhagen.
the world of the dead. But usually we are not told who actually performed the sacrifice.38 The ‘sacrificers’ (those on whose behalf the sacrifice was carried out) of the public rituals were most often high-ranking men who performed on behalf of the people, or the people themselves; but in neither case we are told who actually carried out the ritual.
in general it is not easy to distinguish between the public and the private cults since, in the icelandic sources and primarily in the Íslendingasögur, what we commonly hear is that farmers, most of them chieftains, build a hof (for examples, see Beck 1967: 22), in and around which sacrifices were carried out.39
38 The role of women in the cult in general is treated below, in (è29), and in sundqvist (2007: 56–78). sundqvist argues that the roles of women in the cult were more differentiated than has usually been acknowledged among scholars and that they were as prominent in the public cult as in the private. This seems convincing and in more accordance with the sources than the traditional dichotomy wherein women have been regarded as ritual specialists in the private cult, with men being in charge of the public cult. Not least the name odindisa (è5; figure è19.2) in a swedish runic inscription (sundqvist 2007: 59–60) indicates that women were, indeed, part of the Óðinn cult.
39 Notice, for instance, the information given in Guta saga ch. 1 where it is stated that on gotland there were three ‘levels’ of sacrifices: one for the whole island, one for each of the
‘thirds’ into which the island was divided, and sacrifices for each of the smaller assemblies,
612
Jens Peter Schjødt
However, we do have a few descriptions which clearly belong to the private sphere: namely, that of Vǫlsi in the so called Vǫlsa þáttr (è31) and a description of a blót for the álfar in snorri’s Óláfs saga helga (è31), and also many magical rituals (è26) no doubt had a private character. We shall return to the question of the ‘sacrificers’ in the following chapters.
b) The Sacrificial Object
We do know a lot about sacrificial victims from both archaeological and textual sources. archaeologically it is often the context which allows us to ascertain whether we have found a sacrificial object or not (è27). Therefore, even if it may sometimes be a little problematic to decide the religious significance of such objects (cf. Capelle 2003: 114), we frequently have no doubt as to whether the remains of some animal or human indicate a sacrificial death. for instance, there can hardly be any doubt that the so-called bog bodies from all over the northern germanic areas, dating from the pre-Roman as well as the Roman iron ages, should be regarded as sacrificial victims, clearly showing us that human sacrifices were part of the cultic world of the germani, to which we shall return presently. apart from humans, we also have huge amounts of weapons deposited in bogs, mainly from the southern parts of Denmark and normally referred to as ‘war booty sacrifices’. The weapons of these finds were violently destroyed, which no doubt connects them to some religious ritual.40 The interesting part is that, even though in certain places we have found war gear from rather sizeable armies (in illerup Ådal in eastern Jylland probably as many as two to three thousand individuals; cf. andrén 2014: 92), the human remains of these large armies have not been found. This may indicate that the weapons have been brought from somewhere else, where the battle took place, so that the sacrificers could deposit them in their own sacred places (andrén 2014: 93–95). in the old Norse sources, we also hear of gold and silver being given to the dead freyr ( Ynglinga saga ch. 10) and to Hǫlgi and his daughter Þorgerðr ( Skáldskaparmál p. 60), both of which should no doubt be seen as a kind of sacrifice. as archaeological evidence from the Early iron age, we can mention the bracteates and other gold deposits from wetlands (Näsman 1994). it is also possible, although far from certain, that the gold foil figures (so-called guldgubber), which have often been found in post holes from the later iron age and at clearly on a much more modest scale. it is quite likely that those in charge of these different
‘levels’ of sacrifices would not be of the same rank.
40 for good overviews, see fabech (1991a, 2009) and lund Hansen (2008).
25 – Various Ways of Communicating
613
figure 25.2. Horse skull pierced by a flint
dagger, from a bog at Ullstorp in skåne.
The skull has been dated by Carbon 14 to
the Viking age, whereas the flint dagger
is from the late Neolithic. This find has
been interpreted as a horse sacrifice
carried out with an ancient object several
thousands of years old (stensköld 2006).
Photo: Eva stensköld.
places where we must assume that large public gatherings have taken place (see, for instance, Munch 1991; Watt 1991; adamsen and others 2009), should be seen as a sort of sacrifice. Therefore, such places may also be assumed to have been sacral places, with rituals of various types taking place at certain times.
according to the archaeological as well as the written sources, animals of various kind were the most common sacrificial victims and most valued among them was the horse, which, right from the time of Tacitus ( Germania ch. 10), seems to have played a special religious role; indeed, so special that the pagans could contemptuously be called horse-eaters by the Christians.
similarly, we know that, when the icelandic alþingi in 999/1000 decided to adopt Christianity, one of the prohibitions concerned the eating of horse meat ( Íslendingabók ch. 7, Brennu-Njáls saga ch. 105). in Hákonar saga góða ch.
14 we are told that a horse constituted part of the sacrificial meal, and also in Hervarar saga ok Heiðreks ch. 16 a horse is sacrificed in connection with the inauguration of a king. although the individual sources may all be doubted with good reason, no scholar is likely to deny that the horse was viewed as a very valuable sacrificial victim in PCRN (cf. DuBois 2012: 71–73). as just mentioned, also in Vǫlsa þáttr the horse was closely linked to the sacrifice, although it is not said that the stallion was actually sacrificed. in connection with burials we likewise see from both archaeological and written sources — such as the ladby ship burial in fyn, Denmark; Vendel and Valsgärde in Uppland, sweden; Borre in Vestfold, Norway (cf. for instance Müller-Wille 1978; Wamers 1995); and many others, as well as in ibn fadlan’s account — that horses played an important role in the burials of leaders. Except for Vǫlsa þáttr, all these examples (and many more could be added) concern kings and chieftains, and it is reasonable
614
Jens Peter Schjødt
to deduce from this that there are aspects of a symbolism reaching far back in time, because we know of horse sacrifices specifically in connection with kings and the inauguration of kings from both india (the ashvamedha sacrifice) and ireland (giraldus Cambrensis, Topographica Hibernica ch. 3.25), clearly suggesting a time span reaching back into indo-European times. in both cases, the horse, a stallion in india and a mare in ireland, is apparently identified with a god, so in that sense Vǫlsa þáttr seems a closer parallel than the horse sacrifices taking place in the kingly halls, since, as we shall see, it may be possible to identify Vǫlsi with freyr or some other god (see è31).
Bulls and cows were sacrificed, too, as it is told in Ynglinga saga ch. 15 and 26, while also saxo (1.8.12) says that black oxen are sacrificed to freyr. it seems that, following the outcome of a single combat, the winner sometimes had to kill a bull ( blótnautr), as is told in Kormáks saga ch. 23 and Egils saga ch. 65.
The term blótnautr certainly indicates sacrificial aspects, but beyond that it is not easy to perceive any sacrificial ritual behind this act since, for instance, there is no information about a possible receiver of the victim (cf. also Beck 1967: 104–05). archaeology, however, seems to confirm the importance of cattle in the sacrificial cult. important sites here are Hofstaðir in iceland (lucas and Mcgovern 2008) and Helgö and frösö Church, both in sweden (Zachrisson 2011a: 82; Magnell and iregren 2010).
sheep and rams are also mentioned as sacrificial victims (e.g., Ljósvetninga saga ch. 4 and the stentoften runic inscription: DR 357, Samnordisk runtextdatabas), as are dogs, mentioned by adam of Bremen (4.27), and poultry, mentioned by Thietmar of Merseburg ( Chronicon 1.9).41 That all these animals could be sacrificed is confirmed by archaeology,42 from which it is moreover evident that such animals as elk and even squirrels were sacrificed, which is not mentioned anywhere in the textual corpus (cf. iregren 1989).
We do not know much about the correspondences between the nature of sacrificial victims and the receivers, nor about the correspondences between the kinds of victims and the occasions on which the sacrifices were carried out. However, it seems that some tentative tendencies can be discerned. for instance, the description by Tacitus ( Germania ch. 9) that Mercury received human sacrifices, whereas Hercules and Mars, most likely to be equated with Þórr and Týr respectively, received animal sacrifices, indicates that such corre-41 Many of these animals are also mentioned in ibn fadlan’s funeral description.
42 The archaeological literature dealing with sacrifices is enormous. for fairly recent viewpoints, see the articles in Jennbert and others (2002), and especially Jennbert (2011: 89–117).
Cf. also (è27).
25 – Various Ways of Communicating
615
spondences — not as fixed rules without exceptions, but as tendencies — actually did exist: not any kind of sacrificial victim could be sacrificed to any god. it even makes good sense to speak of a strong continuity here, since Óðinn (who is probably to be equated with Tacitus’s Mercury) appears to be much more strongly connected to human sacrifices than any of the other gods. Perhaps we can also perceive a long-term convention relating to saxo’s mention that freyr is given black oxen, since we know from other cultures (for instance, greece) that the sacrificial victims presented to gods connected to the chthonic world are black. it could perhaps also be argued that horses were particularly associated with freyr, which would fit nicely into the ‘sacred kingship’ pattern (è23). in general, we should expect such systems to have existed, but they are not easy to discern within the sources, whether textual or archaeological, although the problems arise for different reasons in these two types of sources. Perhaps in the future, DNa analyses will help us here.
somehwat similar to animal sacrifices we encounter ‘toasts’ in honour of various gods (e.g., Hákonar saga góða ch. 14). These toasts may also be seen as a kind of gift for the gods, although certainly not as physical objects. genuine libations, the pouring out of liquid as a gift to the gods, are, however, not mentioned in the sources.43
as has been recognized by many scholars, the human sacrifices were probably valuated highest among the pagans,44 and even if some of the descriptions presented by the authors of antiquity and of the Christian Middle ages can be seen as deliberately putting the germani and the scandinavians in a bad light, there can be no doubt that they did take place from the Early iron age (and even earlier) and right up into the Viking age, as is overwhelmingly supported by archaeological finds (è27). There can be no doubt either that in the pre-Viking period most victims were prisoners of war who, as in strabo, were executed, probably as a gift to the war god Wotan/Mercurius.45 This would, again, be well in accordance with the view that the enemy was dedicated to Óðinn, as we shall return to in (è42). We do not know the identities of the humans that were hanged in Uppsala and lejre, as told by adam of Bremen and Thietmar of 43 adam of Bremen ( Gesta Hammaburgensis 4.27) does, indeed, speak of libationes ( libatio,
‘libation’), but we do not know exactly what is meant.
44 in Guta saga ch. 1, we are told that at the main sacrifices for the whole island, humans were sacrificed, whereas in the more local sacrifices only cattle (and food and drink!) were sacrificed, clearly indicating a relative valuation of the sacrificial victims. for archaeological evidence, see (è27).
45 for an overview of the pre-Viking age evidence, see de Vries (1956–57a: i, 408–09).
616
Jens Peter Schjødt
Merseburg, respectively. They could easily be imagined to be slaves or prisoners, however, albeit rather less likely, they might be nobles.46 it does seem, in fact, that some of the pre-Viking age sacrificial victims, such as the bog bodies, were not ordinary working men, judging by the skin on their hands, although this conclusion is not incontestable (Jensen 2003: 186).
Most conspicuous among the human sacrifices are those of the king as these are related in several sources mentioned already. Thus, Dómaldi, Víkarr, Hadingus and others are all killed in ways that are or can easily be regarded as kinds of sacrifice.47 Víkarr and Hadingus are both hanged and we can therefore safely assume that they are sacrificed to Óðinn, who sacrificed himself in the same way ( Hávamál st. 138–41). it has been discussed to which degree we may also regard victims from the lower classes as gifts to Óðinn or whether they should simply be seen as criminals who are punished.48 The question will probably never be solved, but it may be argued that it is not likely that these two ways of viewing persons killed in public were absolutely distinct among the germani themselves. We should rather expect that the spheres of law and religion could very well be indistinguishable (è20). in Kristni saga, it appears that the pagans would sacrifice two men from each quarter of iceland, but they would ‘blóta enum verstum mǫnnum’ (sacrifice the worst men). However, apparently even sons of kings and perhaps of other people also ( Guta saga ch. 1) could be sacrificed, as we are told in Ynglinga saga ch. 25 where King aun sacrificed nine sons to Óðinn in order to secure a long life for himself, and in Flateyjarbók where Hákon jarl sacrificed his son to Þorgerðr Hǫrðabrúðr ( Flateyjarbók i, 191) to obtain victory.49 We do not know whether such sacrifices ever took place in the real world, which is also the case with the sacrifices of kings. The sources are Christian, and of course they are biased in their descriptions of the pagan world. Therefore, it is probably safer to assume that the idea that kings and their sons could be sacrificed existed already in the pre-Christian period, regardless of whether it was actually carried out or not.
46 We should probably not be overly concerned about the argument that the gods would not be satisfied with people of the lower classes. first and foremost, there are many examples all over the world that prisoners were sacrificed to the gods; and even though most sacrifices have an element of contract built-in ( do-ut-des), the ‘value’ of the sacrificed object and that of the return object is never equal.
47 see Nygaard (2016: 18–21) on the possible category of ‘ruler sacrifice’.
48 some of the more important participants in this discussion are Eugen Mogk (1909), Karl von amira (1922), and folke ström (1942).
49 in a version in the icelandic annals the sacrifice is given to Óðinn (Beck 1967: 97).
25 – Various Ways of Communicating
617
a special kind of human sacrifices are those where whole armies are sacrificed or dedicated to Óðinn. We see this in Hervarar saga ok Heiðreks ch. 13, in Styrbjarnar þáttr Svíakappa pp. 70–73, and in other texts. The model here is undoubtedly Óðinn’s own spear throw mentioned in Vǫluspá st. 24, dedicat-ing the valr, ‘the slain’, presumably to Óðinn himself; perhaps it could even be seen as a parallel to the war booty sacrifices which, as we saw, would comprise the war gear of whole armies. What happened to the defeated warriors who survived the battle we can only guess, but as we saw in strabo ( Geography 7.2.3), who wrote his work at the beginning of the first century ce, war prisoners among the Cimbri had their throats cut, probably as gifts to the war god; likewise Procopius, writing in the sixth century, says that among the peoples living in Thule (probably scandinavia), prisoners of war were sacrificed on an altar, as well as by hanging, to ‘ares’, the war god, who in this case can hardly be any other than Óðinn ( History of the Wars 6.15). furthermore, in the almost contemporary account by Jordanes, the goths are said also to sacrifice their war prisoners to ‘Mars’, another version of the war god Óðinn. Jordanes does not say that the prisoners were hanged, but that this is what happened to the armour of the enemy who were defeated ( Getica 5.40). Despite the many differences, this evidence all seems to point in the same direction and to belong within the same discursive space: namely, that war prisoners as well as their gear were sacrificed to the war god after victory in battles as a sort of thank offering,50 and that hanging was in one way or another involved, clearly connecting it to the discursive space of Óðinn.
Perhaps we should also place the scratching of a spear on a person’s deathbed, as related by snorri ( Ynglinga saga ch. 9), within the same discourse. This tradition, it is said, was instigated by Óðinn and was to be followed by the men who would join him in Valhǫll (è42). The problem here is whether what is described is a genuine sacrifice. as has been argued by Jens Peter schjødt (1993, 2008: 173–205), the distinction between initiations and sacrifices are usually far from clear. What we can state, however, is that in gift sacrifices involving humans as the ‘victims’ as well as in initiations, a person is dedicated to a god to whom he will afterwards ‘belong’. in that sense, those who dedicate themselves to Óðinn on their deathbed, or dedicate enemies to him by throwing the spear before battle, can be seen as performing sacrifices to Óðinn as well as ritually 50 a spectacular finding in alken Enge, Eastern Jylland, has revealed human bones from an army of more than a thousand warriors from around the beginning of our era. although we can only speculate, it makes sense to imagine a sacrifice to the war god performed by the victorious army (Holst and others 2016).
618
Jens Peter Schjødt
creating initiates of Óðinn. and perhaps this is observable in certain graves, for instance, in Birka, into which spears have been stuck so that the dead were, in this way, initiated to Óðinn (Nordberg 2002; Nordberg 2003: 280).51
all in all, it seems as if almost anything could be used as sacrificial objects: different kinds of humans, different kinds of animals, different kinds of material objects. as mentioned, there may well have been some system concerning what to sacrifice at which occasions or to which deity, but our sources are not specific enough for us to draw any certain conclusions regarding this, except perhaps pointing to some general tendencies.
c) The Receiver
it cannot be ruled out that some sacrifices — understood in the sense of some object being transformed, most often destroyed or killed — were not thought of as gifts to a deity, or at least not any definite deity (cf. lund 2009). But in our sources it is most often quite clear that a certain god or some other being from the other World is expected to receive the sacrificial victim. We have already seen that the great gods Óðinn, Þórr, and freyr received sacrifices, but a lot of other gods and goddesses also received various kinds of sacrificial victims, among others the collective groups dísir (è58) and álfar (è63). and even the dead, as can be exemplified with the ‘human’ freyr and with Óláfr guðrøðarson to whom sacrifices were given in order to obtain fertility ( Flateyjarbók ii, 7), could be receivers of such gifts.
We should also mention here the legendary cow síbília, mentioned in Ragnars saga loðbrókar (ch 10–11) as belonging to King Eysteinn, said to be so fierce that the enemy of the king will be bewitched when she lows. apparently, the king sacrifices to her for victory. although the account is certainly not trustworthy as it stands, it might still reflect pagan tradition. and there are other examples of sacrifices to various animals: boars ( Hrólfs saga kraka ch. 42), ravens ( Landnámabók ch. H 5) and perhaps horses ( Vǫlsa þáttr). Whether these animals were ‘just’ animals or representatives of some gods is not clear. it is tempting to see the boar and horse as representing freyr, the ravens as representing Óðinn, and cows perhaps as representing freyja. The textual evidence, however, is not clear about these aspects, and we are therefore not in a position to establish a general pattern concerning such representations.
51 Perhaps this can be seen as a parallel to the graves with rings with ‘Thors hammers’
which probably indicate that the dead was in some way linked to Þórr (g. andersson 2005b).
25 – Various Ways of Communicating
619
We also have examples of special places that can be seen as receivers of sacrifice, such as groves ( lundr, Landnámabók ch. H202), rapids ( fors, Landnámabók ch. H313, s355), and mounds ( haugr, Norges Gamle Love, p. 430).52 again, we must ask whether such places were viewed as receivers of sacrifices in themselves or whether they were regarded as linked to certain gods or other other World beings, and again we cannot give any definitive answers.53 We do, nonetheless, know that Tacitus established a close connection between groves and gods ( Germania ch. 39), saying that a certain god rules in a sacred grove venerated by the semnones who performed human sacrifices there. This would clearly indicate that the grove was considered sacred because it belonged to a certain god. in the subsequent chapter (40) we are told about Nerthus whose wagon is placed within a sacred grove on an island. Here, there can hardly be any doubt that the grove is sacred because it is associated with Nerthus. Therefore, it seems most likely that the sacredness or numinosity of certain places in nature had to do with their relation to some divine powers. That means it is no different from the way churches are perceived by most Christians: the building is sacred because it is the house of god and because a number of sacred rituals take place there. We notice here also the words by adam of Bremen that every single tree in the sacred grove is seen as divine because of the death of the victims ( Gesta Hammaburgensis 4.27); these victims are in some way dedicated to a deity, most probably Óðinn, which is why they and the place where they are hanged are sacred.54
finally, it should be mentioned that what is usually thought of as grave goods may also in some cases be regarded more correctly as sacrificial objects, although it is always difficult to interpret the intentions behind depositing things in graves. This is not even clear in the textual sources. if, as an example, we take the description by ibn fadlan of the animals slaughtered in connection with the burial of the chieftain among the Rûs, which is the most detailed description of a ritual that we have, it is not possible to ascertain whether the weapons, fruits, horses, cows, poultry, dogs, and even the slave girl (Montgomery 2000: 16–17) were, from an emic perspective, seen purely as grave goods or whether they were regarded as sacrificial objects; or, which is probably the most likely interpreta-52 Concerning a ‘mound cult’, see Bø (1978: vi, 246).
53 in general, we can here refer to olrik and Ellekilde (1926–51: i, 334–588). This work is a treasure trove for all kinds of folkloristic evidence and is highly recommended. in this connection, also the placename material, especially concerning theoforic placenames, is of utmost importance for the connection between gods and certain places; see Vikstrand (2001) and (è5).
54 for a list of receivers of sacrifice, see Beck (1967: 111–16).
620
Jens Peter Schjødt
tion, some of them are grave goods whereas others are likely victims (è32 and è33). But a further difficulty is that we cannot even take for granted that the participants of the ritual had clear ideas about these things, just as is the case in many rituals across the world, both in the past and in the present. When asked, very few people are able to interpret the symbols and the ‘theological’ notions surrounding the rituals in which they take part (e.g., staal 1975; lawson and McCauley 1990). These problems, which probably cannot be solved empirically but only theoretically, must of course be taken into consideration in dealing with sacrifices from an archaeological as well as a textual perspective.
d) The Return Object
Return objects can be anything wished for among humans. as we see in Hákonar saga góða ch. 14, the toasts are directed towards various gods and for various purposes (è31). Most interesting is the toast for Óðinn with the purpose of securing victory for the king and those for freyr and Njǫrðr for ár ok friðr, ‘good year55 and peace’. although some scholars (e.g., lange 1958: 119; Düwel 1985: 66–69; von see 1988: 84–87) have argued that this formula is not pagan but has snuck in due to Christian influences ( pax et prosperitas), there can hardly be any doubt that at least the semantic content of the formula is genuinely pagan56 and constitutes a substantial argument for the association between the vanir gods and fertility. at any rate, as was mentioned above, it is 55 it has been argued by Åke V. ström (1975: 233–34) that the meaning of ár should more likely be understood as the coming period and thus translated simply as ‘year’. Using comparative evidence he shows that, in many cultures, rituals must be performed in order to secure the continuation of the world. This may be so, but the close link to the vanir seems to indicate that the fertility of the coming year is certainly also at stake. further, the stentoften runic inscription (DR 357, Samnordisk runtextdatabas), referred to above, seems to confirm this semantic content.
56 Hultgård (1993: 242–53) has discussed the formula extensively and concludes that it should be seen as genuinly pagan (‘als echte vorchristliche formeln aufzufassen’, p. 253).
although Hultgård is convincing in his argument, it should also be stated that regardless of whether or not the formula as such is pagan, the content certainly is: it is impossible to imagine that, for a culture like that of pre-Christian scandinavian, good year (fertility for beast and soil) and peace should not be of utmost importance. The notion of friðr should of course not be seen as an abstract or mental state, but is to be taken quite literally. War and plunder must be avoided in ‘our’ land, whereas there are probably no ideas about universal ‘peace’ involved.
War as well as bad harvest is disastrous for the well-being of any agricultural society, but that is important only when ‘we’ are the victims. Plundering elsewhere may be of great benefit to ‘us’, so ‘peace’ is to be interpreted as ‘our peace’.
25 – Various Ways of Communicating
621
sometimes difficult to establish any rules concerning the connection between certain sacrificial victims and certain receivers, except for some general tendencies (human sacrifices to Óðinn, boars and horses perhaps to freyr); however, when it comes to the relation between the receivers and the return objects, we are somewhat better equipped. This is due to the simple fact that the different gods are particularly connected to certain areas of life; they have certain functions, which may well overlap, but are nevertheless important for defining the
‘semantic centres’ (schjødt 2013) of the individual gods. There is, therefore, no point in sacrificing to a certain god for a purpose on which he or she has no influence.
We notice, for example, that Óðinn is usually (there will always be exceptions to any rule) the receiver of sacrifices that have to do with war and kings, whereas there is a clear tendency that the vanir are asociated with fertility and prosperity of various sorts.57 Þórr is harder to relate to a specific functional area, since he appears to have a protective function in general, and it seems as if he receives sacrifices for a great variety of reasons (è41).58 for most of the other gods, we cannot tell with any certainty what they are expected to give back to those who sacrifice to them.
it is even much more problematic when we come to the private cult. as we have already seen, we have very few descriptions of sacrifices carried out within that sphere, and there are no clear indications of what is actually wished for in any of the instances. This being said, however, it should be noted that there can hardly be any doubt as to what were the main concerns among farmers: fertility, good health, and good luck in general.
e) The Reason for Carrying Out Sacrifices
This point is related to d) above insofar as the reason for making a sacrifice is that we want a certain return object, be it victory in battle, good harvest, or something else. The subject will be taken up again in (è30–32) on the individual categories of ritual. We shall therefore deal only briefly with the issue here. in general, we can distinguish between two types of reasons for carry-57 again, we shall refer to the stentoften runic inscription from Blekinge (DR 357, Samnordisk runtextdatabas), in which it is said that a certain HaþuwolAfR gave good year with a sacrifice consisting of nine billy goats and nine stallions. This sacrifice was probably directed to the vanir or other figures related to fertility.
58 for an analysis of the ‘functions’ of and the relation between Óðinn, Þórr, and freyr, see schjødt (2012b).
622
Jens Peter Schjødt
ing out sacrifices — and, indeed, all rituals. There are those which are planned beforehand and can be foreseen for years to come. above we have called them
‘cyclical (calendrical) rituals’. The reason for carrying out this type of sacrifice is that if we do not, all kinds of bad luck may strike us. Perhaps we could even say that, by performing such rituals, we create the summer, the harvest, the ‘year’.
also in connection with passage rituals we sometimes have sacrifices relevant to the future life of the initiates. The other type consists of those carried out in order to cope with a dangerous and unexpected situation, such as illness, war, or bad weather. so where we could say that the first type is concerned with the future, the second type is mainly concerned with critical situations in the present — hence ‘crisis rituals’. as was mentioned above, sacrifice can be seen as a
‘rite’, that is, a part (although often the most important part) of a larger ritual complex, and therefore the reason for the sacrifice cannot be dealt with outside the broader framework of the ritual in question.
f) The Performance of Sacrifices
as mentioned in the introduction to the Nordic sacrifices above, we do not have any descriptions that give us the ‘whole’ picture. Even in the most detailed description that we have, namely that of ibn fadlan, we are not informed about everything that went on. and as was suggested, we do not even know whether the killings in this ritual should be regarded as sacrifices proper (involving receivers and return objects) or perhaps rather as grave goods for the benefit of the dead chieftain.
as mentioned already, sacrifices may be part of different rituals with different aims, and we can be sure that the way the sacrifices were performed was also different. apart from the three saga descriptions from the West Norse area dealing with cyclical rituals, which were mentioned above and to which we shall return in (è31), we have glimpses of other kinds of sacrifices performed in other contexts.
Thus, elements from sacrifices of another kind are sometimes mentioned, although we cannot in these instances see the ritual pattern clearly. for instance, it is related in Kjalnesinga saga that men are sacrificed in a kind of pool. Nothing is said about how they were killed, whether by weapon or by drowning, or perhaps both, but we know from other sources (cf. Gesta Hammaburgensis 4.26) that drowning could be part of divinatory rituals. and of course we cannot escape the supposition that there could be a connection between the bog bodies and other people who were drowned, although this is very uncertain.59
59 it must of course be taken into consideration that all the bog bodies were killed before
25 – Various Ways of Communicating
623
figure 25.3. The bog body from Tollund in central Jylland, dated to about 375–210 bce. The man, who was about forty years old, was killed by hanging and then deposited (‘drowned’) in a bog.
Museum silkeborg. Photo: anders andrén.
anyway, it has been proposed that an idea existed, stemming back from indo-European times, about a so-called ‘threefold death’ (Ward 1970; Näsström 1997; Näsström 2001: 47–55), a sacrifice in which the victim or victims were killed through hanging (or strangulation), stabbing with a knife or a sword, and drowning, with these three methods corresponding to the three functions postulated by georges Dumézil (è11). Exactly this can be observed on some of the bog bodies of the Early iron age, although the interpretations must necessarily remain very uncertain, even if we can be relatively sure that hanging had a special connection to Óðinn. However, it appears that such hanging sacrifices were given to the god more in his capacity as a war god than as a magician (the
‘magical’ aspect of the first function). There is no real evidence for a particular connection between freyr and drowning or Þórr and cutting, and the relation between Dumézil’s three functional scheme and the ‘threefold death’ must therefore remain hypothetical.
space does not allow us to treat all the various forms of sacrificial killings documented in the sources. But one, the hanging, is outstanding and deserves a brief treatment here, although it will be dealt with in other chapters, too, although we have already mentioned some of the classical and early medieval sources referring to hanging, stating that the main receipient of these hanged victims was Óðinn. and hanging, sometimes in combination with the thrust of a spear or some other weapon, seems to occupy a special position in our sources.
We even have a couple of mythic-legendary descriptions of a ritual sequence they were thrown into the bogs. This could well be the case also in the description in Kjalnesinga saga, but hardly in adam’s account.
624
Jens Peter Schjødt
wherein hanging is the central rite, namely, starkaðr’s famous killing of Víkarr, which we shall deal with in (è36) and (è42): in the latter we shall also deal with the god’s own ‘self-hanging’. However, adam of Bremen’s description of the sacrifices in old Uppsala ( Gesta Hammaburgensis 4.27) that also includes hanging does not afford us the opportunity to reconstruct anything like a ritual sequence.60 What we are told concerning the Uppsala sacrifice is that the people sacrifice nine individuals ( capita) of all living males (horses, dogs, humans, and no doubt other species), and through their blood the people attempt to expiate themselves in the eyes of the gods. The bodies, however, are hung in a sacred grove close to the temple. We are not told in any detail about the procedures leading up to the killings and the hangings, and we cannot know, therefore, what the victims actually died from. Most likely, however, they were killed by stabbing or cutting in some way and thereafter hanged in the grove. according to Tacitus, nearly a thousand years earlier, only Mercurius was worshipped with human sacrifices, and, if we accept continuity over so extensive a period, it would indicate that Óðinn was the receiver of these men, which is in turn supported by the fact that they were hanged, exactly like the god himself.61 But adam’s description is not very clear, and it would probably be most realistic to assume that the rituals he was informed about were concerned with all three gods that he mentions, and maybe other deities as well, and that the description includes rites or parts of rites taken from various contexts.
in general we can state that, according to the sources, sacrifice could be carried out in multiple ways: on the one hand, there were sacrifices for all kind of purposes, to different gods and with different participants; and on the other hand, even the ‘same’ sacrifice, for instance, that carried out for the victory of the king, would vary over time and from place to place. The archaeological evidence seems to confirm this assumption: it is hardly possible to find exact similarities, at least when it comes to the finer detail, among even two sacrificial sites (cf. Bemmann and Hahne 1992 concerning mostly weapon deposits).
There may well have been rather close parallels on the ideological level — the ideas lying behind the acts — but the acts themselves would, we must suppose, vary immensely according to natural and political conditions.
60 for a good overview and many interesting analyses, see Hultgård (1997). Chapters 26
and 27 in adam of Bremen’s Book 4 are quoted with translations in (è31).
61 as mentioned above, in strabo’s Geography (7.2.3) for instance, it is related how war prisoners among the Cimbri had their throats cut by priestesses in order to use their blood for prophesying, but we are not told what happened to their bodies afterwards.
25 – Various Ways of Communicating
625
g) The Sacrificial Place
Where did the rituals take place? This question will also be the subject of (è27), so we shall keep it brief here.
already Tacitus mentions ( Germania ch. 9) that the germani did not have
‘temples’ because this was not in accordance with the greatness of their gods.
instead, they would consecrate groves and forests to the gods (cf. Vikstrand 2001: 278–88). in Annals 1.51 Tacitus himself, however, mentions a templum, which in the context must be a building, and archaeological evidence also points in the direction that some buildings from the Early iron age could be ‘cult houses’ of some sort (a.-s. gräslund 2008a: 250–53; see also larsson and lenntorp 2005 and larsson 2006a about the excavations at Uppåkra).
Nevertheless, it may well be true that most of the sacrificial cult took place outside, if only for the simple reason that no houses at that time would probably have room for the killing of many animals, some of them quite large.
Thus, there can hardly be any doubt that natural spaces as well as human constructions were used as sacrificial places. although there has been much debate about where the sacrifices took place, it seems as if much of this debate has been based on false premises. first, it is very likely that there was great diversity, historically as well as geographically and according to their function, which means that sacrifices were not carried out in the same manner and in the same sort of surroundings all through the iron age62 and all across the scandinavian area, not to mention the variations that we should expect with regard to the different purposes of rituals (andrén 2002: 313–18). sacrifices related to fertility cult may sometimes have taken place in the fields, whereas sacrifices linked to war could have taken place in or near the kingly halls. second, the individual sacrificial ritual could well have taken place both inside and outside a cultic house. for instance, it would seem appropriate, as mentioned, that the animals or humans were killed outside, whereas the sacrificial meals were consumed inside. a priori 62 as mentioned above, it has been proposed by Charlotte fabech (1991a) and other archaeologists that the cult places moved around the middle of the first millennium from wetlands to kingly halls, although this distinction should not be taken too literally (cf. fabech 2006: 30; for critique of fabech’s original proposal, see Zachrisson 1998: 118 and Hedeager 1999). another question, which has been debated for a very long time, is the question of cult continuity from the pagan into the Christian era. The main proponent for the rejection of this idea is olaf olsen (1966), but since the 1960s so much archaeological evidence has been found to support that, at least in some places, churches were built at places where pagan rituals had been carried out. for an overview of this debate including more recent evidence, see andrén (2002) and Jørgensen (2009).
626
Jens Peter Schjødt
it would be easy to imagine, and this gains some support from placename material (Brink 1996a, 1997; Vikstrand 2001), that for some rituals, perhaps also sacrificial rituals, movements would be involved from one site to another, from the hall to the natural surroundings (gunnell 2006; Zachrisson 2004b) (è31).
None of our sources informs us in a sufficiently detailed manner about these things for us to obtain certainty, but it is necessary briefly to discuss some of the terms designating arenas for cultic activity.
Vé is related to the verb vígja, meaning ‘to hallow’. Vé, then, should be seen as a hallowed place in a very general sense (Vikstrand 2010a: 58–59).63 it can therefore designate buildings as well as other kinds of places that are sacred, in some but not all cases because sacrifices were carried out there. Vé is thus, as stated by inge Beck (1967: 36), rather the quality of a sacred place in opposition to the ‘normal’, profane space surrounding it. The term does not reveal anything about the ‘architecture’ of sacred places (Vikstrand 2001: 330), and it designates places that can be found everywhere, even among the gods (Beck 1967: 33–36; de Vries 1956–57a: i, 341–43).64
another word often used in connection with sacrifices is hǫrgr, the precise meaning of which is uncertain. on the one hand, it has been discussed whether the hǫrgr was made of wood, as seems to be indicated in Vǫluspá st.
7, in which it says that the gods constructed hǫrgr and hof 65** **out of timber.
on the other hand, the very etymology of the word would indicate that it was made out of stone, perhaps designating a pile of stones somehow used in connection with sacrifices. from a religio-historical perspective the question is not of great importance, and a priori it would seem likely that this sort of construction could be made of various materials according to the actual place where it was raised.66 Much more interesting is the function of the hǫrgr. Most scholars have accepted that it was a kind of altar,67 an interpretation based inter alia on stanza 10 of Hyndluljóð where it is said that Óttarr, the protégé of freyja, has 63 Vígja is thus derived from the root vé. for its etymology and various meanings, see de Vries (1962a: 648–49); Vikstrand (2001: 298–366), and (è5).
64 Vé is, moreover, a part of several personal names (Beck 1967: 77–87) and could indicate that these persons were in some way associated with sacred constructions.
65 These two words often concur (Beck 1967: 3–29), probably both because of the alliterative qualities and because of their semantic content: both designate some sort of sacred building.
66 Jan de Vries tried to solve the problem by suggesting that the hǫrgr was made of stone, but had a wooden roof on top (de Vries 1956–57a: i, 379).
67 others, however, would interpret it as more generally designating simply a cultic place (cf. Näsström 2001: 86).
25 – Various Ways of Communicating
627
figure 25.4. Picture stone from stora Hammars
in lärbro on gotland, dated to the ninth or tenth
century. The third panel from the top depicts a
hanging from two trees and a sacrifice at an altar.
Bungemuseet, fårösund. Photo: anders andrén.
reddened the hǫrgr with sacrifi-
cial blood. if this is correct, we
can easily imagine that in at least
some sacrifices, but hardly all,
the sacrificial victim was killed
on such an altar. Whether or
not the hǫrgr was placed in the
open air or in ‘cult houses’ can-
not be decided. a construction
that could easily be interpreted as
a sacrificial ‘altar’ can be seen on
the gotlandic picture stone from
lärbro, stora Hammars.
The hof, on the other hand,
points to a sizeable building and
was undoubtedly an important
part of public rituals, as is related
in the three sagas referred to above ( Eyrbyggja, Kjalnesinga, and Hákonar saga góða) and elsewhere. it appears that several rites took place inside the hof, but, as mentioned, it is hard to imagine that large animals were killed inside, and we must therefore assume that the sacrificial blood was brought in for sprinkling after the killing. it also appears that statues of divine beings were placed in the hof and that the sacrificial meals were eaten here. it is thus most likely that what is called a templum, for instance, by adam of Bremen, is in reality the same as the hof. Exactly how these buildings were used outside the ritual periods, we do not know, but it is quite likely that they functioned as the manor houses of kings and chieftains.68
other words that probably designate buildings related to sacrifices, perhaps used synonymously with hǫrgr or hof, are blótstaðr (sacrificial place), 68 another word, which may be synonymous, too, is sal/ salr; see Vikstrand (2013b) with many references and Brink (1996a).
628
Jens Peter Schjødt
blóthús (sacrificial building), and goðahús (house of gods) (Beck 1967: 31–32).
Helgistaðr (sacred place), however, should most likely be understood as a sacred place in general and does not have to be a building. in Gylfaginning (p. 17) it is said that the helgistaðr of the gods is by yggdrasill where the gods hold council every day. Therefore, we interpret it as synonymous with vé, a place that is sacred whether or not there is a building there.
apart from these buildings, which we hear of almost exclusively in connection with chieftains or kings and thus as part of the public sacrificial rituals, we also hear of rituals carried out in private houses. in these descriptions, however, we rarely hear about sacrifices of animals or humans, as we shall see in Vǫlsa þáttr (è31). Concerning the álfablót related in Óláfs saga helga, it appears to take place inside a private house, but nothing at all is said about what was sacrificed and — if killings were part of it — where these killings took place (è31).
somewhere in between the sacrificial places created by humans and natural places we have the mounds ( haugar). These were associated primarily with the dead ancestors and can thus be seen as clear evidence of a cult of these ancestors (de Vries, 1956–57a: i, 345–47).
from archaeological evidence in particular, we know that many sacrifices were performed in sacred places in nature where there were not necessarily any buildings, but it is difficult to determine from archaeological remains whether the whole ritual was carried out at the same place. as mentioned, it seems likely that part of the ritual took place at one place and other parts at others: for instance, killings and meals do not have to occur in the same place, as has been suggested for example for Helgö in lake Mälaren, central sweden (Zachrisson 2004b). Therefore, the answer to the question of whether sacrifices were carried out on spots that were constructed by humans or on natural sites appears never to have been one of either/or but rather of both/and. The sources, however, do not have much to say about these matters.
Thus, sacrifices could very likely be carried out by rivers, in groves, and at other sacred places in the natural surroundings where certain other World beings — receipients of the sacrifices — were believed to dwell, probably mostly landvættir. sacrificial victims to rivers were likely drowned, and we hear, for instance in Landnámabók H313, that the remnants from sacrifices should be thrown into the river.
h) The Sacrificial Time
also this point will be treated more extensively in (è28), and we shall merely repeat here what has been stated before: namely, that a blót could be part of
25 – Various Ways of Communicating
629
each of the three categories of rituals mentioned above, that is, rituals of passage, cyclical rituals, and crisis rituals. Except for the cyclical rituals it is not possible to tell exactly when sacrifices took place. as we also saw above, sacrifices were carried out in connection with war, both before and after as offerings of thanks, in connection with the failing of crops, and in connection with all other kinds of crises. Concerning rites of passage, and especially in relation to funeral rituals, the textual sources often tell us about sacrifices, and the archaeological evidence seems to confirm this, although we can, as mentioned, never be sure what ideas lie behind ‘grave gifts’ — whether they should be seen as simply
‘gifts’ meant to ease the passage of the dead person into the other World, or whether they are there in order to ‘bribe’ the dead or other beings of the other World. The distinction is hard to discern, also on a theoretical level, since ‘gifts’
may well be understood as attempts to prevent the deceased from harming the living. in any case, objects given to the other World have the character of sacrifices, even if we are not informed about any return object.
as for the time of performance of both crisis rituals and rituals of passage, whether in private or public spheres, they depend on the circumstances, whereas the cyclical rituals must naturally take place at the same time every year, or with some other type of regularity: for example, adam of Bremen and Thietmar of Merseburg refer to every ninth year when speaking of big public rituals at the central places in Uppsala and lejre.
as we shall see in (è28), the sources do not agree on how many public rituals that can be seen as pan-scandinavian existed. snorri has the following statement in Ynglinga saga ch. 8: ‘Þá skyldi blóta í móti vetri til árs, en at miðjum vetri blóta til gróðrar, it þriðja at sumri, þat var sigrblót’ (a sacrifice was to be made for a good season at the beginning of winter, and one in midwinter for good crops, and a third one in summer for victory). These three public blót feasts were held in the autumn, in December or January, and probably in the spring to mark the beginning of the war season. There may also have been a miðsumarblót, although it is only mentioned by snorri ( Óláfs saga Tryggvasonar ch. 65) and at gói (middle of february to middle of March), the last mentioned perhaps identical to the spring sacrifice (è28; Nordberg 2006a). from the sources, it seems as if the two first feasts, in autumn and midwinter, were primarily concerned with fertility, and thus the vanir and other groups of fertility deities were in focus, whereas the rituals at the opening of the war season had Óðinn as their focus — just as we should expect (for more textual evidence, see Beck 1967: 48–57; cf. also sundqvist 2002: 188–91). However, it also appears that the great blót, described in Hákonar saga góða, was dedicated to both fertility and war gods. Thus, it is hardly possible to find a standard system behind the
630
Jens Peter Schjødt
descriptions in the sagas, although it seems clear that there are some tendencies towards a systematization in regard of the relations between time, purpose, and the receivers of the sacrifices.
Procession
another important issue within the ritual sphere is ‘procession’. although processions are definitely not as common and, from an emic point of view, are probably not as central to the religious life as are sacrifices, they often play an important role in public rituals. They can, of course, take many different forms, but usually their main function, from a phenomenological perspective, is to display some deity. a statue or some other symbol of the god is at certain liminal occasions taken out of the cult house, the temple, or the sacred grove where it is normally placed and displayed to the people. The etic function seems quite clear: namely, confirmation of the communality and strengthening of the solidarity of the people participating in the display of this divine symbol. although the textual evidence from scandinavia does not reveal much about such processions, the Nerthus episode from Tacitus, mentioned in the beginning of this chapter, is surely one example. We shall discuss other possible examples in the following, both from textual and archaeological sources. from an emic point of view, the major function is probably hierophany, the display of the sacred for the public, creating an experience of the numinous.
in general it can be stated that within the comparative study of religion processions have not occupied a prominent position in scholarly works on rituals, perhaps because they have usually been interpreted as merely preparatory acts for the ‘real’ thing, which could be a sacrifice, a ritual death and resurrection, or a sermon.69 as mentioned, however, the appearance of the god in front of the people is certainly one of the more important ways a religion, not least in preliterate societies, is remembered and through which it creates social solidarity.
Processions are, of course, more prominent within public rituals than in the private sphere and, even if they can be performed in all the three main categories of rituals, they appear mostly in cyclical rituals. However, we can safely assume that processions were also part of the rituals of passage concerning coronations of kings and funerals and probably others. This is, for instance, indicated by the so-called Eriksgata, a kind of procession which according to the swedish 69 archaeology, however, suggests that processions have taken place, at least since the Bronze age: see, for example, Rudebeck (2002: 189) and Kaul (2004: 173–78) concerning the Kivik grave and its petroglyphs).
25 – Various Ways of Communicating
631
figure 25.5. a procession of men, women, and carriages on one of the tapestries from oseberg, dated to the early ninth century (Kulturhistorisk museum, oslo no. C 55000_337_1 and C 55000_337_2). Drawing by Mary storm.
Photo: Mårten Teigen, Kulturhistorisk museum, Universitetet i oslo, oslo.
medieval laws had to be performed by the new king. although it cannot be proven that this tradition also existed in pre-Christian times, comparative evidence reveals that this is clearly what we should expect. and processions were no doubt also part of the rituals surrounding funerals, even though they are hardly mentioned in the written sources. However, there are a few hints which may point in the direction of funeral processions, such as the description in Beowulf of scyld’s funeral (28–51) or that of the arrival of the guests for Baldr’s funeral as related in Gylfaginning pp. 46–47 and on the pictures described in Húsdrápa, although it is not stated explicitly that we are dealing with ritualized behaviour in either case. The same goes for the pictures on the tapestry from the oseberg grave; it is hard to interpret them as anything but depictions of some sort of procession.
The Nerthus celebration, which clearly belongs to the cyclical rituals, is described by Tacitus in Germania ch. 40. Here, we are told that a group of tribes in the northern part of the germanic area celebrate Nerthus, identified by Tacitus with Mother Earth ( Terra Mater), who is transported around the entire area on a consecrated wagon drawn by cows and that she is accompanied by her priest. The period during which this lasts is liminal, as we saw earlier, so that people do not initiate battles and they do not carry weapons.
632
Jens Peter Schjødt
figure 25.6. Plan of a ritual road at Rösaring in
låssa in Uppland, dated to the Viking age. The
road, which is 540 m long, is aligned on the east
side with about 105 postholes. it connects a small
house in the north with a grave-mound in the
south. Map: gerhard Winberg and lars löthman.
after this period, the goddess is led by
the priest back to her sacred grove, situ-
ated on an island in the ocean, where she
and her wagon are washed and the slaves
who have carried out this task are subse-
quently drowned. There can be no doubt
that this circumambulation must be con-
sidered a religious procession. This pas-
sage contains a lot of information which
has been discussed by many scholars
dealing with PCRN. Here, however, we
shall note only the processional aspects,
clearly evidencing the existence of pro-
cessions among the germani and help-
ing us interpret certain archaeological
finds, the most prominent among them
being the Dejbjerg wagon (è8), found
in western Jylland and stemming from
the Early Roman iron age (cf. Jensen
2003: 195–202). But other wagons have
been found which may well be inter-
preted as ‘cult wagons’, especially since
their axels were not able to pivot: for
instance, in the famous oseberg grave
(Christensen and others 1992: 119–20).
all in all there is no reason to doubt that
the description by Tacitus describes pro-
cessions that actually took place. good
arguments have been raised — due to
various parallels with this Nerthus cult
— that also the þáttr about gunnarr
helmingr from Flateyjarbók is based on
ritual processions; in the þáttr a statue
25 – Various Ways of Communicating
633
of freyr is accompanied by a priestess as this couple drive around a specific area in sweden. it may also be that the enigmatic story of lytir ( Flateyjarbók i, 579–80) can be connected to such processions (cf. sundqvist 2002: 233–
35).70 if so, that is particularly interesting in the present context, since in the story of lytir the swedish king Erik performs sacrifices in order to make the god appear so that he can ask the god about the future. That would make this an example of the combination of three elements: procession, sacrifice, and divination. finally, we should also mention the description by saxo ( Gesta Danorum 5.16.3) of frotho’s death, in which it is said that his men wanted to keep the death of the king secret in order to prevent the country from fall-ing apart. The men therefore drove him around the country in a ‘royal’ wagon ( regale vehiculum) for three years before they buried him. if we accept a close relationship among, perhaps even an identity of, Nerthus, freyr, and frotho (cf. schjødt 2009b), it appears that these three descriptions are all part of a discourse connecting gods of the vanir type with circumambulations and thus with processions focusing on yearly rituals.
apart from the wagons, which we may well assume to have been parts of processions, archaeology may also help us in locating ‘ceremonial roads’. of course, it is seldom easy to ascertain with what purpose a road was constructed and thus whether it had practical functions or might have been built for purely ritual purposes — or perhaps both. There are, however, archaeological indications that roads consisting of a roadbank should be seen as ceremonial, and, apart from the theoretical probability that processions took place during big, public rituals, what looks like such roads have actually been found — namely, in gamla Uppsala — right next to the mounds of the most famous cult site in scandinavia.71 These roads can hardly be interpreted as anything other than procession roads, probably used during the large gatherings described by adam of Bremen.72 The eastern posthole alignment is about 900 m long and leads from the river samnan towards the royal manor. The postholes are placed 6 m 70 sundqvist apparently accepts that lytir is identical to freyr (2002: 233), which would certainly fit nicely with the whole setup. in this connection it is moreover interesting that processions were part of the cult of st Erik, which is seen by sundqvist as a sort of relict from pagan times (2002: 354–56).
71 for a recent discussion of the role of gamla Uppsala in the pre-Christian North, see Brink (2017) and sundqvist (2018b).
72 The 2012–14 excavations at old Uppsala have not yet been published, but descriptions of the whole area at earlier stages can be found in Duczko (1993–96) and sundqvist and Vikstrand (2013). The monumental posthole alignments that were discovered in 2013 were sensational.
634
Jens Peter Schjødt
apart. They have been interpreted as free-standing posts, and the solid alignments and diametres of the posts of 0.3–0.6 m indicate that they must have reached high up into the air. another alignment to the south leads up to the burial grounds with the huge kings’ mounds and continues east of it. in the postholes there were depositions of jaws and teeth from horses. The bones have been Carbon-14 dated to the late sixth and early seventh centuries, and thus the period of monumentalization of old Uppsala (Beronius Jörpeland and others 2013). a road which must likewise be understood as a procession road has been found in Rösaring in sweden, stretching 540 m and 3.5 m wide, bordered on one side with 125 postholes and leading to a cemetary and for that reason probably used in connection with burial rituals (cf. Damell 1985).
Thus, even if the written sources are meager and the archaeological evidence is not entirely unequivocal when it comes to interpretations, there nevertheless appears to be enough material for us to be able to confirm that processions, known from almost all other comparable cultures, were part also of the rituals within PCRN. We know very little about the details, and, apart from the few sources mentioned above, these will have to be reconstructed from comparative evidence.73
Divination
Divinatory rites seem to be most prominent within the category of crisis rituals, and as such they, like sacrifices and processions, are usually parts of larger ritual settings, which may include these three kinds of rites and many more.
as opposed to the phenomenon of processions, which, as we just saw, is very weakly represented in the written sources, various kinds of divination play a significant role in a variety of textual genres. The saga material mentions a huge number of instances of such practices, no doubt reflecting the importance of divination in the everyday life of the pagan scandinavians.74 The whole idea that it is possible to come to know about the future through various techniques appears to be universal in tribal and folk religious settings of all times, whereas the techniques themselves, however, vary immensely from culture to culture.
Without pushing the religious logic too far, it seems to be a prerequisite for divinatory rites that the future is in some way already decided, that fate is una-73 Very litle has been written on processions in PCRN. The only systematic account is a recent article by simon Nygaard and luke John Murphy (Nygaard and Murphy 2017).
74 a very useful work which mentions all instances of divination in the sagas of icelanders and in Landnámabók is Dillmann (2006); this work discusses various kinds of divination and provides many references to both primary and secondary sources. see especially pp. 29–52.
25 – Various Ways of Communicating
635
voidable, so to speak (è35), and that the rites will allow the privileged person access to knowledge about the future. simultaneously, the germani and the scandinavians as well as all other cultures were actively trying to influence fate.
This may seem a paradox, but it is a paradox that we cannot deny — it is known from belief systems all over the world.
apart from all the other problems pertaining to our sources, there is a particular problem in relation to divination which makes it difficult to judge the pagan attitude towards divination: namely, the fact that omens and other hints about the future are predominanly related in texts composed long after the relevant incidents have taken place. in the literary sources, divinatory features of various kinds, mostly dreams, commonly have the appearance of a literary means by which the author creates excitement for the reader. What usually happens is that a diviner or an interpreter of dreams has foreseen the outcome of a battle or the death of an individual and that it eventually turns out to be true; the emphasis is thus on the result and only rarely on the actual ritual carried out. Nevertheless, we do get a rather clear impression of the variety of divinatory techniques that could be brought into play when important decisions had to be taken, often in connection with some looming or manifest crisis.
We can hardly expect a clear systematization of the various techniques, such as for instance that certain types of divinatory practices were used in connection with certain crises or that males or females were in charge in this or that sort of divination technique. Even so, there do seem to be some general trends.
Before enumerating the various techniques, we shall briefly attempt to create a taxonomy on a more abstract level.
in general, all religious societies have sought knowledge about the future by means of interpreting particular signs as information from the other World.
although all more or less ordinary events — whether they occur in dreams, in the natural phenomena of the surrounding world, or in ritual settings — can be interpreted as divinatory signs, these usually fall into two basic categories: on the one hand, those that appear spontaneously and have not been deliberately sought; and on the other hand, those that have been produced through some magical ritual (è26). in both cases, however, the signs must be interpreted.
some interpretations are culturally determined and are understood by everybody, as when adam of Bremen (4.26, schol. 138) says that during the great ritual feasts in Uppsala, humans were thrown into a well or a stream ( fons), and, if they were not seen again, the wish of the people would come true.75
75 This also shows that divination rites were carried out as part of larger rituals and could be linked to sacrifices in particular (cf. Derolez 1968: 296–97; cf. above about lytir).
636
Jens Peter Schjødt
other incidents require the interpretation of some specialist, as is often the case with dreams. of course many such prophesies should be seen mainly as literary means of creating suspense in our texts, as we just saw, but there is no doubt that they did, in fact, take place in various forms. Thus, the signs of both spontaneous and provoked forms of divination could be culturally determined or would need ‘specialist’ interpretation. another taxonomic feature is the dichotomy between those signs which can be seen, although perhaps not interpreted, by everybody and those that require the diviner to enter a certain state of mind, such as ecstasy, to be able to predict the future happenings. The latter was the case with seiðr rituals and probably also the technique of útiseta, literally ‘sitting out’ (è26), meaning that the diviner or the magician would place herself (it was practised mostly by women, often a vǫlva; cf. Tolley 2009a: 136) somewhere outside and in isolation during the night in order to get inspiration from the spirits.76 People able to reach a certain stage of ecstasy can often, but not always, be classified as belonging to the group of ‘religious specialists’ (è29).
all these forms of divination exist all over the world, whereas the particular techniques are of course different from one culture to the next.77
another general problem with divinations is the question of whether the diviner interprets the signs in an ‘objective’ way or whether he or she actually has the power to manipulate the future. This problem is a variant of the above-mentioned paradox in which fate is on the one hand inescapable, but can on the other hand be manipulated in many ways. in PCRN, it seems as if the attitude of the diviner was of some importance for the way things happened in the future. in Gunnlaugs saga ormstungu (ch. 2), for instance, the interpretation of Þorsteinn’s dream is seen as ‘unfriendly’, as if the interpreter is contributing to the way things will turn out.78 This paradox appears to be present in all sorts 76 However, in Ynglinga saga ch. 7 we are told that Óðinn would sit under the hanged, no doubt in order that they should supply him with knowledge from other worlds, and this may be taken to indicate that also men could perform some sort of útiseta. We also have the term sitja á haugi, ‘to sit on a mound’, indicating that the dead possessed numinous information about the future. although we cannot know for sure, útiseta might be a kind of necromancy. apparently, it was also possible to learn skaldic poetry by sitting under running water or in rapids, as it is evident in Bjárni Kolbeinsson’s Jómsvíkingadrápa.
77 other useful attempts at classifying various divinatory rituals have been carried out by Dillmann (2006: 29–52), Derolez (1968: 274–302), and de Vries (1956–57a: i, 428–38).
Derolez also has a complete list of divinatory techniques to which we shall refer here.
78 The Björketorp inscription (DR 360, Samnordisk runtextdatabas), on a runic stone in Blekinge, probably contains the word úþarba-spá, which may be interpreted as ‘harm-prophecy’ (see Dillmann 2006: 29), indicating that the prophecy in itself has an aspect of bad will from the diviner.
25 – Various Ways of Communicating
637
of divinations: on the one hand, the future is part of the fate of individuals as well as of the whole cosmos and cannot be radically changed; but on the other hand, it can be manipulated by people who know how. The paradox is probably part of the human cognitive apparatus and is hardly subject to anything like a logical solution.
all aspects of the future can be subject to divinatory interpretations. in the North, we have the whole spectrum between the future of an individual, as in Gísla saga Súrssonar ch. 30 concerning the dreams of the hero, and the destiny of the whole world (cf. Raudvere 2003: 58), as is related in Vǫluspá. Between these two extremes we find most of our examples, dealing with subjects that are important to groups of individuals, such as success in hunting, in fertility, or in battle.
But let us turn to some examples of various forms of divination among the germani and the scandinavians. The general term in old Norse, which may cover the whole range of divinatory techniques, is ganga til fréttar, meaning
‘to go to ask for news’, and the phrase is used often in the old Norse sources (cf. Meissner 1917). However, already Tacitus mentions that omens were very important among the germani ( Germania ch. 10; cf. also ch. 7–8). He says: auspicia sortesque ut qui maxime observant: sortium consuetudo simplex. virgam frugiferae arbori decisam in surculos amputant eosque notis quibusdam discretos super candidam vestem temere ac fortuito spargunt. mox, si publice consultetur, sacerdos civitatis, sin privatim, ipse pater familiae, precatus deos caelumque suspi-ciens ter singulos tollit, sublatos secundum impressam ante notam interpretatur.
si prohibuerunt, nulla de eadem re in eundem diem consultatio; sin permissum, auspiciorum adhuc fides exigitur. et illud quidem etiam hic notum, avium voces volatusque interrogare: proprium gentis equorum quoque praesagia ac monitus experiri. publice aluntur isdem nemoribus ac lucis, candidi et nullo mortali opere contacti; quos pressos sacro curru sacerdos ac rex vel princeps civitatis comitan-tur hinnitusque ac fremitus observant. nec ulli auspicio maior fides, non solum apud plebem, sed apud proceres, apud sacerdotes; se enim ministros deorum, illos conscios putant. est et alia observatio auspiciorum, qua gravium bellorum eventus explorant. eius gentis, cum qua bellum est, captivum quoquo modo intercep-tum cum electo popularium suorum, patriis quemque armis, committunt: victoria huius vel illius pro praeiudicio accipitur.
(To divination and the lot they pay as much attention as anyone: the method of drawing lots is uniform. a branch is cut from a nut-bearing tree and divided into slips: these are distinguished by certain marks and spread casually and at random over white cloth: afterwards, should the inquiry be for the people the priest of the state, if private the father of the family in person, after prayers to the gods and with eyes turned to heaven, takes up one slip at a time till he has this on three separate occasions, and after taking the three interprets them according to the marks which
638
Jens Peter Schjødt
have been already stamped on each: if the message be a prohibition, no inquiry on the same matter is made for the same day; if the message be permissive, further confirmation is required by means of divination; and even among the germans divination by consultation of the cries and flight of birds is well known, but their special divination is to make trial of the omens and warnings furnished by horses, in addition to other methods. in the same groves and coppices are fed certain white horses, never soiled by mortal use: these are yoked to a sacred chariot and accompanied by the priest and king, or other chief of the state, who then observe their neighing or snorting. on no other divination is more reliance placed, not merely by the people but also by their leaders and their priests; for the nobles regard themselves as the servants of the gods, but their horses are their confidants.
They have another method of taking divinations by means of which they probe the issue of serious wars. a member of the tribe at war with them is somehow or other captured and pitted against a selected champion of their own countrymen, each in his tribal armour. The victory of one or the other is taken as anticipatory decision.) (pp. 145–47)
some of these different divinatory practices are not known, at least not in any direct form, from the Nordic sources, although there may be some vague hints.
But the casting of lots is known from, for instance, Gautreks saga ch. 7 ( hlutfall) where Víkarr’s men want to decide who should be mock-sacrificed in order to get wind for their ships. although lot casting is probably too common universally for us to speak about a genuine continuity among the germani when it comes to details, it indicates that lot casting in some form took place among the germani for at least a millennium. as described by Tacitus, it is the pater familias who carries out the ritual in private matters, whereas priests are in charge in public matters. We do not know whether this division also obtained in later times, but it seems quite likely that it did. it has been suggested that the signs on each chip of wood ( surculus) could be a kind of runes, although most scholars are sceptical about that (McKinnell and others 2004: 13; de Vries 1956–57a: i, 433). still, the signs ( notae), whether proto-runes or not, needed to be interpreted by the priest or the pater familias, which may indicate that there was no standardized, culturally determined significance to each of them.79
apart from the casting of lots, Tacitus also relates that they perform some sort of ornitomancy. How this is done is not related in any detail, and we have 79 Being open to individual interpretations would, then, indicate a great deal of arbitrari-ness, since not only the priests, whom one can believe had learned the significance of the signs during some rituals, but also the pater familias were able to do the interpretations; and it is not very likely that all such heads of households would have gone through the same rituals, acquiring the same numinous knowledge.
25 – Various Ways of Communicating
639
no clear parallels to that sort of divination from later sources, although sometimes birds do appear to have a special relation to future knowledge. Thus, we are told in Óláfs saga Tryggvasonar (ch. 27) that Hákon jarl made a great sacrifice before his battle against Jarl Óttarr, and ‘þá kómu þar fljúgandi hrafnar tveir ok gullu hátt. Þá þykkisk jarl vita at Óðinn hefir þegit blótit ok þá mun jarl hafa dagráð til at berjask’ (then two ravens came flying, croaking loudly.
Then the earl believed that Óthin had accepted the sacrifice and that it was a propitious time to fight) (p. 167). first, we notice that the divinatory sign is linked to sacrifice, a relation that may have been important in connection with many divinatory practices. second, the very idea that birds — in this instance, not just any kind of birds but exactly birds signifying the war god — in their flight and cries can be interpreted as predicting the future, seems to indicate a continuity from the time of Tacitus, although other possibilities also exist.80
another example could be the titmice in Fáfnismál st. 32–39 who tell sigurðr about what will happen in the future. of course this is not as close a parallel to Tacitus’s statements or to the ravens just mentioned, but it could still indicate that birds were seen as especially associated with knowledge of the future.81
since ornitomancy was common among the Romans, too, Tacitus sim-
ply notes that this practice was also found among the germani. He is much more interested in the hippomancy that he mentions immediately after. The horses here are said to be bred with the purpose of being part of divinations.
again, in later sources, no rituals are described that we can safely assume to be a continuation of the practice that Tacitus relates. Nevertheless, there is overwhelming evidence for the importance of horses in both myth and rituals, as we have seen above, not least in sacrificial rituals, and this is also well known from other indo-European peoples (Derolez 1968: 289). The most famous horse in the sagas is freyfaxi, known from Hrafnkels saga Freysgoða. We are told that Hrafnkel who, as suggested by his appellative freysgoði, has a special relation to freyr, owns a horse to which is attributed a certain degree of sacredness, so that nobody except Hrafnkel is allowed to ride it (ch. 3). This could be compared to Tacitus’s information that the horses taking part in the divination should not be ‘soiled by mortal use’. Even so, there is nothing to suggest that freyfaxi was directly associated with divination.82 Perhaps it is more likely that the horses 80 for instance, some sort of loan from the sámi.
81 and birds, not least ravens, are thought to possess other kinds of knowledge, too, such as the capacity of memory, see lindow (2014a: 43–44), Hermann (2014: 15–17), and Mitchell (2019).
82 However, we know from Gesta Danorum (14.39.10) that hippomancy also existed
640
Jens Peter Schjødt
participating in sacrificial rituals were treated in a special way and that their blood played a role in some form of divination. However, we do not know how they were bred or whether they were treated in a special way prior to the slaughtering, although, again, archaeology may help us here, since it appears that pigs intended for sacrifices may well have been bred in another way than ordinary pigs (Magnell and iregren 2010). We do not know, either, what happened to Tacitus’s horses. Were they perhaps sacrificed after the interpretations of their
‘neighing and snorting’? it cannot be ruled out, since nothing is reported of their fate.83
finally, Tacitus reports of duels as a means of foreseeing the result of battles.84 This form of divination is not found in the old Norse material; although the sagas contain a great many examples of duels ( hólmganga), there are none in which divinatory aspects are obviously involved. a possible example from the mythology may be the fight between Þórr and Hrungnir, where it is said that the outcome is of huge importance for the giants in particular, since Hrungnir is the strongest of them.
There are, however, other kinds of divinations in the Nordic sources than those reported by Tacitus. We have already mentioned the interpretation of dreams, which was probably not connected to any ritual but which shows clearly that the other World could intervene, even without magical and ritual attempts to discover the future.85 We have also mentioned the practice of útiseta, which could well be a kind of necromancy. The wakening and questioning of the dead is widely mentioned in the sources (cf. Boyer 1994: 126–29), and the logic behind this practice appears to rest in the idea that the dead know the other World better than the living do, as we see in Fáfnismál, Baldrs draumar, and many other texts. it must, however, be emphasized that information about the future is only one aspect of the knowledge that can be gained from the dead.86
among the slavs, so it can clearly not be ruled out that this sort of divination existed across the whole of the North European area (cf. słupecki 1994: 29).
83 We do know that horse fightning and horse racing took place, which is also confirmed from archaeological findings, such as those at skedemose on Öland where slain horses were found in a bog (Hagberg 1967b) and as is strongly indicated by the Häggeby picture stone from Uppland, apparently depicting a stallion fight. Whether these fights were part of divination rituals, we cannot know.
84 for other examples from antiquity and medieval sources, see de Vries (1956–57a: i, 429–31).
85 for many examples of dream divination, see Boyer (1986b: 91–95).
86 The idea that the dead can supply the living with information from other Worlds has been treated by ohlmarks (1936), Boyer (1994: 126–29), and briefly in Hasenfratz (2011: 81–82).
25 – Various Ways of Communicating
641
During útiseta, the diviner therefore isolates her- or himself and receives inspiration from undefined spirits which, at least in some cases, seem to be the spirits of the deceased who supply her or him with knowledge, sometimes of the future, sometimes of other things hidden to ordinary people in an ordinary state of mind. We have, as mentioned, many examples of that in the sources, the most famous being the vǫlva in Vǫluspá.
The practice of reading signs from blood is known as far back as the first century bce where, as was mentioned above, strabo ( Geography 7.2.3) tells us that the priestesses of the Cimbri prophesy by observing the blood of war prisoners, clearly as part of a sacrificial ritual. There is no later evidence of this practice,87 but considering the important role that blood plays in many rituals, we should expect that certain divinatory signs could be observed in the blood of sacrificial victims. There are some clear instances where the practice of ganga til fréttar is connected to blót, for instance Landnámabók (sH ch. 7), where it is said that ingólfr prepared a great blót and went to enquire about his destiny, and that the frétt told him to go to iceland. We are not told in any detail exactly how this happened, but the combination of sacrifice and divination could well indicate the importance of the sacrificial blood (cf. Boyer 1986b: 184).
Probably the most famous of all the various ways of coming to know about the future are the seiðr rituals, which are treated in chapters (è22) (è26) and will also be discussed in (è30). suffice it here to say that part of the seiðr complex concerns divination, although many other elements of magic manipula-tions are also involved. When it comes to divinatory seiðr, it appears that only women performed this practice, which is no doubt related to the statement in Ynglinga saga ch. 7 indicating that men could not perform seiðr without being accused of ergi.
There were probably many more ways of looking into the future, some widespread all over scandinavia or the whole germanic-speaking area, others limited to certain areas and others again perhaps individual. as has been stated already, we should not look for any rigid system in the beliefs lying behind all these practices. is it the will of the gods that is revealed, as seems to be the case in Eyrbyggja saga ch. 4 where Þórr is addressed by Þórólfr Mostrarskegg, or is it some impersonal ‘fate’? in all likelihood, there would have been innumerable ways of explaining the ‘theory’ of divination, some broadly accepted and some highly individual, some operating on fully conscious levels and some only subconscious.
87 in his history of the Norman dukes ( De moribus et actis primorum Normanniae ducum) from the beginning of eleventh century, Dudo of saint-Quentin, in his description of the Normans’ pre-Christian past (ch. 2), gives another version of divinatory interpretations related to human victims.
642
Jens Peter Schjødt
Concluding Remarks
The main purpose of this chapter has been to discuss the important role of religious communication and to propose a theoretical model from which we can view the ritual framework in PCRN. further, we have attempted to show some fundamental ways in which communication between worlds could take place in PCRN, emphasizing the ritual dimension. it is not possible in a treatment such as this to be exhaustive on these matters, and other subjects, such as myths, the views on the dead, and many more which are treated in other chapters, could have been taken into consideration, because communication between the world of humans and that of the superhuman was multifaceted and involved many different elements in many different ways. in the following chapters, we shall deal in more detail with some of these subjects, bringing in a wealth of examples from archaeological as well as written sources.