The Historia Norwegiae as a Shamanic Source
Clive Tolley
The earliest account of Sámi (Lappish) – and indeed any – shamanism was recorded by a Norwegian historian in the twelfth century in the Historia Norwegiae (HN). It is, ostensibly at least, of value to students of shamanism both for its age and its apparent reliability as what we might term an ethnographic document. This reliability stems in part from its particularity: unlike most of the later accounts from the seventeenth and eighteenth centuries, which are generalised descriptions of the shaman’s activities, the HN merely presents what was seen by certain Norwegians on one particular occasion (if we trust the narrative).
The text is as follows (Storm 1880: 85–6); the translation is my own:
Page 2
Despite the vividness of this description, it is only right and natural to question the reliability, or at least the bias, of the account. My starting point here is a comment made by John McKinnell (2003: 115) from the perspective of Norse (rather than shamanic) scholarship: ‘This passage probably represents Norse beliefs about Saami magic rather than the reality of it, for the word gandr does not exist in the Saami language.’ This, of course, poses a question for consideration: how far do the Norse perceptions and the description of the event represent something consistent with what we know of Sámi shamanism from elsewhere, on the basis of which a picture, albeit only a partial one, has been built up of how the Sámi themselves perceived their shamanic activities?
The Social function of the shaman
The shaman is an intermediary between this world and the spirit world (Siikala 1978: 321); he communicates between the two worlds in order to resolve a critical situation, and he manifests the presence of the spirits by means of role-play during trance. Clearly, this is a role fulfilled by the shamans of HN.
The shaman exercised various functions (B&H: 15–17), some of which do not appear in HN, namely psychopomp, hunting magician and sacrificial priest, since the particular purpose of the séance here did not match these. The account mentions two other roles of the shaman without exemplifying them in the events described. Foretelling is recognised as a shamanic role in many societies, but amongst the Sámi seems to play a minor role. The emphasis on foretelling in HN is perhaps influenced by the role of witches in the author’s own society; foretelling is recorded as playing a major part in the magic practice known as seiðr (dealt with at length in Strömbäck 1935). Recovering distant objects is ascribed to Sámi shamans, for example in the popular tale of how a Sámi who, during trance, brought a ring from a person’s distant home (B&H: 46). Olaus Magnus (1555: 121) also tells of this skill; it has clearly been a well-known folk-tale motif in Nordic regions since antiquity. Neither of these two roles, whilst they are consistent with the activities of shamans of later times, can be regarded as specifically Sámi as opposed to general Scandinavian characteristics.
The main part of the HN account exemplifies the shaman’s role as doctor. Sickness has two main causes: soul-loss, involving for the shaman a trance journey to the otherworld to retrieve the lost soul, as well as bargaining with the dead who are holding the soul, for example by promising sacrifices; and intrusion of an object or spirit, involving for the shaman the summoning, usually without trance,
Page 3
of the shaman’s helping spirits, who help to remove the intrusion. Whilst the death of the first shaman in the HN account results from a very literal intrusion of an object, in the form of a sharpened stake, this takes place while he is in trance: the activities of both shamans in the account may be classified as dealing with cases of soul-loss.
In addition, the shamanic role of contender is illustrated by the HN account, as it involves contention in the spiritual realm between rival shamans – the hostess, or her soul, is stolen away by enemy shaman spirits, and an enemy shaman spirit transforms itself into stakes (clearly with the aim of injuring whale spirits).
The purposes of the events described by the Norwegian historian are, then, in themselves consistent with Sámi shamanism as recorded later. There are, however, some strange aspects to the event, which I will come to presently. First, however, I present a point-by-point comparison between the Sámi séance as known from later sources and what the HN relates.
The séance
The sequence of the Sámi séance is analysed by B&H (1978: 97–101); I give a summary here. This is a composite analysis, and not all the elements were present in all places and on all occasions. Differences between accounts indicate that séances themselves differed, but the following represents a framework within which these variations occurred.
Page 4
The HN account appears to follow, in outline, the above sequence, and there is, allowing for local and temporal variation, arguably nothing in it which contradicts it.
Page 5
The points of comparison are enough to show that the Norwegian account must represent reasonably accurately an actual course of events. However, some problematic issues arise, partly from the complexity of the particular séance described, and partly from the bias of perception.
1. The set-up of the meeting is odd:
Several possibilities suggest themselves:
Page 6
The last option seems the most likely; this does not, however, exclude the possibility that the séance did not go entirely to plan for the Sámi, and some of the actions may indeed have been emergency measures.
However, these reservations about the honesty of the performance, far from calling into question the authenticity of the account, in fact suggest all the more strongly that the Norwegians reported exactly what they saw, or rather, under the Sámi prestidigitation, were supposed to see, even if we concede that their interpretations may sometimes reflect Norse beliefs.
2. Two shamanic events, as evidenced in later accounts, are intertwined in the sequence of the twelfth-century séance, which the later, perhaps simplified, accounts keep apart:
Presumably in the HN séance the first shaman did send his soul out; the attack on his helping spirit occurred by accident, not design, and really forms a distinct category not otherwise recorded in Sámi shamanism, rather than the normal ritualised contest. In any case, the differences from later Sámi tradition are not necessarily a sign of Norwegian confusion: the distinction between soul and helping spirit (a sort of alter ego) is fluid, and many of the later accounts confuse their roles to some extent.
3. The manner of the stealing of the hostess’s soul – for so we must interpret, from a Sámi point of view, her keeling over as if dead – is unusual in its suddenness, and in the involvement of rival shaman spirits – gandi æmulorum; it is usually the dead or evil spirits that were believed to steal people’s souls in later Sámi belief (B&H, 15); however, such activities on the part of rival shamans are met elsewhere, for example among the Evenks (Anisimov 1963: 107), and there is no reason to doubt the general authenticity of this information. Nonetheless, soul-loss normally manifests itself in the form of chronic illness; the hostess’s sudden demise looks more like the loss of consciousness characteristic of trance, which may again suggest that the hostess had some role to play in a shamanic drama, the precise nature of which eluded the Norwegian onlookers.
4. The obstacle of the stakes is unusual; no parallel appears to exist in the extant records of Sámi shamanism. However, in the Finnish story of Lemminkäinen’s shamanic visit to the feast at Päivölä (‘Sunland’) three obstacles are set in his way, the last of which is an iron fence (Kuusi et al. 1977, no. 34); this fence is to be compared to the spiritual marylya fence set up by the Evenk shaman to protect the clan lands against incursions by spirits sent by enemy shamans (Anisimov 1963: 107) – the posts of the fence were in fact mobile spirit guardians; if an enemy spirit penetrated it, it would bring disease and death. Clearly a similar idea lies behind the HN account, and hints at the complexity of shamanic belief among the Sámi which has left no record in later accounts.
5. The only area where we clearly see Norse conceptions muddying what from a Sámi point of view was actually going on is in the area highlighted by McKinnell: the gandus. According to Sámi belief the events described in HN must have involved the loss or sending out of three souls (the hostess’s, and the two shamans’), yet the writer shows no awareness of the sending out of the free-soul during trance at all. Rather, he sees the séance as a magic ritual for the sending of the gandus on its mission. The events may be elucidated by considering Sámi beliefs about the soul and spirits, and how the Norwegian writer has recast the events into something more comprehensible to him.
Page 7
The soul and shamanic spirits
In common with most peoples of northern Eurasia, the Sámi believed in one or several ‘body-souls’, responsible for the maintenance of life functions, and a ‘free-soul’, which could wander free from the body, for example in sleep (Paulson 1958: 38). Everyone was accompanied through life by various spirits (B&H: 41–3); the shaman was distinguished from others by his ability to contact and make use of these spirits, which were of three sorts (Bäckman 1975: 41–3, 160):
1. Anthropomorphic spirits. These gave the shaman his vocation, and at other times information, and a female spirit succoured the shaman, in particular during his initiation.
2. Theriomorphic spirits, or animal helping spirits. They carried out the shaman’s bidding on his trance journeys, and might act as escorts or steeds. The shaman sometimes imitated, and sometimes described, these spirits’ activities. They were of three main sorts, each with a predominant function: birds (messengers to the anthropomorphic spirits, and guides on trance journeys), fish (guides on journeys to underworld), reindeer (combatants against enemy shaman spirits, with any injuries incurred being reflected on the shaman himself – though Jens Kildal (1943–5: 138–9) notes that any of the spirits could undertake this role.
3. The dead, dwelling underground in a realm ruled over by a powerful old woman. The shaman had to fetch the souls of the sick from this realm.
This categorisation masks a great deal of conceptual fluidity, however; in particular, the dead were often merged with anthropomorphic spirits, and the shaman’s own soul with the theriomorphic spirits: for example, the shaman himself is often conceived as taking on the forms of beasts (I. Olsen 1910: 32, Itkonen 1946: 120). There is clearly a basic identity between the shaman and his spirit helpers (B&H: 100). Even from a Sámi perspective, let alone a Norwegian one, the precise nature of the spirits involved in the HN séance would be hard to pin down. Behind the gandus we may see something vacillating between the shaman’s free-soul wandering during trance and his accompanying helping spirit. The Christian author, however, clearly regards the gandus as an evil spirit quite independent of the shaman (or his soul):
In points 1–3 it corresponds to the Sámi animal helping spirits, though it was anthropomorphic spirits that were consulted for information (B&H: 43). In point 4 it corresponds to Sámi helping spirits or the dead. In point 5 it may represent the Sámi shaman’s own soul, which could travel on the animal spirits as steeds; ships and snow-shoes are not recorded as spiritual vehicles, and these depictions may have served a different purpose from that assigned by the Norwegian writer. Point 6 either represents a distortion of the animal spirit, which had animal form, but not as a result of transformation, and did not assume other forms,
Page 8
or it may represent the shaman himself, who in later tradition could transform himself, and take on the form of various beasts (not just those of the helping spirits), though not, as recorded, stakes. A more sophisticated concept may have underlain this feature, in which the (spiritual) stakes are guarded by shamanic spirits, as in the case of the Evenks.
It thus appears that the Norwegian writer has recast and amalgamated various Sámi spirits, both anthropomorphic and theriomorphic, as well as the shaman’s free-soul and the dead. This spirit he calls a gandus. This is not a Sámi word, but the Old Norse gandr.
Just a few observations on the complex concept of the gandr can be made here. The HN talks of the gandus assuming various forms, just as the gandr could, whereas the Sámi animal spirits did not do so: gandr is used in the sense ‘wolf’ in a number of kennings (Meissner 1984: 100, 102); the connection between ‘wolf’ and ‘sorcerer spirit’ lies in the fact that wolves were witches’ steeds. The world serpent is called Jörmungandr, ‘Mighty gandr’, in Völuspá 47 and in Ragnarsdrápa (Skjaldedigtning B:I:4), which indicates that the wolf was not the only animal that a gandr could appear as.
Some of the earliest occurrences of gandr are in Völuspá: the seiðkona in st. 22 vitti ganda, ‘summoned gandir with a vétt’ (the vétt being some instrument, like a drum, that could be struck: cf. Lokasenna 24, where Óðinn struck a vétt while practising seiðr); in st. 29 Óðinn receives spáganda from the völva: here the word is used in the sense ‘[news from] gandir of prophecy’. Thus one of the main functions of the gandr spirit was to gather information and impart it to the seer(ess) who has summoned it. The emphasis in the concept of the gandr on gathering information I believe furnishes a reason for placing this role at the beginning of the description of the skills the Sámi gandus conferred in HN, whereas in later tradition this is not foremost among the shaman’s roles, and is in any case often performed by the shaman’s own soul wandering rather than by the helping spirits.
The Norwegian writer was led astray by his knowledge of seiðr, the nearest native practice to shamanism, into presenting the séance as one in which the shaman performed certain rites to induce the gandus into effecting particular things, rather than one in which trance took place, during which the shaman sent his free-soul out of his body. The evidence for sending out the free-soul during trance in seiðr is weak: it seems rather to have involved the summoning of spirits to provide information or carry out tasks (seiðr as a source of information is found in later sources frequently; for example, in Fóstbrœðra saga (p. 243) it is said víða hefi ek göndum rennt í nótt, ok em ek nú vís orðinn þeirra hluta, er ek vissa ekki áðr, ‘I have caused gandir to run far in the night, and I have now become wise about those things that I did not know before.’ Eiríks saga rauða gives the fullest account we have of a seiðr séance (pp. 206–9), which involves the summoning of spirits who reveal men’s fortunes to the fortune-teller. Indeed, the points where the Norwegian fails to understand the Sámi concepts of souls and spirit helpers may be useful indicators of areas in which native Norse magical practices differed from the shamanism of their neighbours – though of course the earlier, pagan understanding of such practices may well have been attenuated or lost with the coming of Christianity.
Page 9
The drum
The drum of HN is said to depict four things: water-beasts, reindeer, snow-shoes and a ship, all of them said to be vehicles of the gandus. This interpretation must be approached cautiously. The concept of animal helping spirits travelling on vehicles does not exist among the Sámi; moreover, while the cetus is said to be a vehicle of the gandus when the diagrams are described, the gandus later appears transformed into a cetus. The information seems to represent a vacillation between the animal helping spirits seen as steeds for the shaman’s free-soul, and the free-soul transformed into, or more likely accompanied by, the fish helping spirit.
Later drums depicted a plethora of objects and beings, both spiritual and mundane, with varying purposes, some relating to everyday activities such as fishing, others to the spirit journey of the shaman. Precise interpretation of drum images is a source of much contention (Manker’s magnum opus on the Sámi drum (1938, 1950) remains the standard authority, but his inclination to read pagan spiritual significances into the depictions wherever possible has come under fire in more recent years). The Norwegian writer’s interpretation of the images as vehicles of the gandus at least indicates the probability that some of the images were related to the shamanic journey, but it is likely that he has lumped them all together under this one roof, and not understood their precise functions with clarity. Thus the interpretation of snow-shoes and ship may have been influenced by Norse traditions, where the magical ship Skíðblaðnir was the vehicle of Óðinn or Freyr, and Ullr was the ‘snow-shoe god’ (önduráss).
Conclusion
The Norwegian writer (and probably his informants) did not understand all that occurred in the Sámi séance, the most fundamental matter being the Sámi concepts of the soul and spirits, which have been partially understood and partially recast under the influence of the Norse gandr spirit. Nonetheless, a close examination largely vindicates the details of the description, and often enables us to reconstruct the Sámi conceptions when they have been distorted. The HN account not only confirms much that occurs in later accounts, but adds to them. These accounts were produced when shamanism was moribund, whereas the HN represents a more vital tradition, and hints at features that were later lost or not recorded. It provides a valuable snap-shot (even if one seen through a glass darkly) of an actual séance, as opposed to the usually generalised accounts of later centuries. The misunderstandings shown by the Norwegians are important hints about the differences between the Norse and Sámi magical traditions, whereas the closeness of observation hints at the possibility, in earlier, pagan periods, of Sámi influences on Norse religion.
REFERENCES