Fragments of the Past: How to Study Old Norse Religion

Ardnamurchan boat site / Photo by Jon Haylett, Wikimedia Commons

By Dr. Britt-Mari Näsström
Professor of Religious History
University of Gothenurg, Sweden

Scripta Instituti Donneriani Aboensis 17:2 (1999), 177-185

The Source Problem

Ansgar made an unsuccessful attempt to Christianize the Vikings as early as in the 830s. / From Hamilton, Hugo. 1830. Teckningar ur Skandinaviens Äldre Historia

The state of the sources of the Old Norse religions presents a great problem. Most of them were written down two hundred or even three hundred years after the Christianization of the North, based on an oral tradition going back to the pre-Christian ages. This situation has led to an intense discussion about the Christian influence or even interpolations in the text or what are genuine proofs of Old Norse religion. The discussions apply to nearly every single line in the source material.

Certain motifs seem to originate in a common Indo-European tradition, whereas others leave us in the dark or allude to medieval traditions. The hypotheses about special myths in the texts oscillate noticeably between a view of these as sacred expressions of faith and seeing them as Christian mockery of pagan beliefs. The scholars in the field have for example different attitudes about the works of Snorri Sturluson. Eugene Mogk suggested that they were products of literary novels by Snorri himself; Hans Kuhn saw them as a collection of examples of syncretism; Walter Baetke regarded them as a zealous Christian effort to interpret the myths with reference to Paul’s Epistle to the Romans, 1:18-23. It is stated in this passage, that the Almighty God appeared to all mankind in the creation, but after that some of them declined into paganism (Lindow 1985: 38).

Three Different Points of View

The scholars are divided in three main groups considering the question of the sources and how to value them. There are the deniers, or, perhaps, the very skeptical who take a skeptical, even hypercritical attitude to the study of old Norse religion. This school began with Sophus Bugge in the end of the 19th century, noticing close parallels between the drama of Jesus Christ suffering on the cross and Odin hanging from a tree in the poem Hávamál. He claimed that there must have been an influence of some prevalent Christian topics, which had been transformed into a Nordic context. In the similar way the death of Balder and the mourning of his mother Frigg were similar correspondences (Bugge 1889: 291 ff.).

This school still believes that the notices about the pagan gods, made by Snorri Sturlason and other, anonymous authors, were a compound of stories, mainly emerging of the Old and New Testament. Even rites, such as Snorri’s description of the sacrifice at Hlade, have been regarded as a recital of Israelite sacrifice in the Old\ Testament. They maintain that Snorri Sturlason as a Christian, had no idea of the Old Norse sacrificial ritual, but had copied the only ceremonies he was familiar with, i.e. the sprinkling of holy water andthe narrative about the reddening of the door-post in the Old Testament by the victim’s blood. Neither sprinkling nor smearing of blood has been an exclusively Christian rite in the history of religions, so there are no reasons to presume that such sacrifices never occurred in Old Norse Religion (Düwel 1985: 36 ff.).

The most extreme part of this school still claims that most of the Old Norse myths are free fantasies, composed by skilled Icelandic historians inspired by the account of Adam of Bremen about the sacrifices at the temple of Uppsala. As Adam never visited Sweden, it is easy to prove that the evidences of a pre-Christian pantheon are untrue, they say, neglecting that he got information from the Danish king, Sven Estridsøn, who had been visiting Uppsala for seven years. We may consider his notes about the golden temple and the many male human victims hanging from the trees as utterly incredible without denying his veracity in other descriptions.

The source material is sometimes regarded as being so fragile that the only way of dealing satisfactorily with this problem is to look upon the stories solely as literary phenomena recalling earlier myths and rites. Such a method is used by Margaret Clunies Ross in her recent book with the striking title “Prolonged echoes”. She thus reduces the religion to literary expression, without discussing whetherits rites ever took place (Clunies Ross 1994: passim).

The other point of view, prevalent among historians of religions took the opposite position, which meant that they believed every single word in the written sources as genuine proof of pre-Christian belief. One quotation such as “the winds of the raw heathen ages howl in the family sagas” illustrates the uncritical, sometimes even naïve, attitude of these scholars, who received much contempt and mockery from the denying school.

It goes without saying that the third method must be a critical observation of the source problem, with regard to the influences of Christian and Classical sources while at the same time presupposing a pre-Christian religion in Old Scandinavia. This way is the hard way, between uncritical belief in every source and the hypercritical crushing mill, invalidating everything.

It is true that any attempt wholly to reconstruct the Old Norse religion must fail, but this is not so much due to the lack of sources as to the fact that we are never able to arrive at total insight into, and complete empathy with any religion from ancient times, no matter how many sources we have available. It has never been the task of a historian of religion to effect such a reconstruction, but to pay attention to perceptions and motifs which were important to those who practiced particular religions and then to place them in a social context.

A Polytheistic Pantheon

The primary Norse gods / From Encyclopedia Mythica

It must also be emphasized that the Old Norse religion did not consist of a monolithic or exclusive faith like — for example — Christianity and Islam. It was a polytheistic system, where many gods were worshipped, although the individual preferred to venerate one or a couple of them. As time went by, different gods and goddesses seem to have been worshipped under different names in different regions, like for instance Freyja, whose name in the eastern parts of Scandinavia was Fröja and western Freyja. Frigg who in Snorri’s mythological system became another goddess, seems to have been venerated in the western parts as far as the place names tell us and she has also given her name to sites in Great Britain (Näsström 1995:109).

Considering the study of myths and their provenance another problem must be taken into consideration as any attempt to give prominence to an “original” or “correct” version may plunge us into deep difficulties. The useful methods must, on the contrary, start out from regarding every myth, or fragment of a myth, as carrying a message of some value for the interpretation of the religions. With such a point of departure, we could locate recurrent motifs in the myths, not only belonging to Old Norse religion but in a widened perspective as far back in time as to the ancient Indo-European religions.

Dumézil and the Tripartite Functions

Photo of Georges Dumézil / Wikimedia Commons

This path directs us to the theories articulated by Georges Dumézil and his followers. Unlike contemporary scholars, Dumézil considered Snorri’s Edda, as well as other Icelandic sources, as being exponents of the old Indo-European traditions, making them corner-stones in his well-known hypothesis of the tripartite ideology in the early Indo- European societies.

Dumézil’s method is, as one of his followers puts it, unique and based on two postulates: Primarily, the contention that mythology expresses social values which strengthen social coherence and secondly that a linguistic unity — like the ancient Indo-European Language — suggests a considerable measure of ideological unity, understandable in the terms of comparative method (Dumézil 1959: 38).

His influence on research of comparative religion is immense and has in the field of Old Norse religions affected such scholars as Jan de Vries, Gabriel Turville-Petre and Edgar Polomé among others. Nevertheless, critical attempts to scotch the tripartite system have occasionally been made and one of them by E. Page is often quoted. Still, Page’s arguments are considerably weak and much of the criticism is, as Dumézil indicates in his reply, based on misunderstandings of the text and on misreading (Page 1978-79: 49-69). Some years ago Colin Renfrew attempted to sink the whole system in his much discussed “Archaeology and Language”; his chapter about Dumézil, however, revealed that he had read and understood little of his work (Renfrew 1987: 250-262).

Dumézil revitalized the study of Old Norse mythology and his method proved that the old myths were not novels made by Snorri’s literary activities. The war between the Æsir and the Vanir was regarded as a reflex of a proto-Indo-European myth, where the reconciliation between the classes or “fonctions”, as he preferred to call them, took place as an ideal and not as, some of his opponents still claim, as a historical event changed into a myth.

The Comparative Perspective

Bruce Lincoln used the comparative method in many of his works, which has broadened the view of Old Norse religion. His interpretation of the Indo-European creation myth, giving meaning to the name Ymir as a “twin” as well as the obscure Gods of the Land of the Dead, Yima and Yama and their role in the primeval sacrifice (Lincoln 1975: 121-143).

Jan Puhvel and other followers had pointed out other parallels to the old Norse gods and their myths, using the comparative methods in the Indo-European field (Puhvel 1987: 290 f ). Their approach to Old Norse religion as a part of the Indo-European family of language, no matter where this originally had its dawn, has brought new light on the old problem, though historians and philologists still cling to the Old Norse source per se, uninterested in the comparative perspectives, as they say.

The Functions of Freyja

Nuzzled by her boar Hildisvíni, Freyja gestures to a jötunn in an illustration (1895) by Lorenz Frølich

My own contribution to the field of Old Norse religion was an attempt to make the goddesses in the system visible. Little attention had been paid to them, as Dumézil and other scholars in the comparative field were not interested in goddesses as well as other scholars, who usually saw in her only a Nordic parallel to the Greek Aphrodite, lascivious and sensual.

It is true that Freyj a, like the other gods of the Vanir is voluptuous by nature, yet she is comprised of many varying characteristics. In many of her functions she was the counterpart of Odin, in her knowledge of sejd (witchcraft) and in selector of the dead on the battle field. In her aspect as protector of fertility and especially of love affairs, she fulfilled other functions. As Snorri puts it in Heimskringla, “Freyja had many (varying) temperaments” (Heimskringla 1951: ch. 10).

The only hypothesis that fitted with Freyja, in my opinion, was the tripartite system where she as the great goddess impersonated the three functions, the magico-religious, the warlike and the fertility. She is — to a higher degree — the parallel to the Vedic Sarasvati, the Iranian Anahita and the Roman Juno than to the Greek Aphrodite, though there exist similarities (Dumézil 1970, 300).

Still, one had to ask whether this tripartite appearance of the Great Goddess was a pattern exclusively for the Indo-European ones or if it could be found in other cultures. The Sumerian Inanna, called Ishtar in Akkadian, sometimes the daughter of the mood god Nanna, sometimes of Enlil, the god of the atmosphere or even of Enki, the earth god was the most conspicuous among the goddesses to compare with, as she divides her functions between love and war. These functions are connected with sacred kingship and a third aspect of Inanna is her appearance as the planet Venus, “Inanna of the Sunrise”.

The differences between Inanna and the Indo-European goddesses are distinct, however. Her love-affair with Dumuzi — she has no male spouse permanently connected to her — even her relationship with Dumuzi is rather ambiguous, whereas Freyja, Juno and the other mentioned goddesses are usually linked with at least one male partner, sometimes three, each of them representing one of the functions of the tripartite system.

Furthermore, Inanna’s lust for love corresponds with her lust for land and in this respect she constitutes the ideal goddess for any power-seeking king in the old Mesopotamian society. He consorts with her in a sacred marriage and she is said to stand by his side in battle. Such traces are not found in myths about the Indo-European goddesses, who were not exclusively connected with the kings but with the three different classes of society, priest-kings, warriors and producers.

The Study of Rites in Old Norse Religion

A detail from runestone G 181 in the Swedish Museum of National Antiquities in Stockholm. The three men are interpreted as the Norse gods Odin, Thor, and Freyr.

The fragments of Old Norse religion, based on the oral tradition, written down in the 12th and the 13th centuries, could be studied with other methods than those used by the philologists and historians. Especially in the cases of rites, where the research of anthropologists during the last decades has proved to be useful in the study of such rites as rites de passage and sacrifice.

Like other societies the Old Norse society had its rites of passage, considering baptism, initiation, wedding and funerals. In these we are able to discern the three stages of separation, liminality and aggregation, a pattern elaborated by Victor Turner (Turner 1975: 93-111). This model could also be used on a much discussed rite, called ganga undir jarðarmen in Icelandic and literally, “to walk under the adornment of the Earth”. This ritual appears in several sagas with the meaning of cutting the grassy turf of the earth in two parallel cuts, which were then lifted and propped up by spears. The involved persons then walked under this tunnel

The purposes of the ritual are varied. In Fostbroedra saga it is mentioned as the ritual of sworn brotherhood. The men who had passed the tunnel had to take revenge for each other and regarded each other as brothers. The best example is found in Gisla saga Sursonar, sometimes called the locus classicus of sworn brotherhood. Gisli and his brother prepared for the walk under jarðarmen together with their respective brothers-in-law. According to this text, the ritual was performed so that they walked in under the turfs, mixed blood with each other and let their blood drip into the soil (Gisla saga 1943: ch. 6).

It may seem strange that these two brothers and their brothers-inlaw were going to commit this act, since they were people already firmly joined to each other by both consanguinity and kinship. There must therefore be something more in this ritual, which in this case would have the purpose of uniting the four men. This is also evident in the context as the two brothers suspect that their brothers-in-law were plotting against them. The rite then was thought to become an instrument of keeping peace between them in the future. When they tried to perform the ritual, they could not co-operated in the single acts and they had to walk back to the entrance of the tunnel. This was one of the many omens that forebode the tragedy of Gisli and his brother and at the same time an excellent example of a ritual which failed.

The custom “to walk under jarðarmen” appears as an ordeal in Laxdoela saga, which released the accused from guilt and in Vatnsdoela saga the proud hero becomes humiliated but gains entrance to society instead of excommunication. The ritual survived in folklore with a fourth purpose: to heal sick children, who were drawn though a short tunnel in the earth.

This rite has earlier been interpreted as a returning to mother Earth and rebirth from her womb. In my opinion, this is an example of a rite de passage, distinguishing one kind of status from the other in a concrete and physical form. The rite seems to have had both a religious and a juridical purpose and the walk through the tunnel was an exchange of status as the old sagas tell us: from conflict to friendship, from guilt to innocence, from denunciation to acceptance and as in the late folklore from sickness to health.

Sacrifice is another case where the anthropological method is useful, when the discussion about the value of the sources had put a damper on the research. The recent studies about the meaning of sacrifice, especially in ancient Greece using anthropological and structural methods have extended the perspective, and this is something that also could be done in Old Norse religion.

Finally, the archaeologists are persistently presenting new finds that have to be interpreted. As mentioned earlier, Adam’s description of the hanged victims in the grove of Uppsala has for a long while been regarded as unreliable and invented by himself, in order to show the rudeness and cruelty of the pagan Svear. In 1984 a find under the altar of Frösö revealed the remains of a sacrificial grove, where the victims seemed to have been hanged on the tree. Several animals had been sacrificed there, among them six young bears. No human sacrifice was found, but the bears could have served as substitutes.


These examples of varying methods of overcoming source problems show that it is necessary to put new questions to the source material such as: What was the purpose of the sacrifice? Who sacrificed and who received it? Which sacrifices took place in calendar rites, in rites of passage or in crisis? And what are the areas of purification or avoidance in the sacrificial situation?

A combination of methods encompassing both the comparative and the linguistic aspect provides an opportunity to overcome the difficulties encountered by the students of Old Norse mythology, especially the problems with the sources. As we search for supporting structures in the myths and the mythical fragments, we do not need a hypercritical approach, discussing influences from sources coming from Medieval Europe as those sources are fruits of the same cultural heritage.


Benveniste, Emile
1969 Le vocabulaire des institutions indo-européennes. Paris.

Bugge, Sophus
1889 Studier over de nordiske Gude- og heltesagns oprindelse. Christiania.

Clunies Ross, Margaret
1994 Prolonged Echoes: Old Norse Myths in Medieval Northern Society. Odense: Odense University press.

Dumézil, Georges
—-1959 Les dieux des Germains: Essai sur la formation de la religion scandinave. Paris: Presses universitaires de France. (Mythes et religions, 38)

—-1970 Archaic Roman Religion: With an Appendix on the Religion of the Etruscans. Chicago: University of Chicago Press

Düwel, Klaus
—-1985 Das Opferfest von Lade: Quellenkritische Untersuchungen zur germanischen Religionsgeschichte. Wien: Halosar. (Wiener Arbeiten zur germanischen Altertumskunde und Philologie, 27) Gisla saga

—-1943 Gisla saga Súrsonar. Ed. by Björn K. Thórólfrsson and Gúdhni Jónsson. Reykjavík: Hidh íslenzka fornritafélag. (Izlendsk fornrit) Heimskringla

—-1951 Heimskringla; 3 vols. Ed. by Bjarni Aðalbjarnarson. Reykjavík: Hidh íslenzka fornritafélag. (Izlendsk fornrit)

Lincoln, Bruce
1975 The Indo-European Myth of Creation. (History of Religions, 15)

Lindow, John
1985 Mythology and Mythography. In: Carol J. Clover and John Lindow (eds.), Old Norse Icelandic Literature: A Critical Guide. Ithaca: Cornell University Press. (Islandica, 45)

Näsström, Britt-Mari
1995 Freyja — the Great Goddess of the North. Lund: University of Lund. (Lund Studies in History of Religions, 5)

Page, R. I.
1978-79 Dumézil Revisited. Saga-Book 22: 49-69.

Puhvel, Jaan
1987 Comparative Mythology. Baltimore: John Hopkins University Press.

Renfrew, Colin
1987 Archaelogy and Language: The Puzzle of Indo-European Origin. London: Cape.

Turner, Victor
1975 The Forest of Symbols: Aspects of Ndembu Ritual. Ithaca: Cornell University Press.