Contradictory cosmology in old norse myth and

religion – but still a system?

By eldar Heide

This paper presents a new understanding of the cosmology of pre-Christian

Scandinavian myth. The sources appear to give contradictory information;
for example, the Æsir are located in different places: at the centre of the
world, in the west, in the east, under the sea, and in the sky; Hel is placed
both in the underground and beyond the sea. in recent studies, this has led
to the conclusion that there is no system. The author argues that there is,
and that we misunderstand the passages to other worlds. The otherworld can
be defined as ‘the world beyond what we can access by natural means’. The
starting-point is the realm that is physically accessible to humans, forming a
compressed, wide ‘bubble’ around him/her and the local community, since
our natural range is very wide in the horizontal plane in all directions, but
very short downwards and upwards. Still, people have always imagined that
it is possible supernaturally to transcend this ‘bubble’ through certain pas-
sages. These passages point in many directions from the middle of the ‘bub-
ble’, but the locations of the passages are not identical to the location(s) of
the other world(s), the passages being interchangeable with each other and
often lead to the same (kind of) land/place. The other worlds have interfaces
with this world and with each other, but have no geographical location in re-
lation to this world or to the others – they are simply ‘beyond the passages’,
‘on the other side’. accordingly, the fundamental model may be construed
as an opposition between ‘us’ and ‘the others’, with an ever-changing border.
This system becomes clear when we examine not only the limited old norse
sources, but consider them in the light of the abundant folktales and legends
recorded in post-medieval times throughout northern europe. 1

1. a version of this manuscript was presented at the fifteenth international Saga Con-
ference in Århus, Denmark, in august 2012. it has been partly modified based on
the feedback received there. Thanks to andreas nordberg for commenting on the
manuscript, to Pär Sandin for translating it into english, and to Helen leslie and
Christian etheridge for proofreading. The final version is in all ways my responsi-
bility. The english names of the folktales in asbjørnsen & Moe’s collection have
been translated for this article.

102 Maal og Minne 1 (2014): 102–143

 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

1 Introduction and brief research overview

The conception of cosmology in nordic pre-Christian religion and myth

has been the subject of much discussion. The sources present a vague,
somewhat contradictory picture, which has given rise to a number of dif-
ferent interpretations. Snorri claims that the earth was imagined to be
disc-shaped, surrounded by ocean, with Jǫtunn settlements along the
coasts, inside of which the region of men, Miðgarðr, was supposed to
lie, fenced in with bastions against the Jǫtnar. in the centre of this the
region of gods, Ásgarðr, was finally situated (Gylfaginning 8–9 = Edda
Snorra Sturlusonar 1931: 22 ff.2). Snorri locates the living quarters of the
gods in the heavens, and Hel, the land of the dead, below the earth. at
the centre of the earth stands yggdrasill, the World tree. Scholarship on
nordic cosmology has by and large accepted this model. Handbooks and
other scholarly literature (overview in løkka 2010: 18 ff.) often use a
horizontal model consisting of three concentric circles to describe old
norse cosmology: in the middle, the World tree stands, surrounded by
Ásgarðr, outside of which is the Miðgarðr of men, outside of that is Út-
garðr, home of the Jǫtnar. Steinsland (2005: 98 ff.), for instance, says:
“outside and around Ásgarðr the world of humans unfolds; this place is
called Miðgarðr. as the name implies, the humans live ‘in the middle’;
they are located between the gods and the Jǫtnar. This position tells us
something about nordic man’s experience of life: the humans are crea-
tures living in constant tension between different forces” (my transla-
tion). Many scholars supplement this horizontal ring model with a
vertical axis stretching along the roots, stem and crown of yggdrasill: the
tree reaches from the underworld, up and through this world, and into
the heavens. various mythological creatures inhabit this ‘tree axis’. Some
interpreters (e.g., Meletinskij 1973a, 1973b; Hastrup 1981, 1990) agree
with Snorri that the gods have their place in the heavens, but this is re-
jected by Schjødt (1990: 40 ff.): he believes that this owes to Christian
influence, stating that only Snorri arranges things in this way, not the
poetic edda and Skaldic poetry. Schjødt says that the older sources por-
tray the heavens only as a path in which the gods travel, and not as a
dwelling place. He therefore imagines that the vertical axis consists of
the underworld and the earth’s surface only. another controversial aspect
2. The chapter enumeration of Snorri’s Edda follows the edition of finnur Jónsson,
Edda Snorra Sturlusonar 1931.

103
Eldar Heide

of the cosmological standard model is the number of worlds. Most schol-

ars follow the threefold partition, but structuralists (such as Meletinskij
& Hastrup) consider Miðgarðr and Ásgarðr together as one region, so
that the sum total amounts to two: Miðgarðr and Útgarðr. This world
view will then fit the structuralist notion of human mentality organising
the world in opposite pairs. The structuralists (foremostly gurevich
1969, Meletinskij 1973a, 1973b; Hastrup 1981, 1990) also connect the op-
position between Miðgarðr and Útgarðr with the ‘inland’ and ‘outland’
on a farm, and with further general opposite pairs such as in : out; known
: unknown; centre : periphery; etc. They accordingly link the mythological
macrocosm to the microcosm of everyman.
This model has gained wide impact, especially in the field of archae-
ology, but has faced some opposition in recent years. Clunies Ross (1994:
51) believes that the notion of only one region outside of Miðgarðr does
not agree with the source texts, which speak of several regions, often
nine. She suggests that the evidence favours rather “a spatial conceptual-
ism of a series of territories belonging to different classes of beings
arranged like a series of concentric half-circles, the perimeter of each cir-
cle being imagined as a kind of protective rampart, a garðr.” Moreover,
Clunies Ross observes that the term Útgarðr is hardly documented at all
– it does not appear in poetry and only once in Snorra Edda, pertaining
to the abode of the enigmatic, untypical Jǫtunn Útgarða-Loki. Jǫtun-
heimar is the name for the region of the Jǫtnar in the older sources.3
Brink (2004) goes further in this direction and rejects the structural-
ist, binary model altogether, finding reason to believe the cosmology to
“have been more complex, with a larger number of spheres and poles
than two” (Brink 2004: 297), and enumerating Mannheimar ‘abode of
men’, Þrúðheimr ‘abode of powers(?)’, Jǫtunheimar ‘abode of giants’, Mus-
pellsheimr ‘world of fire (farthest to the south)’; and so on (ibid.: 294).
Brink concludes:

Really it is impossible to try to create a spatial logic out of the myth-

ical rooms and places appearing in Snorri’s narrative. [...] it is not an
improvement, though, to try to structuralise the bits and pieces found
in the Poetic edda into a world system. [...] The ancient Scandinavian
world model was not a logically structured system, but – as so typical
3. vikstrand (2006) discusses the implications of the names Miðgarðr, Ásgarðr and Út-
garðr.

104
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

of oral culture – an unstructured, mutable number of rooms and

abodes [...]. innumerable illogicalities and apparently impossible rep-
etitions appear. (Brink 2004: 296–97; my translation).

Wellendorf independently arrives at the same conclusion (Wellendorf

2006: 52, based on a lecture in 2004). Schjødt also says that “there were
not any consistent ideas about the mythic geography”, although in a foot-
note, without expanding further upon the idea (1995: 23).
on the other hand, the recent dissertation of løkka on ancient
nordic cosmology defends the bipartition. after a close reading of the
most reliable corpus, the mythological eddic poems, she concludes that
they “spring from a cosmology characterized by a basic opposition be-
tween the world of gods and the world surrounding it, an opposition
which primarily seems to concretize the categories of in- and outside” (re-
gardless of whether the outside is explicitly called Útgarðr or not; løkka
2010: 115; my translation). løkka accordingly supports a “cosmological
base model consisting of two primary components, Ásgarðr and its sur-
roundings”, even if “it is obvious that the region outside of Ásgarðr […]
consists of a number of lesser regions”.

2 Contradictory cosmological information

i agree with løkka, and believe that the base model she refers to can be
induced from the primary sources (to which Snorri in this case can only
belong in part, as he, like most modern scholars, is trying to create a sys-
tem), – and that it must be the starting point of any attempt to under-
stand the system of cosmology. However, one cannot escape the criticism
of Clunies Ross, Brink and Wellendorf simply by admitting that the re-
gion outside of Ásgarðr consists of several sub-regions, it is still necessary
to account in greater detail for how the various out-regions are located
within the model. The other problem that Brink and Wellendorf have
observed remains as well: the information in the sources concerning the
location of the different out-regions in relation to each other seems to
deviate far in all directions. i will supply a number of examples of this,
before proposing a manner of understanding the apparent deviations as
part of a logically coherent structure, only on a different level in relation
to our normal thinking.

105
Eldar Heide

2.1 About Hel

Several sources state that Hel, the land of the dead, is located in the earth
or below the earth, yet many and sometimes even the same sources claim
that one may travel in a manner and direction on the earth in order to
get there, while the stated manner and direction may vary to a consider-
able degree. Grímnismál 31 states that Hel lives4 under a root of the World
tree, and Vǫluspá 43 says that Hel is fyr jǫrð neðan ‘below the earth’ (cf.
Vafþrúðnismál 43, Lokasenna 63). in Baldrs draumar 2–3, Óðinn rides
niðr to Hel, but on a foldvegr ‘ground road’, ‘road on the ground’, which
undeniably looks like a direction on the earth. When Hermóðr is travel-
ling to Hel to get Baldr out, Snorri states that the road to Hel goes norðr
ok niðr ‘north and down’, saying nothing of any passage down into the
earth. instead he depicts a landscape through which Hermóðr rides, giv-
ing the impression of travel in the horizontal plane. This also appears in
Helreið Brynhildar (prose preface and first strophe). Baldr, having been
slain, also travels horizontally to Hel, but in a quite different manner.
We are told that he covers the last distance on horseback, across a bridge
over the river gjǫll, but the funeral consisted of him being deposited on
a ship which is put out to sea and set on fire; and we are given the im-
pression that a sea journey is the principal mode of transport (and in ad-
dition, passing through fire. Gylfaginning 33–345). The same impression
is given by the widespread pre-Christian custom of ship burial. in Skír-
nismál 27(–28), too, Hel seems to be located on the horizontal plane: the
Jǫtunn girl gerðr is threatened with being placed outside of the commu-
nity, with her back against the world, ‘turned towards Hel’ (snugga Heljar
til). Here Hel appears to be imagined as located immensely far away, out-
side of everything, similar to at himins enda (see below). There is nothing
to suggest that gerðr is to sit in the underworld when turned towards

4. Hel is the name both for a region and for its ruler.
5. The ‘complete’ story of Baldr is known only from Snorri, but many scholars
(overview in abram 2006) believe that his source value is high at this point, seeing
that a number of alliterated sentences with three stave-rhymes in Snorri’s text sug-
gests that he has followed a subsequently lost eddic poem (e.g. Vex viðarteinungr
einn fyrir vestan Valhǫll; lorenz 1984: 559–60 lists 8 examples). abram (2006) rejects
this view, arguing that alliteration appears also in prose that we know is not based
on poetry. The appearance of alliteration as such in the prose of Snorri is not the ar-
gument, however; what suggests that he here builds on a lost poem is that this part
of Gylfaginning differs from other parts of Gylfaginning by showing a distinctly higher
frequency of this kind of triple rhyme stave sentences.

106
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

Hel; on the contrary she will sit in this manner on an ‘eagle’s tussock’
(ara þúfu á), i.e., in a high location.

2.2 About the Jǫtnar

The information about the location of the Jǫtnar is just as contradictory.
in Skírnismál 10, Skírnir travels across ‘wet mountains’ and through a
‘flickering flame’ to reach the Jǫtunn gymir. This may be in accordance
with the model outlined by Snorri and accepted by most scholars: the
Jǫtnar live in a mountainous region along the sea, surrounding Miðgarðr
like a belt. However, Skírnismál 10 also states that Skírnir passes though
darkness, and this is less apposite. Most sources are also at variance with
Snorri in stating that the Jǫtnar live in the east (løkka 2010: 152 ff.), or
that Þórr travels to the east (í austrveg) in order to slay them (e.g., Hár-
barðsljóð 23 and 29). if so, they do not live in a belt surrounding Mið-
garðr. in addition, other quarters are mentioned: Snorri places the Jǫtnar
in the east several times, but in Gylfaginning 1 and Skáldskaparmál 3 and
26 he says that Jǫtunheimar lie in the north, and in Gylfaginning 23 he
states that the Jǫtunn girl gerðr lives in the north. in Skírnismál, the
sword of freyr is presented to the Jǫtnar (cf. Lokasenna 42), and in
Vǫluspá 52–53, Surtr comes from the south with ‘the sword of the gods’
(sverð valtíva), generally considered to be that same sword, in order to
fight freyr. Discordant with the common ‘belt model’ are also some
sources in which the Jǫtnar appear to live across the sea. in Vǫluspá 51
the Jǫtnar arrive sailing from the east to the battle of Ragnarǫk. in Hár-
barðsljóð (especially strophe 56) it seems that the realm of the Jǫtnar is
separated by a strait from that of gods and men. That the Jǫtnar may
reach the realm of gods across the rainbow bridge Bifrǫst (more of which
later) also suggests that they have to cross water, which is what bridges
normally span. in Hymiskviða 5, Hymir, a typical Jǫtunn, lives east of
Élivágar (fyr austan Élivága), which Snorri interprets as several rivers
(Gylfaginning 5, based on Vafþrúðnismál 31), but which judging from the
name (-vágar) ought to be ocean bays. apparently these are situated a
colossal distance away from the centre of the earth, for the strophe states
that Hymir lives ‘east of the Élivágar, at heaven’s end’ (fyr austan Élivága
... at himins enda). So also in Vafþrúðnismál 37, which reports that the Jǫ-
tunn Hræsvelgr sits in the form of an eagle at heaven’s end (at himins
enda). Something similar is presented by Þrymskviða, in which it is only
possible to reach the Jǫtunheimar through the air. Together, Þórr and

107
Eldar Heide

loki travel in Þórr’s chariot, but when loki is to go by himself, he has

to borrow the eagle’s guise of freyja in order to get there.
other sources locate the Jǫtnar at the centre of the world: in Hár-
barðsljóð 37–39, Þórr encounters Jǫtunn foes on Hlésey / læsø in the
Kattegat, which lies in the centre of south Scandinavia and accordingly
of the old norse world (although in a remote part of this central location,
see below). Several sources also state that the Jǫtunn Hlér / Ægir lives
on this island (Simek 2006: 3, 193f.). in Grímnismál 31, we hear that the
Jǫtnar are located even more centrally: they live under a root of the yg-
gdrasil and accordingly at the very centre of the earth. in this case, the
position on the vertical axis is also different from what i have hitherto
related, i.e., that the Jǫtnar live on the same horizontal plane as the gods,
only far away. other sources, too, locate the Jǫtnar below the earth. in
Skírnismál 35, we learn that the Jǫtunn Hrímgrímnir lives fyr nágrindr
neðan ‘below the corpses’ gates’, that is below Hel; some say that the Jǫt-
nar live inside mountains or below the earth (Hárbarðsljóð 23); even Ægir
on the sandy island of læsø is called a bergbúi ‘mountain resident’
(Hymiskviða 2. in skaldic kennings, Jǫtnar are often referred to by vari-
ants of this.) nevertheless it is also often the case that the Jǫtnar are de-
picted as living on estates up in the daylight (e.g., Skírnismál, Þrymskviða
6, 23; Lokasenna, Hymiskviða 7). in the myth about Óðinn and the Jǫtunn
girl gunnlǫð, keeper of the Skaldic mead, both cases seem to apply at
the same time. gunnlǫð lives in the mountain Hnitbjǫrg, and Óðinn has
a hard time getting to her, but finally manages to have a hole bored
through the mountain, into which he crawls in the guise of a snake
(Gylfaginning 5–6, Hávamál 106. in the Hávamál version it is apparent
that the mead is located at some place under the ground; nú upp kominn,
str. 107). But when Óðinn wants to get out, he does not have to crawl
back the same way, but turns himself into an eagle and flies home, com-
pletely unhindered by the mountain, which would suggest that gunnlǫð
and her family lives above ground, in the daylight. in other myths, too,
gods fly to the Jǫtnar – not only to their realm, but to individual estates,
which are apparently situated above ground (e.g, Þrymskviða, Gylfagin-
ning 3, Skáldskaparmál 27).

2.3 About the gods

it is hard to understand by looking at the sources where exactly the gods,
as a whole, actually live. There are two questions that must be considered

108
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

here: the location of the gods in relation to other mythological creatures,

and their location in relation to humans. The answer to the first question
is usually ‘the opposite’ of whatever information is given regarding the
other mythological creatures; for these are located with the gods as a
starting point. The gods therefore live to the west, north or south of the
Jǫtnar, and on the surface of the earth. But there is also explicit infor-
mation that locates them next to yggdrasill, i.e., in the centre of the
world. it is Snorri in particular who makes this claim, but with support
from Grímnismál 29, who states that the gods keep their assembly place
(þing) by yggdrasill. in Lokasenna 34, it nevertheless looks as if the gods
live somewhere in the far east of the world: loki refers to the myth that
njǫrðr was sent as a hostage from the vanir to the Æsir (cf. Vafþrúðnis-
mál 38–39, Gylfaginning 11), saying that njǫrðr was sendr austr héðan ...
at goðum ‘sent eastward from here to the gods’: ‘here’ is the home of Ægir,
who is the host of the banquet in Lokasenna. This would imply that the
vanir live close to Ægir, and that the Æsir live east of both. There are
also passages locating some of the gods below water: frigg lives in Fen-
salir ‘the fen Halls’ (but yet keeps a falcon’s guise in order to fly ..., Skáld-
skaparmál 27, Edda Snorra Sturlusonar 1931: 105, 110). Grímnismál 7
speaks of a goddess Sága (understood by some as frigg under another
name) who lives in Søkkvabekkr ‘the Sunken Bench’ and sits there drink-
ing with Óðinn as the waves roll above them (svalar knegu / unnir yfir
glymja). The home of Heimdallr is situated in the opposite direction. He
lives at Himinbjǫrg ‘heaven’s mountain’, guarding the rainbow bridge
against the Jǫtnar (Grímnismál 13, cf. 44, Fáfnismál 15, Gylfaginning 6, 7,
8, 15). This would seem to imply that seen from the direction of the Jǫt-
nar, Heimdallr lives on a mountain that reaches up into heaven (and the
gods ‘behind’ Heimdallr on the same height?).
We are allowed to learn more of the location of the gods in relation
to men, but as far as i can tell it is only Snorri who claims that men live
in a belt surrounding the gods, so that this information may be likely to
owe much to his own attempt to create a logically coherent world model.
according to Grímnismál 31, men live beneath one of the roots of the
yggdrasill, with the implication that the gods live upon the earth, i.e.,
above men. But in Hárbarðsljóð 56 the gods live far from men on the hor-
izontal plane, and the same condition is implied by a piece of information
from the arab traveller ahmad ibn fadlan, who describes the funeral of
a Swedish viking chief on the volga in the tenth century. a slave girl is

109
Eldar Heide

to follow the chief into the afterworld, and before she is slain, she is
raised above something that looks like a door frame, through which she
can see the chief seated in paradise.6 as it is a chieftain, one may surmise
that by ‘paradise’, valhǫll is intended. accordingly it is possible to see
this realm by supernatural means, apparently from an aspect in the hor-
izontal plane. Saxo and several old norse sources state or imply that
water has to be crossed, in the horizontal plane, in order to get to valhǫll.
Saxo describes a horse ride from the realm of men, across the ocean to
the home of Óðinn (fisher & ellis Davidson 1979–80, book 1, 6, and 7–
69). in the prose introduction to Grímnismál, two young brothers go
astray on a fishing expedition, landing on a coast where they are wel-
comed by Óðinn and his wife frigg, disguised as peasants. in a prose
section of Frá dauða Sinfjǫtla, the body of Sinfjǫtli is forded across a
fjord – and out of the story – by a mysterious boatman. Presumably he
goes to valhǫll, for Sinfjǫtli was a great hero; and the skaldic poem
Eiríksmál from the tenth century mentions him among the residents
there (finnur Jónsson 1912–15 B i: 165). further examples are given in
Heide (2011: 61–62). The information given in Lokasenna 34 about the
hostage njǫrðr means that the Æsir live (far?) to the east of læsø, an is-
land in the world of men, in the Kattegat. in Helgakviða Hundingsbana
ii 49, on the other hand, valhǫll is ‘west of Heaven’s bridge’ (fyr vestan
/ vindhjalms brúar), and the (human) hero Helgi arrives there by ‘the
flight road’ (flugstígr).
There is also a variant of the scenario that places the gods in heaven.
Snorri locates all the gods there, but this must be owing to Christian in-
fluence, as Schjødt has noted. nevertheless, it seems that the location in
the heavens was one of several variants also in pre-Christian times, as
Mundal (1991: 233, 242, note 4), abram (2003: 55), and Wellendorf
(2006) observe. These scholars present a number of examples, the most
obvious being strophe 21 of egill Skallagrimsson’s Sonatorrek from the
tenth century (finnur Jónsson 1912–15 Bi: 34 ff.): Óðinn has lifted the
deceased son of egill to the realm of the gods (...mank .../ es upp of hóf / í
Goðheim / Gauta spjalli). i may add that this is in accordance with the
scenario where the gods can be reached on Heaven’s mountain across the
rainbow bridge. Friggerocken, literally ‘frigg’s distaff’, the name of the

6. Montgomery 2000: 17–18. The door frame scenario finds a parallel in Vǫlsa þáttr,
as observed by Steinsland and vogt (1981); a woman there asks to be raised above a
door frame, apparently in order to be able to see into the otherworld.

110
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

constellation orion’s belt in Swedish folk tradition, also implies that the
gods may be located in the heavens: the name suggests that frigg was
imagined to be sitting up there spinning.7 as far as i am aware, the name
Friggerocken is not attested before the 19th century,8 but it ought to stem
from pre-Christian times, as it is difficult to imagine why Christians
would invent such a name.9 instead one would expect Marirocken, which
is in fact attested (Sweden: Rietz 1862–67: 165, Zealand: grimm 1953
[1876–78]: 251), but likely to be a younger variant, as there are many ex-
amples of Maria substituting pre-Christan mythological women’s names
(e.g., lid 1946: 18, norn > Maria; grimm 1953 [1876–78]: 251). The place
and municipality Tysnes in Sunnhordland (Western norway) also ap-
pears to give witness to pre-Christian gods connected to heaven. The
name means ‘the headland of the gods’ or ‘the headland of Týr’. (in ad-
dition to Týr, the god’s name, old norse language had a masculine noun
týr meaning ‘god’, pl. tívar; see Marteinn H. Sigurðsson 2009.) The place
is characterized by a special relation to the sun at each of the four solar
turning points of the year. a few minutes after the sun has set behind
the mountains in each evening of either equinox or solstice, it makes an
‘encore’, as it were, appearing again above the mountain, and shines for
a few minutes right upon the Tysnes headland. The effect is most spec-
tacular at the winter solstice, as the sun then appears in a narrow gorge
not far from the headland, so that the sun during these few minutes sends
a precise beam of light right upon a large, man-made stone pile (norwe-
gian røys) situated here, while the rest of the area lies in shadow. inside
the stone pile, archaeologists have found a chamber with clear sacrificial
remains from the late iron age (Rage & agdestein 2007, Heide 2013
footnote 3). We have reason to believe that this ‘heavenly contact’ is the
reason for both the location of the sacrificial stone pile and for the sacral
name attached to the place (– and the unique concentration of place
names bearing witness to pre-Christian cult in the area surrounding the
headland; see olsen 1905 and Heide 2013.)
But in a passage from Saxo, the (human) hero Hadingus visits some-
thing that looks like valhǫll after following an old woman down through
7. in old days rokk(r) was not a term for a spinning wheel, as in Modern Scandinavian,
but a distaff; see Heide (2006: 235 ff. The spinning wheel did not appear until the
late middle ages.)
8. Hyltén-Cavallius 1863–68 i: 237, 305, ii: 124; Rietz 1862–67: 165.
9. Compare the Medieval Christian attempts to root out pagan astronomy, see Mc-
Cluskey 1990.

111
Eldar Heide

the floor near the hearth (fisher & ellis Davidson 1979–80, book 1, 8,
14). Some old norse sources also indicate that Óðinn may be reached
by travelling down through water (Heide 2011: 67–68); some place
names and cult places indicate the same thing. The small lake Odensjön
in Scania, named after Óðinn, if indeed the name is ancient,10 is one ex-
ample. The water of the lake is gathered in a circular, crater-like hole in
the flat landscape; the lake lacks inflows, and in old times it was believed
to be bottomless (Nordisk familjebok 1888: 101). This has a similar char-
acter to the north Sami sáiva ponds (cf., e.g., Pakasaivo in northern fin-
land), which are typically small and without inflow, believed to be
bottomless and containing passageways to another pond rather than the
visible one (Wiklund 1916, Bäckman 1975, Mebius 2003: 82). There is
reason to believe that this passage was considered in the past to be a link
to the otherworld – the noaidis (Sami shamans) most often used the guise
of a fish as transport when they went to the land of the dead;11 they were
said to ‘dive’ when going there (olsen 1910 [etter 1715]: 45, cf. 46; Heide
2006: 232–33), and in southern Sami regions, sáiva – in the form
saavje(aajmoe) – means ‘ancestral mountain’ of a similar type to the old
icelandic Helgafell (Eyrbyggja saga: 19, Landnámabók: 125) and Kaldbak-
shorn (Njáls saga: 46). Judging from its name and from the examples of
the sáiva ponds, there is reason to believe that odensjön was imagined
to be a passageway to Óðinn / valhǫll. The argument is corroborated by
sáiva / saajve being to all appearances a loan from Proto-Scandinavian
saiwa-R ‘lake’ (the etymological ancestor of sea / sjö / See; Weisweiler
1940), which indicates that the ideas of such water passageways existed
in old germanic tradition. This is confusing in that one is able to travel
through water to a mountain visible on earth, but this is also the case in
Eyrbyggja saga and Njáls saga: both Þorsteinn Þorskabítr and Svanr of

10. Stig isaksson (1958: 29) believes that the name Odensjön is a learned invention from
early modern age, but if so, we ought to have heard of an older name. We have not,
and isaksson does not seem to have conclusive arguments. His strongest is that d- in
Oden- is pronounced, while it is not in Scanian legends about odin (Odens jakt). But
there are many examples of peculiarities in the pronunciation of place names, and
very many have had their pronunciation influenced by writing, without being for
this reason learned constructions. There are many examples that an inter-vowel d
because of spelling pronunciation is pronounced in ancient names (e.g. Eide in nu-
merous places in norway), in dialects that normally skip the d in this position.
11. Skanke 1945 [1728–31]: 191–92, 195, cf. Solander 1910 [1726]: 26; cf. Heide 2006:
136 and Skanke 1945 [1728–31]: 195, Jens Kildal 1945 [1730 and later]: 140, Sigvard
Kildal 1807 [1730–t.?]: 456.

112
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

Svanshóll drown before they can enter the mountain. accordingly this
does not run counter to what i have referred to previously so much as it
simply presents further examples of confusing information. The lake
Tissø on western Zealand (Holmberg 1986) also has traits in common
with the sáiva-waters and odensjön. The lake is only a metre deep along
the edges, but about 250 metres from the shore it deepens spectacularly
into a kettle of 10–12 metres’ depth. This is highly unusual in this region,
and it is the closest thing that western Zealand, flat as a pancake and poor
in lakes, has to ‘bottomless’ water with a steep pit depth. This makes it
reasonable to assume that this is some of the reason why the lake was
named ‘Týr’s lake’ or ‘the gods’ lake’. The many deposits of sacrifices
from Merovingian and viking ages found in the lake (outside a king’s
or chieftains’s building complex; Jørgensen 1998, 2002, 2008) also sug-
gest that contact with the gods was sought through the water. (Unfortu-
nately, odensjön has not been subject to archaeological investigation as
far as i am aware.) The name Goðeyjar ‘the god islands’ in Salten (north-
ern norway; Rygh 1905: 209) also indicates that people in pre-Christian
times imagined that contact with the gods was possible down through
water. The islands, which seem to have been a bastion of paganism dur-
ing the conversion of norway (Óláfs saga Tryggvasonar 1941: 324 ff.,
Rǫgnvalds þáttr ok Rauðs; Perkins 2001), are situated in Saltstraumen,
the strongest tidal current in the world, and there is reason to believe
that this is the reason for the name that associates the island with the
gods (and perhaps especially the many powerful whirlpools that occur
in the current; whirlpools have been regarded as passageways to the oth-
erworld in several traditions12).
This is certainly not an exhaustive list, but i have included more ex-
amples of contradictions than previous scholars, as i have been inten-
tionally looking for such examples, whereas the preferred method has
elsewhere usually been to search out information that may as far as pos-
sible be arranged into a coherent pattern. in any case, the above presen-
tation ought to show that it is futile to try to arrange all the cosmological
information found in the different sources into a coherent picture, for
they blatantly contradict each other.

12. Holmberg [Harva] 1927: 78–79, Patch 1970: 21, anttonen 1992: 36, nordberg 2003:
39; cf. ‘sucked down’ in some german tales, Siuts 1911: 54–55.

113
Eldar Heide

2.4 In folk traditions and other religions

This is true not only for (the sources to) old norse religion, but also for
european folk traditions. irish and Welsh traditions, as attested from
early viking age to the present time, exhibit beliefs in something some-
what similar to valhǫll: a rich, lush land of happiness inhabited by a race
similar to the fairy people, which can be reached by heroes and princes
after death. on this realm, MacKillop says:

[The] accounts are frequently ambiguous and contradictory about the

place of the otherworld. it may be unplaceable on human maps, or it
may be identified with a remote island in or under the western seas.
Sometimes an enterprising sailor reaching that remote island may
enter the otherworld. or another adventurer may enter the other-
world by travelling on land to enter mounds or dwelling-places of the
divine; [...] caves, especially the famous one at Cruachain, are often
thought to be routes to the otherworld in all Celtic traditions, and
so are some lakes. (MacKillop 1998: 359–60, similarly in Jones 1930:
52, Macculloch 1964: 49 ff., 114 ff.; cf. Byberg 1970: 181, 198, Carey
2000: 116.)

a similar diversity of possibilities is probably to be found in most folk

traditions, all over the world. Slavic folk tradition exhibits notions about
the dead person having to travel long distances here on earth in order to
reach the land of the dead; or over a rainbow or across the Milky Way;
or up a steep mountain of iron or glass; or sailing after death to a land
beyond the ocean, sometimes imagined to be situated at the place where
the sun sets, sometimes being simply an island; the land of the dead may
also be situated beneath a certain lake, or the dead person may be thought
to inhabit his grave (Ralston 1872: 107 ff.). Siuts has demonstrated this
kind of varation in german folktales (1911: 19–58): in order to get to ‘das
Jenseits’ – the troll or the land of the dead, in many cases – one has to
travel far, and farther than far, maybe through several kingdoms and over
the sea or a big river, or a bridge; or one has to get across a silver moun-
tain or a golden mountain on the way, or across an endless plain, or
through a large, thick, dark forest; in some cases one has to go to the sun
or the moon; or into a hole or shaft in the earth, or a well; far, far down;
or into a hole or through a passageway leaning in a vertical direction or
leading up into the daylight in the otherworld (ibid: 53); or up onto a

114
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

mountain, preferably one that reaches into the sky and is steep and
smooth as glass; or a mountain may open, so that one may enter it. This
kind of variation is found also in norwegian folktales (e.g. asbjørnsen
& Moe 1965 ii: 7 ff., 248 ff., 322 ff.; iii: 55 ff.), and in other ethnic reli-
gions, and partly also in the large book religions (cf. Widengren 1953:
340 ff.). in ancient greek religion one imagined lands of the dead situated
both in the underworld (Hades) and on a western island or islands of
happiness (for the chosen), not unlike that imagined by the insular Celts,
or in or at the grave (Knight 1958: 227). in Christianity, Hell has some-
times been imagined to be situated beyond a large ocean bay and/or
below the land of the living (long 2005: 9454). in Jewish tradition, the
mythological purgatory gehenna seems to be located both under the sea
and/or under the ground; or at the foot of a mountain range (long 2005:
9453 f.).

2.5 Contradictions also within tales

an important point is that the inconsistent variation of locations for the
otherworld is not only found in the tradition at large. Such variation is
found also within the scope of a single tale. We have seen above how
Baldr travels across the sea to reach Hel, even while Hermóðr not long
afterwards can ride there, and even while it is said to lie niðr. We have
seen how Óðinn, having had to bore through the mountain to get into
gunnlǫð’s abode, can fly out of it, unhindered by the mountain. Such in-
consistent combinations may be found also in folk traditions recorded
in later times. The last mentioned inconsistency finds a parallel in the
norwegian folktale “The king’s three daughters in the Blue Mountain”
(asbjørnsen & Moe 1965 ii: 7 ff.). Here the troll abducts the king’s
daughters through the air, but in order to release them, the hero has to
lower himself down through a deep shaft (from the top of a hill, beyond
wastelands and a river guarded by a bear and a lion). When he has slain
the troll, he raises the king’s daughters up through the shaft, but as he is
betrayed he has to find another way for himself, and may finally fly home
on the back of a giant eagle, unhindered by the earth and the hill that he
lowered himself down through. in the north norwegian legend “The
fairies (tuftefolket) at Sandflesa”, the fairies (literally tuftefolket ‘the toft
people’, the people who live under the house foundation), arrive sailing
to a small island in the sea, but they cannot return by ship, since the hero
has thrown steel over their craft. instead they enter a hole in the rock on

115
Eldar Heide

the island, and apparently reach home this way instead (asbjørnsen &
Moe 1965 ii: 52). in a Sami tale from nesseby in eastern finnmark it is
possible to sail, in the horizontal plane, through fog to the sea fairies,
but when their realm is reached, a (dry) land underneath the sea is en-
countered (Christiansen 1920: 19, cf. Heide 2011). in the german folktale
“Das Wasser des lebens” (grimm 1843 ii: 71–78) it is possible without
trouble to ride – very far – to the supernatural part of the world, but in
order to get back again, one has to take a ship across the sea. in another
german folktale, ‘frau Holle’ (grimm 1843 i: 155–58), the stepdaughter
reaches the otherworld by jumping down into a well, but returns in the
horizontal plane through a magical door.

3 A proposal for a solution

3.1 Considerations about the ‘enlarged corpus‘

When i adduce folk traditions recorded in the nineteenth and twentieth
centuries as evidence, this is not because i believe that these traditions
have remained unchanged during the eight to fourteen centuries that have
passed since the advent of Christianity. There is reason to believe that
most folk traditions have changed very much during this period. never-
theless, there is also reason to believe that quite a number of basic notions
stood the test of time well, in particular those that did not directly con-
tradict Christian doctrine. To these belong the cosmological notions in
tales where the plot alternates between where humans live and where
trolls, fairies and other supernatural creatures live (i.e., folktales /
Märchen, and in part legends / Sagen). in some cases we have confirma-
tion that such cosmological notions have been very stable, when legends
known from later years have been recorded in writing in the Middle
ages, such as for instance the gandreið episode in Þorsteins þáttr bǿjar-
magns13 (more of which below). That cosmological notions have remained
relatively consistent can also be observed by comparing folktales and leg-
ends with the mythological texts, because they contain the same cosmo-
logical motifs, possibilities of variation, and internal contradictions.
(Myths are in fact a kind of folk tradition, and the old norse religion

13. Saga af Þorsteini Bæarmagni 1827: 176–78; cf. Jón Árnason 1958–61 [1862-64] i: 101
ff., iii: 162 ff. and Heide 2006: 222 ff. The elf queen legend is so clearly revised in
the þáttr that the later tradition can hardly be based on this version, ibid.

116
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

was folklore, cf., e.g., gunnell 2006, but we are not used to think about
it that way, as Christianity has introduced the idea of an antagonistic re-
lationship between authorized religion and folk belief.) accordingly it
ought to be possible to supplement to some degree the scarce old norse
corpus of texts with cosmological information found in folktales and leg-
ends. Here some might object that it is unnecessary to resort to folk tra-
ditions recorded as late as the nineteenth and twentieth centuries, as
basically the same system appears in legends from the high middle ages,
especially in ‘post-classical’ sagas of icelanders and legendary sagas, some-
times also in the ‘classical’ sagas of icelanders: supernatural creatures are
found in all kinds of remote locations, often at night on distant islands
(e.g., Ketils saga hængs: 172–73) or on a mountain in poor visibility (e.g.,
Bárðar saga: 133 ff.), or in all kinds of remote lands (cf. Røthe 2010,
lassen 2011, arngrímur vídalín 2012). But regarding cosmology, none
of the saga genres is as close to the myths as are folktales. only in folk-
tales, myths and legends, the relation between ‘us’ and ‘the others’ is the
principal motif, with the action based upon an oscillation across the border
between this and the other world. But the encounters with the other-
world in the sagas are also in accordance with the pattern found in folk-
tales and legends from later times. if we then examine this ‘enlarged
corpus’ of myths, and especially the folktales and legends, it will be even
more apparent that it is impossible to tweak and adjust the forthcoming
information so as to arrive at a logically coherent system of the type that
has been construed, with gods in one region of the cosmos, Jǫtnar in an-
other, men in a third and so on. This kind of system is not extant.

3.2 What is the mythological landscape?

nevertheless i have reached the conclusion that there is some kind of
system, since the ‘enlarged corpus’ almost insists upon it being equal
whether one or another manner or road is to be travelled in order to reach
the fairy people, the land of the dead, or whatever place one is going to.
i believe that the reason that we do not discern this system is because we
misunderstand it, or mistake the level of abstraction at which it is con-
strued. Melitinskij (1973a: 56, 50) has used a somewhat similar approach
to the problem when claiming that the horizontal and vertical axes in the
eddic texts are parallel systems of code which may be converted into
each other, so that for instance the travels of Óðinn along horizontal and
vertical vectors in search of secret knowledge may be regarded as variants

117
Eldar Heide

of the same kind of passage (similarly in, e.g., Drobin 1991: 113 ff. and
nordberg 2003: 76 f.). nevertheless this does not solve the problem of
the sources often giving opposite directions to each other, for instance
locating the gods in the west and in the east, in the centre of the world
by the World tree as well as at the end of the Rainbow bridge in the far
west. it looks to me as if we are mistaken as to the very fabric of the
mythological landscape.
Men do not live in the same realm as the gods, as the structuralists
have supposed. We do not live in the ordinary sense in the mythological
world at all, as most scholars seem to take for granted (an exception is
nordberg 2003, e.g. p. 149; there may be others). Certainly some of the
eddic poems (Vǫluspá 4, Hárbarðsljóð 23) do appear to use Miðgarðr as
a designation of a place that is inhabited by both gods and men. This is
understandable since we and the gods are as it were ‘allied’ and share the
same purpose in the cosmos. But when we look at matters more closely,
a different picture emerges: Grímnismál 41 presents Miðgarðr as a place
created for humans only, and the mythology supplies different place
names (overview in løkka 2010: 147), and accordingly an entirely dif-
ferent geography than this world. only exceptionally do men appear
there, and then always when some kind of threshold has been crossed
which is normally impassable for humans – as in Gylfaginning 26, where
Þórr drives with his goats [across the sky] and arrives to a human family,
or when human heroes die and pass over water or other in a mystical
manner in order to reach valhǫll (Heide 2011: 61). it is also only excep-
tionally that gods appear in stories about men, and then only at places or
times that are removed from the normal: long, long ago, as in Rígsþula,
perhaps Grímnismál, and the legendary sagas (see Røthe 2010 and lassen
2011: 152 ff.); beyond the ocean, as in the prose preface to Grímnismál;
on læsø, a small, scarcely populated island far out in the sea (see Heide
2011: 62, 67, 72, 76 ff.); or on a headland jutting out into the ocean in a
gale, as in Reginsmál 15–19. normally, the gods in this world are repre-
sented only by cult places sacred to them and/or carrying their names.14
last but not least, the most dependable sources erect barriers impassable
to men between the realms of gods and humans: the latter may live below
the earth under the World tree (Grímnismál 31) or at walking distances
that are discouragingly long for gods (Hárbarðsljóð 56). The gods and the
14. in a manner, Þórr is in the human realm when he is creating thunder, but at the same
time he is in the heavens; that is not in the human realm at all.

118
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

other mythological creatures do not live in this world; they live in what
scholars usually refer to as ‘the otherworld’. Here, in something that all
scholars are fully aware of, but which i believe not has been taken to its
full consequences bearing on the question of cosmology, the key may lie
to the understanding of the mythological world view.

3.3 What is the otherworld?

‘The otherworld’ i shall define as ‘the world beyond that which is reach-
able by natural means’. The starting point is the flat, very wide ‘bubble’
that constitutes the radius of action for each man and society before mod-
ern technology. in the horizontal plane one may reach far in all directions
around the circle, but down into the earth, rock or water, and up into the
air, only a very short distance. nevertheless, people in all times have
imagined that these are not absolute limits; there is something beyond
them, but only supernatural powers and creatures and sorcerers may
reach there, or from there to here, i.e., pass over the outer borders of our
natural reach. These supernatural travellers cannot pass in any manner
or at any place they might wish, but only through those methods and
passageways that we have seen in our attempt to extract cosmological in-
formation from the mythological texts and folk traditions. Supernatu-
rally, one may reach the otherworld or contact its powers in several ways,
often combined with each other:

– Through the air, flying or walking over a rainbow, usually far and
often into heaven.
– over the ocean.
– Travelling to the end of the world, often to the place of the sun
rising (east) or setting (west, in the ocean).
– Through darkness.
– over high mountains.
– Down into the earth or into mountains, often through natural
‘openings’, like clefts or holes or door-like passageways in the rock.
– into special mounds or mountains.
– Down into (special) lakes, brooks, wells, or bog holes.
– Traveling up or down along the World tree or something that
symbolizes it.
– Through sacrifice or ritual at unusual landscape formations – (spe-
cial) groves, waterfalls, mountains, glacial erratics, caves and crags,

119
Eldar Heide

round ponds without inflows, strong currents, small islands in the

sea, etc. (cf. von Sydow 1926, eskeröd 1947) – or at artificial coun-
terparts to such formations: grave mounds, grave cairns (røysar),
altars or cult houses situated on or near similar formations or at
central locations in human society.

3.4 Passageways
To sum up: to the gods, Jǫtnar or fairies, it is possible to reach by trav-
elling west over the ocean, to the sunrise in the east, towards heaven,
down into a mountain or the earth, or down into water (of a special kind).
But this does not mean that the realm of gods or that of the Jǫtnar ‘is’ in
any one of these places, nor in fact that the information is contradictory.
What has been tended to be understood as the location of these other
realms, seems to me simply to be passageways there, since it is possible
to reach the same realm through several passages, and since they may be
exchanged with each other, as in the myth of gunnlǫð, where Óðinn has
to crawl in through the mountain, but may fly home.
That these constitute passageways and not the otherworld as such
may also be seen in that the realm at which one arrives on the other side
does not have the same character as the passageway, being instead similar
to this world, only with richer qualities, such as greener grass, fatter cows,
greater wealth (and often inverted from this world to a greater or lesser
degree; Holmberg [Harva] 1925, Wiklund 1916: 55–57, Heide 2011). This
is especially obvious in folk tradition;15 whereas in the norse mytholog-
ical texts unfortunately few descriptions are to be found of the places in
the otherworld. one exception is the realm of elves in Þorsteins þáttr bǿ-
jarmagns (see footnote 13 here), taken to be from the end of the thirteenth
century but with much content that is ‘considerably older’ (Power 1993:
675–76).16 The hero and his companion jump into a river and pass down
through some sort of fog, until it brightens and they arrive at a place
where a waterfall falls from some cliffs, and they see a large settlement
and a castle. inside, a party is going on where people are drinking wine
from silver cups; in the high seat sit what appear to be a king and queen;
and everything looks as if it is made of gold and silver. it becomes clear

15. e.g., Jones 1930: 58 ff., grimm 1843 i: 155–158 (“frau Holle”), Straubergs 1957: 84.
16. Two parchment mss. (+ younger paper mss) are extant, one dating from ca. 1450–
75 (aM 343, 4to) and one from around 1550 (aM 510, 4to. Cf. Ordbog over det norrøne
prosasprog. Registre 1989: 452–53).

120
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

that at least one of the party guests is a fairy (huldumaðr), and since the
episode is a legendary saga version of the elf queen legends attested on
iceland from the early nineteenth century (see footnote 13), it is apparent
that this is the realm of the elves. Perhaps this description is only ‘semi-
mythological’, since the elves keep a marginal position in old norse
mythology, but the mythological texts that tell of gods and Jǫtnar supply
many enough clues to allow the conclusion that the pattern is the same
as in folk tradition and Þorsteins þáttr: Óðinn and Sága ‘drink happily’
beneath the waves, so they too apparently sit in a dry world below a water
passageway, not in a world of water that mixes with the mead in their
cups. in Vǫluspá 43, Hel is not in the earth, but ‘below the earth’ (fyr jǫrð
neðan), which probably means ‘on the other side of a passageway leading
down through the earth (or through water?)’.17 Beyond the aerial passage-
way, the Jǫtunn Þrymr keeps a farm, as everyone else, but there is obvi-
ously more wealth in his world; for his cows have golden horns, and he
has gold in abundance (Þrymskviða 23). We are not told how things are
constituted at gunnlǫð’s place beyond the passageway through the
mountain or air, except that the wealth obviously is greater there than in
this world, since she sits on a golden chair (Hávamál 105).

3.5 Only two realms in focus at the time

i believe that to a great degree there is a basis for a model in which the
realms of gods, men and Jǫtnar circularly surround each other with the
World tree in the middle, corresponding on a cosmological level to the
tuntre (‘courtyard tree’), gardstun (‘courtyard’), innmarka (‘inland’) and ut-
marka (‘outland’) on a farm (Modern norwegian forms). To be sure, the
placement of men as a belt in-between the gods and the Jǫtnar ought to
be rejected as a construction of Snorri, but the rest seems to be correct.
in iron age agricultural society, most people inhabited a personal envi-
ronment that often had a farmyard tree in the middle, surrounded by
17. The general system is not contradicted by the controversy regarding whether Hel
specifically is ‘a place with richer qualities’ or rather the opposite. Snorri describes it
as a pale, cheerless place (Gylfaginning 34), but in this, Christian influence quite cer-
tainly can be discerned. Baldrs draumar presents another vision of Hel; one where a
banquet is prepared and the floors and benches are littered with rings and gold (str.
6). in Snorri’s version of the Baldr myth (Gylfaginning 49), Hel also seems like an
ordinarily pleasant realm (cf. nordberg 2003: 72). in a 10th-century skaldic stanza
by egil Skallagrimsson, Hel is called Heljar hásalar, ‘the tall hall of Hel’ (finnur Jóns-
son 1912–15 B i: 43 [lausavise 5], Bergsveinn Birgisson 2001: 5), i.e., something de-
sirable.

121
Eldar Heide

houses and the cultivated land, outside of which lay the areas that were
uncultivated but still resourceful, housing powers over which one did
not exercise control. This is in accordance with the world of gods, having
the World tree in the middle, surrounded by the homesteads of the gods,
outside of which are all ‘the others’, whom the gods do not fully control,
but which represent important resources and potential. This is the model
at which løkka 2010 arrives, entirely correctly in my opinion – except
that i believe that it stands in need of revision in one important aspect:
The other realms do not keep a (mytho-)geographical location in relation
to the realm of gods, nor in relation to each other. each realm – for in-
stance Hel and the individual Jǫtunn homesteads – is instead closed
within itself, like a ‘bubble’ (of the flat type i have described above). The
different realms are not situated in one or another direction from other
‘bubbles’, nor inside others, but they have interfaces with, and passage-
ways to, other ‘bubbles’. They are not located in any geographical coor-
dinates, but simply ‘beyond the passageways’. it may seem strange that
such a system could work without the realms having a geographical lo-
cation in relation to each other, but we must remember that mythological
geography is always tied to myths, and subordinate to narratives. Then
it works, since in the myths, only two realms are in focus at the time, or
to be more precise: the realm which one occupies and another one, which
is accordingly ‘the other’ seen from the point of view of the first one. as
one enters the other, that realm becomes ‘the present realm’, from which
one may again enter ‘the other’, that is yet another world. Here are some
examples: in Gylfaginning 26, Þórr drives with his he-goats [across the
sky] and arrives at the farm of a human family, spending the night there,
before travelling austr across the great ocean to the realms of the Jǫtnar.
in Frá dauða Sinfjǫtla (a prose tale transmitted with the poetic edda),
the human hero Sinfjǫtli dies. it seems that he then comes to the gods
after having crossed the sea (Heide 2011: 61), and there he may have wit-
nessed the dead Baldr being sent across the sea to Hel. Before this, in
Baldrs draumar, Óðinn wants to find out if it is true that Baldr is going
to die, and for this purpose rides on the magical horse Sleipnir, making
the foldvegr rumble, niðr to Hel, the land of the dead. But he does not
stop there; he takes another just as large a step, riding east before the en-
trance, coming to the grave of a vǫlva, whom he raises from death with
spells and interrogates about the future. – njǫrðr is a vanr, having come
a long time ago from the vanir to the Æsir (Vafþrúðnismál 38–39,

122
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

Lokasenna 34, Gylfaginning 11), and he follows the gods along to a ban-
quet at Ægir’s place. The home of Ægir in the sea is then in the other-
world as seen from the realm of gods, Ásgarðr; but Ásgarðr beyond the
sea in the east is in the otherworld as seen from the realm of the vanir.
The vǫlva in the grave is in the otherworld seen from Hel; Hel is in the
otherworld seen from valhǫll (to which Sinfjǫtli arrives and from which
Baldr departs); and valhǫll is in the otherworld seen from the realm of
men (from which Sinfjǫtli has departed). Similarly the human family that
Þórr visits is in the otherworld as seen from the realm of gods; and the
Jǫtnar beyond the sea is in the otherworld as seen from the human fam-
ily. Which realm is ‘this’ and which ‘the other’, and which realm borders
on which, are perpetually shifting; and if one attempts to place all the
realms in relation to each other, the result will only be confusion. But as
long as the focus remains on each separate narrative, featuring one inter-
worldly relation at the time, no problem will ensue.

3.6 The passageways are re-used

one thing that clearly indicates a system of the suggested kind is that re-
gardless of where one is and regardless of what version of the otherworld
one is to reach, the same methods or passageways are always used in
order to reach the target. Here are some examples: læsø / Hlésey is a
small island in the middle of the Kattegat, and accordingly a place to
which notions of the otherworld have often been tied in norse literature
as well as in later folk traditions (see, e.g., Byberg 1970 & Heide 2011).
in folk tradition the fairies and similar creatures often live on similar is-
lands; and in Egils saga einhenda 1954 (365) it looks as if the hero Ásmundr
comes to Óðinn after jumping into the sea at læsø (Heide 2011: 67, 76
ff.). But it is not only to humans that læsø is a place of contact with the
otherworld; this is the case also for the gods. Þórr meets some Jǫtunn
crones on læsø (Hárbarðsljóð 37–39), and the gods meet the sea Jǫtunn
Ægir / Hlér there (from where the name Hlésey comes. Lokasenna, Grím-
nismál 45, cf. Simek 2006: 193–94, Edda Snorra Sturlusonar 1931: 78).
The geographical quarters are also ‘recycled’: we have just seen that both
the realm of gods and that of the Jǫtnar are situated beyond the sea in
the east; that is, seen from the home of Ægir and from the human family
at which Þórr stayed. The home of Ægir and the realm of the vanir ought
then, seen from the home of the Æsir, to be located to the west over the
sea (above 2.2, cf. Ólafur Briem 1985 i: 325), but in Helgakviða Hund-

123
Eldar Heide

ingsbana ii 49, the hero reaches Valhǫll travelling to the far west (across
the sea? See above, 2.3). The direction down is ‘recycled’ in the same man-
ner: generally it is possible to go down to the elves (footnote 13), or the
fairies, or dwarfs, and occasionally Jǫtnar, but Grímnismál 31 locates the
humans in this direction, seen from the gods; and in Saxo, the (human)
hero Hadingus goes down into the ground next to the hearth in a house
and reaches something that looks like valhǫll (fisher & ellis Davidson
1979–80, book 1, 8, 14). The many cult places that are connected to spe-
cial geographical formations (like Tissø) also suggest that one imagined
that the gods could be contacted in a downward direction. not least
often, people have attempted to get into contact with gods and other
powers though springs and wells; and the same method was successfully
applied by Óðinn – that is, a god – when he sacrificed one of his eyes to
a Jǫtunn who is associated with the well beneath the World tree (Vǫluspá
27–28, Simek 2006: 281–82). Simultaneously there are myths that locate
some of the gods themselves under water (above, 2.3). The motif ‘im-
mensely far away’ is also often ‘recycled’; that passageway you can always
travel, no matter what version of the otherworld you want to go to. also,
‘death’ and the ‘land of the dead’ are often ‘recycled’: from there Óðinn
attains secret knowledge, like sorcerers (McKinnell 2005: 200 ff.), –
only he cannot get it from that land of the dead which he himself inhabits
(i.e. valhǫll); he goes instead to Hel; he raises corpses from the grave;
and he speaks with bodies dangling from the gallows (above 3.5; Há-
vamál 157, Ynglinga saga 18). Óðinn accordingly uses the same points of
contact with the land of the dead as are available to humans with special
abilities. This holds true in general: the gods use the same points of con-
tact with the otherworld as humans do, even in cases where it appears to
make no sense for the gods: for they have cult places and cult buildings
(Vǫluspá 7, Grímnismál 13 and 16, probably Hyndluljóð 1); and in
Lokasenna, it seems that the gods attend the same kind of inviolable, cul-
tic symposia as the humans use in their religious cult – even while the
gods do not have gods.18
This last phenomenon, and the fact that no matter in which world
one is situated, the other worlds and the passageways to them are always
placed in the same directions and at the same kind of geographical for-

18. in Vafþrúðnismál 38, the many hof and hǫrgar over which njǫrðr is said to be master
may be located in the human world, but in the mentioned eddic stanzas, it is clear
that the cult place / building is located among the gods.

124
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

mations in the landscape, can only be explained to stem from an ethno-

centric generalization from the point of view of the ‘bubble’ each man
and society is surrounded by, as described above. as we humans – if we
possess special abilities – may reach the otherworld by travelling west
on the ocean (to the place where the sun sets?), east to heaven’s end
(where the sun rises?), heavenward, over faraway mountains, down
through openings in the ground; or contact the otherworld though sac-
rifice on cult locations, etc., so has to be the case also among the others
– the gods, elves, Jǫtnar, etc. The other realms are generalized from the
present one, with the same points (or types of points) of contact with
other realms as are imagined to exist in the present realm.19

3.7 The fundamental model of understanding

This ought not to be particularly controversial in itself. But it implies
that one did not imagine a unity of any greater kind than that relative to
the individual ‘bubble’; there was no notion of a unity of several ‘bubbles’,
but only of something else, of similar kind, outside the limits of the present
‘bubble’. it would then be a misinterpretation to attempt to unite the dif-
ferent realms into one complex model.
But if one is able to go in all directions from the gods to the other
realms, surely they must nevertheless – collectively – be situated as a
‘bubble’ outside of the ‘god bubble’, so that all realms may yet be de-
scribed with one single model? Regarded from the realm of the gods,
being then ‘this world’, this appears to be the case. But seen from the
realm of the Jǫtnar, or from any one of the other realms, the opposite
relationship is likely to be true: the realm of the gods is situated around
each one of these realms, as one may go there in any direction from
where one is situated at present, through the same kind of passageways
that we have seen time and again. on the other hand, it would seem to
misrepresent matters to say that all of the other realms collectively lie
wrapped around the realm of the gods, for only one of them at the time
does so. When it appears that the other realms collectively lie wrapped
around the realm of the gods, or this world, that is because there is only
one set of passageways out of this world.
accordingly, i reject the analysis of Meletinskij, according to which
the gods live in heaven and the Jǫtnar in the horizontal periphery, which
19. This may be compared with the way nomadic Sami created a ritual microcosm sur-
rounding each new camp (Rydving 2004: 100–101).

125
Eldar Heide

is the same as the underworld, where Hel is situated. nor do i think that
Schjødt’s analysis is on target (1990), which states that the gods are lo-
cated at the centre of the earth’s surface, Hel in the underworld, and the
Jǫtnar in the horizontal periphery. To associate the gods exclusively with
heaven must be attributed to Christian influences upon Snorri, as Schjødt
has made clear, even if this location was probably one of several variants
also in pre-Christian times. But to place the gods in an opposite relation
to the underworld is also problematic since frigg lives in Fensalir and
Sága in Søkkvabekkr, Ásmundr comes to Óðinn by jumping into the sea
at læsø, humans throw sacrifices to tívar or Týr down into lake Tissø,
and so on. (also one cannot claim that the humans are situated in an op-
posite relation to the underworld, when Grímnismál 31 locates us in this
very region.) it is not even necessarily the case that Hel is always located
in the underworld (as Schjødt claims, 1990: 47); there are examples of
Hel being simply in a place terribly far away, in the horizontal plane (ii
above). also, an outright opposition between heaven and the Jǫtnar does
not seem to exist. it appears to be unproblematic that the eyes of the Jǫ-
tunn Tjatse become stars (Hárbarðsljóð 19, Skáldskaparmál 4, Edda
Snorra Sturlusonar 1931: 81), and the vault of heaven was created from
the skull of the primordial Jǫtunn Ýmir (Gylfaginning 5). Christianity
has probably influenced us both directly and indirectly to see a model
with the gods’ enemies in the underworld and the gods situated upon
high. Snorri is an important source, and he claims this to be so; and from
our Christian background we are used to seeing it this way.
To me, it looks as if two ‘universal directions’ are essential to norse
religious cosmology: away in the direction of the otherworld, from a cen-
tre that is dependent on the circumstances, and back. This may be heavily
reminiscent of Meletinskij, who claims that down and away horizontally
mean the same thing, each in its own system of code, and that they may
be ‘translated’ into each other. But there are two important distinctions.
The first is that away may be in any direction, not just down or horizon-
tally, but up as well, and not just in the case of the gods. i mentioned an
example of this in section 2.1 above: in Skírnismál 27, the Jǫtunn girl
gerðr is threatened with having to sit outside of society, with her back
on the world, ‘turned towards Hel’ – on an eagle’s tussock, that is on a
steep, preferably high, inaccessible crest. We find another example in
Fjǫlsvinnsmál, where the Óðinn-hero Svipdagr has travelled long roads,
arriving at a Jǫtunn homestead where the desired Jǫtunn maid resides

126
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

on the top of a mountain (str. 35–36 & 49).20 This is one of the manners
in which she is located out of reach; trolls in the folktales are often both
terribly far away and situated on a high mountain (above, 2.4). Probably
this is the idea also behind Heimdallr’s residence, Himinbjǫrg, which we
considered above in part 2.3 That loki hides at the top of a mountain in
Gylfaginning 36, and that Þjazi and after him his daughter Skaði reside
on mountains, should probably also be understood in the same way, as
well as perhaps the general tendency of the Jǫtnar to reside in mountain-
ous regions. The other difference from Meletinskij’s model is that the
centre does not have to be Ásgarðr; the location of the centre is depend-
ent on the myth and situation. When the gods attend a banquet at the
Jǫtunn Ægir’s place in Lokasenna; or Helgi is to go to valhǫll from his
wife in the realm of men (Helgakviða Hundingsbana ii 49); or people in
the human realm make sacrifices to the gods or send one of their own
deceased to them, the centre is located at Ægir’s or somewhere in the
human realm, and the realm of the gods is instead ‘beyond the passage-
ways’: in the far east (Lokasenna 34), or beyond Heaven’s bridge in the
west (Helgakviða Hundingsbana ii 49), or beyond the ocean (human boat
funeral), or in the underworld (human cultic practice: the names Fensalir
and Søkkvabekkr accordingly make better sense seen from the point of
view of humans worshipping frigg and Sága than seen from the realm
of the gods). The relevant directions are accordingly not away from an
absolute centre and back again, but away from and back to the centre as
defined by the present circumstances, the starting point being, as
sketched above, the ‘bubble’ surrounding each human being and society
stretching in all directions as far a distance as ordinary human reach al-
lows. But the same goes for any realm, not just the world of men. When
it looks like other beings than gods properly live at the end of the world
and in the underground, this is probably because most myths are re-
garded from the point of view of the realm of gods. if they had offered
a perspective towards the realm of the gods, from any one of the other
realms, there is all reason to believe that it would be the realm of gods
that would most often have appeared to be situated at the end of the
world or in the underground.21

20. The idea that Fjǫlsvinnsmál is a late pastiche is on close examination not firmly based
on textual evidence: see Heide 1997.
21. The reason for supernatural powers being more often connected with the under-
ground, through passageways such as springs, caves, clefts, special mounds, etc., than

127
Eldar Heide

This proposed model of understanding may be relevant to the inter-

pretation of sacrificial findings and graves. if the underlying idea is that
several passageways exist out of the ‘bubble’ and into the otherworld,
and accordingly several methods may be used in order to get there or in
contact with the powers there, no essential difference needs to be read
into for instance cremation, corporeal burial, horse and ship burials, or
various kinds of sacrifices. at least the different kinds of cultic practice
need not imply different cosmological beliefs. Certainly the point of view
that i propose does not imply that the choice between a sacrifice in water,
on a mound, at a mountain or of another kind, or the choice between a
burial in one or the other manner is necessarily irrelevant. That the var-
ious kinds of sacrificial or funereal methods are variations on the same
theme – i.e., manners of reaching, or getting in contact with, the other-
world – does not rule out the possibility that some powers were partic-
ularly connected to certain passageways (e.g. female powers to springs,
wells and water, due to elementary symbolism), or that the choice of fu-
nereal method may have some significance. for instance, one may sur-
mise that men earning their living from the sea would prefer a ship burial
or boat funeral, whereas horsemen would choose a horse burial. This
would be an ‘extended parallel’ to the practice that women were usually
buried with women’s tools and men with men’s tools and weapons; and
a direct parallel to the artefacts found in graves that answer to the dead
person’s profession, for instance in the grave of a smith the particular
tools of his trade. if this is accordingly the case, i.e., that the choice of a
certain ritual is not necessarily significant, but may be so, a parallel can
be observed in the periphrastic practice of skaldic poetry. normally it
makes no difference if a ship is called a ‘deer of sail’ or a ‘reindeer of the
waves’, but in certain instances the poet can be shown to have chosen a
heiti or kenning in order to add particular connotations.

with the sky, may be that a great part of the background of the idea of the otherworld
are mirror images in water, which show us just that, another world beneath us: see
Holmberg [Harva] 1925 and Heide 2011. other locations to which supernatural pow-
ers are especially connected are also inciting to the imagination or attract attention,
but in different ways: the place where the sun rises or sets; the rainbow; far away,
tall mountains, etc.

128
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

4 Problems and paradoxes

4.1 Otherworld creatures on the border

in my opinion, there are strong arguments that support the model of un-
derstanding outlined here, but this does not mean that it is entirely un-
problematic. Something that may look as an objection is that it is not
always clear that the creatures in the otherworld inhabit universes of their
own, as for instance the elves do in Þorsteins þáttr bǿjarmagns. This is es-
pecially the case when mounds or mountains that resemble peat houses,
or caves, are not depicted as passageways to a different universe beyond
them, but quite simply as homesteads,22 with outdoor areas belonging to
the present world (the world of humans). This scenario we find in for
instance Bergbúa þáttr and also regarding the elf hill in Kormáks saga
(288). elves live inside it (hóll ... er álfar búa í), and people from the human
settlement that lies adjacent to it sacrifice an ox to the hill, painting it red
with the blood of the sacrifice for the purpose of healing a man. often
such mounds or hills are situated outside of human settlements, though,
in an undefined or unknown location far into the forest or high in the
mountains. This is the case in for instance Göngu-Hrólfs saga (199–200),
Þorsteins þáttr uxafóts (351–54) and Þorsteins þáttr bǿjarmagns (Saga af
Þorsteini Bæarmagni: 176–78), and also concerning the Dwarf stone in
Þorsteins saga Víkingssonar (12–13). in the folktales as well, several trolls
do actually live in cliffs or mountains (e.g. asbjørnsen & Moe 1965 ii:
233, 39; iii: 7–13, 117–22), and not in another world beyond a passageway
in a mountain. They may also live in ordinary farms in the daylight (e.g.
ibid: 189–207). in such cases, they nevertheless live beyond obstacles of
other kinds, for instance far into the woods, beyond a large lake, or ter-
ribly far away, and are in this respect situated in ‘the otherworld’, as most
of the elf hills and Dwarf stones are. But the elf hill in Kormáks saga is
placed right next to the human settlement. So is the stone in which the
ármaðr / spámaðr of Koðrán eilifsson lives in Kristni saga and Þorvalds
þáttr víðfǫrla (Biskupa sögur i: 8–9, 62 ff.), and there are several parallels
in folk traditions.
22. Some of the background to this is probably that many houses, especially peat houses
(and especially of the goahti kind. north Sami; norwegian gamme, Swedish kåta),
in old times resembled spherical mounds, cf. lid 1942: 143. larger houses also often
had peat roofs and peat or stone walls, and could therefore look like rectangular
mounds. Caves have been used as homesteads, especially by hunters and shepherds,
until the present times.

129
Eldar Heide

The problem here can be formulated in the following way: some of

the creatures of the otherworld do not exclusively belong there. Perhaps
notions around the deceased inhabiting mounds and other kinds of graves
may aid our understanding of how this adds up. it seems that even
though there was a belief in mythological lands of the dead such as Hel
and valhǫll, this was not incommensurate with the notion that the
deceased inhabited their graves. in Helgakviða Hundingsbana ii 51, a
burial mound is accordingly called draughús ‘ghost house’. in Land-
námabók (102 ff.), Ásmundr atlason composes a strophe of poetry from
inside his grave mound about his existence there, and the legendary sagas
in many places speak of deceased people living on inside their grave
mounds (e.g., Egils saga einhenda: 338, Hálfs saga ok Hálfsrekka: 96,
Hrómundar saga Gripssonar: 410 ff.). as many scholars have observed,
perhaps most lucidly nordberg (2003: 73 ff.), a ‘mythological
amalgamation’ of the graves and the mythological lands of the dead takes
place. Here are some examples: in Ynglingatal 30 (finnur Jónsson 1912–
15 B i: 12), Hel is called hallvarps hlífi-nauma ‘shelter wife of the (grave)
mound’ – as if she is present in the grave (mound) itself. in Hervarar
saga ok Heiðreks (17), as Hervǫr conjures angantýr up from the dead in
his grave, and the grave mound opens, he says hnigin er helgrind ‘the gate
of Hel has opened’ – as if the entrance to the grave mound is also the
entrance to Hel. a near identity between the grave and the land of the
dead was also seen above, in the myth in which Óðinn visits the grave of
a vǫlva that is situated in Hel itself – in Baldrs draumar. (a more
extensive treatment of this can be found in Quinn 1994.) in Helgakviða
Hundingsbana ii, it is in the context of Helgi returning from valhǫll to
the mound in order to meet his beloved that his grave mound is called a
draughús. Perhaps the connection between the grave and the land of the
dead was imagined to consist of an option for the dead to ‘commute’ be-
tween the two, or more diffusely in a simultaneous presence of some
kind at both places. Similar notions occur even today in everyday Chris-
tianity – a Christian widow, for example, believes that her husband is in
heaven, but if she wishes to speak to him, she typically visits his grave.
in such a ‘holistic’ comprehension of the grave and the land of the dead,
the grave becomes a kind of ‘border post’ between this and the other
world. it is situated in this world, often at a central location or viewpoint,
and we may assume that the deceased in the grave is following what is
happening in this world among his progeny, of which Hœnsa-Þóris saga

130
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

(46) is one example. if the deceased returns, his or her ‘haunt’ is in this
world (e.g., Eyrbyggja saga 1935: 93 ff.). nevertheless, the grave, as the
residence of the deceased and as a closed space beneath the earth, is ob-
viously in the other world, and from there passages lead further to the
collective / mythological lands of the dead. Such an idea of a constant
connection between the grave and the mythological land of the dead is
probably inherent in the notions of grave mounds that kept green or bare
throughout the winter (Bárðar saga: 121, Gísla saga Súrssonar: 57, Ketils
saga hængs: 173; maybe others).
i believe that mounds or stones which are residences of elves
(fairies), and which are placed at central locations in human settlements,
should also be considered as the same kinds of border posts as these grave
mounds. Perhaps people imagined that they bordered on a collective
realm of elves beyond them, but this is not crucial to the argument. The
important thing as regards to my proposition is that such borderline cases
do not contradict the idea that a greater realm of elves existed in a uni-
verse of its own, in the same way as the idea of the dead residing in their
own graves does not contradict the notion of separate collective lands of
the dead situated in universes of their own. My investigation does not
concern such borderline cases but the problem of how to understand the
(mytho-)geographical relations between the different mythological races,
when the sources state that they inhabited realms of their own.
it may also be significant that the beings who appear to live in
mounds, cliffs, mountains, caves, etc., and have their outdoor areas in
this world, seem especially to belong to what used to be referred to as
the ‘lower’ mythology, which perhaps more properly (and less evolution-
istically) could be called ‘local mythology’, as opposed to the mythology
that was to a large degree common to all nordic or even north germanic
areas.

4.2 Basic elements common to all realms?

another difficult question is whether the different realms were actually
imagined as entirely complete in themselves and separate. i have indi-
cated above evidence that suggests that this was the case. But the extreme,
and therefore the most unitive, elements in the universe sometimes con-
tradict this notion.23 The Miðgarðr Serpent encircles the world seen from
23. for this question i have tried to extract some cosmological evidence out of Alvíssmál,
where Þórr questions the dwarf alvíss (‘all-wise’) about different names of for in-

131
Eldar Heide

the point of view of the gods (Gylfaginning 31), and from the point of
view of the Jǫtnar (Hymiskviða), and from the point of view of the men
(as seen in skaldic kennings; see Meulengracht Sørensen 1986: 271,
Bergsveinn Birgisson 1997). as it is the same serpent in all three cases,
we could say that it encircles all three worlds at the same time, and if so
there is a greater system than simply individual ‘bubbles’. Still i believe
that this is a modern conclusion that overlooks the mind-set of the past.
Just as pre-modern people imagined the dead to be present at the same
time in the grave and in a collective mythological land of the dead, he
may also have regarded each realm as complete in itself, yet surrounded
by the same serpent, without reflecting on this being problematic. This
follows in a manner from the other realms being generalized from the
present one, i.e., the ‘bubble’ we are surrounded by: ‘just as things are
here, so are they also in the other realms (only qualitatively greater)’. This
paradox – that the same Miðgarðr Serpent encircles different universes
– is of a similar type to that of the other realms surrounding the realm
of the gods seen from the point of view of the latter, but the realm of the
gods most likely surrounding each of the other realms seen from the
point of view of them.
a similar case is presented by the sun. in Gylfaginning, the Jǫtunn
master builder demands among other things the sun and the moon as
payment for building a wall around Ásgarðr (Edda Snorra Sturlusonar:
45–47, cf. Vǫluspá 21–26), and this indicates that the Jǫtnar have the same
sun as the gods. But the question is whether this is not simply the case
of the other realms being generalized from the present one. in folk tra-
dition, there are several examples of a unique sun in the otherworld (e.g.,
“The king’s three daughters in the Blue Mountain”, asbjørnsen & Moe
1965 ii: 7 ff.). This has to be the case when the otherworld is reached
through a passageway down through the ground, and also when
Þorsteinn Bǿjarmagn goes to heimrinn niðri (see above) – there is no ex-
plicit mention of a sun, but light appears when the passageway of water
and fog is exited (Því næst birti þeim fyrir augum). on the other hand, when
one only has to travel terribly far in order to reach the troll, the same sun

stance great and important things in the universe, but it is not clear whether the same
earth, sun and moon are present in all the realms, or if each has its own separate ver-
sions, or if the questions exclusively consider the earth, sun and moon of the realm
of the gods.

132
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

is above one’s head all the time, and this means that this and the other-
world are not separate ‘bubbles’ or universes. accordingly, Hel has its
own sun and universe when it is described as situated below the ground,
but when it is located terribly far away in the horizontal plane, it has the
same sun as the realm of the gods, and shares its universe.
To observe such contradictions is subtlety of a modern academic
kind. Clear notions about these things probably did not exist, nor the
idea that such notions should be necessary. Still i believe that the ‘bubble
model’ best captures and describes the system that yet exists: ‘This world
stretches to the outer limits of our normal reach, in all directions, and
beyond it lies the otherworld, which is basically similar, but with quali-
tatively greater properties, and which may be reached by supernatural
means through certain anomalous locations that constitute passageways.’
This system entails paradoxes, but is still what may be abstracted from
most of the sources.24
another question is to what degree the system outlined here was
something that people in general saw, or reflected upon. as the reader
may have observed, i have not, even after having formulated a theory
about it, been able to say consistently that everything is just passageways;
i sometimes say that the different mythological races live in this or that
direction. This is of course because the sources often express matters in
this way. Did not people then actually imagine that the Jǫtnar quite sim-
ply lived in the far east? Well, the sources also state that the gods live in
the east, and that the gods live in the west, and that the sea Jǫtunn Ægir
lives in the west, and so on. Would not accordingly ‘beyond what you
and i can reach’ be a simpler way to understand the notion that lay behind
these formulations, the passageways to the beyond being ‘re-cycled’ again

24. Some may perhaps say that the World tree has to be common to all of the realms,
but there is hardly ground for this supposition, as the World tree is first and foremost
simply a centre, and little universes have centres as well. Therefore it is reasonable
to believe that each world had its own central location – if it was at all structured in
a centralizing manner. one might imagine that the Jotnar had a central location with
a World tree of their own, but as løkka (2010: 242) observes, fragmentation is one
of the characteristics of Jotunn society; i.e., that it lacks a central organisation. Rather,
the Jotnar have many ‘world trees’, the farmyard tree of each Jotunn farm constituting
a centre of its own, just as each Jotunn homestead seems to be in a ‘bubble’ of its
own in the sources. (The model that i outline in fact means that one central location
did not exist in our world (the world of humans), only local and regional centres, in
part with symbolical equivalents of the World tree, as for instance the sacred grove
in Uppsala [Adam av Bremen 1984 [1075]: 224, Buchner & Trillmich 1961: 470] and
irminsūl of the Saxons [Simek 2006: 222].)

133
Eldar Heide

and again, because this manner of thinking was useful and economical
as long as only one version of the beyond featured in any given case?
How aware, or unaware, any given person was, is not, to my mind, cru-
cial.

Bibliography

abram, Christopher. 2003: Representations of the pagan afterlife in me-

dieval Scandinavian literature. Doctoral thesis, University of Cam-
bridge.
—. 2006: “Snorri’s invention of Hermóðr’s helreið”. in John McKinnel
et. al. (ed.): The Fantastic in Old Norse / Icelandic Literature. [Durham]:
[University of Durham]. 22–31.
Adam av Bremen. Historien om Hamburgstiftet och dess biskopar, 1984
[1075]. Translated by emanuel Svenberg. Comments by Carl fredrik
Hallencreutz et. al. Stockholm: Proprius.
anttonen, veikko. 1992: “The Concept of pyhä (Sacred) in Pre-Christian
finnish Religion.” in Mihály Hoppál and Juha Pentikäinen (ed.):
Northern Religions and Shamanism. ethnologica uralica 3. Budapest:
akadémiai Kiadó. 31–38.
arngrímur vídalín. 2012: The Supernatural in Íslendingasögur. A theoret-
ical approach to definition and analysis. The nordic Middle ages series
1. Reykjavík / Kopenhagen: Tower Press.
asbjørnsen, P. Chr. and Jørgen Moe. 1965: Samlede eventyr. 3 volumes.
oslo: gyldendal.
Bárðar saga [Snæfellsáss]. 1991. in Harðar saga. Íslenzk fornrit 13. edition
by Bjarni vilhjálmsson and Þórhallur vilmundarson. Reykjavík: Hið
Íslenzka fornritafélag. 99–172.
Bergbúa þáttr. 1991. in Harðar saga. edition by Bjarni vilhjálmsson and
Þórhallur vilmundarson. Reykjavík: Hið Íslenzka fornritafélag. 441–
50.
Bergsveinn Birgisson. 1997: Miðgarðsormur. B.a.-ritgerð í íslensku, Há-
skóli Íslands, vor 1997. Unpublished.
—. 2001. Hel. Heimildagildi dróttkvæðra kenninga um fornt hugarfar.
Cand.philol-verkefni í norrænum fræðum, University of Bergen.

134
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

Biskupa sögur I. 2003. Íslenzk fornrit 15. Ritstjóri: Jónas Kristjánsson;

Sigurgeir Steingrímsson, Ólafur Halldórsson og Peter foote gáfu út.
Reykjavík: Hið íslenzka fornritafélag.
Brennu-Njáls saga. 1954. Íslenzk fornrit 12. edition by einar Ól. Sveins-
son. Reykjavík: Hið íslenzka fornritafélag.
Brink, Stefan. 2004: “Mytologiska rum och eskatologiska föreställningar
i det vikingatida norden.” in anders andrén et. al. (ed.): Ordning mot
kaos. Studier av nordisk förkristen kosmologi. vägar till Midgård 4.
lund: nordic academic Press. 291–316.
Buchner, Rudolf and Werner Trillmich. 1961: Quellen des 9. und 11. Jahr-
hunderts zur Geschichte der hamburgischen Kirche und des Reiches. Fontes
saeculorum noni et undecimi historiam ecclesiae hammaburgensis necnon
imperii illustrantes. ausgewählte Quellen zur deutschen geschichte
des Mittelalters freiherr vom Stein-gedächtnisausgabe 11. Berlin:
Rütten & loenig.
Byberg, Jan e. 1970: “Dei lukkelege øyane” i norsk folketradisjon. Ei motiv-
gransking. 2 volumes. Hovudfagsoppgåve i nordisk, University of Ber-
gen. Pdf at http://eldar-heide.net/ -> anna fagleg.
Bäckman, louise. 1975: Sájva. Föreställningar om hjälp- och skyddsväsen i
heliga fjäll bland samerna. Stockholm studies in comparative religion
13. Dr. philos. thesis, University of Stockholm. Stockholm: almqvist
& Wiksell.
Carey, John. 2000: “The location of the otherworld in irish Tradition.”
in Jonathan M. Wooding (ed.): The Otherworld Voyage in Early Irish
Literature. An Anthology of Criticism. Dublin: four Courts Press. 113–
19.
Christiansen, Reidar Th. 1920: “litt om nordlandsk folketradition”.
Håløygminne 1. 12–26.
Clunies Ross, Margaret. 1994: Prolonged echoes. Old Norse myths in me-
dieval Northern society. Volume 1: The myths. The viking collection.
Studies in northern Civilisation 7. [odense]: odense University
Press.
Drobin, Ulf. 1991: “Mjödet och offersymboliken i fornnordisk religion.”
in louise Bäckman et. al. (ed.): Studier i religionshistoria. löberöd:
Plus Ultra. 97–141.
Edda Snorra Sturlusonar. 1931. Udgivet efter håndskrifterne af Kommis-
sionen for Det arnamagnæanske legat ved finnur Jónsson. Køben-
havn: gyldendal.

135
Eldar Heide

Egils saga einhenda ok Ásmundar berserkjabana. 1954. in Fornaldar sögur

Norðurlanda iii. edition by guðni Jónsson. [Reykjavík?]: Íslendin-
gasagnaútgáfan. 322–65.
eskeröd, albert. 1947: Årets äring. Etnologiska studier i skördens och julens
tro och sed. nordiska museets handlingar 26. Stockholm: nordiska
museet.
Eyrbyggja saga. 1935. Íslenzk fornrit 4. edition by Matthías Þórdarson
and einar Ól. Sveinsson. Reykjavík: Hið íslenzka fornritafélag.
finnur Jónsson. 1912–15: Den Norsk-islandske skjaldedigtning. 4 volumes.
København: gyldendal.
fisher, Peter and H. R. ellis Davidson. 1979–80: The history of the Danes.
2 volumes. [Saxo grammaticus: Gesta Danorum]. Translated by Peter
fisher, edited by Hilda ellis Davidson. Cambridge: Brewer.
Gísla saga Súrssonar. 1943. in Vestfirðinga sǫgur. Íslenzk fornrit 6. edition
by Björn K. Þórólfsson. Reykjavík: Hið íslenzka fornritafélag. 3–118.
grimm, Jacob. 1843: Kinder- und Hausmärchen. vollständiges ausgabe.
erster Band. leipzig: Philipp Reclam jun.
—. 1953 [1876–78]: Deutsche Mythologie. 3 volumes. Basel: Benno
Schwabe et.co. verlag.
gunnell, Terry. 2006: “How elvish were the Álfar?”. in John McKinnel
et. al. (ed.): The Fantastic in Old Norse / Icelandic Literature.
[Durham]: [University of Durham]. 321–28.
gurevich, elena aronovna. 1969: “Space and Time in the Weltmodell of
the old Scandinavian Peoples”. Mediaeval Scandinavia 2. 42–53.
Göngu-Hrólfs saga. 1954. in Fornaldarsögur Norðurlanda 3.edition by
guðni Jónsson. [Reykjavík?]: Íslendingasagnaútgáfan. 161–280.
Hálfs saga ok Hálfsrekka. 1954. in Fornaldar sögur Norðurlanda ii. edition
by guðni Jónsson. [Reykjavík?]: Íslendingasagnaútgáfan. 93–134.
Hastrup, Kirsten. 1981: “Cosmology and Society in Medieval iceland”.
Ethnologia Scandinavia. 63–78.
—. 1990: Island of anthropology. Studies in past and present Iceland. The
viking collection vol. 5. [odense]: odense University Press.
Heide, eldar. 1997: Fjølsvinnsmål. Ei oversett nøkkelkjelde til nordisk my-
tologi. Unpublished master thesis in old norse studies. Pdf at
http://eldar-heide.net/ -> Kultur- og religionshistorie.
—. 2006: Gand, seid og åndevind. Dr. art. thesis, University of Bergen.

136
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

—. 2011: “Holy islands and the otherworld: Places Beyond Water.” in

gerhard Jaritz and Torstein Jørgensen (ed.): Isolated Islands in Me-
dieval Nature, Culture and Mind. CeU medievalia 14. Budapest /
Bergen: Central european University / Centre for Medieval Studies,
University of Bergen. 57–80.
—. 2013: “Sola og gudane på Tysnesøya”. Chaos. 49–57.
Hervarar saga ok Heiðreks. 1954. in Fornaldar sögur Norðurlanda ii. edi-
tion by guðni Jónsson. [Reykjavík?]: Íslendingasagnaútgáfan. 1–71.
Hœnsa-Þóris saga. 1938. in Borgfirðinga sôgur. Íslenzk fornrit 3. edition
by Sigurður nordal and guðni Jónsson. Reykjavík: Hið Íslenzka
fornritafélag. 1–47.
Holmberg [Harva], Uno. 1925: “vänster hand och motsols”. Rig. 23–36.
—. 1927: Finno-ugric, Siberian [Mythology]. The Mythology of all Races
iv. editited by Canon John arnott et al. Boston: Marshall Jones
Company.
Holmberg, Bente. 1986: “Den hedenske gud Tyr i danske stednavne.” in
vibeke Dalberg and gillian fellows-Jensen (ed.): Mange bække små.
Til John Kousgård Sørensen på tresårsdagen 6.12.1985. navnestudier 27.
København: Reitzel. 109–27.
Hrómundar saga Gripssonar. 1954. in Fornaldar sögur Norðurlanda ii. edi-
tion by guðni Jónsson. [Reykjavík?]: Íslendingasagnaútgáfan. 405–
22.
Hyltén-Cavallius, gunnar olof. 1863–68: Wärend och Wirdarne. Ett
försök i Svensk Ethnologi. 2 volumes. Stockholm: norstedt.
isaksson, Stig. 1958: Skånes ortnamn. Serie A, bebyggelsenamn. Del 14,
Onsjö Härad. Utgivna av Dialekt- och ortnamnsarkivet i lund. lund:
gleerupska universitetsbokhandeln.
Jón Árnason. 1958–61 [1862–64]: Íslenzkar þjóðsögur og ævintýri. Nýtt safn.
6 volumes. Árni Böðvarsson og Bjarni vilhjálmsson önnuðust útgá-
funa. Reykjavík: Bókaútgáfan þjóðsaga.
Jones, T. gwynn. 1930: Welsh folklore and folk-custom. london: Met-
huen.
Jørgensen, lars. 1998: “en storgård fra vikingetid ved Tissø, Sjælland –
en foreløbig presentation.” in lars larsson and Birgitta Hårdh (ed.):
Centrala platser, centrala frågor: samhällsstrukturen under Järnåldern.
acta archaeologica lundensia 8:28. Stockholm: almqvist & Wiksell.
233–48.

137
Eldar Heide

—. 2002. “Kongsgård – kultsted – marked. overvejelser omkring Tis-

søkompleksets struktur og funktion.” in Kristina Jennbert et. al. (ed.):
Plats och praxis. Studier av nordisk förkristen ritual. vägar till Midgård
2. lund: nordic academic Press. 215–47.
—. 2008. “Manor, cult and market at lake Tissø.” in Stefan Brink and
neil S. Price (ed.): The Viking world. The Routledge worlds. london:
Routledge. 77–82.
Ketils saga hængs. 1954. in Fornaldar sögur Norðurlanda ii. edition by
guðni Jónsson. [Reykjavík?]: Íslendingasagnaútgáfan. 149–81.
Kildal, Jens. 1945 [1730 and later]: “afguderiets Dempelse og den Sande
lærdoms fremgang [...].” Nordlands og Troms finner i eldre håndskrif-
ter. nordnorske samlinger 5, 2. edition by Marie Krekling. oslo:
Brøgger. 97–152.
Kildal, Sigvard. 1807 [1730s?]. “nogle observationer om Trommen eller
Runebommen, med en kort Beskrivelse over de hedenske afguder
blandt finner og lapper ...”. Det skandinaviske Litteraturselskabs Skrif-
ter 2/1807. edition by Jan Werner abrahamson. His title: Efterretning
om Finners og Lappers hedenske Religion. 449–475.
Knight, W. f. Jackson. 1958. “The after-life in greek and Roman an-
tiquity”. Folklore 69. 217–36.
Kormáks saga. 1939. in Vatnsdœla saga. Íslenzk fornrit 8. edition by einar
Ólafur Sveinsson. Reykjavík: Hið íslenzka fornritafélag. 201–302.
Landnámabók. 1968. in Íslendingabók. Landnámabók. Íslenzk fornrit 1.
edition by Jakob Benediktsson. Reykjavík: Hið íslenzka fornritaf-
élag. 29–397.
lassen, annette. 2011. Odin på kristent pergament. En teksthistorisk studie.
København: Museum Tusculanum.
lid, nils. 1942. “gudar og gudedyrking.” Religionshistorie. nordisk kul-
tur 26. Stockholm: Bonnier. 80–153.
—. 1946. Light-mother and earth-mother. Studia norvegica 1:4. oslo.
long, Bruce J. 2005. “Underworld.” in lindsay Jones and Mircea eliade
(ed.): The Encyclopedia of Religion 14. 2nd edition. 1st edition 1987, The
Encyclopedia of Religion, editor in chief Mircea eliade. Detroit:
Macmillan/Thomson gale. 9451–58.
lorenz, gottfried. 1984. Gylfaginning. Texte zur forschung. Snorri Stur-
luson. Texte, Übersetzung, Kommentar von gottfried lorenz.
Darmstadt: Wissenschaftliche Buchgesellschaft.

138
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

løkka, nanna. 2010. Steder og landskap i norrøn mytologi. PhD thesis,

University of oslo.
Macculloch, John arnott. 1964. “Celtic Mythology.” in louis Herbert
gray (ed.): Celtic. Slavic. The Mythology of all Races 3. Boston: Mar-
shall Jones Company. 1–213, 333–50, 365–88.
MacKillop, James. 1998. Dictionary of Celtic mythology. oxford: oxford
University Press.
Marteinn H. Sigurðsson. 2009. Tyr: The One-Handed War-God
Toronto: University of Toronto Press
McCluskey, Stephen C. 1990. “gregory of Tours, Monastic Timekeep-
ing, and early Christian attitudes to astronomy”. Isis 81. 9–22.
McKinnell, John. 2005. Meeting the other in Norse myth and legend. Cam-
bridge: D.S. Brewer.
Mebius, Hans. 2003. Bissie. Studier i samisk religionshistoria. Östersund:
Jengel.
Meletinskij, e. 1973a. “Scandinavian Mythology as a System”. Journal of
Symbolic Anthropology 1(1). 43–57.
—. 1973b. “Scandinavian Mythology as a System”. Journal of Symbolic An-
thropology 1(2). 57–78.
Meulengracht Sørensen, Preben. 1986. “Thor’s fishing expedition.” in
gro Steinsland (ed.): Words and objects. Towards a dialogue between ar-
chaeology and history of religion. oslo. 257–78.
Montgomery, James e. 2000. “ibn faḍlān and the Rūsiyyah”. Journal
of Arabic and Islamic Studies 3. 1–25.
Mundal, else. 1991. “forholdet mellom myteinnhald og myteform.” in
gro Steinsland et. al. (ed.): Nordisk hedendom. Et symposium. odense:
odense universitetsforlag. 229–44.
nordberg, andreas. 2003. Krigarna i Odins sal. Dödsföreställningar och
krigarkult i fornnordisk religion. PhD thesis, University of Stockholms.
Nordisk familjebok. Konversationslexikon och realencyklopedi 12. 1888.
Stockholm: gernandts boktryckeri-aktiebolag.
Óláfs saga Tryggvasonar. 1941. in Heimskringla i by Snorri Sturluson. Ís-
lenzk fornrit 26. edition by Bjarni aðalbjarnarson. Reykjavík: Hið
íslenzka fornritafélag. 225–372.
Ólafur Briem. 1985. Eddu-kvæði I–II. With comments. Reykjavík: ve-
röld.

139
Eldar Heide

olsen, isaac. 1910 [etter 1715]. Om lappernes vildfarelser og overtro. Kilde-

skrifter til den lappiske mythologi ii. Det Kgl. norske videnskabers Sels-
kabs skrifter. 1910. no 4. edition by Just Qvigstad. Trondhjem.
olsen, Magnus. 1905. Det gamle norske ønavn Njarðarlǫg. Christiania vi-
denskabs-Selskabs forhandlinger for 1905. no.5. Christiania: i Com-
mission hos Dybwad.
Ordbog over det norrøne prosasprog. Registre. 1989. København: Den arna-
magnæanske kommision.
Patch, Howard Rollin. 1970. The Other World According to Descriptions
in Medieval Literature. new york: octagon.
Perkins, Richard. 2001. Thor the Wind-Raiser and the Eyrarland Image.
viking Society for northern Research text series Xv. london:
viking society for northern research.
Power, Rosemary. 1993. “Þorsteins þáttr bœjarmagns.” in Phillip Pul-
siano (ed.): Medieval Scandinavia. An Encyclopaedia. new york /
london: garland Publishing. 675–76.
Quinn, Judy. 1994. “The Gýgr, the Völva, the Ásynja and her lover:
Meetings with Remarkable Women in the eddic Poems.” in Katrina
Burge and John Stanley Martin (ed.): Treasures of the Elder Tongue.
Fifty years of Old Norse in Melbourne. [Melbourne]: University of
Melbourne. 137–48.
Rage, Knut and Svein ove agdestein. 2007. “Det førhistoriske anlegget
på Todneset – horg, gravkammer eller ein kultplass for soldyrking?”.
Sogeskrift for Tysnes.
Ralston, W.R.S. 1872. The songs of the Russian people as illustrative of
Slavonic mythology and Russian social life. london.
Rietz, Johan ernst. 1862–67. Svenskt dialektlexikon. Ordbok öfver svenska
allmogespråket. lund: gleerup.
Rydving, Håkan. 2004. The end of drum-time. Religious change among the
Lule Saami. 1670s–1740s. Historia religionum 12. 3rd, revised edition.
1st edition PhD thesis 1993, Uppsala University. [Uppsala]: University
of Uppsala.
Rygh, Karl. 1905. Norske Gaardnavne 16. Nordlands amt. Kristiania: fa-
britius.
Røthe, gunnhild. 2010. I Odins tid. Norrøn religion i fornaldersagaene.
Hafrsfjord: Saga bok.
Saga af Þorsteini Bæarmagni. 1827. in Fornmanna sögur 3. Útgefnar ad
tilhlutun hins norræna fornfræða fèlags. Kaupmannahöfn. 175–98.

140
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

Schjødt, Jens Peter. 1990. “Horizontale und vertikale achsen in der vor-
christlichen skandinavischen Kosmologie.” in Tore ahlbäck (ed.): Old
Norse and Finnish religions and cultic place-names. Scripta instituti Don-
neriani aboensis 13. Åbo. 35–57.
—. 1995. “The Ship in old norse Mythology and Religion.” in ole
Crumlin-Pedersen and Birgitte Munch Thye (ed.): The Ship as symbol
in prehistoric and medieval Scandinavia. Publications from the national
Museum. Studies in archaeology & history 1. Copenhagen: national
Museum of Denmark, Department of archaeology and early His-
tory. 20–24.
Simek, Rudolf. 2006. Lexikon der germanischen Mythologie. Kröners Ta-
schenausgabe Band 368. Dritte, völlig überarbeitete auflage. Stutt-
gart: Kröner.
Siuts, Hans. 1911. Jenseitsmotive im deutschen Volksmärchen. Teutonia. ar-
beiten zur germanischen Philologie 19.
Skanke, Hans. 1945 [1728–31]. “epitomes Historiæ Missionis lappo-
nicæ. Pars Prima anlangende de nordske lappers Hedendom og Su-
perstitioner.”. Nordlands og Troms finner i eldre håndskrifter.
nordnorske samlinger 5, 2. edition by o. Solberg. oslo: etnografisk
museum. 179–224.
Solander, Carl. 1910 [1726]. “Kort relation, om den danska Missions tils-
tånd uti norrige [...].” in edgar Reuterskiöld (ed.): Källskrifter till lap-
parnas mytologi. Stockholm: ivar Hæggströms boktryckeri. 18–27.
Steinsland, gro. 2005. Norrøn religion. Myter, riter, samfunn. oslo: Pax.
Steinsland, gro and Kari vogt. 1981. “aukinn ertu Uolse ok vpp vm te-
kinn”. Arkiv för nordisk filologi 96. 87–106.
Straubergs, K[arlis]. 1957. “Zur Jenseitstopographie”. Arv. Nordic Year-
book of Folklore 13. 56–110.
Sydow, C. W. von. 1926. “Det ovanligas betydelse i tro och sed”. Folk-
minnen och folktankar. 53–70.
Þorsteins saga Víkingssonar. 1954. in Fornaldarsögur Norðurlanda 3. edi-
tion by guðni Jónsson. [Reykjavík?]: Íslendingasagnaútgáfan. 1–73.
Þorsteins þáttr uxafóts. 1991. in Harðar saga. Íslenzk fornrit 13. edition
by Bjarni vilhjálmsson and Þórhallur vilmundarson. Reykjavík: Hið
Íslenzka fornritafélag. 339–70.
vikstrand, Per. 2006. “Ásgarðr, Miðgarðr, and Útgarðr. a linguistic ap-
proach to a classical problem.” in anders andrén et. al. (ed.): Old
Norse religion in long-term perspectives. Origins, changes, and interactions.

141
Eldar Heide

An international conference in Lund, Sweden, June 3–7. 2004. vägar till

Midgård 8. lund: nordic academic Press. 354–57.
Weisweiler, Josef. 1940. “Seele und See. ein etymologischer versuch”.
Indo-Germanische Forschungen 57. 25–55.
Wellendorf, Jonas. 2006. “Homogenity and heterogenity in old norse
cosmology.” in anders andrén et. al. (ed.): Old Norse religion in long-
term perspectives. Origins, changes, and interactions. vägar till Midgård
8. lund: nordic academic Press. 50–53.
Widengren, geo. 1953. Religionens värld. Religionsfenomenologiska studier
och översikter. 2. omarbetade och utökade upplagan. Stockholm:
Svenska Kyrkans Diakonistyrelses Bokförlag.
Wiklund, K.B. 1916. “Saivo. Till frågan om de nordiska beståndsdelarna
i lapparnas religion”. Le monde oriental. 45–74.

Samandrag

Denne artikkelen presenterer ein ny måte forstå kosmologien i førkristen

nordisk mytologi på. Kjeldene ser ut til å gje motstridande opplysningar;
til dømes blir æsene plasserte på ulike stader: i midten av verda, i vest, i
aust, under havet og i himmelen; Hel blir plassert under bakken og bort-
om havet. På grunn av slike ting har ein i nyare granskingar konkludert
med at det ikkje er noko system. Denne artikkelen argumenterer for at
det er det og at vi misforstår passasjane til andre verder. Den andre verda
kan definerast som ‘verda bortanfor det vi har tilgang til med naturlege
middel’. Utgangspunktet er sfæren som var tilgjengeleg for menneska
før moderne teknologi og som har form som ei kjempestor, flattrykt
boble: Bortover har vi svært lang rekkevidde, heile sirkelen rundt, men
nedover i jord, berg og vatn, og oppover i lufta, har vi svært kort rekke-
vidde. Til alle tider har folk likevel tenkt seg at denne avgrensinga ikkje
er absolutt; det finst noko bortanfor, men berre overnaturlege makter og
skapningar og trolldomskunnige menneske kan fara dit eller koma deri-
frå, gjennom visse passasjar. Dei peikar i mange retningar ut frå ‘bobla’,
men passasjane er ikkje det same som Den andre verda, for dei kan skif-
tast ut med kvarandre og fører ofte til same slag stader. Dei andre verdene
har kontaktflater med denne verda og med kvarandre, men ligg ikkje på
nokon stad som kan plasserast i høve til denne verda eller andre. Dei er
berre ‘på den andre sida’, bortom passasjane. Den grunnleggjande mo-

142
 ConTRaDiCToRy CoSMology in olD noRSe MyTH anD Religion

dellen ser dermed ut til å vera ein motsetnad mellom ‘oss’ og ‘dei andre’,
med ei stadig skiftande grense. Dette systemet kjem til syne når vi un-
dersøker ikkje berre dei avgrensa norrøne tekstane, men ser dei i lys av
det kolossale korpuset av eventyr og segner nedskrivne i nord-europa
etter mellomalderen.

eldar Heide
associate professor in norwegian
Bergen University College
P.o. Box 7030
no-5020 Bergen
norway
eldar.heide@hib.no

143