Académique Documents
Professionnel Documents
Culture Documents
Series Editors
Volume 150
By
Olof Sundqvist
LEIDEN | BOSTON
Sundqvist, Olof.
An arena for higher powers : ceremonial buildings and religious strategies for rulership in late Iron Age
Scandinavia / by Olof Sundqvist.
pages cm. — (Numen book series : studies in the history of religions : ISSN 0169-8834 ; Volume 150)
Includes bibliographical references and index.
ISBN 978-90-04-29270-3 (hardback : alk. paper) — ISBN 978-90-04-30748-3 (e-book : alk. paper)
1. Scandinavia—Religion. 2. Religion and politics—Scandinavia—History. 3. Scandinavia—Kings and
rulers—Religious life. 4. Scandinavia—Antiquities. I. Title.
BL863.S85 2015
293—dc23
2015034775
issn 0169-8834
isbn 978-90-04-29270-3 (hardback)
isbn 978-90-04-30748-3 (e-book)
Preface ix
List of Figures xi
Abbreviations xiii
Part 1
Prolegomena
1 Introduction 3
1.1 Religious Ruler Ideology in Previous Research 7
1.2 Towards a Model for Analysing Religious Strategies for Rulership at
Ceremonial Buildings 14
1.3 Approach and Methods 20
1.4 Outline and Hypotheses 29
Part 2
The Regional-Contextual Approach
Part 3
General Features
Part 4
Regional Diversity
Part 5
Epilegomena
16 Conclusions 539
Bibliography 547
Index 617
The original aim of the present research project, which was financed by
Riksbankens Jubileumsfond in Sweden (2008–2011), was to investigate the pre-
Christian and medieval cultic buildings in Scandinavia from an ideological
perspective. The specific purpose was to study how rulers used these buildings
to express and reinforce their political power. The study involved a long-term
perspective, AD 550–1200. The purpose of this long-term perspective was to
investigate continuity and change in the context of the conversion from the
pre-Christian halls and cultic buildings to their replacement by churches in
Scandinavia.
The project followed to a great extent the initial project plan, but some limi-
tations were defined as work progressed. The empirical study was delimited to
the Lake Mälaren area, Trøndelag and Iceland. The project also shifted focus
somewhat. In the present monograph, the regional perspective has gained
greater weight than the diachronic dimensions of the study. Focus has been
placed on the pre-Christian contexts, specifically on the issue of regional simi-
larities and differences/variations. The diachronic dimension of the study, i.e.
the development from pre-Christian to Christian contexts, was published in an
article at Brill 2011, with focus placed on the Mälaren region.1
I am indebted to many scholars for giving me valuable advice and construc-
tive criticism during the preparation of the present book. First and foremost
I express my gratitude to Professor Bo Gräslund (Uppsala), who has been
standing by my side from the first manuscript of this work to the final produc-
tion. Also my former “Doktor Vater” Professor Anders Hultgård (Uppsala) has
offered me a great deal of good advice and encouragement, as well as my friend
and colleague Professor Michael Stausberg (Bergen), who also convinced me to
publish the results of my project in an international publication. My dear col-
leagues, Associate Professor Torun Zachrisson (Stockholm), Professor Anne-
Sofie Gräslund (Uppsala), Associate Professor John Ljungkvist (Uppsala) and
Associate Professor Per Vikstrand (Uppsala), have all contributed with valu-
able advice and comments, particular regarding archaeology and onomastics.
They have indicated additional materials, described excavation sites, answered
questions and discussed problems. I am also indebted to some other schol-
ars who also have generously shared their profound knowledge, especially
when working at the Centre for Advanced Study (CAS) in Oslo (2007–2008), but
also later; in particular Professor Gro Steinsland (Oslo), Professor Jon Viðar
Sigurðsson (Oslo) and Professor Jens Peter Schjødt (Aarhus). Gro and Jon Viðar
gave valuable comments on the entire manuscript at an earlier stage. I am
very thankful for that. Some parts of this text have also been discussed in the
higher seminars at the University of Gävle (with valuable comments from e.g.
Professor Peder Thalén and Professor Jari Ristiniemi) as well as in the higher
seminars at Stockholm University (with good advice and reflections from
e.g. Professor Peter Jackson, Associate Professor Andreas Nordberg, Associate
Professor Erik af Edholm, Associate Professor Marja-Liisa Keinänen; Asso-
ciate Professor Susanne Olsson; Doctor Ferdinando Sardella; Doctor Stefan
Larsson and Doctor Niklas Foxeus). I am also grateful to Doctor Merete Røskaft
(Trondheim), who showed me some of the central places and cultic sites when
travelling in Trøndelag, Alan Crozier (Lund) for correcting my English, Maarten
Frieswijk (Leiden) and Judy Pereira (Boston) for editing the text, and Professor
Anders Andrén (Stockholm) for providing me with photos. Many thanks also
to Professor Neil Price (Uppsala) and Professor Terry Gunnell (Reykjavík) for
answering my questions. For all shortcomings and faults that still may remain
in the text I am alone responsible.
The project (and partly the printing of the book) was made possible by
grants from Riksbankens Jubileumsfond, while language correction of the
manuscript was financed by the University of Gävle. Their support is hereby
gratefully acknowledged. I am also grateful to Brill, for publishing my work.
Olof Sundqvist
Fårösund, July 2015
16b Detail of the picture stone from Lärbro (Hammars I). Photo: Olof
Sundqvist 335
17a The ring from Forsa in Hälsingland, Sweden. Photo: Bengt A Lundberg.
ATA, Riksantikvarieämbetet 378
17b Detail from the ring from Forsa. Photo: Bengt A Lundberg. ATA,
Riksantikvarie-ämbetet 378
18 The door ring from Delsbo church, Hälsingland, Sweden. Photo:
Unknown. ATA, Riksantikvarieämbetet 381
19a Þórr’s hammer made of silver from Läby, Uppland. SHM 385
19b Þórr’s hammer from Mickels, När, Gotland. SHM 385
20 Fire-steel-shaped pendants from Borg, Östergötland. SHM 385
21 The Häckelsäng ring, found in Hamrånge parish in Gästrikland.
Gävleborgs museum 397
22 A picture stone from Tängelgårda, Gotland. SHM 399
23a Gold foil figures from Hauge-Tu, Rogaland, Norway. The Museum in
Stavanger 410
23b Gold foil from Krokek, Östergötland. SHM 410
24a and 24b Iron door rings from Uppåkra, Scania, Sweden. Photo: Lunds
universitets Historiska museum 423
25 The Sparlösa stone (Vg 119), Västergötland. Photo: Bengt A Lundberg ATA,
Riksantikvarieämbetet 424
26 The small figurine from Kymbo, Västergötland. SHM 428
27a The three royal burial mounds from Old Uppsala. Photo: Marie-Louise
Sundqvist 434
27b The burial mound at Alstadhaug, Trøndelag. Photo: Olof Sundqvist 435
27c The burial mounds at Bertnem, Trøndelag. Photo: Olof Sundqvist 435
28 The distribution of theophoric place-names in the Mälaren region. Made
by Fideli Sundqvist and based on Per Vikstrand 2001 517
Introduction
The great Icelandic writer and chieftain Snorri Sturluson (1179–1241) narrates
in his Ynglinga saga (a part of his masterpiece Heimskringla, c. 1230) the story
of King Ingjaldr of Svetjud and the funeral feast (erfi) he gave for his father
King Ǫnundr at the pre-Christian royal site of Uppsala in present-day Sweden.1
According to Snorri, Ingjaldr prepared a hall in no wise smaller or less stately
than the hall called Uppsalr, and called it “the Hall of Seven Kings”. In this
hall seven high-seats were erected. Then the king sent messengers through all
of Svetjud, inviting seven kings, earls and other prominent men. All but one of
the kings arrived at Uppsala, where they were given seats in the new hall.
Snorri than describes in general terms the rituals usually performed at a
royal funeral feast in pre-Christian contexts. It was the custom, for instance,
when a funeral feast was prepared to honour a departed king or earl, that the
one who prepared the feast and was to be inducted into the inheritance was
to sit on the step before the high-seat until the beaker called the bragafull was
brought in; and then he was to stand up to receive it and make a vow, then
quaff the beaker, whereupon he was to be inducted in the high-seat which his
father had occupied. Then he had come into his rightful inheritance as his suc-
cessor. This duly took place here; and when the beaker was brought in, King
Ingjaldr stood up, seized a large drinking horn, and made the vow that he
would increase his dominion to twice its size in every direction, or else die.
Then he emptied the beaker. Later the same night Ingjaldr said to his retain-
ers that they should bring fire into the new hall and set it aflame. There the six
kings burned and all their troop, and those who came out were quickly killed.
After that, King Ingjaldr put under himself all those realms that the six kings
had had, and he took tribute from them.
This text cannot be regarded as a source for historical reconstruction in a
general sense. Several fictitious, or literary elements may be at stake in this
account, for instance, Ingjaldr’s project of making a kingdom under a single
monarch by means of burning the other district kings in the new hall. Actually
we do not really know whether King Ingjaldr ever lived in Svetjud in ancient
times, or if he held a funeral feast for his father Ǫnundr at Uppsala. Anyhow,
there are some details in the text which seem to be based on more solid
1 Ynglinga saga 36, Heimskringla I, Ísl. Fornr. 26, 66–67. This survey is based on Hollander’s
translation. This text will be quoted in full and analysed in detail in ch. 13.
in Late Iron Age Scandinavia. They will not primarily be focused on in this
study.
The present investigation focuses on the following questions:
– who controlled, owned and managed the halls, cultic houses and sanctuar-
ies, and what relations did they have with the ruling elites in the three areas?
– where were the ceremonial buildings and sanctuaries located and what
did they look like, and in what sense can they be connected to the leading
groups in society?
– which ritual objects found, and ritual activities deduced, in and outside
these buildings, may have had a connection to the religious strategies for
legitimacy, authorization and empowerment?
onstrate and manifest their power, in both a concrete and a symbolic sense,
and as a place where they took tribute from their subjects. Since the political
structure and other contextual aspects partly differed in the three regions, the
religious strategies for gaining legitimacy and authorization at the sanctuaries
also varied to some extent in these areas. Conditions in the Mälaren area and
Trøndelag resembled each other more than circumstances in Iceland.
The present work is to a considerable degree based on textual studies.
Since the questions posed in this study concern Late Iron Age ceremonial
buildings and material aspects of ruler ideology, archaeological sources and
place-names are equally important source materials. In what follows, previous
research on religious ruler ideology in Scandinavia will be surveyed (i.e.
the theory of sacral kingship), also noting some methodological problems
and theoretical issues. I will then present my own perspectives, approaches,
essential concepts and hypotheses. The concept of “religious strategies for
rulership” will be discussed below. A thorough presentation of the pre-
Christian ceremonial buildings, their history of research, and the terminology,
will be given in ch. 4, as well as the evidence of them in the sources referring
to the investigation areas.
12 Schück 1904.
13 See e.g., Schück 1904; von Friesen 1932–34; F. Ström 1954.
14 Grønbech 1997 (1909–12).
15 See e.g., Kern 1954 (1914), 3–45; Schramm 1937, 14, 28, 140–178; Hauck 1950; Höfler 1952;
1959; Hoffmann 1976, 1–22; de Vries 1956–57, §§242, 274–276. On this theory, see also ch. 13
below.
16 Åke V. Ström (1959), for instance, stated thus: “That kingship in Old Scandinavia was
entirely sacral, is nowadays considered as a matter of fact.”
17 E.g. Kuhn 1971. Cf. Lönnroth 1963–64; 1968; 1986.
1.1.4 The Static and Abstract Character, and the Method of Using
Texts Only
In early studies, the pattern of the sacral kingship concept was thus more
or less imposed upon the Scandinavian material. Scholars looked almost
22 Schück 1904, 248–306; von Friesen 1932–34; F. Ström 1954, 7ff., 32ff.
23 Steinsland 1991, passim; 2000, 57ff.; 2011b, 16f. See also her new criticism against the “sacral
kingship theory” in Steinsland 2011a, 16, note 3.
24 Steinsland 1991, 310ff.; 2011b, 57.
25 Näsström 1995, 200.
26 E.g. Phillpotts 1920; F. Ström 1954.
desperately for evidence to fill out the pattern, which was sometimes bor-
rowed from totally different areas and periods. Because of this wide synchro-
nous method, they often made broad generalizations based on weak data and
they also overlooked possible regional and temporal variations not only in
Scandinavia, but even in all Germanic areas. Their method was built on a tacit
theoretical assumption that religion, ideology and culture were homogeneous
over the entire Scandinavian or Germanic area. It also seems as if they felt
that time had no influence on the development of the ruler ideology. When,
for instance, Folke Ström only found evidence of the “pagan notion” of the
king’s luck (Königsheil) in the medieval Scandinavian prose (and slight evi-
dence in skaldic poetry), he supplied this lack of data by bringing attestations
from Ammianus Marcellinus’ text about the Burgundian rulers written in the
fourth century, which actually had a quite different context from the former
materials.27 This wide synchronous method was repeated by Åke V. Ström
on the same issue, where he brought together information from Ammianus
Marcellinus with some utterances made by King Gustav I (Vasa) at the “Riksdag”
in Västerås 1527, recorded by Peder Swart.28 In this case too, the contexts of
the sources were quite different. It thus seems as if the application of the
“sacral model” to the Germanic sources has sometimes resulted in reconstruc-
tions which totally overlook possible regional and diachronic variations. Folke
Ström’s and Åke V. Ström’s approaches to the notion of king’s luck are good
examples here. It is also typical that scholars who applied the sacral model
mainly investigated written sources, and often disregarded other types of
27 Ammianus Marcellinus XXVIII 5,14. F. Ström 1968. According to Baetke (1964) and
Lönnroth (1963–64) the notion of a pre-Christian king’s luck in the Scandinavian sources
was affected by the medieval Christian notion of the king’s divine grace (rex dei gratia)
or “charisma”. Criticism of this position is voiced by Hallberg 1966; 1973. See further ch. 12
below. It should be noticed that Folke Ström in previous investigations sometimes
took regional variations into account. In, for instance, his Diser, nornor, valkyrjor:
fruktbarhetskult och sakralt kungadöme i Norden (1954) he made a distinction between the
cult of dísir in Western and Eastern Scandinavia. In this study he nevertheless interprets
the skaldic poem Ynglingatal with support of the universal theory of sacral kingship and
thus reproduced “the general pattern” of this institution in his interpretation of this lay.
Ström was actually aware about the problem of continuity, change and regional variation
within the Norse religion. In his handbook Nordisk hedendom, for instance, he stated that
Norse religion, may have varied over time and space, and that we are facing a serious
problem because of the uneven distribution of the literary sources, which almost all have
a West-Nordic background. F. Ström 1985, 7f. Cf. Nordberg 2012, 126f. In my opinion, Ström
was anyhow one of those scholar who overlooked possible local and temporal variations
related to ruler ideology, by using unilaterally the pattern of the sacral kingship theory.
28 Å. V. Ström 1975, 268f.
sources. This fact also contributed to the static character of the reconstruc-
tions, since the written materials alone do not allow chronological analyses of
pre-Christian conditions in Scandinavia.
In the present study I will argue that the sources indicate that the religious
strategies for authorization and empowerment at ceremonial buildings in Late
Iron Age Scandinavia may have varied in different regions and contexts.
29 See e.g., Jón Viðar Sigurðsson 2008a, 22–31; Krag 2008, 645–651; Lindkvist 2008; P. H.
Sawyer 1991, 3–10.
30 See, however, Steinsland 2011a; 2011b; Jón Viðar Sigurðsson 2011; Sundqvist 2011a; 2012b.
31 See e.g., Steinsland 1985a; 1985b; Enright 1996; Näsström 2001b; 2009; N. Price 2002;
Raudvere 2003; Dillmann 2006; Sundqvist 2007.
instance, that queens and other female leaders (ON húsfru, gyðja, vǫlva) may
have played important ritual roles during cultic feasts in the aristocratic halls.
Thus these females too used religious strategies to gain authority. These power-
ful women may be attested notably in the archaeological materials,32 but they
seem to appear in other types of sources as well.33 In chapters 5 and 9 I will
argue that female cultic leaders played a significant ritual and ideological role
within the contexts of halls, cultic buildings and sanctuaries.
essential ingredients in the “sacral kingship package”, i.e. the Romantic theo-
ries associated with Frazer and Mannhardt.36 The concept has thus in certain
contexts developed into a communicative obstacle.
There is thus a clear need to reconsider the entire issue of rulers, religion and
ideology in ancient Scandinavia. The application of the sacral kingship con-
cept to the Scandinavian context suffers from a number of methodological
flaws, and its “patternism” has sometimes led to a disregard of the evidence of
the sources themselves. Hence, several reorientations must be made. Firstly,
when using the comparative approach to this theme, not only similarities but
also differences or variations should be sought.37 We must look for unique
aspects of the religious ruler ideology and not only discuss the general features
which have been associated with the sacral model. The comparisons should
also be based on historical studies of primary sources. Secondly, we must go
beyond the intrinsic static character of the model, in order to take temporal
and regional variations into consideration.38 Not only texts but also other
types of sources must be applied in this enterprise. Thirdly, since the formation
of “sacral kingship” or ruler ideology concerns the relationship between reli-
gion and political structures, decontextualized analysis must be avoided. This
implies an increased demand for general cultural competence on the part of
the scholar.
Even if the phenomenological category of “sacral kingship” in many senses is
problematic, it has at least been useful as a heuristic instrument, for instance,
when framing a research area and posing questions to the materials. If this
concept is to be dismissed, it must be replaced by another analytic concept
which is more adjusted to conditions in pre-Christian Scandinavia, and also
to the object of the present study, i.e. the ruler-ideological dimensions of cer-
emonial buildings and sanctuaries.39 In my opinion an alternative term could
be proposed, namely “religious strategy/strategies for rulership, (or politi-
cal authorization)”.40 This concept can be applied to different social, gender
and religious contexts. It can, for instance, be applied to a chieftain (hǫfðingi,
hersir, goði) in a local settlement district, as well as to an earl ( jarl) or a king
(konungr) ruling over people in a regional dominion, or even to a “federation
der jeweiligen Ausprägung des Königstums in den einzelnen Kulturen und fordert ein
bei weitem feineres Differenzierungssensorium sowohl für den interkulturellen Vergleich
als auch für Entwicklungsstufen innerhalb ein und desselben Kulturraums.” Ahn 1992, 18.
39 Alternative concepts to “sacral kingship” have been presented in previous research on
pre-Christian Scandinavian religion. The concept of ruler ideology, for instance, has been
discussed in this context. See Steinsland 2011a. Cf. Jón Viðar Sigurðsson 2011, 70. In the
research project “Ideology and Power in Viking and Middle Ages” at CAS in Oslo 2007–2008,
led by Steinsland, where also I took part, the concept of “religious ideology of rulership” was
often applied. See also Sundqvist 2011a. Steinsland (2011a, 4) defined ideology as a “system
of thoughts which expresses a world-concept and which permeates culture as a whole”.
The concept referred to “the intentional use of myths, rites or narratives by individual
groups with the aim to legitimise a certain type of rulership” (Steinsland 2011a, 4). For good
overviews of the concept of ideology, see e.g., Geertz 1993 (1973); Heywood 1998; Freeden
1996; 2003; Žižek 2012 (1994). See also the articles in Žižek (ed.) 2012 (1994), Mapping
Ideology. I will not use “ruler ideology” as a central concept or as a kind of super structure
in the present study. Instead I will apply the term “religious strategy/-ies for rulership” since
it fits my regional and contextual perspectives well. See also Sundqvist 2012b, 234–236. The
concept of “ideologic strategies” has been used in political science. See e.g., Schweitzer 1962.
On the concept “strategy”, see further below.
40 In present study the concept “rulership” refers to “a position in which one rules or has
sovereignty over others”, while the term “ruler” refers to “a person who rules or governs;
a person who exercises dominion or controlling power over others”. A ruler could appear
on different social levels and be of different genders. It could be a king, a queen, an earl, a
local chieftain, or a powerful female etc. See Sundqvist 2002, 11.
king”, i.e. an overlord or a “king of kings”, who controlled several groups living
in a wide area. It can also be used when investigating ideological aspects of
powerful women (dróttning, húsfrú, gyðja), appearing on different social levels
in society. Another advantage of this concept is that it can easily be used in the
plural. Thus in one and the same cultural context, such as ancient Scandinavia,
we may speak about different religious strategies for rulership. Most important
for the present study is that this concept is more suitable when applied to the
theme of ceremonial buildings than the term “sacral kingship”.
The noun authorization is related to the term authority (see below). It refers
to the action or fact of authorizing or being authorized. In addition, the related
term ruler-empowerment will occasionally be applied as well as the concept of
legitimacy. The latter concept was coined as a scientific tool by the sociologist
Max Weber and has since played an important role in political sociology and
political science.41 Scholarly treatments have often emphasized the subjective
aspect of legitimacy; for there is no such thing as objective legitimacy. Whether
the authority, the institution, the action and so forth are to be considered legit-
imate always depends on the beliefs of the acting subjects. Legitimacy may
be derived from different sources, such as law, charisma, tradition, history,
democracy, religion, etc. These sources of legitimacy may be widely shared by
different groups in a society as common values and qualities, but they may also
lead to conflicts.
Religion may legitimate non-religious as well as religious actions and insti-
tutions. For instance, the ruler’s divine character may legitimate both his role
as military commander and his function as an officiant in the public cult.
Closely connected to the concept of “legitimacy” is the concept of “authority”.
In previous scholarly traditions authority was considered as an entity. Weber
introduced an ideal typology with three categories: traditional, charismatic
and legal-rational authority.42 In agreement with later generations of scholars I
view authority from a more discursive and contextual perspective.43 Authority
is always established by means of the premises that are prevalent in a spe-
cific context. It may be achieved by means of law, charisma, tradition, history,
democracy (see above), but also in many other ways.
When using the concept of strategy I refer to a more or less deliberate plan
of action or policy designed to achieve a major or overall aim.44 Tentatively
41 Weber 1971 (1922). See also Adams & Mikelson 1987; Kehrer 1998.
42 Weber 1971 (1922). Cf. Waida 1987.
43 See e.g., Bloch 1989; Lincoln 1994.
44 Such strategies are of course embedded into the ruler’s understanding and interpretation
of the situation. The assessment of the ruler’s agency was also embedded in their
there are four possible religious strategies for raising the political leaders
above common people, gaining legitimacy and political authority appearing
at the ceremonial buildings. It should be noted that there are no watertight
bulkheads between the strategies; they actually overlap each other. They could
appear in different combinations at different contexts. When the individual
strategies appear in combinations they may actually support each other in
order to achieve the major aim, i.e. to gain legitimacy, political authority or
empowerment:
(1) By means of the specific relation to the mythic world. Legitimacy and/or polit-
ical authority may be achieved by means of the ruler’s close relation to the
mythic world. The character of this relationship varies in different contexts.
The following are some types of relations that may be tentatively discerned,
although it is possible that additional types of relations exist: (a) the ruler may
be considered divine by his or her own means and power; (b) the ruler may be
seen as an incarnation of the deity or the mythic being, and represent him/
her in ritual contexts. In such cases the ruler may be worshipped as a mythical
being, although his/her power actually derives from the divinity or the mythi-
cal being; (c) the ruler may be considered as a descendant of a mythical being;
(d) the ruler may be the deputy or delegate of the divinity on earth, i.e. just a
weaker variant of types a–c; (e) the ruler may be the most important mediator
in the relationship between the human and the mythical worlds, i.e. he or she
may act as the (highest) cult leader (see below). As noted, the list consists of
a hierarchy, where the “sacredness” of the ruler seems to be more intensified
in the upper half (a–c), while the lower part (d–e) reflects relations which put
less stress on the “divine status” of the ruler.45 The main purposes of the ruler’s
close relation to the mythic world are in all cases to mark him/her off from the
rest of the community and to give his/her authority a divine sanction.
The ruler’s close relation to the mythic world may be expressed in sev-
eral senses at ceremonial buildings and sanctuaries. Certain narratives pro-
claiming these ideas, such as myths or legends, may be recited under ritual
observation, for instance by cultic leaders in these buildings and places. Such
performances probably had a striking effect on the audience. There were also
other ways of communicating these ideological notions. The seat of the ruler
in the ceremonial building, for instance, may be construed as an interface with
cosmology or religious worldview. In some contexts the rulers may have been more
tactical than strategical, when aiming at short-term aims, or when having to improvise for
example when facing situations of crisis.
45 For a similar typology, see Köhler 2004.
the Other World, i.e. a liminal and ritual place intended for people who had a
specific relation to the divinities. There may also be other ritual objects (e.g.,
cultic images) or structures, appearing either inside or outside these types of
buildings, which may indicate that the ruling family had a certain relation-
ship to the divine world. The sanctuary including the ruler’s seat, may, for
instance, be deliberately constructed as a reflection of the cosmos, or arranged
in accordance with the mythical topography. Thus the sanctuary emphasizes
that the ruler was situated close to the centre of the world (axis mundi), per-
haps even inside the mythical world, where divine beings dwelled and could
be encountered.
(2) By performing central roles in (religious) rituals. The ruler’s legitimate right
to exert power in different fields of society is also related to his/her roles in
cults, (religious) rituals and ceremonies. These roles, as noted above, may be
connected to his/her specific relation to the mythic world. Since the ruler is
the mediator between the human and the mythic world, he or she is also the
one who must perform the sacrifices and so on. The rituals of the ruler can
be described as “political rituals”, i.e. they comprise those ceremonial prac-
tices that specifically display and promote the power of the ruler.46 They do
not simply give form to the power, they actually construct it, and orchestrate
the ruler’s position in society and the cosmos. Coronations, consecrations and
inaugurations, for instance, are important rituals for such constructions. These
types of performative actions often take place in a ceremonial building at the
cultic site. Most likely the properties of such “sacred rooms”, and also the archi-
tecture of the rooms/places themselves, contribute to creating the dignity of
these “political rituals”. The sanctuary is thus the perfect arena and setting for
such activities, which may include processions, ambulatory rituals, drinking
ceremonies, as well as gift-giving ceremonies and the like.
(3) By using (religious) symbols. In the inauguration ritual the ruler may receive
certain (religious) symbols which have a similar intention to the rituals of the
ruler, i.e. to raise the ruler above his/her fellows, and/or to relate him/her to
the mythical world. These symbols of power may consist of certain clothes,
insignia and other paraphernalia. They may also be constituted by specific geo-
graphic places and sites, cultic buildings and halls, ritual structures and seats,
monumental graves and so on. Ceremonial buildings, for instance, are some-
times located (symbolically) at elevated places in the landscape, indicating
dignity and sovereignty, but also a position in the universe. The monumental
architecture of these buildings, the interiors, including the ritual objects, can
thus be concrete or symbolic expressions of power. Like the “political rituals”
these buildings likewise have the function of displaying and demonstrating the
political power, and they construe borders between the ruler and the rest of
the society.47 These symbols attempt to demonstrate that the values and forms
of social organization are naturally derived from the way the world or the cos-
mos is organized. Therefore symbols (and rituals) are more effective than coer-
cive force when securing a particular social order and hierarchy.
(4) By controlling the cultic organization or institution. In order to carry out the
strategies shown above (1–3) efficiently, the ruler must control the cultic organ-
ization in different ways. The ruler may, for instance, be considered as the
owner or the manager of the cultic sites and ceremonial buildings. Thus his
or her involvement is necessary for performing the public cults. He or she may
also be the one who controls and organizes the cultic leaders at these sites,
and sometimes he or she may even be regarded as the highest cultic leader. In
such cases the ruler may have a clear view of the activities taking place during
public gatherings, and give priority to ones that may be advantageous to the
political power. The ruler may also, in different ways, encourage or even force
the members of the cultic community to take part in the common religious
feasts at the sanctuaries. The organization of cult may also be seen as one of
the ruler’s duties. He/she may only be legitimate as long as he/she maintains
and secures the cultic activities.
47 Bell 1997, 129–135, 155–159; Gilhus & Mikaelsson 2003, 132ff.
52
The reason I have chosen not to take southern Scandinavia and Denmark into
consideration in this study is twofold: (a) there are few written sources informing about
ceremonial buildings from this area; (b) this study needs some kind of delimitations.
Since much interesting archaeological evidence referring to ceremonial buildings has
been found in southern Scandinavia, I will nonetheless use these sources as comparative
materials.
53 On the problems of “insiders” and “outsiders” in the study of religion, see Riesebrodt 2007
and Stausberg 2009.
54 See Hultgård 1996, 25–28; 2003b, 436–37. Jens Peter Schjødt (2012, 263–265) uses the
concepts “contemporary” and “medieval sources”. My division of sources is actually a
simplification of the actual case. There are sources which fit badly into this categorization.
The Eddic poems, for instance, could perhaps be classified as a third group, since the
dating of them often is so uncertain. For good and more detailed overviews of the sources,
see e.g., Steinsland 2005, 35–66 and Schjødt 2008, 85–107. Cf. Sundqvist 2002, 39–62.
55 Meulengracht Sørensen 1991a, 225.
56 Mundal 2008, 1.
high source value. Also most of the Eddic poetry must be considered as direct
sources, even if there are some exceptions.57 It must essentially be seen as
memorised oral literature, “although it must have been more open for improvi-
sation and changes than skaldic poetry”.58 Most of the Eddic poems have thus
been changed during the oral transmission, but their fundamental content
may have been mostly pre-Christian.
Place-names and runic inscriptions are further direct sources, since they
derive directly from the Late Iron Age society. They give information about the
old rituals and belief systems from an insider’s perspective. However, scholars
working with these source categories must also face some serious source criti-
cal problems.59 Anyhow, these sources contain linguistic materials that some-
times have escaped distortion by secondary traditions.
The indirect sources are made up of prose narratives and descriptions trans-
mitted by people who were not adherents of the ancient religion. Their authors
lived outside the Viking Age society and their writings give only indirect infor-
mation about the ancient Scandinavian cult system. Oral traditions appear-
ing in these sources have often been reworked by rhetorical embellishment,
artistic inspiration, ethnographic clichés, and subjectivity. Most of the sources
belong to this category. They comprise the Sagas of Icelanders from the Middle
Ages, Kings’ Sagas, medieval laws from Sweden, Norway and Iceland, and eccle-
siastical texts in Latin, for example, written by Adam of Bremen and Rimbert,
Arabic reports, and finally as historiographical accounts from the first century
AD to the Middle Ages, such as those of Tacitus and Jordanes.
The mythical and historical writings of Snorri Sturluson (1179–1241) must
also be considered as indirect sources, since they were written by a Christian
more than two hundred years after the official conversion of Iceland.60 Snorri’s
texts have therefore been felt to be almost useless when discussing ancient
Scandinavian religion and society.61 This type of source-critical approach has
deep roots in the study of Old Norse philology and religion as well as in the
general study of history.62 Sophus Bugge was one of the first to claim that
the Old Norse myths and heroic narratives transmitted by Snorri were
57 Individual Eddic poems may have a medieval background, but many of them were
probably composed during the Viking Age. See Mundal 2004, 222.
58 Mundal 2008, 2. Cf. Clunies Ross 1994, 20–33.
59 See e.g., Vikstrand 2001; McKinnell & Simek 2004; Düwel 2008.
60 On Snorri’s person and authorship, see e.g., Vésteinn Ólason 2008.
61 For example Baetke 1964.
62 See e.g., Lauritz Weibull 1948; Curt Weibull 1964; Baetke 1950; 1951; 1964; Holtsmark 1964;
Olsen 1966; Düwel 1985.
traditions which reach back to the Viking Age.70 Some traditions have been
reworked by Snorri and other medieval writers. This reworking is no reason
to dismiss them altogether. Snorri and the medieval writers probably had
access to more direct sources than the modern historian has, and most likely
they were better suited to interpret them.71 Even if Snorri’s texts and the sagas
themselves are dated to the thirteenth century, their meaning and content may
be much older.
By using a comparative method which combines source criticism and phi-
lology, ancient structures may be detected in the materials—structures which
can be interpreted by setting them in a wider context of sources—and devel-
oped alongside knowledge which we have from the auxiliary disciplines, that
is, archaeology, runology and toponomastics.72 When such comparison is not
possible (due to scarcity of direct sources) the absence of similar ideas and
customs in contemporary Christian culture is here taken as a support for the
reliability of medieval statements on ancient Scandinavian tradition.
70 On the transition process from oral tradition to literacy in Old Norse society, see e.g.,
Clunies Ross 1994; Quinn 2000; Gisli Sigurðsson 2004; Brink 2005; Mundal 2008; T. M.
Andersson 2008; Lincoln 2012 and 2014.
71 Preben Meulengracht Sørensen (1991b, 244) made the following observations when
considering Snorri’s narrative about Hákon the Good and sacrifices in Trøndelag: “History
is not the data extracted from the sources, but the syntheses created out of the data, and
historical writings are not the reestablishment of past reality, but the creation of historical
probabilities. In this perspective the medieval narrative historiographic texts should be
considered primary sources, since their authors had better access to authentic sources
than the modern historian and were presumably better suited to interpret them.”
72 See e.g., Steinsland 2000, 16–18, 101–29; 2005, 35–66. Several contemporary scholars have
defended the value of Snorri’s writings and other medieval prose texts as sources for
ancient Scandinavian religion and culture. See particularly Georges Dumézil 1986 (1948),
61–83). Cf. Hultgård 1993 and Dillmann 1997; 2008.
75 See e.g., Miller 2005 and 2010; Meskell 2005; Henare et al. 2007; Meyer et al. 2010; Engelke
2012; Stausberg 2012; Carp 2014.
76 Henare et al. 2007, 1.
77 Henare et al. 2007.
78 Cf. Engelke 2012, 213.
79 Cf. Miller 2005.
In the present study the methodological and theoretical parts of the “stance
to materiality” will be followed to some extent. It must be considered as a
fact that religion includes objects, artefacts, images, gestures, words, ways of
speaking, i.e. the materiality of religious life. For some people in Late Iron Age
society the sanctuaries, the burial mounds, cultic images, the sacrificial blood,
and the ritual structures were probably experienced in a direct way, without
reflecting on symbolic dimensions, deeper meanings or abstract aspects. For
them the dichotomy between materiality and immateriality never existed.
Most likely, however, some people who had access to religious and cultural
knowledge also interpreted the ritual objects at the sanctuaries as semiotic
signs reflecting more abstract religious and ideological aspects.81 That religion,
culture and objects in ancient Scandinavia often were symbolically encoded
may be attested in several ways. One example is the language and diction in
the Viking Age skaldic poetry, and the metaphors called ON heiti ‘appellations’
and kenningar ‘kennings’.82
When studying and interpreting religious strategies for rulership at cere-
monial buildings in ancient Scandinavia we cannot reject the possibility that
some sanctuaries, ritual constructions and objects were intentionally made as
signs referring to more abstract ideas, such as myth and cosmology, at least by
people who had access to cultural and religious traditions and probably some-
times also were the producers of these sanctuaries. Most of the people repre-
senting the elite probably had the ability to interpret the materiality of religion
in that way. For ordinary people, however, these ritual structures were prob-
ably perceived in a more direct way. Perhaps ontologies related to things could
differ for one and the same individual on different occasions and in various
contexts. The things themselves may have dictated a plurality of ontologies.
In the present study, hence, I will argue that sanctuaries and images may
represent or refer to a mythic world and deities, although the materiality of reli-
gion, in the first place, probably gave form to the immaterial ideas. Most likely
the ritual structures and ceremonial buildings of the elite were symbolically
encoded in some way, since the entire Scandinavian culture was imbued with
metaphors and “kennings”. These symbolic dimensions were also part of the
religious strategies used for legitimacy. Whether all people approaching these
sanctuaries noticed them is, however, uncertain. It was most likely due to their
gender, age, social status and also if they had earlier encountered such milieus.
Since this study has a thematic outline, some repetitions of source instances
are inevitable. As far as possible, however, such disturbing reiterations will
be avoided. In what follows I will present the outline, the general hypothe-
sis as well as some sub-hypotheses. In Part 2, chapter 2, the social, political,
and historical contexts of the three investigation regions—the Mälaren area,
Trøndelag and Iceland—will be discussed in more detail. I will argue there
that the social and political structure differed in these areas, and it also varied
over time. In both the Lake Mälaren area and Trøndelag the power structure
was originally decentralized. However, as time went, a process of centraliza-
tion and power concentration took place. During the early Viking Age over-
lords, warrior lords, High Kings, and powerful earls appeared in these regions,
with ambitions to exercise domination and power over several local chieftains.
In Iceland, however, the process towards a concentration of power never took
place during the Viking Age. The early Icelandic society and its “constitution”
were built on a decentralized system, where power was split between many
minor chieftains. Why the settlers of Iceland never established a society with a
king, a military or administrative centre is uncertain.
In my opinion, these regional differences between the Mälaren area,
Trøndelag and Iceland had consequences for the formation of the religious
strategies for political authority and legitimacy. As indicated by a case study
presented in chapter 3, the religious strategies for rulership differed in these
regions in a substantial sense. The kings called Ynglingar and the Earls of Lade
in Svetjud and Norway claimed, according to Old Norse sources and Latin
texts, that their families had a divine descent from Freyr or Óðinn, while the
Icelandic chieftain family called Þórsnesingar more humbly argued that they
were the deity Þórr’s “dearest friend”, and therefore took care of his cult. The
kings and the earls among the Svear and the Þrœndir thus had more bombastic
terms than the Icelandic chieftains when expressing ideological matters. The
centralizing tendencies which were observed in the societies of the Mälaren
area and Trøndelag had thus generated a type of mythic strategy which was not
as visible in the more decentralized Icelandic chieftain society. Chapter 4 will
concentrate on ceremonial buildings, and the evidence of such buildings in the
sources relating to the three investigation areas. A preliminary statement will
be formulated, i.e. that more monumental Iron Age buildings (including rich
contextual finds made of, for instance, gold) appeared in Svetjud and Trøndelag
compared to the ones occurring in Viking Age Iceland. Monumentalization of
power was thus more accentuated in the former regions (i.e. the Lake Mälaren
area and Trøndelag).
The general hypothesis for the present study is that the social-political con-
ditions and other contextual aspects in the three investigation areas affected
the use of religious strategies for gaining legitimacy and political authority
observable in connection with the halls, cultic houses and sanctuaries in each
area. There is a link between the political and historical conditions in these
areas and the formation of religious strategies for rulership as they appear
at public cultic sites. It is argued, therefore, and will be shown in Part 4 (ch.
11–15) of this study, that the ideological dimensions of the sanctuaries and the
strategies for legitimacy and empowerment in Svetjud and Trøndelag often
resemble each other, while the situation in Iceland to some degree diverges
from the other regions. Some religious strategies at sanctuaries in the former
regions included a monumentalization of power and rulership and were also
related to the idea that certain royal or noble families of the Svear and Þrœndir
claimed that they were of divine descent and thus raised above other people.
Such strategies and mythic-cosmic notions were constituted by, for instance,
gold foil figures and golden bracteates (with images of mythical beings), which
often appear in the context of the ruler’s high-seat in the ceremonial build-
ings. Certain monumental graves may also be associated with such a strategy.
Thus certain impressive (symbolic) objects or structures related to some of the
ceremonial buildings and sanctuaries of Svetjud and Trøndelag emphasized a
close relationship between the rulers and the mythical beings. In Svetjud and
Trøndelag the halls and cultic houses were also used by the kings and earls as
arenas for more specific royal rituals, which were never performed in Iceland.
The royal inheritance ceremony performed by Ingjaldr, as presented in the
introduction to this chapter, is an example of one such performative ritual.
These actions could also be seen as ritual strategies for gaining legitimacy and
constructing authority at the highest level in the Scandinavian societies. There
were also certain martial hall milieus in Norway and Sweden (and Denmark),
where warrior lords with warrior bands appeared quite frequently. These halls
were, for instance, decorated with weapons. Such martial hall buildings were
less common in Iceland, since the institution of the comitatus did not exist
there. In this connection, it is suggested that the implicitly martial ideology of
the eschatological Valhǫll mythology was very important when constructing
power in Svetjud and Trøndelag, but less essential among chieftains in Iceland.
Such mythical traditions may have their Sitz im Leben in these hall contexts.
In Part 4, but also in ch. 5 (Part 3), it is argued that regional differences also
are observable in connection with the organizational dimension of ceremonial
buildings. Inter-regional chieftainly sanctuaries, comparable to the ones in
Uppsala, Lade, and Mære, never existed in Iceland. In my opinion this may be
due to the lack of kings and earls in Iceland. In Svetjud and Trøndelag it seems
as if there were ceremonial buildings and sanctuaries at several social levels in
society, which were controlled and used by chieftains, earls and kings in order
to gain legitimacy. Some of them were local, while others had a regional or
inter-regional significance. In Iceland the ceremonial buildings were erected
on the local farms and most of them played a role only for local groups. It was
thus the local chieftains who organized public cult there at the local ceremo-
nial buildings. There were no supreme leaders of Iceland, and no central power
to organize cult at central sanctuaries for all Icelanders. I completely agree
with Bjarni F. Einarsson when he states that: “Icelandic Iron Age society, with
no king and no centralized power, did not need all-encompassing religious
buildings or large monumental burials. It was the family of the farmstead that
was of central importance.”83
In Part 3 (ch. 5–10) of this study, I will suggest that the strategies for political
legitimacy at ceremonial buildings and sanctuaries in the three areas not only
diverged, but to a large extent also resembled each other. In my opinion, the
members of the lower elite in Norway and Svetjud also had religious functions
as leaders of public cult. Some of these chieftains went to Iceland, because
of the political situation and other conditions in their homelands. In Iceland
they appeared as cultic leaders and erected ceremonial buildings in a similar
way as they had done previously back in their homelands. We may reckon with
several reasons for building sanctuaries. One reason may be related to eco-
nomic aspects. It seems as if rulers and cultic leaders in all three investigation
areas, and at different social levels in society, took tribute from the people who
83 Bjarni F. Einarsson 2008, 178. Terry Gunnell (2001, 14) has also emphasized some differences
in the constitution of ceremonial buildings and rituals between Norway, Denmark and
Sweden, on the one hand, and Iceland, on the other hand.
attended the public cultic feasts at the sanctuaries. But mostly the sanctuary
was used by the elite as a place where they could gain legitimacy by religious
means, such as appearing on a stage which reflected a mythic-cosmic symbol-
ism. I will give examples of such strategies in Part 3 below.
To sum up, when discussing the religious strategies for political legitmacy
and authority at ceremonial buildings and sanctuaries in the present study, a
regional-contextual approach is applied. These buildings and sanctuaries will
be investigated in three regions, namely Svetjud, Trøndelag and Iceland. In this
investigation similarities between the three regions will be sought in Part 3
(ch. 5–10), and regional differences in Part 4 (ch. 11–15). In Part 2 (ch. 2–4), the
regional-contextual approach will be discussed further. The main sources will
also be presented there.
The purpose of the present chapter is to outline and discuss the social, politi-
cal and historical contexts of the three investigation regions: the Lake Mälaren
area, Trøndelag and Iceland. First I will discuss the development of political
power during the Late Iron Age in the societies of the Svear and Þrœndir. In
both the Mälaren area and Trøndelag a tendency towards power concentration
and political centralization is visible in the sources. A central power emerged
in both regions, which gradually led to more institutionalized kingdoms in the
Late Viking Age or the Early Middle Ages. This development will be compared
with evidence referring to Viking Age Iceland, where the social and politi-
cal structure seems to have been quite decentralized until the breakdown of
the Free State in 1262/64. In Iceland many chieftains ruled, while power in the
Mälaren area and Trøndelag was focused on kings and earls.
Very little is known about the development and structure of the Late Iron
Age society of the Svear. There are some mentions of the Svear in classical
and ecclesiastical sources as well as in Anglo-Saxon traditions. Otherwise Old
Norse texts, archaeological finds and place-names must form the base of infor-
mation when seeking knowledge of these matters. Only a rudimentary image
of the Late Iron Age Mälaren society can therefore be outlined. However, some
general tendencies connected to the political, religious, ideological and social
structures are accessible through some written sources.1
1 See e.g. Rimbert’s Vita Anskarii, Adam’s Gesta Hammaburgensis Ecclesiae Pontificium, and
perhaps Ynglingatal.
2 The name of the people “Svear” appears as early as in Tacitus’ Germania 44 (c. 100 AD). He
calls them Suiones (= PG *sueonez > OSw svear). On Svetjud and the Svear, see e.g. P. H.
geographic extent of this area, but it has been suggested that it comprised
the Mälaren region, i.e. Uppland, Södermanland, Västmanland and parts of
Närke. This area was delimited by a wide belt of forest towards the south and
the south-west. Beyond these woodlands was Götaland, i.e. the land where the
people called Götar lived. The unification of Svealand and Götaland probably
took place late, perhaps not before the Late Viking Age or the Early Middle
Ages (the eleventh or twelfth century).3
There are signs of an organized society in the Mälaren region during the Late
Iron Age. Trade, handicrafts and import of high-status goods provide evidence
of chieftainships as early as the Bronze Age. During the Middle and Late Iron
Age the economy was based on cattle-breeding and cultivation. From ancient
times Scandinavians were well-known for their skins and furs; Jordanes, for
example, wrote about the Suetidi (the Svear) who supplied Romans with sap-
phire furs.4 Adam also comments on their rich goods.5 Archaeological finds
from the Migration Period show vigorous iron production in the northern part
of the area.6 Handicrafts, especially gold-working, were highly developed. Lake
Mälaren was surrounded by extensive arable lands. Fishing was important not
only along the coasts, but also in lakes, rivers and streams. Hunting was essen-
tial for the food supply, but also for the production of fur pelts. In the Early
Middle Ages, but probably also earlier, markets for the exchange of products
produced in different areas were common.
During the Late Iron Age there was strong economic growth in Svetjud, vis-
ible in trade centres, such as Helgö and Birka, and long-distance trade.7 This
development led to an increase in population, observable in cemeteries and
Sawyer 1991; Sundqvist 2002; T. Andersson 2004; 2005; Arrhenius 2004b; Ljungkvist 2006;
N. Blomkvist et al. 2007. See also the distinction between Svíþjóð and Svíaveldi in Snorri’s
Heimskringla, as described by T. Andersson 2004.
3 Adam of Bremen distinguishes Götar from Svear, but paradoxically he also includes the area
of Götar in Sueonia. See e.g. Nyberg 1984, 312–323.
4 Alia vero gens ibi moratur Suehans, quae velud Thyringi equis utuntur eximiis. Hi quoque sunt,
qui in usibus Romanorum sappherinas pelles commercio interveniente per alias innumeras
gentes transmittunt, famosi pellium decora nigridine. Hi cum inopes vivunt, ditissime vestiuntur.
Getica 21. Jordanes is probably referring to arctic fox furs. See T. Zachrisson 2011b, 123f.
5 Omnia enim instrumenta vanae gloriae, hoc est aurum, argentum, sonipedes regios, pelles
castorum vel marturum, quae nos ammiratione sui dementes faciunt, illi pro nihilo ducunt.
Adam IV,21.
6 Hyenstrand 1974, 153f.
7 See Holmqvist et al. 1961–64, 29; K. Lamm 1999; Clarke & Ambrosiani 1993, 65ff.
settlements.8 Local centres can be attested in this area through the presence
of certain graves and place-names. Towns arose gradually during the Viking
Age, e.g. Birka and Sigtuna. They were often important for the region and its
economic life.
8 It has recently been argued that the increase in population started as early as Roman Iron
Age. See e.g. B. Gräslund 2004.
9 See an overview in e.g. Lindkvist 1995, 10–12.
10 See Lindquist 1944; Baetke 1964, 169f.
11 E.g. Lindkvist 1995; 2008; 2009; P. H. Sawyer 1991.
12 See e.g. P. H. Sawyer 1991; Hyenstrand 1996; Lindkvist 2008. Recently scholars such as
T. Andersson (2004; 2005) and B. Arrhenius (2004b), have opposed a late dating of
kingship in Svetjud. They argue that it should be dated at least to the Late Iron Age.
13 See e.g. Lindkvist 1995; P. H. Sawyer 1991.
14 See Hyenstrand 1996; Ljungkvist 2006.
15 See e.g. Hyenstrand 1996.
16 Nylén & Schönbäck 1994, 128.
17 Ljungkvist 2006, 65–72.
18 See e.g. Lindqvist 1936, 14f., 18–36; Hyenstrand 1974, 116f.; Ljungkvist 2006; Bratt 2008.
In the upper stratum of society, among the nobility, there were probably
many officials supporting the petty kings and chieftains, such as cultic leaders
(e.g. goðar) and warriors (rinkar, þegnar, karlar). Below the elite there were free
farmers, craftsmen, sailors, and also unfree people, called thralls (þræll sg.).
Stefan Brink has argued that the prehistoric society in Svetjud not only was
divided in a vertical sense, but also in a horizontal, spatial one.19 The spatial
or horizontal division of society was based on small settlement districts (Sw
bygd), which in some sense were tied to some larger, regional formations
(Sw land), which in its turn had ties with a social formation of an inter-regional
kind (Sw rike).
For each and every one of these settlement districts, regional formations,
and inter-regional units, there were one or more focal sites which may be char-
acterized as “central places”, i.e. “sites or small settlement structures that have
had some function or significance exceeding the particular site or settlement,
in other words, some kind of ‘power’ over a wider area”.20 The local rulers and
chieftains of the Svear thus seem to have resided at the central places in the
settlement districts during the Late Iron Age. Not only archaeological finds but
also place-names help to identify such sites. Tuna names, for instance, refer
to ancient centres in the settlement districts and they are often attached to
boat graves with rich finds.21 It has been suggested that the distribution of the
tuna names shows that such sites most likely played an administrative role in
a central-place complex.22 Originally some tuna names may have just referred
to farms of local chieftains, while others may have had a function as a central
place within the administrative district.23 Specialists in place-names have even
suggested that the Late Iron Age system of central places (including habita-
tions) named Tuna, surrounded by one or more sites with sacred names, was
transformed into royal sites at the end of the prehistoric period.24 “The clearly
pre-Christian nature of this system can be regarded as an indication of an
indigenous king in the Late Iron Age in the Mälaren Valley” (on cultic organiza-
tion, see further ch. 5 and 14 below).25
There is actually evidence indicating that the people living in the different
settlement districts of the Mälaren region comprised a political unit including
a common organization during the Late Iron Age. These regions, e.g. Uppland,
Västmanland, Södermanland, may have had a common law, or better in pre-
historic society: common legal customs with a law-speaker. These regions
probably had common assembly places. The names of the “folklands” in the
Mälaren area, including the concept of hund, may also indicate regional units
and a common organization. The medieval “folklands” (OSw folkland sg.) of the
Svear were called OSw Tiundaland, Attundaland and Fiædhrundaland.26 They
were made up of several hund or hundare. On linguistic grounds, Thorsten
Andersson has suggested that the hund division goes back to the Early Viking
Age or perhaps even earlier.27 The younger term hundare appears in the Late
Viking Age runic inscription at Vallentuna church (U 212) (c. 1050).28 This sys-
tem may thus reflect some kind of common military organization in Late Iron
Age Svetjud, also indicating a political unit among the Svear people.29 At the
top of such system we may assume that some type royal institution was in
charge (see further ch. 15).30
The term husaby, visible in place-names, may also be associated with a com-
mon royal organization in the Mälaren region with roots in the Late Iron Age.31
Admittedly the concept of husaby itself is probably secondary,32 denoting a
medieval royal estate or farm belonging to the medieval crown lands called
“Uppsala Öd” (Uppsala auðr). The connection of husabyar with Iron Age
mounds and other ancient monuments, however, indicates that at an earlier
stage these places may also have functioned as seats of rulers.33 It is of course
possible that royal power confiscated these important chieftain farms during
the Late Viking Age and turned them into crown lands. But it is also possible
that they were connected even earlier to an ancient bona regalia system,34
and thus may have constituted a well-organized pre-Christian kingship in the
Mälaren area.
Written sources indicate, however, that during the eleventh century there
were still several petty kings and chieftains ruling over local districts in Svetjud.
In the passage about the Uppsala cult Adam of Bremen tells us that all groups
of the Svear, from all districts (omnium Sueoniae provintiarum), gathered at
Uppsala. Kings (reges) and peoples (populi) had to send their gifts to the “tem-
ple” there.35 In my opinion, this passage indicates that an early central religio-
political power used the public cult and the important cultic site of Uppsala
to take tribute from the chieftains and the people living in the settlement dis-
tricts of Svetjud. I will develop this argument below in Part 3 and Part 4.
34 See Hyenstrand 1974, 103–118, Cf. N. Blomkvist et al. 2007, 167.
35 Adam, IV,27.
36 Evidence of high kings of the Svear may be seen in the expression or title rex sueonum in
the ninth century Vita Anskarii 19, 26. Adam of Bremen (1075) calls the tenth-century king
Eiríkr inn sigrsæli, Hericus igitur rex Suedorum, II,38. See also scholion 24, which mentions
Hericus rex Sueonum. The term kunungr (ON konungr, OSw kununger, konunger) is
evidenced in Viking Age runic inscriptions (Ög 136; Sm 42; U 11; Vg 40). Peterson 1994,
28. We do not know what this concept exactly referred to in these early sources; Brink
1990a, 55. Neither is the etymology clear. It has been suggested that the word konungr is
derived from the stem of ON konr ‘(distin guished) man, son’ (from PG *kunja- ‘family, kin,
descent’), with the patronymic suffix ‑ung-, and thus means ‘son of a distinguished man’.
Hellquist 1957, 496f. De Vries (1977, 326) has ‘Mann von göttlicher Abkunft’. In such cases
the etymology indicates a strategy of legitimacy, whereby the ruler had to be descended
from a noble or divine family. When I use the analytical concept of “king” in the present
study I refer to a person who was at the top of the ruling elite.
37 See e.g. Vikstrand 2000.
38 Vita Anskarii 26–27. This thing institution most likely restricted the king’s power.
According to Rimbert the Svea king Olof told Ansgar: “It is our custom that the control of
public business of every kind should rest with the whole people and not with the king”
(Sic quippe apud eos moris est, ut quodcumque negotium publicum magis in populi unanimi
voluntate quam in regia constet potestate). Vita Anskarii 26. Trans. Robinson.
39 Vita Anskarii 27.
40 Vita Anskarii 11. Cf. Norr 1998, 159f., 165; T. Zachrisson 2011a, 100f.
41 Lönnroth 1977, 7–16; Lindkvist 2008, 669.
42 Þonne æfter Burgenda lande wæron us þas land þa synd hatene ærest Blecingaeg & Meore &
Eowland & Gotland on bæcbord, & þas land hyrað to Sweon. The Old English Orosius. (Ed.)
Bately, 16. Trans. Page.
43 Vita Anskarii 30.
and cultic centre in Old Uppsala.44 For instance, the Viking Age skalds quite
often relate the rulers of the Svear (e.g. the royal family called the Ynglingar)
to Uppsala,45 and the medieval saga authors describe them as Uppsalakonungr
who sat on konungsstóll at Uppsala.46 Similar information appears in indepen-
dent medieval sources from eastern Scandinavia, where we meet the expres-
sions kunungr i Upsalum and upsala kunungr.47 Archaeology supports the
assumption that Old Uppsala was a royal centre during the Late Iron Age.48
This site is one of the most complex archaeological monuments in Scandi
navia. The finds are sufficient for it to be considered a high-ranking central
place during the entire Late Iron Age. The royal mounds (including rich finds),
boat graves, traces of a hall at Kungsgården and house terraces, the long row
of posts, appearing both south and north of the place, exclusive handicrafts
44 See e.g. Lindkvist 2008, 669. About Uppsala, see B. Gräslund 1993; Arrhenius 1995; Duczko
1996a; Sundqvist 2002; 2013; Ljungkvist 2006; 2013; Göthberg 2008; Frölund & Göthberg
2013; A.-S. Gräslund 2013; Vikstrand 2013; T. Zachrisson 2013; Beronius Jörpeland et al. 2013.
45 In the ninth-century skaldic poem Ynglingatal the Norwegian skald Þjóðólfr says that Aun
st. 18 (13) and Aðils st. 21 (16) died at Uppsala, while Óláfr trételja st. 26 (21) is mentioned
as sá áttkonr frá Uppsǫlum “that relative from Uppsala”. Dómarr burned at Føre (við
Fýri brann) st. 9 (6), i.e. a place close to Uppsala. About Vanlandi it says: sá brann á beði
Skútu “at the banks of [possible] Skutån” st. 6 (3). Skúta is most likely a small tributary
of the River Fyris located in the parish of Skuttunge (OSw Skuttungi), just a couple of
miles north of Old Uppsala. Cf. Calissendorff 1986, 35, 55; Vikstrand 2004b. References
from both Ynglingatal and Ynglinga saga are cited from Bjarni Aðalbjarnarson’s edition
of Heimskringla I in Ísl. Fornr. 26. The numbers of stanzas in Ynglingatal are counted
from this edition with the numbers of Noreen’s edition in brackets. There is actually a
new edition of Ynglingatal and some skaldic poems appearing in the Kings’ Sagas (see
SkP I–II). This edition appeared for me unfortunately at such a late stage in the writing
process that it was not possible for me to use it in this book properly. On Ynglingatal, see
SkP I, 3–60.
46 According to Snorri Sturluson, several Svíakonungar of the Ynglinga dynasty were
connected with Uppsala and the surrounding area. See e.g. Ynglinga saga (c. 1230), in
Heimskringla I, Ísl. Fornr. 26, 25, 28f., 31f., 41, 45–49, 52, 57–59, 63, 65–67, 72–74. So also
historical kings, see e.g. Óláfs saga Helga, in Heimskringla II, Ísl. Fornr. 27, 109f. As early
as c. 1120 Ari inn fróði explicitly called an ancient Svea king: Aðísl at Uppsǫlum. See
Íslendingabók, Ísl. Fornr. 1, 27.
47 The judicial text Landamæri I mentions a kunungr i Upsalum. SSGL 1, 67f. Landamæri II
has the expression Upsalakunungr. The provincial Gotlandic narrative, Guta saga, which
also is independent of Icelandic traditions, tells about upsala kunungr. Guta saga. (Ed.)
Peel, 6.
48 See e.g. Lindqvist 1936; Duczko 1993b; 1996a; 1996b; 1996c; 1996d; 1997; Ljungkvist 2005;
2006; 2008a; Ljungkvist et al. 2000; Rahmqvist 1986; Göthberg 2008. A history of research
and bibliography about Uppsala has been written by Duczko 1993a.
indicate the presence of a ruling stratum at least from the sixth century up to
at least the Early Viking Age. Excavations at the church plateau in 2005 indicate
that Late Iron Age settlements also existed under the church.49 Huge post-holes
and hearths were discovered, indicating a large house dated to the Migration
Period. Remains of (pig) bones were also found in two hearths, probably reflect-
ing sacrificial meals from the same period.50 Archaeological analyses show that
the site has been settled since the Roman Period. It was an important site from
at least the Migration Period and probably up to the Middle Ages, though it
may have experienced some troughs during the Late Viking Age, when the
kings of the Svear converted to Christianity. Arguments from settlement his-
tory, however, support the notion that Uppsala remained a royal site in the
Viking Age and the Early Middle Ages. During the twelfth century Old Uppsala
was regarded as crown land and one of the largest villages in the region. Sigurd
Rahmqvist’s analysis indicates that the western part of Old Uppsala was a royal
demesne even before the Middle Ages. The ancient name of the crown lands,
Uppsala Öd (ON Uppsala auðr ‘the wealth of Uppsala’), suggests that Uppsala
was the very core and the centre of these estates.51 Old Uppsala maintained its
symbolic significance during the Early Middle Ages. In 1164 it became the arch-
bishopric of the new church province in Sweden. The remains of the royal saint
Eiríkr were placed in the newly built cathedral there.52
In Adam’s description Eiríkr’s son, King Olaph (ON Óláfr sœnski), was a
very pious Christian, and hostile to paganism.60 For instance, he was eager
to destroy the pagan temple at Uppsala. Adam states that Óláfr founded
the bishopric in Skara. He let Unwan consecrate Thurgot as the first bishop
there.61 Scholars have previously argued that this passage shows that Óláfr lost
power in Svealand and moved to Västergötland.62 On the other hand, Óláfr may
have effected a diplomatic compromise.63 He could continue to be the king
of the Svear although he had cut his connections with Uppsala and the old
cult. His influence in Uppland remained. It is actually quite evident that Óláfr
was also the king of the Christian Svear living in the southern parts of the
Mälaren area.64 Numismatic evidence supports this and even connects him
with Sigtuna. It seems as if King Óláfr had a royal mint there. On the coins
produced in Sigtuna, Óláfr is called rex svevorum “the king of the Svear” and
rex Situm “king in Sigtuna”.65
According to Icelandic traditions the Uppsala lineage ended with King
Steinkell (r. 1060–1066); Óláfr sœnski’s sons Anundr Jacob (r. 1022–1050) and
Eymundr (r. 1050–1060) were the last kings of this family.66 Adam is somewhat
uncertain regarding Steinkell’s relationship to Eymundr, and states that he
either was the nephew/grandson or the stepson of the old king (nepos an privig
nus regis).67 According to Hervarar saga he was the son-in-law of Eymundr
(hann atti dóttur Eymundar konungs).68 This text says that Steinkell’s father
was Rǫgnvaldr the Old, perhaps an earl from Götaland.69 Sources indicate that
Steinkell had interests in Uppland; according to Adam he supported the foun-
dation of the diocese of Sigtuna, with Adalvard the Younger as bishop.70 When
60 Adam II,58. On the baptism of Óláfr, see Palme 1959, 69–72; Hellström 1996, 15–41.
61 Adam II,58.
62 Cf. P. H. Sawyer. 1991, 32.
63 Cf. Hallencreutz 1997, 125–127.
64 In the Old Norse contemporary poem, Óláfsdrápa sœnska (c. 1020), the skald Óttarr
praised Óláfr as a king of the Svear: “Host-Baldr defends field, few kings have such ability.
Óláfr gladdens the eagle [with corpses]. The lord of the Svear is outstanding.” Skj B1, 267.
My trans.
65 Malmer 1996, 99–111.
66 Cf. Óláfs saga helga, in Heimskringla II, Ísl. Fornr. 27, 152, 156.
67 Adam III,15–16.
68 Hervarar saga . (Ed.) Turville-Petre, 70. Cf. (ed.) Jón Helgason, 159.
69 Rǫgnvaldr is mentioned by the skald Sigvatr in a contemporary poem and several times by
Snorri. See e.g. Heimskringla II, Ísl. Fornr. 27, 28, 85, 88ff.
70 Adam IV,29. Cf. P. H. Sawyer 1991, 34f. Hallencreutz interpreted this passage as an
expression of Adam’s historico-theological dualism. Hallencreutz 1993, 56.
Adalvard wanted to burn down the temple at Uppsala, Steinkell stopped him
and declared that it would be punished with immediate death and that he,
Steinkell, would be driven from the kingdom for bringing malefactors into the
country.71 Steinkell’s son, King Ingi (r. 1080–1110) was a devoted Christian, and
had probably his base in Götaland.72 Perhaps it was during his reign the pagan
cult at the sanctuary of Uppsala finally was destroyed.
The political and social structure of the Late Iron Age society of the Svear was
initially decentralized. Gradually the people living in the settlement districts,
the lands and the regions of the Mälaren area, were united in an unstable inter-
regional federation under a common ruler, and a loose empire or kingdom
emerged.73 This federation leader, called the “king of Svear”, probably had a
specific relation to Uppsala, where a political, royal and cultic centre early
emerged. Perhaps Adelsö and Fornsigtuna (and later Sigtuna) became royal
sites too during the Viking Age. A more stable and institutionalized Christian
kingdom first emerged in the eleventh or twelfth centuries in the Mälaren area.
At that time Svealand and Götaland also constituted a united kingdom.
Compared to ancient Svetjud, the source situation of Late Iron Age and medi-
eval Trøndelag is good. There are several Old Norse traditions referring to this
area. Most of them were written down in the thirteenth and fourteenth centu-
ries, but some poetic traditions are contemporary with the Viking Age society
there. In addition, there is archaeological and toponomastic evidence from
Trøndelag, which helps to reconstruct the contextual aspects of this study.
areas. For a majority of the people living in Trøndelag, however, farming was
the major industry/activity, not least in the inner parts of the Trondheimsfjord,
but also in the southern parts of the region, at Gauldalen, Orkdalen and
Neadalføret. The rich Late Iron Age farms of the province are located at places
where farming conditions were good. In the mountains hunting was also
common.
The name Trøndelag is a compound including the folk-name Þrœnda- (gen.
pl. of Þrœndir) and the word lǫg ‘law’.76 Today Trøndelag consists of an exten-
sive region called Nord- and Sør-Trøndelag. In more ancient times Trøndelag
was more limited and equivalent to the area called Frostuþingslǫg, i.e. the areas
situated close to the Trondheimsfjord. The Old Norse names Þrœndalǫg and
Frostuþingslǫg indicate that the organization of society and the affinity of the
people living in this region were connected to a common law (ON lǫg). Most
likely the common judicial assembly was held at Logtu (Lagatún),77 on
the Frosta peninsula, in the central parts of the fjord (the river). The age of the
thing-organization is, however, uncertain, but most scholars think that it at
least goes back to the Viking Age.78
In the Viking Age, Trøndelag consisted of eight “shires” or “folks” (ON fylki sg.):
four situated in Inn-Trøndelag (Eynafylki, Sparbyggvafylki, Verdœlafylki,
Skeynafylki) and four in Ut-Trøndelag (Strindafylki, Stjórdœlafylki, Gauldœlafylki,
Orkdœlafylki).79 Most likely the farmers connected to each fylki were ruled by
a chieftain. These chieftains resided at great farms situated at central places
in the landscape (No sentralgårder).80 At some of these central places local
assemblies were held, as well as public cultic activities. In Inn-Trøndelag such
76 A more ancient name of the province is Þróndheimr. See Nyman 2002, 143.
77 The first element is pl. gen. of lǫg ‘law’. The second element is tún ‘courtyard, enclosed
field’, see Røskaft 2003, 210.
78 See e.g. Sveaas Andersen 1977, 59. It has, however, been discussed whether Frostathing or
Eyrathing was the most ancient thing of the Þrœndir, and whether the inner parts or the
outer parts of Trøndelag were the central settlement districts. According to Jørn Sandnes
(1967) the inner parts of Trøndelag were most ancient and thus Frosta was the old
central place for all Þrœndir. In the Late Viking Age, however, Eyra became the new
assembly place, because of the new political power appearing at Lade, i.e. the earls of
Lade and the central royal power. In historical time the Eyrathing became a konungsþing.
It was also located close to the new ecclesiastic centre of Niðaróss. See also Krag 2007.
79 On the fylki system, see e.g. Sveaas Andersen 1977, 58f.; Røskaft 2003, 168ff.; Brendalsmo
2006, 40f. It has also been discussed whether Romsdal, Nordmøre and Naumadalen were
also included in the judicial “Trøndelag”. On this discussion, see e.g. Krag 2007, 269.
80 For a careful analysis of the central places and the political power in Trøndelag, see
Røskaft 2003.
locals often were designated with names containing the element haugr.81
The fylki system with the central places was probably very important for the
organization of the local society.
The smallest unit of society in Trøndelag was the settlement districts called
ON bygðir or héruð. The society was thus organized in segmentary fashion.
The people living in the settlement districts were grouped together in the
“shires” (fylki), which in turn were grouped together to form a “law” (lǫg), i.e.
Þrœndalǫg. The term “law” (lǫg), thus intimates not only the authoritative dis-
course that regulated social practice, but also the territory in which that
discourse prevailed and the social group whose community was based upon
their submission to it, i.e. the Þrœndir (on the cultic organization of Trøndelag,
see ch. 5 and 14).
It has been suggested that the political structure of Trøndelag during the
Viking and the Early Middle Ages should be described as a “farmer democracy”
(No bondedemokrati), where the ruler only was a primus inter pares among
the farmers.82 This theory had its base in Snorri’s description in Heimskringla
of political conditions in Trøndelag. Today most scholars argue that this
society was differentiated in both a social and economic sense.83 The farm-
ers in this region were not a homogeneous group. The concept of hǫldr in the
Frostathings-Lov designates a small farmer elite group who owned land (cf. ON
óðal). These free men had strong political power on a local level of the soci-
ety. There were also other free men who did not own land. They were called
árborinn menn (cf. bóndi in the Gulaþingslǫg). There were also unfree people
called þrælar ‘slaves’ at the very bottom of the society.84
Norway, and in Hálogaland, where they originally came from, more precisely
Andøya. There was also an alliance between the earls of Lade and the Danish
royal power during the Late Viking Age, with a short break during the reign
of Earl Hákon. Lade (ON Hlaðir or Laðir) was the central place and main resi-
dence of the Earls in the tenth century. It was situated at the mouth of the river
Nið (Nidelva), not far from the present Trondheim (ON Niðaróss). For purposes
of communication it had a very good location. The name may be connected
to the verb (h)laða ‘to load’ and perhaps it designates ‘the place where things
were loaded [on boats]’.86 Thus Lade originally was connected to trade and
communications.87
Most likely Lade also was an important ruler site. Documents from the
eighteenth and nineteenth centuries report that ancient grave mounds and
monuments were located there, along the waterside. When Gerhard Schøning
visited the place in 1774 he observed several ancient monuments at “the
church of Lade”: “Norden for Hlade, paa den slette Mark, har været en rund,
med store Steene omkringsat Plads, som rimeligst har været en Domplads.”88
“Strax sønden for [Kirken] har man oppløiet en Hob Kull og Steene, tildels
smukt udarbeidede.”89 Klüwer noted at the beginning of the nineteenth cen-
tury “nogen næsten jævnede Kæmpehouge, hvori er fundne Vaaben, Ringe og
Spænder af Guld, [dette] er ogsaa Alt hvad der nu fines tilbage af det gamle
navnkundige Hlade.”90 These burial mounds are not preserved today. Only a
few archaeological finds have been made there, such as a Merovingian point
Scandinavia, too. It has, for instance, been suggested that earls ruled Götaland in the tenth
and early eleventh century. Foote & Wilson 1980, 135, critically considered by Gahrn 1988,
139–150. In historical sources referring to later periods, we hear that Earl Birger lived and
ruled in Östergötland around the 1250s. RSw iarl is also attested in two Swedish Viking Age
inscriptions, though both refer to Danish and Norwegian rulers. Sm 76 and U 617. Peterson
1994, 26. The etymology of iærl is ‘noble man’ (Sw ‘förnäm man’). Hellquist 1957, 418f. See
also Ebel 2000, 29ff.; Düwel 2000, 32ff. The primary meaning of the concept jarl is thus ‘a
distinguished, or noble man’, but in Viking Age Scandinavia it seems as if it was almost
restricted to men of high rank, who might be independent rulers or subordinate only to
kings. In present study I translate jarl to ‘earl’, even if the modern concept actually has
somewhat different connotations compared to the ancient term. Cf. Cleasby & Vigfusson
1957 (1874), 323. On the concept of jarl, see also P. H. Sawyer 1991; T. Zachrisson 2011a.
86 An alternative interpretation is ‘Etwas aufgestapeltes’. Sandnes 2001; cf. Sandnes &
Stemshaug 2007, 278.
87 Sveaas Andersen 1977, 226ff.
88 Gerhard Schøning II:6.
89 Gerhard Schøning II:4.
90 Klüwer 1823, 44.
100 According to the saga traditions it was Óláfr Tryggvason who founded the site, see e.g.
Óláfr saga Tryggvason 70, in Heimskringla I, Ísl. Fornr. 26, 318: Óláfr konungr fór liði sínu
út til Niðaróss. Þá lét hann reisa þar hús á Niðarbakka ok skipaði svá, at þar skyldi vera
kaupstaðr . . .
101 See Røskaft 2003, 106f.
102 Brendalsmo 2006, 521–525
103 An overview of the early research on “rikssamlingen” in Norway is given by Sveas Andersen
1977, 40ff. For a more recent account, see e.g. Krag 2000, 44–80. Cf. Jón Viðar Sigurðsson
2011 and most recently Lincoln 2012, 52–62 and 2014.
104 See Haralds saga ins hárfagra, Heimskringla I, Ísl. Fornr. 26, 94–149.
105 See Krag 2000, 44ff.; 2008, 647.
In these areas an inner network was created, by family ties, within the highest
social circles of chieftains. However, the political structures in the new king-
dom were weak and the areas around Viken, in the south, were still during the
tenth century under the influence of Danish kings. It was only during the reign
of Earl Hákon of Lade (r. 970–995) that Norwegian rulers became more inde-
pendent in relation to the Danish kings.
As in other Scandinavian areas, the shift in religion was very important
for the political and social development of Norway in the Viking Age. As
early as the tenth century Christianity had reached most parts of the coastal
regions. Hákon the Good (r. 934–961) was the first king who tried to establish
Christianity in Norway, although he met some resistance from the conserva-
tive farmer elite in Trøndelag. It was therefore not until King Óláfr Tryggvason
(r. 995–1000) and Óláfr Haraldsson (r. 1015–1030) that the whole country was
converted.106 The canonization of King Óláfr contributed to legitimize his
successors to the throne and it also helped to establish the Christian kingship
in Norway and unify the country.
In previous research, the relationship between local aristocracy in Trøndelag
and central royal power has been much focused on.107 Heimskringla reports
that the pre-Christian chieftains and farmers of Trøndelag were quite inde-
pendent of the central royal power in the tenth century. The Þrœndir of Inn-
Trøndelag, for instance, resisted Hákon the Good’s attempt to Christianize
the region. Even if King Hákon had good relations with Earl Sigurðr of Lade, the
latter and his son Earl Hákon remained pagans. Earl Hákon, who was a highly
conservative and confident pagan, had a great conflict with the Christian sons
of King Eiríkr blóðøx (King Hákon’s brother). Earl Hákon’s sons (Eiríkr and
Sveinn), however, converted to Christianity at the beginning of the eleventh
century. Gradually the central royal power and the official Church gained more
political and religious influence over several groups in Trøndelag. The opposi-
tion between them and the elite of the Þrœndir seems to have remained, how-
ever. It is visible in the two systems of organizing the churches in the province.
One system was connected to the fylkiskirkjur and it was controlled by the
bishop and the king. The other system was connected to the hǿgendiskirkjur,
and represented the local elite. These two competing systems were maintained
until at least the thirteenth century.108
106 See however Sæbjørg Walaker Nordeide (2011 and 2012) who by means of archaeological
finds has established a more differentiated image of the Christianization of Trøndelag.
107 See e.g. Helle 1974; Røskaft 2003, 20.
108 Brendalsmo 2006, 147–226, 285.
During the Early Middle Ages Trøndelag was developed into one of the most
important parts of Norway. In Fagrskinna it is described as hǫfuð Nóregs,109 and
Historia Norwegie calls it patria principalis.110 The ceremony of taking the king
(konungstekja) now took place at the Eyrathing. Perhaps this role of Trøndelag
was inspired by the fact that St Óláfr’s remains were placed in the shrine of
Niðaróss (today Trondheim), after his death at Stiklastaðir in 1030. According
to tradition, a lot of miracles took place in connection with the saint’s relics.
Soon Niðaróss became both an ecclesiastic and a royal centre. As early as 1075
Adam of Bremen refers to Niðaróss as the most important ecclesiastic centre
(metropolis civitas) of Norway.111 Several churches and monasteries were built
there as well as royal settlements. One of the most important churches was
Kristkirken, which later was made into the cathedral of Niðaróss (finished
in the thirteenth century). In 1152–1153 Niðaróss became the archbishop seat
and the centre of Norwegian church province.112
It seems as if the people of Trøndelag, the Þrœndir, were gathered early on into
a political unit with a common law. The political structure of the early Viking
Age society seems to have been decentralized and the political power was split
among many hands. There were several independent local chieftains ruling
in the settlement districts. During the tenth century a regional elite appeared
with ambitions to control the whole province, namely the Earls of Lade. Most
likely they were supported by the kings from the south-western Norway. The
descendants of King Haraldr converted to Christianity in the middle of
the tenth century. Gradually the central royal power and the official Church
gained more political and religious impact in Trøndelag. In the twelfth
century, however, they were still opposed by some elite groups of the Þrœndir,
who claimed local independence from central power.
Iceland was settled in the period 870–930, which is usually called “the settle-
ment period” (landnámatíð). There are several medieval texts describing this
and other important processes of Early Iceland, such as the political develop-
ment, the constitution of the “Free State”, and the introduction of Christianity.
The most important sources for these processes and developments are
Landnámabók, Ari inn fróði’s Íslendingabók, the law-book Grágás, the Sagas
of Icelanders, and the Bishops’ Sagas. In addition to these written sources we
also have place-names and archaeological finds illuminating the contexts of
the present study.
North Atlantic Drift (Gulf Stream) gave propitious conditions for fishing,
whale hunting and living in general.119 The single farm was the main feature of
the settlement pattern. No villages or towns developed there during the Middle
Ages. Many farms were moved or even completely abandoned in the early
phase, until the settlers had found the most favourable places to live at.
There are several causes behind the emigration from Norway to the Atlantic
islands. One major cause often emphasized is King Haraldr Finehair’s politics
at the end of the ninth century.120 He challenged the long-standing tradition
of local independence, and sought control over the greater part of the country.
According to the thirteenth-century sources, King Haraldr also levied prop-
erty taxes on men who traditionally not had paid any land taxes because they
owned their lands, i.e. he disturbed the age-old customs of the family-based
landholding, called ON óðal. Many landowners reluctant to accept Haraldr’s
demands therefore left Norway, and went to Iceland, but also to the Shetlands,
the Orkneys, the Hebrides, England, Scotland and Ireland. When they came
to their new lands they settled quite extensive areas and distributed land to
family members, relatives, friends and protégés. One example of such a set-
tler is Skalla-Grímr Kveld-Úlfsson (the skald Egill Skalla-Grímsson’s father)
in Egils saga, who was forced to leave Norway because of his conflicts with
King Haraldr, and therefore settled at Borgarfjǫrðr in western Iceland.121 He
built his farm at Borg, in the Mýrar area, and distributed land to his fellows
there. Landnámabók reports in laconic statements about Ǫrn í Arnarnesi. He
was a renowned man and a friend of Geirmundr Hjǫrsson. He left Rogaland in
Norway because of King Haraldr’s overbearing power (hann fór af Rogalandi
fyrir ofríki Haralds konungs). He took land in Arnarfjǫrðr, in western Iceland.122
Probably this short passage is more reliable than most of the stories about
Skalla-Grímr in Egils saga. In Landnámabók we find roughly 20 similar state-
ments about people who left Norway because of Haraldr’s politics.123 Even
if the composer(s) of Landnámabók and the saga authors exaggerated King
Haraldr’s greed and influence on the Icelandic landnámsmenn,124 there is most
119 See Byock 2001, 26, 31; Jón Viðar Sigurðsson 2008b, 572.
120 See the overview in Jón Viðar Sigurðsson et al. 2005, 95f., but also in Byock 2001, 82ff.
121 Egils saga, Ísl. Fornr. 2, 5–80. Cf. Landnámabók S29–30, Ísl. Fornr. 1, 68ff.
122 Landnámabók S134; H106, Ísl. Fornr. 1, 176.
123 Jón Viðar Sigurðsson et al. 2005, 95.
124 It should be noted that Ari makes no mention of such factors in Íslendingabók. However, it
is not unlikely that some people left Norway because of disputes with the king. See
Vésteinn Ólason 1998, 24.
likely some kind of truth behind this information.125 Of course there were also
other factors behind the emigration from Norway. Population growth and lack
of agricultural lands in Norway may have played a certain role in this process.126
Also prospects of good fishing and hunting in the new lands in the North
Atlantic may have attracted some of the emigrants, as well as good conditions
for stock breeding and cultivation of crops.127
The early settlers, of course, had several problems when arriving in Iceland.
They had to make the limited habitable area of the empty land prosper.
However, they managed to establish a society with a general assembly as early
as the tenth century, and soon it functioned as a single island-wide community.
Viking Age Iceland was in many senses a decentralized, stratified society, with
a blend of pre-state features and organized state institutions.128
Around 965, however, when the country was divided into four quarters,
it was expanded with three new chieftaincies in the Northern Quarter
(Norðlendingafjórðungur).145 In order to maintain the balance between all
quarters, the Eastern, Western and Southern Quarters (Austfirðingafjórðungur,
Vestfirðingafjórðungur, Sunnelendingafjórðungur) were also given three addi-
tional goðar each, resulting in a total of 48 chieftaincies represented at the
General Assembly. Hence, each quarter had 12 chieftains represented at this
annual meeting.
In each quarter spring assemblies (sg. várþing) were also held. In the Eastern,
Western and Southern Quarters there were three such assemblies, while the
Northern Quarter had four spring assemblies.146
All 48 chieftains constituted together the Law Council (lǫgrétta). Each chief-
tain had two “assembly men” (þingmenn) as advisors. Hence the total num-
ber of men in the Law Council was 144. In addition to them the Lawspeaker
(lǫgsǫgumaðr) and later the two bishops of the country took part in the Law
Council. The power was thus distributed among many hands in this system.
2.3.4 Christianization
In Íslendingabók Ari mentions that Christianity was preached and accepted
as the official religion in Iceland during King Óláfr Tryggvason’s reign.147 The
king sent a priest called Þangbrandr to Iceland who taught Icelanders about
Christianity and baptized them. Ari also states that the final decision concern-
ing the conversion was taken at the General Assembly in 999 or 1000. In this
process the Lawspeaker Þorgeirr Ljósvetningagoði Þorkelsson played a signifi-
cant role. In Ari’s opinion, it was because of him that the conflict between the
Christians and the heathens at the assembly resulted in a peaceful outcome.148
He let the Christian party win, although some concessions were made to the
pagans. For instance, everybody could continue to eat horse meat and perform
infanticide by exposing unwanted children in the woods, according to ancient
customs. People were also still allowed to perform sacrifices (blót) to the old
gods as long as they did so in secret.
145 Whether these systems with 36 or 39 chieftaincies really existed in Viking Age Iceland is
much debated. See e.g. Jón Viðar Sigurðsson 1999.
146 Jón Viðar Sigurðsson 2008b, 573.
147 Íslendingabók 7, Ísl. Fornr. 1, 14ff.
148 For the peaceful nature of the introduction of Christianity in Iceland, see Maurer 1855–56;
Jón Hnefill Aðalsteinsson 1998; Orri Vésteinsson 2001; Jón Viðar Sigurðsson 2003.
It has been much debated why the conflict at the General Assembly had this
outcome.149 In this debate the role of the goðar has been discussed. About 150
years ago the German legal historian Konrad Maurer argued that the system
of the goðar had a pagan and religious origin.150 He based his argument very
much on the etymology of the concept goði (see above). Maurer’s thesis has
been contradicted with the claim that if the system of goðar had a pagan and
religious origin, it would never have survived the introduction of Christianity at
the General Assembly 999/1000.151 This argument has its turn been challenged.
Recently scholars have thought that it was precisely because of the goðar’s
control over the pagan cult in Iceland that it was possible to introduce the new
religion via a resolution at the General Assembly.152 In religious questions the
goðar were dominant over their subjects, i.e. the farmers and the þingmenn.
Thus the farmers had to accept the decisions of the chieftains.
It is well known that many goðar became Christian priests when Iceland con-
verted to Christianity in the year 1000 (or 999). They erected private churches
on their farms in much the same way that the pagan chieftains had erected hof
buildings there previously. Ulrich Stutz argued that the Eigenkirchenwesen was
built on a pagan Eigentempelwesen in Iceland.153 In his theory the system of
goðar was also involved. The goðar thus maintained their traditional authority
over different aspects of society in the twelfth and thirteenth centuries. They
probably constituted a small-scale local elite, who exerted some power over
Christian cult, politics and trade. The influence of the goðar in ecclesiastic
matters in Iceland was later opposed by the Church.
The bishopric of Skálholt was founded in 1056, and in 1106 the see of Hólar
was established. Before 1104, Iceland, like the rest of Scandinavia, belonged to
the archdiocese of Hamburg-Bremen, and from 1104 to 1152–53 that of Lund.
In 1152–53 the archbishopric in Niðaróss was founded in Norway. This archdio-
cese included the Norse settlements in North Atlantic islands. One important
reason for founding the see at Niðaróss was the intentions of the reform move-
ment (libertas ecclesiae), i.e. to free the church from secular influence and to
set the Scandinavian churches more directly under the authority of Rome.
The archbishop of Niðaróss demanded complete control over the Ice-
landic Church. In order to achieve this, alliances were forged with the Icelandic
bishops. The Niðaróss bishop gained support from Þorlákr Þórhallsson, who
became bishop of Skálholt in 1178. After a conflict with one of the leading chief-
tains of Iceland, Jón Loptsson, however, Þorlákr relinquished these claims, and
did not raise them again. He continued as bishop of Skálholt until his death
in 1193, when Páll Jónsson (of the powerful Oddaverjar family) was elected as
his successor.154 In Hólar the bishop Guðmundr Arason was elected in 1201.
He was a strong supporter of the reform movement and established bonds
with Niðaróss. Because of this he had many conflicts with the chieftains.
The conflicts between the Church and the chieftains in Iceland were
thus related to the new centralization tendencies of the Church. According
to Icelandic tradition, power and religious concerns had to be connected to
the chieftains. The authority of these concerns thus had to be divided among
many hands. It was not until the middle of the thirteenth century, and in con-
nection with the breakdown of the Free State, that the Icelandic Church was
transformed into a bishop’s church, which better suited the general structure
of the papal church created in contemporary Europe.
In this chapter I have argued that the political structure and society in the
different parts of Late Iron Age and medieval Scandinavia varied. Both in
the Mälaren region and in Trøndelag the power structure was originally decen-
tralized. However, as time went, a concentration of power and a centralization
process took place. During the early Viking Age overlords, High Kings and power-
ful earls appeared in these regions, with ambitions to exercise domination and
power (ON ríki) over several local chieftains. At times such overlords consti-
tuted loose kingdoms. In the second half of the Viking Age more stabilized
As noted in chapter 2, the social and political structure differed in the three
investigation areas, and it also varied over time. In the following chapter I
will delve into the question of the religious strategies for rulership in these
three areas, although the investigation here will be limited to the mythical-
narrative dimension and a case study. I will investigate and compare traditions
connected to three representative ruling families in the investigation areas,
namely the Swedish-Norwegian Ynglingar, the Norwegian Earls of Lade, and the
Icelandic chieftain family called the Þórsnesingar. These traditions appear in
Old Norse skaldic poetry, Kings’ Sagas, and Sagas of Icelanders, and also
in Latin texts. None of the descriptions may be regarded as strictly “historical”,
but we may at least reach medieval opinions in them about these pre-Christian
rulers and their religious-mythic strategies. By means of the skaldic poetry we
may also grasp some attitudes prevalent in Viking Age society. It will be argued
that the kings called the Ynglingar and the Earls of Lade had a quite similar
religious strategy, while the Icelandic chieftains used another strategy. It is sug-
gested that these differences are related to the variations in the social-political
structures in the three areas. This conclusion has formed the basis for the gen-
eral hypothesis of this study.
In what follows I will concentrate on the Ynglinga family and study their
mythic strategies for political authority and legitimacy. It will be argued that the
real core of this strategy was the notion that these kings descended from
the gods. It will also be argued that this mythic-cosmological strategy of the
Ynglingar might very well reflect historical conditions in Late Iron Age Svetjud.
3.1.1 Ynglingatal
The oldest traditions about the “Ynglingar” appear in the skaldic poem
Ynglingatal. This poem has been preserved to the present via manuscripts of
Snorri Sturluson’s prose text Ynglinga saga (in Heimskringla) (c. 1230), where
it is quoted.1 According to Snorri, it was composed by Þjódólfr inn fróði ór
Hvini, who was King Haraldr Finehair’s skald sometime towards the end of
the ninth century.2 It is composed in Þjódólfr’s native tongue, and recounts the
reigns of twenty-nine rulers. Ynglingatal has twenty-seven stanzas. Each
stanza briefly describes the deaths, burials and sometimes burial places of the
kings. It seems as if the poem is made up of three distinct units.3 The introduc-
tory eight stanzas concern mythical and/or heroic kings living in the Mälaren
region. There are then thirteen stanzas about legendary kings of the Svear, with
names beginning with a vowel. Finally, the last six stanzas describe six possibly
historical Norwegian kings living in the areas around Vestfold.
It has been commonly held among scholars that Snorri’s information con-
cerning the dating of Ynglingatal is reasonable, that is, c. 890.4 This dating,
however, was challenged by Claus Krag in 1991.5 He argued that there are medi-
eval Christian values and ideas present in the poem indicating an anachro-
nism. I have previously presented my objections and those of other scholars
1 There are three major witnesses of Ynglinga saga. Two of them are later copies of the medieval
parchment codices Kringla and Jǫfraskinna. (1) K—AM 35 fol. is a copy of Kringla made by
Ásgeirr Jónsson (c. 1700), and provided with corrections made by Árni Magnússon. (2) J1—
Manuscript AM 37 fol. is the oldest and best copy of Jǫfraskinna, unfortunately defective,
made by Bishop Jens Nielssön (c. 1567–68). J2—AM 38 fol. is a copy of Jǫfraskinna made by
Ásgeirr Jónsson 1698. (3) F—Codex Frisianus, AM 45 fol., was written by an Icelander c. 1325.
Cf. Noreen 1925, 195–197. Wessén 1964, v–vii; Bjarni Aðalbjarnarson 1979 (1941), 2.
2 Ynglinga saga, in Heimskringla I, Ísl. Fornr. 26, 4.
3 See Wessén 1964. John McKinnell (2010) argues that the poem is made up of three or four
distinct blocks, connected by two “linking” figures, which probably reflects different oral
sources.
4 See e.g. Åkerlund 1939.
5 This opinion was not new in the research on Ynglingatal, both Bugge (1894, 108–153) and
Neckel (1908, 389–421) have made similar statements earlier, but Krag (1991 and 2009)
emphasized it further.
against Krag’s arguments for a late dating,6 and I will not repeat them here. In
my opinion, there is no reason to abandon the traditional dating of Ynglingatal.
Ynglingatal is usually regarded as a genealogical poem. It was composed
in honour of King Rǫgnvaldr, a minor ruler in Vestfold, Norway, son of Óláfr
Guðrøðarson (called digrbeinn and/or Geirstaðaálfr in later texts, see below)
and a relative of King Haraldr, and it includes twenty-eight of his forefathers.
The name Ynglingar is not attested in the poem, and does not occur before
Ari’s list of this family (Þessi eru nǫfn langfeðga Ynglinga . . .) in Íslendingabók,
dated to c. 1130.7 It thus seems as if the medieval reception regarded this poem
as a genealogy of one family.
In his thesis, Inn i skaldens sinn, Bergsveinn Birgisson rejects the idea that
Ynglingatal reflects a “historical genealogy” of the family of “Ynglingar”.8 He
conceives the poem as a níð since the kings mentioned in it died ignoble deaths.
According to Bergsveinn Ynglingatal was composed by a skald who was acting
within the hirð of King Haraldr,9 and the polemics in it were directed against
Danish royal power ruling in the area around Viken in southern Norway, and
also against Swedish kings. I agree with Bergsveinn that the poem may not
reflect one “historical genealogy” and that some of the rulers die a humili-
ating death. In my opinion, however, not all of the kings in the poem have
these disgraceful deaths and thus an interpretation of all of Ynglingatal as a
níð is somewhat misleading.10 I think that Ynglingatal may have been com-
posed from several genealogical traditions deriving from eastern Scandinavia
and elsewhere.11 Þjódólfr revised these traditions to suit his purpose. In my
opinion, however, the poem was early conceived as a genealogy of one fam-
ily. Ari obviously made such interpretation of it in his Íslendingabók,12 and
6 See e.g. Sundqvist 1995; 2002, 43–52; 2005b; 2007, 83–85. Also critical of Krag’s thesis are,
e.g. Fidjestøl 1994; Dillmann 2000a; Skre 2007b; Bergsveinn Birgisson 2007; McKinnell
2010; and Marold in SkP I, 5f.
7 Íslendingabók, Ísl. Fornr. 1, 27f.
8 Bergsveinn Birgisson 2007. Krag (1991) likewise argues that the Ynglinga dynasty was a
medieval fabrication.
9 A similar interpretation has been made by Lönnroth 1986. Svanhildur Óskarsdóttir (1994,
768) characterized the poem as grotesque rather than ironic.
10 See the discussion in McKinnell 2010, 34ff. and Marold in SkP I, 7f.
11 See Sundqvist 2002, 47–52; 2007, 86–88.
12 See Ísl. Fornr. 1, 27f. A genealogical account of these kings also appears in Historia
Norwegie (c. 1160–1175). See (eds.) Ekrem & Boje Mortensen 2003. In this Latin text
several kings are presented as cuius filius “his son”, emphasizing that it was a genealogy
concerning one family. So too does Snorri in his Ynglinga saga. In ch. 12, for instance, he
states that “Sveigðir succeeded to the realm after his father” (Sveigðir tók ríki eptir fǫður
sinn). Ynglinga saga, in Heimskringla I, Ísl. Fornr. 26, 27.
probably also Eyvindr Finnsson drew such conclusions when he composed his
genealogical poem Háleygjatal (c. 990) by taking inspiration from Ynglingatal.13
Most likely Þjódólfr also intended to create a genealogical poem when he com-
posed Ynglingatal. There are some references in the text which indicate this.14
In stanza 9 (6), for instance, King Dómarr is called Fjǫlnis niðr “Fjǫlnir’s rela-
tive”. Hence, Þjódólfr demonstrated that Dómarr was related to Fjǫlnir, who
appears at the very top of the list, in stanza 4 (1). In stanza 13 (10) the brothers
and kings Alrekr and Eiríkr are designated with the kenning Dags fríendr
“Dag’s kinsmen”. The kenning thus intimates that the brothers were related to
King Dagr, who is mentioned earlier in the poem, in stanza 11 (8). When King
Álfr in stanza 14 (11) is called Dǫglingr “descendant from Dagr”, Þjódólfr empha-
sized that he too was related to King Dagr. About King Óláfr trételgja, stanza
26 (21) says that he was “that offspring of well-known men from Uppsala”
(sá áttkonr frá Uppsǫlum). In this study I therefore call the kings appearing
in Ynglingatal “Ynglingar”, even though I am aware of the possibility that
the “genealogy” in it is only a construction made by Þjódólfr, and should not
be considered historical.15 Þjódólfr’s intention was most likely to make his lay
as a genealogical poem over one family.
ing the Norwegian pedigree to the well-known Uppsala dynasty, it has been
argued, Þjódólfr was able to glorify the petty rulers of Vestfold.18
At any rate, Þjódólfr seems to have had access to older traditions about the
Ynglingar when he composed the poem. This is indicated by expressions
he used in the poem such as “often I have asked the wisest men [about where
Dómarr’s burial place was located]”;19 “I have been informed [about Dagr’s
death]”;20 and “I learned [that King Aðils’ life was to finish by means of a vile
witch]”.21 These terms could of course be literary or conventionally expressions
without any factual historical background. There are however other argu-
ments supporting the notion that Þjódólfr built his poem on information from
older (poetic) traditions about the Svea kings and their deaths. In what follows,
the arguments for this and for an eastern origin of these traditions are set out.
Elias Wessén referred to the names in Ynglingatal which reflected Upplandic
conditions, as support for the eastern origin of the Ynglinga traditions.
Personal names with Ing-, Ingi-, equivalents of ON Yngvi, for example, in st.
10 (7), were more common in eastern Scandinavia than in Norway during this
period.22 They appeared especially in the realm of the Svear. Lars Hellberg
has proposed that there is a group of place-names in the Mälaren region
(Sweden), Ingeby (four places), Ingespjuta and Ingeberga, which can be associ-
ated with the names Ynglingar and Yngvi. These place-names have as the first
element OSw Inge (< *ingi). According to him, this element is derived from a
designation *ingvi (linguistically speaking the eastern Scandinavian equivalent
of the ON personal name Yngvi) which was regarded as the Svea ruler’s honorific
(tígnarnafn).23 Inge is thus explained as a designation for the king of the
1978–79, 52f.; McKinnell 2010, 28f. Birgit Arrhenius (2004a) has, however, recently argued
that grave gifts in Óttarr’s mound, in Vendel parish, indicate a real connection between
Norwegian chieftains and the elite of the Svear.
18 This is contested by Bergsveinn Birgisson (2007), who sees the poem as a nið. See above.
19 Ok ek þess opt/ of yngva hrør/ fróða menn/ of fregit hafðak, . . . Ynglingatal 9 (6).
20 Frák, at Dagr/ dauðaorði,/ frægðar fúss, . . . Ynglingatal 11 (8).
21 Þat frák enn,/ at Aðils fjǫrvi/ vitta véttr/ of viða skyldi . . . Ynglingatal 21 (16). See further
Wessén 1964, xiv. Wessén states that the Swedish part of the poem was built on eastern
traditions. Also Bergsveinn Birgisson (2007, 341, 415f.) argues that there were traditions
among the Svear about ancient rulers’ deaths, which Þjóðólfr had access to when he
created his poem.
22 Wessén 1924, 59, 64–67.
23 In Ynglinga saga 10 and 17 Snorri writes: “Freyr was also called Yngvi. The name of Yngvi
was for a long time afterwards kept in his line as an honour-name (tígnarnafn). His
race was thereafter called Ynglingar. . . . Everyone in their line was always called Yngvi or
Ynguni, and all of them Ynglingar.” Heimskringla I, Ísl. Fornr. 26, 24, 34f.
which was customary in Norway during the Viking period”. In the Swedish gen-
erations of Ynglingatal there are several stanzas showing “an intimate familiar-
ity with Swedish customs and beliefs—about the burial customs which seem
to have prevailed among the Swedish royal dynasties ever since the period of
migrations”, i.e., the cremation customs.34 Stanza 26 (21), for instance, men-
tions that the fire swallowed King Óláfr the Woodcutter’s body by the lake. Also
Vanlandi (st. 6 (3)) and Dómarr (st. 9 (6)) seem to have been cremated after
death in places around Uppsala, i.e., Skutån and Föret. The most important
monuments in the Mälaren region in the Early Merovingian Period are the
Old Uppsala mounds and the mound in Vendel (Ottarshögen), which reflect
cremation funerals of people belonging to the highest strata of Scandinavian
society.35 The prerequisites for a mighty royal family are evident in the Uppsala
area in the sixth–seventh centuries.
In my opinion, evidence indicates that the poem Ynglingatal was com-
posed by the Norwegian skald Þjódólfr sometime around AD 900. He probably
had access to traditions of eastern Scandinavian origin. These traditions had
flourished among the Svear at least during the ninth century and they were
his sources for the Swedish generations. It seems reasonable to suppose that
Þjódólfr received the accounts of the Swedish ancestors in the form of an active
poetic tradition.36 Runic inscriptions confirm that both poetry and genealogi-
cal knowledge existed in eastern Scandinavia during the Viking period.37 For
instance, the Rök inscription (Ög 136) from ninth-century Östergötland con-
tains an eight-line stanza in fornyrðislag. It tells about ÞioðrikR, whoever he
was: “He lived nine generations ago.” This demonstrates an interest in genea-
logical matters.
Even if Þjódólfr partly built his poem on eastern traditions, he could
have revised them and added ideas and ideological aspects reflecting the
contemporary society of western Scandinavian. The six last stanzas of
the poem obviously refer to cultural conditions prevailing in Early Viking
Age Vestfold.38 In any case, I am inclined to view Ynglingatal as a primary
34 Lindqvist 1936, 302–307, 351. According to Ljungkvist (2005 and 2008a), the Uppsala
mounds should be dated to the Merovingian Period.
35 Cf. Ljungkvist 2008a; Bratt 2008.
36 J. Turville-Petre 1978–79, 55.
37 See e.g. the runic inscriptions from Norra Sandsjö in Småland (Sm 71), where Ärnvard
names himself and five paternal ancestors, and Malsta in Hälsingland (Hs 14), where
Frömund names himself and six paternal ancestors. See e.g. Jansson 1987, 100f.; Hübler
1996; Sundqvist 2002, 151ff.; T. Zachrisson 2002.
38 Wessén (1964, xvii) noted that the naming custom among the Norwegian (or Danish)
kings is different from that of the Swedish kings. This indicates that the last six stanzas
were based on other traditions than the rest of the poem.
source for religious and ideological aspects of Svetjud (and southern Norway)
during the Late Iron Age as well. When Þjódólfr in his poem gave the rulers
of Svear certain attributes, he probably was convinced that these nobles in
eastern Scandinavia were described in such a manner. I am therefore inclined
to take these attributes as indications of a religious-mythic ruler strategy of the
pre-Christian Svea kings.
entire dynasty is called þróttar Þrós niðkvísl, ‘the kin-branch of the powerful
(potent) man [i.e., the god]’.
The Svea king Dyggvi is called allvaldr Yngva þjóðar, ‘the ruler over Yngvi’s
[the Ingvaeon’s] people’ (st. 10 (7)). We do not know if Þjóðólfr was referring
to a heiti/title of a ruler, a god or a mythical ancestor with the name Yngvi.43 If
he regarded Yngvi as a mythical being, it would appear that the entire group
was being lauded for its divine descent. Similar ideas are attested elsewhere
in Scandinavia and the Germanic area.44 Originally, not only royal houses
but also groups and peoples could have been praised as the descendants of
gods. This strategy may have been gradually monopolized by noble families.45
Expressions and denominations of the kings found in Ynglingatal, therefore,
clearly indicate that members of the Ynglingar were praised for their divine
origin.
to Freyr when he is having a tryst with Gerðr.52 The name Þrór, however, is not
restricted to gods. It occurs as the name of a dwarf in Vǫluspá 12 and in
Gylfaginning 14. Hence, Þrór cannot be interpreted as Óðinn or any other par-
ticular god.
Freyr is the only god explicitly mentioned in Ynglingatal in connection with
the kings’ ancestry.53 The medieval reception of the old traditions likewise
held Freyr as the ancestor of the Ynglingar. Snorri, for instance, made use of an
authentic tradition in Ynglinga saga when he separated Njǫrðr and Freyr from
Óðinn’s kin and made the Vanir deities into ancestors of the Uppsala family.54
According to Ari and Historia Norwegie, Freyr/Froyr was one of the progenitors
of the Ynglingar.55 Nothing is said about Óðinn there. Saxo also regards Frø
(i.e., Freyr) as the Svea people’s special cult god. He tells us that the great men
of the Svear were regarded as the sons of Frø:
The most valiant of the Svear were Ari, Haki. . . . Indeed, they were kins-
men [or close friends] of the divine Frø and faithful confederates of
the gods.56
52 Wessén (1964: 77) argued that the mythical name Þrór in Ynglingatal referred to Freyr,
since he was the forefather of the Ynglingar. Þrór was probably also a male name. It
appears in Swedish runic inscriptions. See Källström 2010b. Marold suggests that Þrór in
Ynglingatal refers to ‘an unidentified divine being’. SkP I, 56.
53 It is very uncertain whether we should interpret týss ǫ́ttungr (st. 14) as “descendant of
Týr”, as suggested by e.g. Norr 1998, 85 and Marold, SkP I, 31f. Cf. Noreen 1925, 203, 235;
Wessén 1964, 66.
54 In the “Prologue” of Snorra Edda, there is another version: “And Óðinn took with him a son
of his whose name was Yngvi, who became king in Sweden, and from him are descended
the family lines known as the Ynglingar.” It has been argued that the “Prologue” was never
a work by Snorri, but a late addition. It has also been shown that genealogical ideas in
Snorra Edda were clearly influenced by British regnal lists. Cf. von See 1988, 18–30. See also
Faulkes 1978–79; J. Turville-Petre 1978–79.
55 Íslendingabók, Ísl. Fornr. 1, 27; Historia Norwegie IX. (Ed.) Ekrem & Mortensen, 74f.
56 At Sueonum fortissimi hi fuere: Ari, Haki. . . . Qui quidem Frø dei necessarii erant et fidissimi
numinum arbitri. Saxo 8.3.11. Text Friis-Jensen. Trans. Fisher. See also Saxo. 3.2.13; 9.4.1.; 6.5.10.
57 Wessén 1923, 1ff.; Hellberg 1986a, 54; T. Andersson 1992c, 247ff.; Vikstrand 2001, 55ff., 72ff.
58 Vikstrand 2001, 69f.
Frö- may directly or indirectly be connected to tuna places, i.e., ancient admin-
istrative centres in the settlement districts. Frustuna (< *Frøstuna), in Frustuna
parish, Södermanland, is one example.59 Thus, the place-names in the Mälaren
area support a connection between ruling power, organization of society, and
the god Freyr.
If we combine the information in Ynglingatal with information found in
other sources, it seems reasonable to assume that Freyr was regarded as the
father of the Ynglingar, as early as in the Viking Period. The designation of
Freyr as blótgoð svía also supports the assumption, although it is evidenced in
late sources.60
59 Cf. K. A. Holmberg 1969; Hellberg 1986a, 50–54; T. Andersson 1991, 201f.; SOL, 328ff.
60 See e.g. Óláfs saga Tryggvasonar, Flateyjarbók I, 339 and Ǫgmundar þáttr dytts, Ísl. Fornr.
9, 114.
61 Ynglinga saga, in Heimskringla I, Ísl. Fornr. 26, 24. Trans. Hollander, modified.
62 Cf. Gylfaginning 37.
63 See Sundqvist 2002, 167. Cf. Hultgård 1994, 78.
64 See ch. 1.
65 Steinsland 1991; 2000. Scholars have supported her theory, e.g. Herschend 1996; Munch
2003b; Wickström 2004.
66 Steinsland 1991; 2000. In a more recent article, Steinsland (2011b, 21ff., 58) hesitates
regarding her former view of Freyr and Gerðr as the divine parents of the Ynglingar,
since the Ynglinga genealogy might be seen as a construct from thirteenth century
(cf. Bergsveinn Birgisson 2007).
67 The text is quoted below. Adam mentions that the cruel Hákon had sprung from Yngvar
and from a giant. Haccon iste crudelissimus, ex genere Inguar et giganteo sanguine
descendens, . . . Adam II,25. This passage may, on the other hand, be a secondary
interpolation.
68 E.g. La Farge 1994; Hultgård 1994; Clunies Ross 1994; Motz 1996; Lönnroth 1997; Krag 2001;
Sundqvist 2002; Frank 2007; Cöllen 2011.
69 Hultgård 1994.
70 Motz 1996.
71 Heimskringla I, Ísl. Fornr. 26, 14f.
72 Flateyjarbók I, 21–24. The name Fornjótr could be explained ‘Ancient-Jutlander (or Giant)’,
‘Early-User, Early-Destroyer’, ‘One-who-enjoys-sacrifices’, or ‘Ancient-Screamer’. Lindow
2001, 119.
73 Motz 1996, 79.
temja skyldi
svalan hest
Signýjar vers.79
The wedding motif may also be emphasized by the kenning and designation
of Skjálf, loga dís. This poetic expression has been interpreted as a genitive
plural of a plural tantum ON *log ‘marriage’ (cf. OIr luige ‘oath’; Goth liugan
‘to marry’; OFris loga ‘to get married’). The second element in the kenning,
dís, means either ‘woman’ or ‘female mythical being’. Thus the kenning may be
interpreted as “marriage woman, female mythical being”.80
The medieval reception of Ynglingatal regarded Skjálf as the wife of Agni.
Historia Norwegie, for instance, states that Agni was killed by his own wife.81
Ynglinga saga mentions that the hanging of Agni took place in the evening
after a banquet, perhaps the wedding feast.82 Skjálf’s identity has been debated.
Folke Ström argued that she is identical to the goddess Freyja, i.e., the divine
mother of the royal Uppsala family.83 According to him, this tradition concerns
the universal ideology of sacral kingship, where the divine king after a hieros
gamos was sacrificed to his own consort, that is, the fertility goddess.
A similar pattern, according to Ström, was also present in the traditions
about King Dómaldi. In Ynglingatal 8 (5) Dómaldi was sacrificed/slain (sóa)
by the Svear, because they were eager for crops (árgjarn). In order to find the
torquis. Noreen 1925, 201, 227f. Also D. A. H. Evans connected taur(r) with Lat torques or
thesaurus ‘necklace’. Evans 1981, 92f. Cf. McKinnell 2010, 29 and SkP I, 23f.
79 “I think that it is strange/ if Agni’s men/ thought that/ Skjálf’s marriage plans were
convenient,/ when “the wedding dis” [Skjálf]/ raised aloft/ the ruler [King Agni]/ with
the golden neck-ring,/ he who at Taur,/ had to tame/ Signý’s husband’s [Hagbarðr’s]/ cold
horse [= the gallows].” My trans.
80 This interpretation is supported by Noreen 1925, 226f.; F. Ström 1954, 40f.; Evans 1981,
105. The kenning could also be interpreted plainly as Loga dís ‘Logi’s sister’, i.e. the inter
pretation Snorri made. See e.g. Wessén 1964, 62f. Some scholars have felt that the kenning
is almost impossible to explain, see e.g. Åkerlund 1939, 89; Marold 1983, 114, 121; SkP I, 24.
81 Istum uxor sua iuxta locum Agnafit propriis manibus interfecit suspendendo ad arborem
cum catena aurea “[Agne], whose wife dispatched him with her own hands by hanging him
on a tree with a golden chain near a place called Agnafit.” (Ed.) Ekrem & Mortensen, 76f.
82 Ynglinga saga, in Heimskringla I, Ísl. Fornr. 26, 37–39. Folke Ström (1954, 40) interpreted
the banquet as a wedding feast.
83 F. Ström (1954, 39ff.) supported this interpretation by references to Wessén’s onomastic
investigation, since Wessén regarded Skjálf as a divine mother of the Skilfingar, i.e.
perhaps a name of the ancient ruling family at Uppsala. See e.g. Wessén 1924, 54. The
connection between Skjálf and the Skilfingar is, however, much debated. See e.g. Noreen
1925, 201, 225; Åkerlund 1939, 89; Baetke 1964, 125–139; Evans 1981. Steinsland 1991, 203ff.;
Gade 1985.
recipient of this sacrifice, Ström consulted Historia Norwegie, where it says that
Dómaldi “was hanged by the Svear as a sacrificial offering to Ceres to ensure
the fruitfulness of the crops”.84 Ström stated that Ceres was an interpretatio
romana of Freyja. The divine king had failed to maintain a positive relationship
with this fertility goddess, i.e., to bring prosperity to his land, and was therefore
sacrificed to her.85 In my opinion, Ström’s interpretation leans too much on
the general paradigm of the sacral kingship theory as produced by Frazer and
others. The complete pattern, including both the hieros gamos motif and the
royal sacrifice, was never discovered in the textual traditions relating to one
single king, but instead by compiling different traditions referring to several
kings. When it comes to the Agni traditions, for instance, the royal sacrifice
is not clear. The traditions only say that Agni was hung by his wife Skjálf. In
the Dómaldi tradition, on the other hand, the marriage motif is totally missing.
It is hard to deny, however, that the earthly kings in Ynglingatal are often
related to female mythical beings, as Ström suggests. These females either
kill or take the kings to the realm of death. It seems as if the rulers had inter-
course with them there. In stanza 10 (7) Dyggvi’s death occurs in symbolic
language with sexual undertones.
Kveðkat ek dul,
nema Dyggva hrør
glitnis Gnǫ́
at gamni hefr,
þvít jódís
Ulfs ok Narfa
konungmann
kjósa skyldi,
ok allvald
Yngva þjóðar
Loka mær
of leikinn hefr.86
84 [Cujus filium] Domald Sweones suspendentes pro fertilitate frugum Cereri hostiam
obtulerunt. Historia Norwegie IX. (Ed.) Ekrem & Mortensen, 74f.
85 Baetke (1964) opposed this interpretation as well as the existence of divine kings in
Scandinavia. According to him, the ruler gave good crops and fertility among cattle only
when he maintained the sacrifice. In a later essay F. Ström (1968) argued that Domaldi
was sacrificed because he had lost his “luck” (Königsheil). This theory has also been
contested. See e.g. Lönnroth 1986. Cf. Sundqvist 2002, 243ff. and the literature cited there.
86 “I do not tell lies,/ that Glitnir’s woman/ has Dyggvi’s corpse/ for pleasure./ The “horse-
dis” of/ Ulfr and Narfi/was choosing/ a king-man./ Loki’s maiden/ had the ruler/ of Yngvi’s
people/ to play with.” My trans. based on Wessén 1964.
The identity of the woman in the stanza is debated.87 The kenning jódís Ulfs
ok Narfa may however facilitate the identification, especially as the poet also
explicitly refers to this being as Loka mær, “Loki’s daughter”. It must be Hel.88
She takes Dyggvi’s body “to play with” and “for pleasure”. Dyggvi’s departure
from this life is thus described through sexual allusions.89 Dyggvi comes to
the realm of death for an erotic tryst with Hel.
Gro Steinsland has argued that the theme of Eros and death was crucial
to ruler ideology in general in ancient Scandinavia. Death is represented as a
hieros gamos, a holy wedding between the dead ruler and a female representa-
tive from the realm of death. This theme can also be seen in several stanzas
of Ynglingatal.90 In the Norwegian parts of Ynglingatal 29 (24), for instance,
a mythical female Hveðrungr’s maid [i.e., Hel]91 invited the third chief-
tain [i.e., Halfdan] to a tryst92 beyond the world.93 Similar death symbolism
may also occur in iconography. Picture stones from Gotland, dated to about
AD 700–900 and appearing in aristocratic milieus, depict a man on a horse being
welcomed by a woman holding a cup. The pictures are often interpreted as
the scene of the dead arriving in Valhǫll, welcomed by a valkyria; she gives the
dead warrior a horn with mead. The ships on these stones are usually inter-
preted as the craft which take the dead to the realm of death. One example is
the Tjängvide stone, which contains both the ship scene and the Valhǫll scene
(fig. 1).94 Erotic elements have been noticed; for instance, the shape of the
stone has been interpreted as phallic. But perhaps the further interpretations
of the sword hilts of the men in the ships or on the horse as erect penises are
exaggerated.95 At any rate, we can accept that these picture stones can involve
instances where death is described in erotic terms.96
In my opinion, the relation between the Svea kings and the mythical females
in Ynglingatal, and in other sources, comprises clear features of a mythic-cos-
mic ruler strategy (i.e., propaganda). The ruler was supplied in death, as in life,
with riches. He not only had weapons, food and drinking horns for his feast
in the Other World, but also partners for intercourse. In the realm of death
the Ynglinga kings came to a place where not only banquets but also erotic
relations were common. The departed leader continued to be wealthy and suc-
cessful in the cosmic world, so much so that even mythical beings were ready
to have intercourse with him. Perhaps this was only one aspect of the Valhǫll
mythology glorifying the life beyond, as typically represented in Viking Age
warrior society (see further ch. 15 below).
kings are related to the mythical world in other ways, too. It seems as if
the kings’ deaths in Ynglingatal were sometimes related to “mythical
marriages”, where the rulers were sexually united with mythical females.
97 Haralds saga ins hárfagra, in Heimskringla I, Ísl. Fornr. 26, 99.
98 F. Ström 1981.
(1)Viljak hljóð
at Hǫ́ars líði,
meðan Gillings
gjǫldum yppik,
meðan hans ætt
í hverlegi
galga farms
til goða teljum.
. . .
(3) Þann skaldblœtr
skattfœri gat
ása niðr
við jarnviðju;
þás þau meir
í Manheimum
skatna vinr
ok Skaði byggðu.
Snorri comments on these stanzas thus: “Njǫrðr married a woman who was
called Skaði. She would not have intercourse with him, and later married Óðinn.
They had many sons. One of them was called Sæmingr. . . . Earl Hákon the
Mighty reckoned his pedigree from Sæmingr.”100 According to these passages,
99 “(1) I desire silence/ for Har’s ale [Óðinn’s mead = poetry]/ while I raise/ Gilling’s payment
[= the mead],/ while his descent/ in pot-liquid [the mead = poetry]/ of gallows-cargo
[=Óðinn]/ we trace to gods. . . . (3) That distributor/ of gifts [= Earl Sæming],/ Æsir-kin
got,/ [the one who is] worshipped by skalds [=Óðinn],/ with giantess [= Skaði];/ the time
when they,/ the men’s friend [=Óðinn] with Skaði,/ in Manheim. (4) . . . of the bone of
the sea [the stone’s],/ and sons many,/ the ski-goddess,/ had with Óðinn.” My trans. This
“restored text” is taken from Skj. B1, 60–62. For the actual witnesses, see Skj. A1, 68ff. See
also SkP I, 195–213.
100 Njǫrðr fekk konu þeirar, er Skaði hét. Hon vildi ekki við hann samfarar ok giptisk síðan Óðni.
Áttu þau marga sonu. Einn þeira hét Sæmingr. . . . Til Sæmings talði Hákon jarl inn ríki
langfeðgakyn sitt. Ynglinga saga, in Heimskringla I, Ísl. Fornr. 26, 21f.
Earl Hákon saw Óðinn as his divine forefather. Some earls in Háleygjatal are,
however, said to be descended from Freyr or Yngvi. Earl Hákon Grjótgarðsson, for
instance, is called Freys ǫ́ttungr “Freyr’s descendant”. Snorri also stated in other
passages that the Lade Earls descended from Yngvifreyr or Ingunarfreyr. Some
scholars have therefore argued that Óðinn’s position at the top of the genea-
logical list is a sign of western European influence, i.e., the Anglo-Saxon royal
lists.101 It has also been argued that Eyvindr applied a learned euhemeristic
approach since he stated that Óðinn and Skaði lived in Manheimar, instead of
Goðheimar, which would have been their natural home.102 Like the Anglo-Saxon
authors, Eyvindr also interpreted the old deities as human beings. This has like-
wise been regarded as a sign of Western and Christian impact on the poem.
In my opinion, it is not necessary to see any Anglo-Saxon influence on
Háleygjatal. Even if Snorri made a euhemeristic interpretation of the Óðinn-
Skaði myth, it is far from certain that Eyvindr also had this perspective.
Admittedly Eyvindr stated that Óðinn and Skaði lived in Manheimar. But
we do not know how Eyvindr conceived the mythological topography and how
he designated the different places located there.103
The idea that Óðinn was regarded as the ancestor of the Lade Earls may
very well be based on old native traditions from Trøndelag or Hálogaland.104
The skaldic poem Vellekla, for instance, reports that Earl Hákon was called
Yggs niðr, “Óðinn’s relative”. In Hákonardrápa, Hákon’s conquest of the land
is described symbolically as a marriage between the earthly ruler and the
wild land, here represented as a mythical female (see below). In this meta-
phoric language Hákon is also identified with Óðinn. A similar symbolism is
also evident in Háleygjatal 15.105
It should be noted that Eyvindr in the introduction to Háleygjatal men-
tioned that Óðinn begat Sæmingr (the mythical father of the Lade Earls) on
Skaði. According to Gro Steinsland, this foundation myth resembles the one
found in Ynglingatal (and Yngling saga). The ruling families described in these
poems both derived from a peculiar marriage between a god and a giantess.106
I agree with Steinsland that this type of marriage is probably evident in the
101 Bede, for instance, made Woden (Óðinn) into an ancestor of ancient British kings in
genealogies. See Bede, I,15. See also J. Turville-Petre 1978–79, 63; Faulkes 1978–79, 96.
102 Cf. Bjarni Aðalbjarnarson in Heimskringla I, Ísl. Fornr. 26, xxxii. Faulkes 1978–79, 97f.
103 On interpretations of Manheimar in this context, see Steinsland 2011b, 30 and SkP I, 200.
104 Cf. Wessén 1924, 34f.; F. Ström 1981, 447; Steinsland 2011b, 32f., 39.
105 See Sundqvist 2007, 101ff.
106 See Steinsland 1991. It is somewhat uncertain whether we should classify Skaði as a
giantess. See Näsström 1995, 51f.; Lindow 2001, 268ff.; Simek 2006, 286f. In a genealogical
sense she is undoubtedly a giantess, since she is the daughter of Þjazi. The denomination
of her in Háleygjatal, jarnviðja, also indicates that she was a giantess. Steinsland 2011b, 29.
genealogy of the Lade Earls. However, it is not completely safe to place the
giantess Gerðr as a mythical mother at the upper end of the “Ynglinga gene-
alogy”, even if it is a good suggestion (see above). Anyhow, there is no doubt
that both Ynglingatal and Háleygjatal clearly indicate that both the “Ynglinga
kings” and the Lade Earls claimed that their families were of divine descent on
the male side.
107 “(3) The war-ship’s brisk lord/ with the veracity of the sword/ entices under him Þriði’s
[=Óðinn’s] beloved,/ whose hair is the foliage of pine trees./ (4) Therefore I think that
the thrower of the spear/ is very unwilling to leave Auðr’s glorious sister [= Earth, the
land] alone;/ the land prostrates herself /under the ring-waster./ (5) The outcome of
the union/ was afterwards this quick-witted intimate friend of kings/ gained possession in
marriage of Ónarr’s/ forest-grown only daughter./ (6)The commander of ships [= ravens
of the harbour]/ was able to attract to himself/ Báleygr’s [=Óðinn’s] broad-faced bride
[= the land, Norway]/ with the help of the sword’s powerful discourse.” Text and trans.
F. Ström 1981, 452f. Cf. Skj. B1, 147f. Only nine half-stanzas are preserved of this poem. It has
been argued that the sixth stanza should be the first in this sequence. See F. Ström 1981.
fertility goddess called Þórgerðr Hǫlgabrúðr.115 This hieros gamos was part of
an old local ruler ideology used by the conservative Earls of Lade in Trøndelag
and Hálogaland.
There are reasons to believe that Hákonardrápa reflects structures of an
ancient mythic-cosmic ruler strategy in Trøndelag. That the wedding symbol-
ism referred to aspects of inauguration seems likely. By transferring the his-
torical event of Hákon’s conquering of Norway to a mythical scene, and by
identifying the ruler with Óðinn and the land with a female mythical being,
the perspective is moved from the microcosmos to the macrocosmos. The local
perspective thus fades and the ruler’s activity acquires a cosmic dimension.
Placing the earl in a mythical context was an effective means in homage poetry
and was an instrument of political propaganda for Earl Hákon.116
The mythical wedding in Hákonardrápa is thus styled differently from the
symbolic marriage in Ynglingatal. In Ynglingatal, the deaths of the rulers
are described as a tryst between the earthly ruler and a mythical female, while
Hákonardrápa depicts Hákon’s victory and inauguration as a sexual union
between the earl/Óðinn and the land/mythical female. Since the contexts of
Ynglingatal and Hákonardrápa varied, as did the themes they were built on
(death and inauguration respectively), the mythical strategy has acquired dif-
ferent conceptions in these poems.117
At any rate, there are some striking similarities between the mythic-cosmic
ruler strategy found in Ynglingatal and the poems praising Earl Hákon. Both
the Ynglingar and the Earls of Lade counted themselves as descendants of the
gods, Freyr and Óðinn respectively. It also seems as if the members of both
families related themselves sexually to female mythical beings.
115 Roberta Frank (1978, 62ff.) has suggested that the hieros gamos myth behind Hákonardrápa
concerns Freyr and Gerðr. This idea is rejected by F. Ström (1983, 72) since the poem
obviously refers to Óðinn. In a later study Frank (2007) is sceptical of the hieros gamos
theory when applied to Old Norse skaldic traditions.
116 Cf. F. Ström 1981; 1983. See also Steinsland 1991.
117 A similar strategy to that seen in Hákonardrápa also appears in connection with the kings
of the Svear and Danes. Around 1018 Óttarr svarti praised King Óláfr (sœnski) for seizing a
kingdom: “The ruler takes for certain battle-Freyr’s [Óðinn’s] desirable wife [Jǫrð =land],
loveless, in the east. The all-powerful commander of men has a good life.” Skj. B1, 267. The
verb taka has here the double meaning ‘conquer, have sex with’. See Frank (2007, 182f.) for
this example and skaldic references to Knútr the Great.
advice about what he should do, whether to make peace with the king or leave
the country. He was advised to go to Iceland. Þórólfr had a figure of Þórr carved
on one of his high-seat posts. When he arrived in Iceland he threw these posts
overboard and declared that he would settle at any place where Þórr chose to
send the posts. The posts reached land at a ness in Breiðafjǫrðr. Þórólfr called
that place Þórsnes, and a river close to that place was called Þórsá and the
whole area was consecrated with a fire ritual. Not far from there he established
his farm Hofstaðir, where he also erected a hof, and placed figures of gods
and other sacred objects in it (see further chapter 4 below). In his old age
Þórólfr married a woman called Unn. They had a son called Steinn. Þórólfr
dedicated the boy to his friend Þórr and gave him the name Þorsteinn. That boy
was called Cod-Biter (þorskabítr), and he inherited his father’s farm. He mar-
ried a woman called Þóra. She gave birth to a son, who was sprinkled with water
and given the name Grímr. “Þorsteinn dedicated this boy to Þórr, calling him
Þorgrímr, and said he should become a hofgoði.”123 Þorgrímr’s son was Snorri
the goði. He too was a cultic leader and cared for a hof located at Helgafell.
Eyrbyggja saga and Landnámabók also mention another son of Þórólfr, i.e.,
Hallsteinn Þorskafjarðingagoði.124 In the Landnámabók version Hallsteinn’s
close relation to Þórr is also emphasized.125 When he settled in Þorskafjǫrðr
he sacrificed to Þórr in order to get his high-seat posts. Soon a huge tree was
washed ashore at his site. This tree was used to make high-seat posts. It thus
seems as if several male members of the Þórólfr family were goðar and devoted
to the cult of Þórr. Evidence in Eyrbyggja saga indicates that Þórólfr himself
was regarded as a goði, and it seems as if his office was inherited by his off-
spring (see ch. 5).
123 . . . þann svein gaf Þorsteinn Þór ok kvað vera skyldu hofgoða ok kallar hann Þorgrím. Ísl.
Fornr. 4, 19.
124 Eyrbyggja saga, Ísl. Fornr. 4, 136. See also Landnámabók S85, Ísl. Fornr. 1, 126.
125 Landnámabók S85, H73, M26, Ísl. Fornr. 1, 124–126.
126 The naming custom in Eyrbyggja saga is most likely fictitious. See Vikstrand 2009.
more than Freyr”.127 The chieftain Þorkell the High from Þverá made a prayer
to Freyr in Víga-Glúms saga ch. 9: “Freyr, who for a long time has been my con-
fidant (fulltrúi) and received many gifts from me and well rewarded them . . .”128
It has been argued that the linguistic usage in these texts was based on biblical
expressions and that the saga authors had the Christian relationship between
man and God as a model.129 It has also been suggested that, for instance, the
concept of fulltrúi,130 which sometimes expresses an individual devotion to
a pagan deity in Old Norse prose,131 is based on Christian ideas.132 This posi-
tion has however been contradicted. Other scholars have felt that even if these
terms were first applied to pre-Christian conditions by authors of the twelfth
and thirteenth centuries, it is likely that memories of personal devotion to
the old deities were passed on by oral tradition into later centuries.133 Some
scholars argue that a personal relationship between man and god in Iceland
is evidenced in the tenth-century skaldic poetry.134 In the Icelandic skald
and chieftain, Egill Skalla-Grímssonar’s Sonatortek, for instance, we read thus:
Átta ek gott
við geirs dróttinn,
gerðumk tryggr
at trúa hánum,
áðr vinátt
vagna rúni,
sigrhǫfundr,
um sleit við mik.135
127 Hrafnkell elskaði eigi annat goð meir en Frey, . . . Hrafnkels saga Freysgoða, (ed.) Jón
Helgason, 2. Cf. Ísl. Fornr. 11, 99.
128 Freyr, sagði hann, er lengi hefir fulltrúi minn verit, ok margar gjafar at mér þegit ok vel
launat, . . . Víga-Glúms saga 9, Ísl. Fornr. 9, 34.
129 See e.g. Maier 2003, 36. He refers to Exodus 33,11; John 11,11; James 2,23.
130 ON fulltrúi m. ‘confidant, true or completely trustworthy friend’.
131 See e.g. Eiríks saga rauða 8, Ísl. Fornr. 4, 224.
132 See e.g. Zernack 1998.
133 Hultgård 2008, 213.
134 See e.g. Å. V. Ström 1975, 198f.; 1990, 374f. On “friendship” between man and god in pre-
Christian Scandinavia, see also Jón Viðar Sigurðsson 2010, 122ff.
135 “I was on good term with the lord of the spear [Óðinn], I grew trustful, believing in him
[Óðinn], until the friend of chariots [?], the prince of victory [Óðinn], broke friendship
with me.” Text and trans. E.O.G. Turville-Petre 1976. Cf. Skj. B1, p. 37. See also Skj. A1, p. 43.
Whether this stanza really indicates that Egill previously had a close relation
to “the lord of the spear [Óðinn?]” is, however, somewhat uncertain.136
In any case, I agree with those scholars who accept that a personal devotion
to a pagan god may have existed in more ancient time. For there are several
instances where the phrase trúa á einum “believe in someone or something”
seem to indicate a personal relation between man and pagans gods.137 Some
of them appear in Eddic lays, which probably were composed in Iceland. In
Harbarðzljóð 34, for instance, Hárbarðr states: “Then I should believe in you
[Þórr]” (Ec mynda þér þá trúa). According to Åke V. Ström this phrase confirms
a personal piety in Norse religion: “here we really have ‘une foi personelle’ in
pagan time”.138 In Hyndlolióð 10 we read:
The expression “Óttarr has always trusted (or believed) in Asynior” may be an
ancient expression even if the poem is quite young. At least the cultic context
of this expression indicates a pagan setting. Anyhow, it seems as if Óttarr was a
devoted worshipper of the goddesses called Asynior (ON ásynior pl.).
Thus it is not impossible that the idea that the Þórsnesingar in western
Iceland considered themselves as close friends of Þórr may be based on old
oral traditions. The close personal relationship to a god may even have legiti-
mated the Þórsnesingar’s central position in public cult. It seems as if some
members of this family were goðar, who protected the hof and organized
the common cult and sacrificial feasts.140 This type of strategy is quite differ-
ent from the one found among the Ynglingar and the Lade Earls. Relating to
the tentative taxonomy presented in ch. 1 above, the West Icelandic chieftains
may be regarded as the most important mediator in the relationship between
the human and the mythical worlds, i.e., they perform as the (highest) cultic
leaders. They are in some sense symbolically close to the divine world, although
they are not regarded as divinities or praised for having a divine origin and the
like. Instead they are regarded as a “close friends” of the god, with a cultic rela-
tionship, and not a genealogical one.141
It thus seems as if the religious ruler strategy of the Icelandic chieftain family
was different from the ones used by the Ynglingar and the “Lade Earls”. The
kings and earls of Svetjud and Norway seem to have applied a more “bom-
bastic” or “monumental” religious-ideological strategy than the chieftains
in Iceland. They argued that their families were of divine descent, while the
Þórsnesingar just stated that they were “friends” of the god. The reasons for
these differences may be related to various aspects, such as the heterogeneous
source materials, but also other contextual matters. In what follows I will con-
sider some possible explanations.
The families investigated in this study represent three different types of
political structures or systems, namely kingdom, earldom and chieftaincy.
They also represent different geographical areas. The Ynglingar may be con-
nected to the Mälaren region and Vestfold, the Lade Earls to Trøndelag, and the
Þórsnesingar to the area around Snæfellsnes, at Breiðafjǫrðr in western Iceland.
In Svetjud and Norway we may see a concentration of power and some tenden-
cies towards a more centralized political system during the Late Iron Age, with
the appearance of some great royal and noble dynasties.142 In Iceland power
141 See Jón Viðar Sigurðsson 1999, 187. It should be noted that goðar also existed in Norway,
Svetjud and Denmark and thus this ruler strategy may have existed there too. See e.g.
Sundqvist 2007, 30–34. In late sources Earl Hákon, for instance, was also regarded
as Þorgerðr Hǫlgabúðr’s best trusted friend (fulltrúi), see Þáttr Þorleifs jarlaskálds in
Flateyjarbók I, 213f. But the genealogical relation to the gods seems to be exclusive for
certain noble families who mainly resided in Svetjud, Norway and Denmark. Occasionally
Icelanders claimed in euhemeristic accounts that they were related to these families.
See e.g. Íslendingabók Ísl. Fornr. 1, 27f. This, however, is not a common strategy among
Icelandic goðar, as we see from Landnámabók, Íslendingasögur and Íslendingaþættir. Jón
Viðar Sigurðsson (2011, 83) writes: “Ari is the only Icelandic chieftain claiming such noble
origin. Why he created this pedigree is still an unsolved puzzle, but for Ari, as well as for
the Oddaverjar, it was probably important to stress that the Breiðfirðingar were of royal
blood.”
142 See e.g. Lindkvist 2008; Krag 2008.
was split among many hands during “the settlement period” and “the saga age”
(c. 870–1050), and there was a large number of Icelandic chieftains.143 It was
not until the second half of the Commonwealth period (c. 1050 to 1262/64)
that the number of chieftaincies was reduced and we may see a concentra-
tion of power.144 The early Icelandic chieftains often traced their genealogies
back to a Norwegian hersir, i.e., a type of local chieftain who belonged to the
lower elite.145 It has often been argued that such a chieftain was regarded as
primus inter pares. He had to secure his own positions by creating alliances and
economic wealth, so he could support a small group of men at his farm and a
larger group in times of conflict. Such alliances were created by friendship and
mutual giving of gifts.
It was probably not possible to establish a “political power” in Iceland com-
parable to the one held by the great dynasties in Norway, Svetjud and Denmark.
Iceland had only a thinly spread population and there were no members of
royal and noble families present to make such claims. Perhaps there were also
strong feelings of independence and pride among those settlers who had emi-
grated from Norway. Kings and earls were regarded by them as, at best, “a nec-
essary evil”.146
During the Early Viking Age, Norway consisted of over twenty great and per-
haps even a hundred minor chieftaincies.147 At the end of the ninth century
the number of chieftaincies was reduced. The kingship of Haraldr included the
breakthrough of a new type of political structure. It seems as if Haraldr and
his sons had ambitions to control all Norway.148 As a reaction to this ambi-
tion, some of the minor Norwegian chieftains moved to Iceland, where they
gradually established the system of chieftaincies and the new constitution. It
is probable that the pre-Christian kingdom in Svetjud also had a rather highly
structured political organization.149 It is uncertain, however, when the Svear
first constituted a political unit. Early society in Svetjud seems to have been
divided into several groups under independent chieftains. A concentration of
143 Jón Viðar Sigurðsson (1999; 2008a, 63f.) has argued that there were a large number
of chieftains in Iceland, possibly between 50 and 60 during the “saga age”, and new
chieftaincies were created continuously. Cf. Byock 2001, 3f., 13f., 63–80.
144 Jón Viðar Sigurðsson 1999, 39–83; 2008a; 2008b.
145 Cf. Phillpotts 1912–13, 276ff.; Vésteinn Ólason 1998, 29f.
146 Vésteinn Ólason 1998, 27–30.
147 Jón Viðar Sigurðsson 2008a, 22ff.
148 It should be mentioned, however, that King Haraldr’s kin only controlled Vestlandet c. 900.
Danish kings had the power over Østlandet and the Earls of Lade controlled Trøndelag
and northern Norway. See e.g. Krag 2008 and Jón Viðar Sigurðsson 2008a, 28.
149 Lindkvist 2008, 669.
power in this area may have started as early as the sixth or seventh century,
when the three royal mounds were erected at Uppsala.150 It is likely that, at
times, the chieftains and groups were united in a loose alliance under a federa-
tion leader or High King.151 Such a king or overlord could occasionally exer-
cise power over several groups and later also over territory. As time passed,
this alliance of groups was stabilized and kingship developed in the Mälaren
region, with a royal and cultic site in Uppsala.152 Thus, the political structures
in Svetjud and Norway resembled each other in the Viking Age, while Iceland
was different.
In my opinion, there is a link between the political and historical condi-
tions in Norway, Svetjud and Iceland and the formation of religious strategies
for rulership in these three areas. The kings of Norway and Svetjud as well
as the Earls of Lade had certain characteristics which few others possessed in
these societies.153 By means of religious symbols, descent and wealth, as well
as monumental buildings and structures, these royal and noble lines raised
themselves over other influential families. They claimed, for instance, that
their families were of divine descent. Perhaps such a strategy in a previous his-
torical phase was related to most of the chieftains in Norway and Svetjud, but
when the powerful “kingdoms” and “earldoms” emerged, such characteristics
were more exclusively used by kings, earls and the very top elite.154
It is also possible that the notion of the rulers’ divine descent only was
applied by some of the greatest dynasties in Late Iron Age Scandinavia, such
as the Skjǫldungar (of Denmark), the Ynglingar, the Earls of Lade and perhaps
the Earls of Orkney.155 Since no kings or earls from these dynasties played
a central role in early Icelandic society, this strategy rarely appeared there.156
Perhaps this religio-political strategy was even discouraged in the early
Commonwealth society of Iceland, since it was associated with oppressive
rulers who dominated in the old homeland, the Norwegian kingdom. Perhaps
also “the Icelandic constitution” discouraged political centralization. Most
likely the historical and socio-political contexts affected the formation of the
religious strategies for gaining political authority in the three areas. Probably
other contextual aspects were involved too, but the suggestions presented
above may at least have played a certain role.
In the present chapter a case study has been presented where the specific
religious strategies for rulership were investigated in relation to some noble
kindreds as narrated in written sources. It included the royal family named
Ynglingar from the Mälaren region, the Earls of Lade from Trøndelag, and the
chieftain family named Þórsnesingar from Iceland, i.e., representatives from
the three regions addressed in the monograph. The kings and earls from the
Mälaren region and Trøndelag claimed to have a divine origin. This strategy
was not seen among the chieftains in Iceland. They argued, however, that
they were the deity Þórr’s dearest friend. The relationship between rulers and
mythical beings in the former areas could be conceived as “genealogical”, while
the latter (found in Iceland) was “cultic”. The conclusion was that the central-
izing tendencies which were observed in the societies of the Mälaren region
and Trøndelag had generated a type of ideological strategy, which included the
monumentalization of power and rulership and the notion of divine descent.
Such a strategy was not as visible in sources referring to the more decentralized
Icelandic chieftain society.
Kári, Frosti, and Snær. See Orkneyinga saga, Ísl. Fornr. 34, 3. Cf. Steinsland (2011b, 48ff.)
who emphasizes the hieros gamos motif at the top of this genealogy.
156 Vésteinn Ólason (1998, 29f.) has made a similar interpretation. Gro Steinsland (2011a, 6)
has recently argued: “Kinship was probably more important for the elite than for other
classes of population—note that for example Germanic titles of rulers are derived from
kin-groups. The importance of kin for the elite can also be seen in the many genealogies:
on the level of kings and earls ancestors were counted back to the gods and other non-
human powers. By contrast, chieftains do not claim descent from mythical forces; they
were linked to the gods in different ways, for example via initiation.” As noted above, Ari
traced his family back to the Ynglingar and a divine descent in his euhemeristic account.
But this is an exception in the Icelandic context. Óttarr in Hyndlolióð is descended from
the gods (er frá goðom qvómo) although he was born into a hersir family (hersborinn).
On the other hand, the poem states that he also was a relative of the royal Skjǫldungar,
Ynglingar and Skilfingar.
1 J. Grimm 1835.
2 E.g., Keyser 1847; Thümmel 1909.
supported with archaeological evidence.3 For almost sixty years there was
close consensus among scholars that the ancient Scandinavians had “temples”
for their cultic activities.4
However, Olaf Olsen’s source-critical investigation, Hørg, hov og kirke (1966),
constituted a radical reconsideration. Olsen argued that the descriptions
of the “temples” (hof sg.) in the Old Norse traditions could not be supported
by the archaeological evidence discovered at so-called hof-lands (hovtomter)
in Iceland. He believed that the written descriptions were projections of the
medieval authors’ and saga scribes’ experiences of Christian churches. Olsen
proposed that the “temples” excavated in Iceland, such as the one at Hofstaðir,
were in fact the dwelling-houses or halls of the chieftains, and as such mul-
tifunctional: ceremonial feasts would probably have taken place there, but
also many non-religious activities. In other parts of Scandinavia too, cultic
feasts were celebrated in the multifunctional halls situated on the magnates’
farms. According to Olsen, we have no reliable evidence of specific pre-
Christian “temples” in ancient Scandinavia, except in (Old) Uppsala. The
pre-Christian cultic activities were in general performed outdoors, at “natural
sanctuaries” (naturhelligdomme), such as trees, wells, stones and cairns.
Occasionally a type of simple building was erected at such sites, called ON
hǫrgr. Olsen argued that these structures may have been some kind of pre-
cursors to the earliest Norwegian stave-churches. He also concluded his the-
sis with the statement that we have no evidence for cultic place continuity
in ancient Scandinavia, i.e., that the earliest churches were situated at pre-
Christian cultic sites: “vi . . . må hævde, at teorien om, at de ældste kirker
fortrinsvis er pleceret på hedenske kultsteder, ikke hviler på et forsvarligt
grundlag.”5 According to him, (Old) Uppsala cathedral was the only certain
example in Scandinavia where we have such cultic place continuity.6
Olsen’s conclusions have been contradicted by several scholars. It has been
argued, for instance that Olsen generalized too much when he stated that cul-
tic place continuity did not exist in ancient Scandinavia.7 His results regarding
this specific issue were actually only based on materials from Denmark, i.e.,
sources which need not be valid for other areas in Scandinavia. Both in the
Mälaren area and in Trøndelag there is clear evidence that early churches occa-
sionally were erected at pre-Christian cultic sites.8 It has also been stated that
Olsen was inconsistent in his source criticism, for instance when he rejected
the indications of hof as a cultic building in the Eddic and skaldic poetry, but
accepted hǫrgr as a pagan structure basing his statement on equivalent source
materials.9
Despite this criticism, Olaf Olsen’s dissertation exerted a strong influence
on research in the following decades, also amongst archaeologists discussing
Late Iron Age halls. The archaeologist Frands Herschend, for instance, argued
that the halls could be interpreted as multifunctional buildings attached to
nobility and royalty.10 According to him, the hall was the room for leadership
in an economic as well as a military sense already in the fifth century. The finds
discovered in them, for instance gold foil figures (Swedish guldgubbar) and
ceremonial glass, indicate that also cultic acts and rituals were occasionally
performed there. Such halls had been erected at aristocratic central places
of varying local, regional and trans-regional significance, such as Dejbjerg,
Gudme and Lejre in Denmark, Uppsala, Vallhagar (Gotland) and Slöinge in
Sweden and Huseby (Tjølling) and Borg (Lofoten) in Norway (fig. 2a and 2b).
Historians of religions were also inspired by Olsen’s study, arguing that the hof
buildings mentioned in the Kings’ Sagas and Sagas of the Icelanders should be
interpreted as multifunctional halls.11
The most recent scholarship, however, indicates that the debate concern-
ing specific cultic buildings must be resumed. New archaeological finds sug-
gest that more exclusive cultic houses may have existed in Late Iron Age
Scandinavia. At several excavation sites in Denmark and Sweden, traces of
8 See e.g., Lidén 1969; Sandnes 1987; A.-S. Gräslund 1992; Jørgensen 2009. See also the self-
criticism by O. Olsen 1996: “Jeg ville være mere åben over for muligheden af, at man på
centrale hedenske kultpladser har rejst kirker som udtryk for ecclesia triumphans . . . Jeg
ville have brugt ordet ‘magtkontinuitet’ om stormændenes kirkebyggeri på deres gårde.
Her er det magten og ikke kulten, der er det centrale element.”
9 See Sandnes 1964; 1987, 145, who also refers to Rostvik’s (1967) critical investigation of
hǫrgr.
10 Herschend (1993: 182; 1998: 16) has pointed out certain features connected to these
buildings: they belong to big farms; they consist of one room with a minimum of posts;
they are singled out by their position on the farm; their hearths are not used for cooking,
nor do they facilitate handicrafts; artefacts found in these houses are different from those
found in the dwelling part of the main house on the farm.
11 See e.g., Hultgård 1996. Terry Gunnell (2001) has argued that the hall was temporarily
transformed into a sacred building at particular times of year, for example during
calendrical feasts. See further ch. 8 and 9 below.
Figure 2a
The great Merovingian period hall in Old Uppsala,
c. 50 metres at length. Ljungkvist & Frölund 2015 MS.
Figure 2b Southern plateau in Old Uppsala, where the hall was erected.
Photo: Anders Andrén.
Figure 3 The cultic house at the Viking Age chieftain farm of Borg,
in Östergötland, Sweden. The building was erected on a stone-paved
courtyard where some 75 kg unburned bones were discovered in
association with a rock. Beside the rock there was a deposit of almost
100 amulet rings.
Illustration: Richard Holmberg.
Figure 4 A cultic house at the central place of Uppåkra, outside Lund, Scania, a
possible cultic house (13.5 × 6 m) was explored. This house was founded
during the Roman Iron Age, and was rebuilt in several phases up until the
Viking Period. The finds indicate that this house may have been a more
exclusive and specific cultic building.
Illustration: Loic Lecareux.
first been built during the Roman Iron Age, and was rebuilt in several phases
up until the Viking Period. The peculiar finds, for instance the gold foil figures,
a glass bowl and a gilded silver beaker, indicate that this house may have
been a more exclusive and specific cultic building.
4.1.2 Large Residences and Small Cultic Houses: Some Views in Recent
Research
In a very important article, the Danish archaeologist Lars Jørgensen argued
that a certain “pattern in terms of the organization of the central, prestigious
area on the one hand, and of the closest hinterland on the other hand” could
be detected at aristocratic residences (central places), and classical sites in
southern Scandinavia, such as Gudme, Uppåkra, Sorte Muld, Tissø, Lejre,
Toftegård and Järrestad.14 At all these sites the main building (the actual
14 Jørgensen 2009, 349ff. Cf. Christensen 1991, 1997; Söderberg 2005
Certain terms in the Old Norse written sources, such as hǫll, salr, veizlustofa,
hirðstofa and skáli have sometimes been associated with the Iron Age hall
buildings found at excavations in Scandinavia.24 In Old Norse texts it seems
as if the term hǫll was mostly applied to royal contexts in Norway, Svetjud and
Denmark, while the banqueting buildings at chieftains’ sites in Iceland were
usually called stofa or skáli.25 A royal hall (or hall room) in Norway, Svetjud and
Denmark could also be described as a veizlustofa.26 In connection with the
smaller and more specific cultic buildings, scholars have searched for other
Old Norse concepts which may correspond to the archaeological finds. These
smaller buildings have sometimes been regarded as equivalents to the hǫrgr
appearing in the textual sources, while the fenced and ritual areas around them
have been interpreted as a hof.27 Even if these interpretations are not totally
unreasonable,28 they have weak support in the ancient Icelandic literature.
The idea that hǫrgr designates a house can only be seen in a few (problematic)
texts. The Codex Regius manuscript of Snorra Edda, for instance, states that the
gods built a hall (salr). This was the hǫrgr that belonged to the goddesses. This
building (hús) was called Vingólf.29 However, if we go to the manuscript called
Codex Upsaliensis, the hǫrgr in the same passage designates a structure inside
the hall, not the building itself.30 The other instance where hǫrgr refers to a
house is in Kong Sverrers Christenret. According to this text, it is forbidden to
build a house and call it a hǫrgr.31 The interpretation of this reference has also
24 See Brink 1996a; 1999b; Meulengracht Sørensen 2003; Sundqvist 2005c; 2007; 2009a.
25 Cf. Carstens 2012 and 2014.
26 See e.g., Fagrskinna, Ísl. Fornr. 29, 300.
27 See e.g., Jørgensen 1998; 2002; Andrén 2002, 315f.; Söderberg 2005, 109, 195f. In the Old
Norse traditions more specific ritual buildings, (“cultic houses”), seem to be designated
hof, goðahús and blóthús, while hǫrgr refers to an open-air or outdoor “sanctuary” or
“altar”. The Old Norse terms vé and perhaps *al seem to be more comprehensive concepts
corresponding to the operational terms “sanctuary” or “shrine”. See e.g., Vikstrand 2001;
Sundqvist 2005c; 2007; 2009a. For a critical discussion on *al as ‘sanctuary’, see Brink 1992;
Vikstrand 2001, 192; Elmevik 2004; Melefors 2014. On Goth alhs, see Jackson 2014a.
28 See e.g., Rostvik 1967; Vikstrand 2001, 211, 260.
29 Annan sal gerðu þeir, þat var hǫrgr er gyðjurnar áttu, ok var hann allfagr. Þat hús kalla men
Vingólf. Gylfaginning (Cod. Reg. Edda). (Ed) Faulkes 1988, 15.
30 Annan sal gerðu þeir er hǫrgr var í . . ., Gylfaginning (Cod. Ups. Edda). (Ed) Heimir Pálsson
2012, 24.
31 ef maðr . . . læðr hauga eða gerer hus oc kallar horgh. Here quoted from Kong Sverrers
Christenret 79. NGL 1, 430.
a designation. It is also possible that the meaning of the term varied at differ-
ent places in Scandinavia and was also altered over time. Scholars who have
searched for terms from the ancient Scandinavian languages to describe the
archaeological finds discovered at cultic sites have sometimes not taken into
consideration the problems the linguistic materials involve. The varying and
occasionally unclear meaning of hof makes it hard to connect this term to a
specific type of find or structure found at archaeological excavations. Because
of these conditions it is very important to formulate operational concepts
which may be applied as analytical instruments, when analysing ceremonial
buildings and sanctuaries in the present study.
37 A Latin Dictionary gives the following meanings of this world: ‘a space marked out’; hence,
in partic., in augury, ‘an open place for observation, marked out by the augur with his
staff’. In addition to this meaning the dictionary also gives following: 1) ‘an open, clear,
broad space’, 2) ‘a consecrated or sacred place, a sanctuary’, in part., ‘a place dedicated to
some particular deity, a fane, temple, shrine’. A Latin Dictionary, (eds.) Lewis & Short 1879.
Cf. Georges 1951II, 3050.
38 van der Leeuw 1956, 448ff.; Ringgren & Ström 1984 (1957), 28; Heiler 1961, 131ff.; Widengren
1969, 328ff.; Hinnells 1995, 517; Eliade 1991 (1949), 6ff.; 1987 (1957), 58–65; 1974 (1949), 371ff.
39 See e.g., Eliade 1991 (1949), 12f., 16f.; Widengren 1969, 344.
restricted area, and sometimes only the religious specialists have access to this
room or space. Ritual objects are also kept there, such as the images, sacrificial
tools and ceremonial equipment. These objects may not be used by common
people.
The synthetic concept “temple” has been applied to many cultures all
around the world, but perhaps particularly to sources describing ceremonial
buildings in the Mediterranean area. “Solomon’s Temple” in Jerusalem is one
example. It was built on a Canaanite pattern, and included a long building
which was opened on one of its sides.40 Inside the Temple, the building was
divided into three parts; a vestibule, a room for worship and the place usu-
ally called the “Holy of Holies”, i.e., the room reserved to Yahweh, and the Ark
of Covenant.41 In the second Temple, during post-exile period, this room was
only opened to the High-Priest, alone, once a year, on the Day of Atonement.42
According to the Old Testament, the Temple in Jerusalem was modelled on a
divine prototype.43
There is evidence that in their early history the Indo-Europeans (includ-
ing the ancient Scandinavians) did not build temples or make images of their
deities.44 Ethnographical observations made by Greeks and Romans support
these assumptions. According to Herodotus, Persians sacrificed on the highest
mountains and in an open space.45 Strabo describes feasts and sacrifices per-
formed under the open sky, on cliffs facing the sea with caverns underneath.46
Ovid reports that Romans worshipped their gods in consecrated groves:
“Under the Aventine there lay a grove black with the shade of holm-oaks;
at the sight of it you could say, ‘There is a spirit here.’ ”47 It seems, however, as if
the Greek and Roman people, as well as the Persians, gradually started to build
temples. Well-known is the temple of Apollo in Delphi. The ὀμφαλός (ompha-
los) or ‘navel’ situated in the holy area (ἄδυτον or aduton) symbolized the cen-
tre of the world.48 In Rome the term Lat templum refers to an inaugurated
rectangular area, which had been marked out by an augur ‘sage, diviner’, by
means of divine signs, in order to form the sanctuary according to the heavenly
40 Ringgren 1989, 90f.; cf. de Vaux 1980, 312ff.; Smith 1987, 47–73.
41 1 Kings 6; Fritz 2002.
42 Lev. 16:15; Ezek. 9:7.
43 1 Chron. 28:19.
44 For this aniconic tendency, see e.g., Widengren 1969, 339–346.
45 Herodotus I:131–132.
46 E.g., Strabo XI 7,5. See also Widengren 1965, 131, 161.
47 lucus Aventino suberat niger ilicis umbra, quo posses viso dicere, ‘numen inest’. Fasti III,
295f. Trans. Frazer. Widengren 1969, 340; cf. van der Leeuw 1956, 446f.
48 S. R. F. Price 1993, 135f.
prototype. The earthly templa were considered as loci effati, i.e., they were
relieved of sins. A templum often had a building with a large room (cella). The
sacrificial altar (ara) was situated in front of the building. The Roman term
harmonizing best with the concept of temple is actually Lat aedes sacra ‘tem-
ple, house of god(s)’.49 In the mystery cults too, temples and roofed sanctuaries
were used in cultic contexts. In the Mithras cult, for instance, underground
sanctuaries (Lat spelaeum, templum, crypta, fanum) were used, with an assem-
bly hall, a central gangway with benches on both sides, and at the very front,
round sculptures were situated together with the relief representing Mithras
killing the bull (tauroctonus).50 The sanctuary mirrored the cosmic cave created
by Mithras in primordial time.51 These Mithraic sanctuaries are attested in the
Germanic area.52
There is no doubt that some type of ceremonial buildings also existed in
Late Iron Age Scandinavia. Since the analytic concept of “temple” (from Lat
templum) was coined in a specific cultural context and arouses strong asso-
ciations of Greek-Roman Antiquity, and the Classical world (including e.g.,
Israel, Mesopotamia, Egypt, China, Persia and India), and in addition, the
empirical situation still is very unclear in ancient Scandinavia (see above),
I will use more neutral and tentative terms in this study when describing
such buildings. As noted above, archaeology indicates that two types of ceremo-
nial buildings existed in the pre-Christian context, i.e., (1) the multifunctional
and aristocratic banqueting buildings, and (2) the (smaller) more specific cultic
houses. In this work I will designate the multifunctional buildings “halls”, while
the more specific ritual structures will plainly be called “cultic houses”/“cultic
buildings”.53 In the hall building there might have been certain places or areas
which were regarded as ritual spaces more permanently, such as the high-seat
(hásæti). Otherwise the hall was probably only temporarily transformed into a
sacred building at particular times of year, for example during the calendrical
feasts.54 The cultic house/cultic building was more permanently regarded as a
sacred place.
49 Tomlinson 1987, 386ff.; Beard 2002, 22f., 87ff., 121ff., 196ff., 253ff.; Rüpke 2001, 180ff.; Golzio
2002.
50 Clauss 1990, 51ff.; Turcan 1993, 74ff.
51 Porphyry, De antro nympharum, 6.
52 Vermaseren 1956–1960 II, 47–143; Clauss 1992, 96ff.
53 In my opinion, we need more information from archaeology about these small cultic
buildings, before we may describe them as “temples”.
54 Cf. Gunnell 2001.
55 Historians of religions usually make a distinction between rite, ritual and ceremony. Spiro
and Rydving regard ritual as “the generic term for any kind of cult behavior, regardless
of its degree of elaboration or complexity”, while they understand rite as “the minimum
significant unit of ritual behavior”. They describe ceremony as the smallest configuration
of rites constituting a meaningful ritual whole” and ceremonial “the total configuration of
ceremonies performed during any ritual occasion”. See Spiro 1982, 199; Rydving 2011a,
37. The concept of “ceremonial building” refers to a building where rites, rituals and
ceremonies sometimes take place. Cf. Torun Zachrisson & Margareta Kempff Östlind
2006, who use the Swedish term “ceremoniella rum” in similar contexts.
56 For a discussion of these operational concepts, see e.g., Rydving & Kristoffersson 1993;
Nordberg 2010; 2011; 2014. See also Per Vikstrand’s (2001, 20–29; 2011, 315f.) general
descriptions of sacral places.
57 In the field of archaeology it has been emphasized during the last few decades that the
terms “private” and “public” may be problematic in such contexts, since a lot of things
that we today relate to the public sphere actually took place during the Viking Age in the
banqueting hall of the ruler or chieftain, which was located at his private farm. Anne-
Sofie Gräslund (2001) writes thus: “Det är emellertid tveksamt om man bör använda
begreppen [privat och offentlig] för vikingatiden med den innebörd de senare kommit
att få. Sannolikt var dessa två sfärer mycket mer inflätade i varandra, åtminstone så till
vida att mycket av det som skedde på gården, särskilt i stormansgårdarnas hallbyggnader,
i hög grad tillhörde det offentliga livet.” I agree with Gräslund’s opinion about these
terms, but still I will occasionally use them, even though I am aware of the fact that public
sanctuaries in ancient Scandinavia almost always were located on private farms.
That folk [the Svear] has a very famous temple called Uppsala (Ubsola),
situated not far from the city of Sigtuna (Sictona). In this temple, entirely
58 On Adam and his work, see e.g., Adam of Bremen, Historien om Hamburgstiftet och dess
biskopar, 1984.
59 P. H. Sawyer 1991, 16–19.
60 Nyberg 1984, 302–307.
61 Hultgård 1997, 9–15.
62 Hallencreutz 1997.
decked out in gold, the people worship the statues of three gods, in such
wise that the mightiest of them, Thor (Þórr), occupies a throne in the
middle of the room intended for ceremonial banquets. Wodan (Óðinn)
and Fricco (Freyr) have places on either side. The significance of these
gods is as follows: Thor, they say, presides over the air, which governs
the thunder and lightning, the winds and rains, fair weather and crops. The
other, Wodan—that is, the Furious—carries on war and imparts to man
strength against his enemies. The third is Fricco, who bestows peace and
pleasure on mortals. His likeness, too, they fashion with an immense
phallus. But Wodan they chisel armed, as our people are wont to repre-
sent Mars. Thor with his sceptre apparently resembles Jove. The people
also worship heroes made gods, whom they endow with immortality
because of their remarkable exploits, as one reads in the Vita of Saint
Ansgar they did in the case of King Eric.63
In scholion 138 we get some information about the surroundings of the temple:
Near this temple stands a very large tree with widespreading branches,
always green, winter and summer. What kind it is nobody knows. There is
also a spring at which the pagans are accustomed to make their sacrifices,
and into it to plunge a live man. And if he is not found, the people’s wish
will be granted.64
63 XXVI: Nobilissimum illa gens templum habet, quod Ubsola dicitur, non longe positum ab
Sictona civitate. In hoc templo, quod totum ex auro paratum est, statuas trium deorum
veneratur populus, ita ut potentissimus eorum Thor in medio solium habeat triclinio; hinc et
inde locum possident Wodan et Fricco. Quorum significationes eiusmodi sunt ‘Thor’, inquiunt,
presidet in aere, qui tonitrus et fulmina, ventos ymbresque, serena et fruges gubernat. Alter
Wodan, id est furor, bella gerit hominique ministrat virtutem contra inimicos. Tercius est
Fricco, pacem voluptatemque largiens mortalibus. Cuius etiam simulacrum fingunt cum
ingenti priapo. Wodanem vero sculpunt armatum, sicut nostri Martem solent; Thor autem
cum sceptro Iovem simulare videtur. Colunt et deos ex hominibus factos, quos pro ingentibus
factis immortalitate donant, sicut in Vita sancti Ansgarii legitur Hericum regem fecisse.
Translation of Adam’s text by Francis J. Tschan 2002, but slightly modified here.
64 Scholion 138: Prope illud templum est arbor maxima late ramos extendens, semper viridis in
hieme et aestate; cuius illa generis sit, nemo scit. Ibi etiam est fons, ubi sacrificia paganorum
solent exerceri et homo vivus inmergi. Qui dum non invenitur, ratum erit votum populi.
This scholion as well as scholia 139–141 are preserved in the oldest manuscript called A
2 (“the Leiden manuscript”). It is dated to c. 1100. See Hultgård 1997, 15. The last sentence
in scholion 141 (Hoc sacrificium fit circa aequinoctium vernale) is, however, only attested
late in the tradition of manuscripts, more precisely in manuscript A 3 from 1434. Hultgård
A golden chain goes round the temple. It hangs over the gable of the build-
ing and sends it glitter far off to those who approach, because the shrine
stands on level ground with mountains all about it like a theatre.65
For all their gods there are appointed priests to offer sacrifices for the
people. If plague and famine threaten, a libation is poured to the idol
Thor; if war, to Wodan; if marriages are to be celebrated, to Fricco. It is
customary also to solemnize in Uppsala, at nine-year intervals, a general
feast of all provinces of Svetjud. From attendance at this festival no one is
exempted. Kings and people all and singly send their gifts to Uppsala and,
what is more distressing than any kind of punishment, those who have
already adopted Christianity redeem themselves through these ceremo-
nies. The sacrifice is of this nature: of every living thing that is male, they
offer nine heads, with the blood of which it is customary to placate gods
of this sort. The bodies they hang in the sacred grove that adjoins the
temple. Now this grove is so sacred in the eyes of the heathen that each
and every tree in it is believed divine because of the death or putrefac-
tion of the victims. Even dogs and horses hang there with men. A certain
Christian informed me he had seen seventy-two miscellaneous bodies
suspended there. Furthermore, the incantations customarily chanted in
the ritual of a sacrifice of this kind are manifold and unseemly; therefore,
it is better to keep silence about them.66
1997, 30. In this study I consider the notes (scholia) 138–141 as parts of Adam’s text, even if
it is somewhat uncertain.
65 Scholion 139: Catena aurea templum circumdat pendens supra domus fastigia lateque
rutilans advenientibus, eo quod ipsum delubrum in planitie situm montes in circuitu habet
positos ad instar theatri.
66 XXVII: Omnibus itaque diis suis attributos habent sacerdotes, qui sacrificia populi
offerant. Si pestis et fames imminet, Thor ydolo lybatur, si bellum, Wodani si nuptiae
celebrandae sunt, Fricconi. Solet quoque post novem annos communis omnium Sueoniae
provintiarum sollempnitas in Ubsola celebrari. Ad quam videlicet sollempnitatem
nulli prestatur immunitas. Reges et populi, omnes et singuli sua dona transmittunt ad
Ubsolam, et, quod omni pena crudelius est, illi, qui iam induerunt christianitatem, ab illis
se redimunt cerimoniis. Sacrificium itaque tale est: ex omni animante, quod masculinum
est, novem capita offeruntur, quorum sanguine deos [tales] placari mos est. Corpora autem
In scholion 141 Adam mentions how many days the sacrifices were solemnized.
He also reports about the nature and number of the victims, as well as when
the feast took place:
Feasts and sacrifices of this kind are solemnized for nine days. On each
day they offer a man along with other living beings in such number that in
the course of the nine days they will have made offerings of seventy-two
creatures. This sacrifice takes place about the time of vernal equinox.67
When surveying all Adam’s notices on Uppsala, including the scholia, it seems
as if the cultic site there comprised several ritual elements and cultic places.
Besides the temple with the three statues of gods and the holy grove where the
sacrificial objects were hung, a specific holy tree and a well are mentioned. In
the latter place the pagans are accustomed to make their sacrifices. One scho-
lion also says that the shrine stands on level ground with mountains all about
it. Whether Adam in this case is referring to the royal mounds in Old Uppsala
is debated.
suspenduntur in lucum, qui proximus est templo. Is enim lucus tam sacer est gentilibus, ut
singulae arbores eius ex morte vel tabo immolatorum divinae credantur. Ibi etiam canes
et equi pendent cum hominibus, quorum corpora mixtim suspense narravit mihi aliquis
christianorum LXXII vidisse. Ceterum neniae, quae in eiusmodi ritu libationis fieri solent,
multiplices et inhonestae, ideoque melius reticendae.
67 Scholion 141: Novem diebus commessationes et eiusmodi sacrificia celebrantur. Unaquaque
die offerunt hominem unum cum ceteris animalibus, ita ut per IX dies LXXII fiant animalia,
quae offeruntur. Hoc sacrificium fit circa aequinoctium vernale.
68 As early as the 1920s Elias Wessén (1924, 131) wrote: “Omdömena om hans [Adam’s]
vederhäftighet och tillförlitlighet äro emellertid mycket olika. Den ene forskaren är
benägen att ta för gott allting, som kan stödja Adams auktoritet; den andre möter hans
framställning med mycket stor misstro.”
69 See e.g., de Vries 1956–57, §290. See also Nils Lid (1942, 86), who stated: “Ein må gå ut frå at
dei einskilde ting Adam fortel om, har eit faktisk grunnlag.” Folke Ström (1985, 79) wrote
thus: “I en berömd och ofta citerad skildring har Adam av Bremen återgett ett ögonvittnes
berättelse om de offer och ceremonier som utspelades i Uppsala vart nionde år”, while
Anne Holtsmark (1992, 17) argued: “Detta är en autentisk skildring av gudadyrkan i
Norden i slutet av den hedniska tiden.”
information, but his informants were eyewitnesses. One of them was the Danish
King Svein Estridsson (Sveinn Ástriðarson). He had lived among the Svear
for a long period (perhaps around the 1030s and 1040s) and was probably
most familiar with their customs, at least at that time. More recent research
is more sceptical.70 Scholars now argue that Adam’s description of Uppsala is
permeated with rhetorical adornments and missionary strategies. In the book
Uppsalakulten och Adam av Bremen this attitude is emphasized. According to
the editor of the book, Anders Hultgård, the text is a result of a literary process
in which the informants’ reports have been reworked by rhetorical embellish-
ment, ethnographic clichés and subjectivity. There are, for instance, two classi-
cal narrative genres present in Adam’s account, called evidentia and ekphrasis
(descriptio). Both aimed at sharpening his description. Elements from Adam’s
own imagination were added to make the narrative vivid and clear, for instance
in the description of the temple with the golden chain. Even if Hultgård’s criti-
cism is sharp, he reaches a fairly positive conclusion regarding Adam’s text.71
It has also been argued that the missionary strategy of Hamburg-Bremen
diocese may have affected Adam’s description of Uppsala and the religion
of the Svear. Carl Fredrik Hallencreutz, for instance, has shown that Adam
applied an Augustinian perspective when describing the pagan cult at
Uppsala. This included a historico-theological dualism, in which the civitas dei,
i.e., Christianity, was opposed by the paganism of Uppsala (civitas terrena).72
This tendency must be taken into consideration when evaluating the text. The
Augustinian perspective adds secondary elements to Adam’s account. Just
before describing the cult at Uppsala, he has this to say about Svetjud: “There
are Amazons, and Cynocephali, and Cyclops who have one eye on their fore-
heads; there are those Solinus calls Himantopodes, who hop on one foot, and
those who delight in human flesh as food, and as they are shunned, so may
they also rightfully be passed over in silence.”73
These examples show that Adam’s text must be treated with great care.
Adam had a specific purpose with his text. He wanted to legitimize the German
mission in Svetjud. Influences from the Bible are also present in Adam’s text.
The description of the gold-decorated temple, for instance, reminds us of the
portrayal of Solomon’s temple in the Old Testament.74 There are thus elements
in Adam’s text which may be regarded as unreliable.
The most radical criticism of Adam’s text has been formulated by the histo-
rian Henrik Janson. He states that the passage about the temple of Uppsala is
not aimed at pagan rituals, but as an allegorical satire against the Gregorians
and the affiliated Church (called gallicana ecclesia), which according to Janson
existed at Uppsala in the mid-eleventh century.75 In the middle of the elev-
enth century the archdiocese of Hamburg-Bremen was in a bitter feud with
the pope in Rome, Gregory VII. When Adam wrote about pagani in Sweden he
thus did not mean people of the ancient Scandinavian religion, but the mem-
bers of the affiliated Roman Church. The temple in Uppsala with its triclinium
actually referred to the Lateran palace and its triclinium majus. The three gods
mentioned by Adam were fictive and functioned as figurations of the sins that
the pope and the curia comprised: Wodan (Óðinn) represented fury (furor),
Thor (Þórr) pride or arrogance (superbia) while Fricco (Freyr) personalized
fornication and pleasure (voluptas).
4.3.3 Adam and the Problem with Christian Runic Stones in Old Uppsala
In my opinion, most of the recent source-critical contributions on Adam’s text
have been good for the research in this area. That theology, mission strategies
and rhetorical aspects affected his description is quite plausible. But some of
the criticism has gone too far, for instance Janson’s position. Janson denies that
a pre-Christian cult at all existed in Svetjud during the second half of the elev-
enth century. According to him, there are no sources supporting such a cult
in this area at that time. On the other hand, there are plenty of runic inscrip-
tions from this period testifying that Christianity was adopted by the Svear in
Uppland, with its centre in Old Uppsala.
I agree with Henrik Janson that the erection of runic stones in Uppland dur-
ing the eleventh century probably reflects the propagation of Christianity in
this province. But it is no obvious fact that Uppsala was the centre of these
activities, as suggested by Janson.76 As several scholars have noted, there are
surprisingly few runic inscriptions in Old Uppsala parish compared with the
neighbouring parishes.77 In Sveriges runinskrifter nine runic stones (U 978–986)
are mentioned from Old Uppsala parish,78 while of the neighbouring parishes
Vaksala has eighteen, Ärentuna thirteen and Bälinge as many as thirty-seven
runic stones. In Sveriges runinskrifter five of the runic stones are said to be (or
have been) located right beside the church and the vicarage: U 978, U 979,
U 980, U 981† and U 986†.79 It is exactly in this area that scholars have expected
to find the cultic site.80
The well-preserved U 978 is today mortared into the southern apse wall
of the church, on the outside. We do not know its original site. According to
Peringskiöld it had previously been placed on the altar, as a table, inside the
church. The stone is decorated with a cross and is dated to the first half of
the eleventh century.81
The stone U 979 lacks runes. It is raised against the western gable, outside the
church. According to Peringskiöld it was previously the threshold of the main
entrance of the church. It was decorated with a ship which had a Christian
cross as a sail. The picture has, however, been worn away.82
U 980 is now preserved only in fragments, three pieces. It was previously
mortared into the oven in the vicarage bakery. We do not know its original
site. Peringskiöld’s assumption that the stone “warit uprest någonstädes ute på
backen wid Kongsgårds Högarna” is according to Wessén and Jansson unprov-
able (“ett löst antagande”).83 The stone is depicted in Bautil (1750) standing in
front of the mounds, but this was just a guess. In fact the stone was broken
in pieces when it was discovered on the ground at the vicarage at the begin-
ning of the eighteenth century. On stylistic grounds it can be dated to
the second half of the eleventh century.84
U 981† does not exist today. We are dependent on the information from
Johannes Bureus and Johan Rhezelius living in the seventeenth century.
Seminghundra härad, is much higher. See A.-S. Gräslund 1987. Cf. Thompson 1975, 163;
R. Palm 1992, 98–100.
77 Cf. Sundqvist 2002, 112ff; Lerche Nielsen 2003; A.-S. Gräslund 2013.
78 Besides the nine stones there is one recently discovered runic stone from Ulva kvarn. See
Lerche Nielsen 2003.
79 Wessén & Jansson 1940–58, Vol. 4, 126–144.
80 A.-S. Gräslund 1997, 108.
81 A.-S. Gräslund 1997, 108. Wessén & Jansson 1940–58, Vol. 4, 126–129. U 978 can be dated to
c. 1025–1050 (Pr 2). See also A.-S. Gräslund 2013.
82 Wessén & Jansson 1940–58, Vol. 4, 129–132.
83 Wessén & Jansson 1940–58, Vol. 4, 132.
84 A.-S. Gräslund 1997, 108. U 980 can be dated to c. 1075–1100 (Pr 4). A.-S. Gräslund 2013.
According to them it was located in the wall of the vicarage bakery.85 Michael
Lerche Nielsen has recently argued, on good grounds, that U 981† never
existed.86 According to Lerche Nielsen “Bureus inscription [U 981†] is a con-
tamination of two inscriptions: U 980 and U 986†.”87 U 986† is now lost, but a
picture from the seventeenth century indicates that it originally stood by the
brook Samnan, situated on the royal manorial lands.88 It was thus not located
at Kungsgården as suggested by Sveriges runinskrifter.
None of these runic stones, therefore, are known with certainty to have
been raised in the area around the cathedral, the vicarage or the central parts
of the royal estate during the eleventh century. They may have been brought to
this place when the Romanesque church was erected.89 Old Uppsala thus can-
not be viewed as an “epicentre” for the raising of runic stones in Uppland, as
Janson does.90 The statement made by the runologist Michael Lerche Nielsen
harmonizes better with reality: “Gamla Uppsala är anmärkningsvärt fattigt på
runstenar.”91
There are also many other obscurities related to Janson’s hypothesis.92 One
fundamental problem with his theory is the question why Adam must criti-
cize the pope and Rome in an ulterior way, i.e., in an allegorical form. Adam
actually worked among like-minded people, who had the same attitude as he
had against Rome. His supreme leader, Archbishop Liemar, and the royalists
demonstrably rejected the pope at a synod in Worms 1076.93 Adam actually
had nothing to fear when criticizing Gregory and Rome. Janson’s suggestion
must therefore be regarded as far-fetched and unconvincing.
at least one or two decades later than the date of issue. This tomb reveals that
people in Old Uppsala at least until the middle of the eleventh century were
buried in accordance with the ancient pre-Christian practice.101 Likewise, cre-
mation graves in five smaller piles “in the cemetery beside the Kings’ Mounds”
(SHM 23316) indicate pre-Christian burial customs. They were examined by
Greta Arwidsson and Gunnar Ekholm in 1944. In two of these cremation tombs
double combs (Sw dubbelkammar) were found, suggesting that the graves
could be from the Late Viking Age, eleventh century or even younger.102
Old Norse sources confirm the idea that parts of Svetjud were Christianized
only at a late date.103 The story about Blotsveinn in the U-version of Hervarar
saga might reflect a pagan reaction somewhere in central Uppland during King
Ingi Steinkelsson’s reign 1080–1110.104 This account should not be interpreted
literally. Some details about Blotsveinn, for instance, are not historically reli-
able. On the other hand, we cannot rule out that the story reflects a real conflict
between King Ingi and those people who still remained at and defended the
“Uppsala temple”.105 The tradition about King Ingi and Blotsveinn also appears
in Orkneyinga saga, which is usually dated to the early thirteenth century (see
further ch. 9 below).
We also have information in the legend of St Botvid (Vita sancti Botvidi)
about a pagan revolt as late as the 1120s in the Mälaren area.106 This vita is pre-
served in a thirteenth-century manuscript, but most likely it was formulated
as early as the twelfth century.107 Other written sources likewise support the
existence of pagan cult in Svetjud in the late eleventh century.108
grave, that is, a form of burial which undoubtedly perceived as a pre-Christian custom.
But we cannot rule out that a syncretistic situation existed in Old Uppsala during the mid
eleventh century. See further below.
101 Personal communication from Hans Göthberg, October 2012. See also A.-S. Gräslund 2013.
102 A.-S. Gräslund 2013.
103 Cf. L. Lönnroth 1996, 141–158.
104 Hervarar saga. (Ed) E. O. G. Turville-Petre, 70f. Cf. (Ed) Jón Helgason, 160. Lönnroth
1996, 150f.
105 “Blotsven har förmodligen aldrig regerat vårt land . . . Däremot är det inte alls omöjligt att
berättelsen om honom trots allt bygger på minnen av någon autentisk konflikt mellan
den kristne kung Inge och de hedniska krafter som ännu höll fast vid Uppsalatemplet.”
Lönnroth 1996, 151.
106 SRS I, 377ff.
107 Schmid 1931, 109; N. Blomkvist et al. 2007, 186.
108 See DS, no. 25; SRS II, 389–404. See N. Blomkvist et al. 2007, 186.
109 According to Torun Zachrisson (1998, 126–164, 224f.), the Christian runic stones in Uppland
were raised by people belonging to a Christian community mainly resident in Sigtuna.
110 Adam IV,27. The Latin text is quoted above.
111 Cf. Hultgård 1997, 29; Göthberg et al. 2010, 34.
112 A.-S. Gräslund 2013.
113 There are many Icelandic stories of those who wavered in the faith during the transition
period. In Landnámabók (S218, Ísl. Fornr. 1, 250) we hear about Helgi the Lean, who was
very mixed in the faith. He believed in Christ, but invoked Þórr when travelling and facing
difficult business (Helgi var blandinn mjǫk í trú; hann trúði á Krist, en hét á Þór til sjófara
ok harðræða). Another who wavered in the faith was Óláfr Tryggvason’s poet Hallfreðr,
whose contemporary poetry from 990–1000 expresses the faith conflict. See Skj. B1,
147–163. See also Ljungberg 1938.
In the Mälaren area there are at least seventeen place-names containing the
name Þórr, sixteen with Freyr and twelve with Óðinn.123 In my opinion, it is
no coincidence that Adam mentions these particular gods, since they were the
most worshipped ones in the surroundings of Uppsala. Archaeological finds
may also support the existence of cult images or symbols representing these
gods in pre-Christian Svetjud. A small Viking Age bronze statue discovered
in Rällinge, Södermanland, for instance, has a phallic shape and might repre-
sent Freyr (see fig. 13 below). The phallic shape resembles Adam’s description
of Freyr in the “temple”: “His likeness, too, they fashion with an immense phal-
lus (cum ingenti priapo)” (see ch. 7 below).
Adam says that Thor (Þórr) has as his attribute a sceptre and thus he resem-
bles Jove (Thor autem cum sceptro Iovem simulare videtur).124 Þórr’s attribute
was probably not a sceptre in the Mälaren region during the Viking Age. In
this passage Adam has applied an interpretatio romana when comparing
Þórr with Jove. The most important symbol of Þórr was his hammer. It prob-
ably appears on two Viking Age runic stones from Södermanland, the Åby
and Stenkvista stones (see fig. 5). On the Altuna stone (eleventh century),
Uppland, we may see the mythical scene when Þórr with his hammer tries to
catch the Miðgarðsormr (see fig. 6). On this image the deity’s feet are pushed
down through the bottom of the boat, in exactly the same manner as Snorri
describes in Gylfaginning 48: “Then Þórr got angry and summoned up his ás-
strength, pushed down so hard that he forced both feet through the boat, . . .”125
There are plenty of Þórr’s hammer symbols in Uppland (see fig. 19A). Some of
these symbols are also found in Old Uppsala, for instance several Viking Age
Þórr’s hammer-amulets.126 Thus we have good evidence that at least Þórr was
worshipped in Old Uppsala and Svetjud.
Old Norse traditions, appearing for example in the Kings’ Sagas, also sup-
port Adam’s idea that pre-Christian Uppsala was an important cultic site as
well as a common assembly place among the Svear. Snorri Sturluson writes in
Óláfs saga helga 77 (1230):
123 See Vikstrand 2001, 55ff., 115ff., 141ff., 167ff. It has been argued that the god Ullr belonged to
a more ancient stratum of deities. See e.g., T. Andersson 1992c, 256.
124 Adam IV,26.
125 Gylfaginning 48: Þá varð Þórr reiðr ok færðisk í ásmegin, spyrndi við svá fast at hann hljóp
báðum fótum gǫgnum skipit . . . Cf. also Meulengracht Sørensen 1986.
126 Personal communication from Lena Jörpeland, Hans Göthberg and others in January
2013. One Viking Age Þórr’s hammer symbol was found in June 2015 in the area called
Kungsgården at Old Uppsala, close to the place where the hall building was found.
Personal communication from the excavator John Ljungkvist.
Figure 5
The Stenkvista runic stone, Södermanland (Sö 111).
Photo: Bengt A Lundberg. ATA,
Riksantikvarieämbetet.
Figure 6
The Altuna stone, Uppland (U 1161).
Photo: Pål-Nils Nilsson. ATA,
Riksantikvarieämbetet.
Svíaveldi. Óláfs saga Helga, in Heimskringla II, Ísl. Fornr. 27, 109. Trans. Hollander. See also
Ynglinga saga 34.
129 See e.g., Ynglinga saga, in Heimskringla I, Ísl. Fornr. 26, 23, 41, 65f.
130 Adam’s text was known in fourteenth-century Iceland. One passage of it is translated into
Old Norse in AM 415 4to (c. 1310) and in Flateyjarbók (c. 1390). But Snorri was probably not
acquainted with it. Cf. Lassen 2011, 122f.
131 On my view of the dating, provenance and source value of Ynglingatal, see Sundqvist
2002, 41–52. The designation dísarsalr also occurs in Hervarar saga 7, (ed.) E. O. G.
Turville-Petre, 30.
132 . . . ante edem Diane, dum ydolorum sacrificia faceret, equo lapsus expirauit. Historia
Norwegie IX, (eds.) Ekrem & Mortensen 2003, 76ff.
133 The author of Historia Norwegie was probably acquainted with Adam’s work. See Ekrem
& Mortensen 2003. Most likely, however, Snorri never used Historia Norwegie for his
presentation of Aðils.
was important in Uppsala can be seen in the name of the fair there, Disthingen
(OSw Disaþing).134 It was held during the spring and might be identical with
the calendrical feast mentioned by Adam (at the vernal equinox) and Snorri
(in the month of gói). Place-names also indicate that the dísir were worshipped
at specific sanctuaries in the eastern Scandinavian area. The name Disevid
in Östergötland, for example, has been interpreted as ‘the vi sanctuary of
the dísir’.135
Most recent research discusses whether the term templum in Adam’s text
really refers to an exclusive religious building. Evidence indicates that Adam
was actually referring to a multifunctional hall building when using this term in
this context. This building was not only intended for religious rituals and sym-
bols, but also for other purposes, such as a dining room at banquets, a room for
political-judicial meetings and a gathering place for the hirð. The room inside
the “temple”, where the idols were placed, is called triclinium by Adam. In
classical and mediaeval Latin it has both the meaning of ‘dining-room’ and
‘room for ceremonial banquets’.136 It should also be noted that the Latin term
used for designating the sacrificial rituals in the ceremonial building, libare,
actually refers to drinking ceremonies. It thus seems as if Adam’s “temple” resem-
bles the banqueting halls described in the medieval Old Norse texts. In Hauks
þáttr hábrókar, for instance, a hall (hǫll) located in Uppsala is mentioned. In this
hall King Eiríkr made libation sacrifices to the image of the god called Lytir.137
There is archaeological evidence for at least one banqueting hall in prehistoric
Uppsala. On Södra Kungsgårdsplatån, just north of the church, post-holes and
a stone structure belonging to a Merovingian Period hall (50 × 12 m) were found
(see fig. 2A and 2B).138 Also on the northern plateau (Norra Kungsgårdsplatån)
a large three-aisled building was recently found.139 It was smaller than the
hall on Södra Kungsgårdsplatån, but also dated to the Merovingian Period.
It seemed to have been a very exclusive workshop.140 We may recognize the
134 See e.g., Upplandslagen (1296 AD). SSGL 3: 274−275. See Naumann 1984.
135 See Sundqvist & Vikstrand 2014. This name is debated, however. See Elmevik 2005.
136 See Dillmann 1997, 65–69, 72.
137 Óláfs saga Tryggvasonar, Flateyjarbók I, 579−580. The text is quoted in ch. 9 below.
138 Nordahl 1993; 1996; Hedlund 1993; A.-S. Gräslund 1997. Scholars have also argued that the
last element in the name Uppsala refers to ‘banqueting halls’ (pl.). E.g., B. Gräslund 1993;
Brink 1999b, 38−39, 48−49, note 15; Herschend 2001, 39ff. For a recent treatment of the
name Uppsala, see Vikstrand 2013. New excavations at Old Uppsala in the summer of 2011
indicate that the hall was 50 metres in length.
139 Personal communication from the archaeologist and chief excavator at Old Uppsala,
Hans Göthberg, August 2011 and e-mail from John Ljungkvist, October 2011.
140 Personal communication from John Ljungkvist.
141 Cf. Ljungkvist 2013. A seventh century pendant of gold was found close to the hall in June
2015. Personal communication John Ljungkvist.
142 Personal communication from John Ljungkvist, January 2013.
143 Nordahl 1996, 63, 73.
144 Nordahl 1996, 60–62.
145 See Göthberg 2008. He has also shown that the houses indicated by means of ground-
penetrating radar (see Alkarp & Price 2005) cannot be confirmed by the archaeological
investigations made at the church in 2005.
146 Beronius Jörpeland et al. 2013; Sundqvist 2013, 89f. and personal communication Per
Frölund 2012 and Jonas Wikborg 2013 and 2014. This long row of posts could also indicate
the existence of a procession road.
this area, at least one ceremonial building and several great burial mounds
were located during the Merovingian Period and Early Viking Age. Perhaps the
large amount of deposited amulet rings (roughly 80), knives and bones, which
were found at the Fire Station in Old Uppsala in the summer of 2012, indicate
that the consecrated area was large and also extended to the east. These new
findings occurred in connection with a settlement with a large Late Iron Age
burial ground (with over 90 cremation graves).147
147 Personal communication from Lena Beronius Jörpeland and Hans Göthberg 2013.
148 See Herschend 1995; 1998; K. Lamm 2004; J. P. Lamm 2004; Arrhenius 2011; Lund Hansen
2011. For an overview of Helgö in general, see K. Lamm 1999.
149 Herschend 1995. Arrhenius suggests that the house was divided into four or five rooms.
See Arrhenius 2011, 20.
Figure 7a
The hall at Helgö. Find distribution in the hall on Foundation
Ia at Helgö. (a) Fragments of filigree glass claw and cone
beakers = open triangle. (b) Guldgubbar = open square;
weapons = filled square. (c) Knives = filled triangle; loom-
weights and wetstones = a dot. (d) Density map of the kind
of distribution. The first isarithmic curve marks areas with at
least 1 find per m2. The most dense square metre contains
8 finds, and the entire house (125 m2) contains 90 finds. The
‘H’ marks a presumed position of the high seat. After
Herschend 1998.
Figure 7b
Gold foil from Helgö. SHM.
Olof Sundqvist - 978-90-04-30748-3
Downloaded from Brill.com04/16/2019 08:51:54PM
via McGill University
Ceremonial Buildings And Sanctuaries 129
corner of the banqueting room there was a concentration of finds. This area has
been interpreted as the place of the high-seat.150 Most of the gold foil figures
and the expensive glass were found there. In front of the assumed “high-seat”
was also the hearth, which may have had a ritual significance during the cere-
monial banquets which took place in the hall. In my opinion, there is no doubt
that the hall at Helgö had a cultic significance. Besides the hall, an open-air
ritual site was found at the foot of a bare rock. Arrowheads, ceramics and food
were deposited there under a Viking Age triangular stone setting, i.e., a “tricorn”
(Sw treudd) (fig. 12A).151 This sacrificial place was used during Merovingian
Period too. There were also other large residences (longhouse VI and hall III A2)
during the Migration and Merovingian Periods located beside the cultic hall
(I A) at Helgö, indicating a situation similar to what is seen at Gudme, Funen
and Uppsala. During the Viking Age there were two hall buildings, I A and V A.
According to Birgit Arrhenius, several finds at these buildings indicate that
they should be interpreted as cultic. She argues, for instance, that the traces of
a rectangular platform in V A should be interpreted as something equivalent
to the Old Norse seiðhjallr, used during divination rituals (ON seiðr).152 She
also suggests that the post which stood in post-hole 49 in hall building I A was
a large “phallus post”.153 In addition to this, Arrhenius proposes that some
kind of ceremony also took place at the early boat-shaped building III A1 in
Building Group 2. The boat kept in this building was used during sacrifices to
the gods, especially Freyja. It was filled with sacrificial objects and then carried
in procession to the lake. In my opinion, some of Arrhenius’s suggestions must
be considered very hypothetical.
The place-name Helgö (OSw *Hælghø) nevertheless indicates that this place
undoubtedly was an important cultic site.154 This name has been interpreted
as ‘the holy island’ or ‘the island where peace prevails’.155 It has been argued
that the name indicates that certain activities took place at this island, such as
trade and cult, which were both protected with peace.
Sanda in Central Uppland (see ch. 5 below) and the island of Birka in
Mälaren (see ch. 6 below) also have remains of Late Iron Age halls, appear-
ing in contexts of finds indicating cult.156 Similar indications of Iron Age halls
(sometimes on terraces) have also been found at several other central places
in the Mälaren region, for instance at Hovgården on Adelsö, at Fornsigtuna,
Granby-Hyppinge in Vallentuna, Karsvik in Bromma, Runsa in Upplands
Väsby, Barksta in Västmanland, Valsgärde and Vendel in Uppland.157 In these
cases it is mainly the construction and/or the location of these buildings which
indicates that they may have had a possible cultic or ceremonial function in
connection with banquets. At the hall of Husby, Glanshammar, Närke, dated to
the Migration and Merovingian Periods, there was a demarcation or palisade
around the building.158 This demarcation, and the gold foil figure discovered to
the west of the hall, support the idea that this building may have had a ritual
function. Perhaps a hall was also located at the same spot as the monastery in
Eskilstuna. Adjacent to a row of stone-edged post-holes two gold foil figures
were discovered, dated to c. 700.159 A gold foil figure from the eighth or ninth
century was also found at the hall building of Svintuna, in Östergötland.160
There is also slight evidence of more specific cultic houses in the Mälaren
region.161 To the north of the Middle and Late Iron Age hall in Lunda,
Södermanland, remains of a small building (3 × 6 m) were found together with
two phallic figures.162 A third figure was found in the yard south of the hall.
According to the archaeologists the small house should be interpreted as a spe-
cific cultic house, comparable to the structures found at Tissø and Järrestad.
In the yard to the north of the hall, several pits were discovered, probably
intended for ritual cooking. Þórr’s hammers, amulets and miniature scythes
156 Åqvist 1996; Holmquist Olausson 2001; Holmquist Olausson & Kitzler Åhfeldt 2002.
157 Brunstedt 1996; Damell 1991; 1993; A. Carlsson 1997; Hedman 1989; Ringsted 2005;
Karlenby 2005; Norr & Sundkvist 1995; Arrhenius 2000; Skre 2007a; Olausson (ed.) 2011
and Olausson 2014.
158 It should be noted that several halls (“house II” and “house III”) may have succeeded each
other. Ekman 2000, 25.
159 See Sune Zachrisson 1963 and 1998, 13f. The archaeologist Torun Zachrisson (the daughter
of Sune Zachrisson) has orally suggested that there are reasons to interpret the finds in
Eskilstuna as a hall.
160 Herschend 1998.
161 Indications of a more specific three-aisled cultic house may have been found at Alby in
Botkyrka parish, Södermanland. See Hjulström 2008.
162 G. Andersson et al. 2004.; Skyllberg 2008. Cf. L. Larsson 2011.
were also found in this yard. A grove was situated on a hill approximately 200
metres to the west of this ritual area. Archaeological finds show that rituals
were performed there too.
163 For a critical investigation of these terms, see Vikstrand 2001; 2002, 129ff.; 2004a.
164 It should be noted that the word hof could also mean ‘farm’ or ‘elevation, height’. The
latter meaning is preserved in Norwegian dialects. Sandnes 1992, 261; Vikstrand 2002, 132.
See also below.
165 Vikstrand 2001, 256ff., 267ff. There are nine places called Hov (simplex) in the Mälaren
region, seven in Uppland, one in Södermanland; one in Västmanland. On the word ON hof
as a ceremonial building in different contexts, see also Sundqvist 2009a.
166 See Vikstrand 2000.
167 See the discussion in Vikstrand 2013.
168 In this region there are forty-nine place-names which include the term vi.
169 Vikstrand 2001, 298ff.; 2002, 129f.
170 Bäck et al. 2008.; Hållans Stenholm 2010; and on the place-name, see Vikstrand 2010.
No ceremonial building was found, however. Also very interesting is the Viking
Age site of Götavi in Närke.171 The last element of the name, -vi, indubitably
designates a holy place.172 Archaeological finds also indicate that this site was
an important outdoor cultic place during the Viking Age.
To sum up, the combination of a monumental ceremonial hall and royal burial
mounds in Uppsala indicates that this site was an important sanctuary in the
Late Iron Age. In my opinion, there are no reasons to rule out that this place
had a similar function in the middle and at the end of the Viking Age, even
if the archaeological evidence does not clearly indicate this. If Uppsala had
lost its significance during this period, one must ask why it was so important
to establish an episcopal see there during the 1120s, and later make this place
into the archiepiscopal see in 1164. My only answer is that Uppsala had an
enormous symbolic significance for the Svear during the entire Viking Age.
The symbolic capital that Uppsala constituted was also useful for the church.
Some information in Adam’s description may thus be based on authentic con-
ditions, for instance that Uppsala was an important sanctuary for the Svear.
Written sources, place-names and archaeological finds indicate that pre-
Christian halls and cultic buildings also existed at other places in the Mälaren
area. At certain places, such as Helgö, we may see the pattern of a large resi-
dence situated beside a small hall building. Perhaps there was a functional dif-
ference between these buildings (see above).
173 Heimskringla I, Ísl. Fornr. 26, 166–73. On Heimskringla, see Whaley 1999.
Snorri also states that Earl Sigurðr was generous and that he once made a
great sacrificial feast at Lade, defraying all expenses himself. He supports his
narrative with a stanza from Kórmakr Ǫgmundarson’s Sigurðardrápa (960 AD)
where Earl Sigurðr is praised.
In the frame story Snorri recounts how King Hákon the Good had had a
Christian upbringing in England. He was unenthusiastic about the pagan
cult in Norway, and wanted to convert the Norwegians gently. The farmers of
Trøndelag opposed him and at the Frostaþing required the king to take part in
the rituals: “The farmers said that it was their wish that the king should make
sacrifice to procure for them good crops and peace, as his father had done.”175
Later, during the autumn, the king came to the annual sacrificial feast at Lade:
174 Trans. Hollander (somewhat modified). Íslenzk Fornrit has this text: Sigurðr Hlaðajarl var
inn mesti blótmaðr, ok svá var Hákon, faðir hans. Helt Sigurðr jarl upp blótveizlum ǫllum af
hendi konungs þar í Þrœndalǫgum. Þat var forn siðr, þá er blót skyldi vera, at allir bœndr
skyldu þar koma, sem hof var, ok flytja þannug fǫng sín, þau er þeir skyldu hafa, meðan
veizlan stóð. At veizlu þeiri skyldu allir menn ǫl eiga. Þar var ok drepinn alls konar smali
ok svá hross, en blóð þat allt, er þar kom af, þá var kallat hlaut, ok hlautbollar þat, er blóð
þat stóð í, ok hlautteinar, þat var svá gǫrt sem stǫkklar, með því skyldi rjóða stallana ǫllu
saman ok svá veggi hofsins útan ok innan ok svá støkkva á mennina, en slátr skyldi sjóða til
mannfagnaðar. Eldar skyldu vera á miðju gólfi í hofinu ok þar katlar yfir. Skyldi full um eld
bera, en sá, er gerði veizluna ok hǫfðingi var, þá skyldi hann signa fullit ok allan blótmatinn,
skyldi fyrst Óðins full—skyldi þat drekka til sigrs ok ríkis konungi sínum—en síðan Njarðar
full ok Freys full til árs ok friðar. Þá var mǫrgum mǫnnum títt at drekka þar næst bragafull.
Menn drukku ok full frænda sinna, þeira er heygðir hǫfðu verit, ok váru þat minni kǫlluð.
Heimskringla I, Ísl. Fornr. 26, 167f.
175 Bœndr segja, at þeir vilja, at konungr blóti til árs þeim ok friðar, svá sem faðir hans gerði.
Heimskringla I, Ísl. Fornr. 26, 170.
the sign of the cross over it. Then Kárr af Grýtingi said “Why does the king
do that? Doesn’t he want to drink of the sacrificial beaker?” Sigurðr jarl
made answer, “The king does as all do who believe in their own might and
strength, and dedicated his beaker to Þórr. He made the sign of the ham-
mer over it before drinking.” People said no more about it that evening.
Next day when people had seated themselves at the table, the farmers
thronged about the king, saying that now he must eat the horse meat.
That, the king would not do under any condition. Then they asked him to
drink the broth from it. He would not do that, either, and they came near
to making an attack on him. Sigurðr jarl said he would help them come to
an agreement, asking them to cease their tumult; and he asked the king
to gape with his mouth over the handle of the kettle on which the smoke
of the broth from the horse meat had settled, so that the handle was
greasy from it. Then the king went up to it and put a linen cloth over the
handle, and gaped with his mouth over it. Then he went back to his high-
seat, and neither party was satisfied with that.176
Later, at Yule, the king visited Mære, where the chieftains and farmers held
ceremonial meals. On this occasion he was forced to eat a few pieces of horse-
liver. He also drank the toasts without making the sign of the cross.
176 Um haustit at vetri var blótveizla á Hlǫðum, ok sótti þar til konungr. Hann hafði jafnan fyrr
verit vanr, ef hann var staddr þar, er blót váru, at matask í litlu húsi með fá menn. En bœndr
tǫlðu at því, er hann sat eigi í hásæti sínu, þá er mestr var mannfagnaðr. Sagði jarl, at hann
skyldi eigi þá svá gera. Var ok svá, at konungr sat í hásæti sínu. En er it fyrsta full var skenkt,
þá mælti Sigurðr jarl fyrir ok signaði Óðni ok drakk af horninu til konungs. Konungr tók við
ok gerði krossmark yfir. Þá mælti Kárr af Grýtingi: ’Hví ferr konungrinn nú svá? Vill hann enn
eigi blóta?’ Sigurðr jarl svarar: ’Konungr gerir svá sem þeir allir, er trúa á mátt sinn ok megin
ok signa full sitt Þór. Hann gerði hamarsmark yfir, áðr hann drakk.’ Var þá kyrrt um kveldit.
Eptir um daginn, er menn gengu til borða, þá þustu bœndr at konungi, sǫgðu, at hann skyldi
eta þá hrossaslátr. Konungr vildi þat fyrir engan mun. Þá báðu þeir hann drekka soðit. Hann
vildi þat eigi. Þá báðu þeir hann eta flotit. Hann vildi þat ok eigi, ok var þá við atgǫngu.
Sigurðr jarl segir, at hann vill sætta þá, ok bað þá hætta storminum, ok bað hann konung
gína yfir ketilhǫdduna, er soðreykinn hafði lagt upp af hrossaslátrinu, ok var smjǫr haddan.
Þá gékk konungr til ok brá líndúk um hǫdduna ok gein yfir ok gekk síðan til hásætis, ok líkaði
hvárigum vel. Heimskringla I, Ísl. Fornr. 26, 171f. Trans. Hollander, somewhat modified.
177 See Olsen 1966, 59ff.
In a general sense the criticism by Olsen, Walter, Düwel and Baetke180 is well-
founded. Düwel was, for instance most likely right in his conclusions about
the terms hlautteinn and hlaut. However, sometimes this source criticism has
been somewhat simplified and without nuances, especially when stating that
Snorri had no access to ancient sources on these rituals. Preben Meulengracht
Sørensen gave, in my view, a more balanced account.181 He agreed with Olsen
and Düwel that some of the details in Snorri’s hof descriptions cannot be veri-
fied and thus it is not a trustworthy image of the pagan cult, as previous histo-
rians of religions sometimes claimed. He also agreed with these scholars that
these texts must be related to Snorri’s Christian view of history. However, this
does not mean that Snorri was not at pains to convey a reliable image of the
pre-Christian sacrificial cult. Meulengracht Sørensen also wrote thus concern-
ing Düwel’s critical opinion of the terminology that Snorri applied in his text:
At de centrale religiøse ord ikke med sikkerhed kan føres tilbage til
hedensk tid, er ikke et tilstrækkeligt argument for, at de begreber, de
udtrykker, ikke har været hedenske. Det samme gælder de parallelste-
der, som Düwel påviser i den kristne litteratur. De er ikke enbetydende
med, at Snorri har lånt sin fremstillings enkeltdele derfra. Begge forhold
viser kun, at Snorri har udtrykt sig ved hjælp af kristen tankegang og
terminologi.182
Most scholars interpret the expression valdr vés as “the ruler of the sanctuary”,
i.e., Earl Sigurðr. Klaus Düwel has rejected this interpretation and suggested
that the sequence valdr vés vægja should be interpreted as “der Beherrscher
des Thingplatzes der Schwerter (= Schilde) = der Kriger Sigurðr”—that is, “the
warrior Sigurðr”.187 According to Düwel, there is nothing supporting an asser-
tion that Snorri knew about ancient traditions of the ritual feasts (blótveizlur)
in Trøndelag; he had only support for the idea that Sigurðr was regarded as gen-
erous. Düwel’s position, however, has been contradicted. Skilled philologists
such as Ottar Grønvik and François-Xavier Dillmann have put forward strong
arguments in favour of the old interpretation of valdr vés as “the ruler (protec-
tor) of the sanctuary”.188 Grønvik relates vægja to kind ‘offspring’ and interprets
the former word as a hapax legomenon of a noun vægir (m.) ‘godfearing man,
that is, a man who is full of fear (for the deity)’.189 Vægja (gen. pl. of vægir)
is interpreted as a nomen agentis of the verb vægja ‘turn aside’ (Norwegian
186 “You do not need to bring with you/ neither basket nor tankard/to the generous man./
The gods deceived Þjazi./ All men should avoid opposing/ the keeper of the sanctuary,/
because he makes the chieftains glad./ The ruler battled for the gold.” Heimskringla I, Ísl.
Fornr. 26, 168. See also Skj. B1, 69f. My trans.
187 Düwel 1985, 14–17.
188 Norwegian: “véets (templets, hovets) høvding, hersker”. Grønvik 1989, 82–90. See also
Dillmann 1997.
189 Norwegian: “gudfryktig mann, d.e. mann fylt av vægje, age (for guddommen)”.
‘vike til side’). Accordingly, the half-stanza is interpreted as: “Which offspring
of godfearing men will oppose the ruler (protector) of the sanctuary?’190
In my opinion, Grønvik’s interpretation is quite convincing. It seems as if
Snorri at least had some support for his description of ceremonial meals at cul-
tic sites in Trøndelag, by means of the tenth-century stanza in Sigurðardrápa.
The first half-stanza mentions that nobody needed to bring food to the ban-
quets arranged by Sigurðr, because of his generosity. Since the other half of
the stanza indicates that Sigurðr was regarded as a protector of the sanctuary,
we may suppose that the first half of the stanza actually referred to a religious
feast, i.e., a blótveizla. The manager and agent of this feast was thus the earl
himself.
Other skaldic poems may likewise support Snorri’s information that
Norwegian earls and kings took care of the public cult in Trøndelag.191 According
to Vellekla 15–16 (c. 990), which was quoted by Snorri, Earl Hákon of Lade also
continued to perform cultic duties as his father had done. He thus restored the
sanctuaries which had been destroyed by the sons of Eiríkr. A similar cultic
role is also applied to King Hákon the Good in Hákonarmál 18 (c. 960). When
the king died the skald praised him as follows: “Then it was made known how
well that king had protected sanctuaries (þyrmt véum) when all those who
have domination and power [the gods] bade Hákon welcome.” It thus seems as
if Hákon became an apostate or turned to a tolerant attitude towards pagan-
ism. This poem too was quoted by Snorri.192
Snorri thus had other sources, besides Kormákr’s stanza, when writing about
the pagan cult in Trøndelag. Some of his sources consisted of older prose tradi-
tions. Ágrip (1190 AD), for instance, tells us that the Þrœndir in Mære required
King Hákon the Good “to sacrifice as other kings used to do or we throw you
out of the country, if you do not follow us in this thing”.193 The story about the
horse liver also occurs in this text, but in another version and in much sparser
terms.194 Fagrskinna, written sometime between Ágrip and Heimskringla,
also reproduces this tradition briefly. At the assembly in Mære “the people of
190 Norwegian: “Hvilken ætling av agefylte (gudfryktige) menn/ vil vel strides med (sette seg
opp mot) templets herre?” In the first half-stanza, Sigurðr’s generosity at feasts is praised
and the stanza thus indicates a pre-Christian tradition of ceremonial banquets in Lade
led by Earl Sigurðr.
191 See also Jón Hnefill Aðalsteinsson 1998, 62ff
192 Heimskringla I, Ísl. Fornr. 26, 193–197.
193 . . . báðu hann blóta sem aðra konunga í Nóregi, ‘ella rekum vér þik af ríki, nema þú gerir
nekkvern hlut í samþykki eptir oss.’ Ágrip, Ísl. Fornr. 29, 8.
194 . . . at hann biti á hrosslifr, ok svá, at hann brá dúki umb ok beit eigi bera, en blótaði eigi
ǫðruvís. Ágrip, Ísl. Fornr. 29, 8.
Trøndelag gave him [King Hákon] two things to choose between. Either he
should sacrifice as kings used to do before and in such a way maintain old
law/custom for good crops and peace, or else they would throw him out of the
country”; “because of this threat and out of love for his friends the king yielded
to their demands and sacrificed.”195 Compared to Snorri’s version, Fagrskinna
presents the events at Mære as more of matter-of-fact statements. The author
of Fagrskinna has a sober style and avoids much of the wonders and the super-
natural elements that Snorri was so fond of.196
Several other Old Norse texts report that both Lade and Mære had well-
known pre-Christian hof sanctuaries during the Viking Age.197 According to
them, these buildings seem to have had a function and significance similar
to the hof sanctuaries in Snorri’s accounts. Most likely it was widely known
knowledge among learned medieval Icelanders that Lade and Mære were pre-
Christian cultic sites.
Snorri Sturluson’s Hákonar saga góða and other medieval prose thus indicate
that Lade and Mære were important pre-Christian cultic sites with ceremonial
buildings. Snorri’s description of these sites and the cultic activities that took
place there has been partly distorted by Judaeo-Christian elements. In my opin-
ion, Snorri had some ancient traditions available when describing these sites,
primarily from skaldic poetry, but also from prose traditions. Archaeological
sources also indicate that some of his information is quite sound (see below).
Therefore, we should not discard all of his information about ceremonial
buildings. We will return to detail in Snorri’s accounts and other texts below, in
the thematic analysis.
195 . . . ok á því þingi gørðu Þrœndir konunginum tvá kosti, at hann skyldi blóta eptir vanða
enna fyrri konunga ok fylla svá en fornu lǫg til árs ok friðar, elligar mundu þeir reka hann
af ríkinu, ef hann vildi <eigi> . . ., þá gørði hann eptir bœn þeira ok blótaði. Fagrskinna, Ísl.
Fornr. 29, 80.
196 The diplomatic role of Earl Sigurðr in Hákonar saga góða does not appear in the two
older prose versions and may be seen as a supplement by Snorri himself. Perhaps Snorri’s
positive description of the earls was connected to his friendship with Earl Skúli (of Lade),
whom he met in Tønsberg sometime around 1220. Steinsland 2000, 112.
197 See e.g., Landnámabók S 297, H 258, Ísl. Fornr. 1, 307f.; Óláfs saga Tryggvasonar eptir Odd
munk Snorrason, Ísl. Fornr. 25, 193 (AM 310 4to) and 278ff.; Brennu-Njáls saga 87, Ísl. Fornr.
12, 210.
Apart from some Late Iron Age weapons there is not much indicating high sta-
tus at this site.198 Gerhard Schøning’s and Klüwer’s reports from the eighteenth
and nineteenth centuries indicate that ancient grave mounds and monuments
were situated there when they visited the place (see ch. 2 above). These graves
and monuments are not visible today. Anyhow, the written sources and the
fact that Lade in the Middle Ages was regarded as crown land support
the notion that the Hlaðajarlar really controlled a public sanctuary at this
site. The fact that a fylkiskirkja for Strindafylki was erected at this site in the
eleventh or the twelfth century also supports this conclusion, since cultic place
continuity is attested at other places in Trøndelag (see below).
Mære in Inn-Trøndelag is mentioned as a pre-Christian cultic site sev-
eral times by Snorri and in other Old Norse sources.199 The important text
Landnámabók, for instance, which may have been composed as early as the
twelfth century, describes a Viking Age cultic leader (hofgoði) and his cer-
emonial building (hof ) situated in Mære.200 Not only written sources but also
archaeological finds produce good evidence of a pre-Christian ceremonial
building at Mære. In 1966–1967 archaeological excavations were conducted
under the floor of Mære church. These excavations were led by Hans-Emil
Lidén.201 The church, which is dated to the last part of the twelfth century,
is placed on the top of a hill, east of the Borgenfjord, called “Mæreshaugen”
(fig. 8a). In the prehistoric period Mæreshaugen was an island in the fjord, which
completely surrounded Inderøya.202 Under the stone church, vestiges of an
early wooden church were discovered. It was surrounded by a churchyard, with
about sixty graves. At the church the excavators also found remains of two pre-
Christian buildings. The oldest one was a building with wattle-and-daub walls
and dated to the Migration Period. The function and character of this building
is uncertain. The later one was a Viking Age building with huge post-holes. The
form of this house, its dimensions and orientation are unknown, however.
The only traces of it were remains of an assumed “palisade wall”, lumps of burnt
clay with twigs and branches, fire-cracked stones and post-holes with remains
of timber. The post-holes formed irregular circles with diameters varying from
80 to 100 cm. In the post-holes some 19 (or 23)203 gold foil figures were found,
which probably should be dated to the Early Viking Age (fig. 8b). Most of them
Figure 8a
The church of Mære, on the top
of a hill, east of the Borgenfjord,
called “Mæreshaugen” in
Trøndelag.
Photo: Olof Sundqvist.
Figure 8b
Gold foils from Mære, Norway.
Figure 8c
Gold foil from Borg, Lofoten, Norway.
were related to a structure which was interpreted as the place of the high-seat.204
On these figures a couple was depicted, standing against each other, a man
and a woman. Whether these symbols shall be regarded as sacrificial objects
or some kind of mythical insignia for aristocratic families is uncertain (see
ch. 11 below). At this site other finds were made which may likewise be associ-
ated with pre-Christian ceremonial feasts, such as pieces of pottery and glass,
quantities of animal bones and some characteristic cooking stones.
Place-names may also support the idea that Mære was an important pagan
cultic site in Inn-Trøndelag. Close to the church of Mære there is a place called
Hælgjin, ‘the holy meadow’ or rather ‘the holy meadows’.205 The composite
name Mære (ON Mærin) has the same last element vin ‘meadow’ as Hælgjin,
while the first element mærr, according to one interpretation, means ‘famous’,
referring to the famous and holy meadow at this site, where specific rituals
were performed during the religious gatherings.206
It may therefore be suggested that the Viking Age house at Mære was some
type of pre-Christian ceremonial building. Whether this house was a multi-
functional hall or a cultic house is almost impossible to say. Lidén suggested
that the Viking Age building should be related to the ceremonial buildings
called hof in the written sources. As noted above, the term hof in the written
sources may have semantic variations in different contexts. Nevertheless, the
place indicates high status with the gold finds. The medieval farm was consid-
ered as crown land in the Early Middle Ages.207 The finds at Mære also indicate
that we have a clear example of cult-place continuity, that is to say, the early
Christian church was erected on the pagan cultic site.208 Lidén even suggested
that the Viking Age building where the gold foil figures were found was turned
into a church after the conversion. Christian graves were found associated with
this building, which could not be related to the wooden church.
As far as I know, there is little archaeological evidence of other ceremo-
nial buildings in Trøndelag similar to Mære. In the summer of 2010, however,
“a well-preserved cultic site” was found at Ranheim in Sør-Trøndelag. According
204 This interpretation was later reconsidered by Lidén (1999, 2, 38f.). He rejected the high-
seat interpretation.
205 Dillmann 1997, 71, not 43. Critically considered by Sandnes 1987, 151.
206 Dillmann 1997, 63f. The adjective mærr means ‘fræg [berømt], gjæv, gild’. See Heggstad et
al. 1993. Sandnes & Stemshaug (2007) state that the first element is uncertain. It could be
an ON adjective meaning ‘herleg, gjæv’ or perhaps a noun mæri meaning ‘grense’.
207 Røskaft 2003, 143f. See ch. 5 below.
208 Whether this pre-Christian ceremonial building was used as a church after conversion is
debated. Cf. Lidén 1969, 20ff., 30ff.; 1996; Olsen 1969, 26f.
to the excavator, Preben Rønne, this site consisted of a flat, roughly circular
stone cairn about 15 metres in diameter and 1 metre high.209 This cairn was
interpreted as something equivalent to an Old Norse hǫrgr and dated to
c. AD 400. Just beside the cairn was an almost rectangular building with strong
foundations, and a processional avenue marked by two stone rows. It was
argued that this house was an Old Norse hof sanctuary and dated to the period
AD 895 to 990. The posts belonging to the house had been pulled out and all of
the wood removed, and the cairn had been carefully covered with stones and
clay. Afterwards, the whole site had been entirely covered with earth, presum-
ably when the people there converted to Christianity.210
One very early hall building has been discovered at Hovde, Ørland, in Sør-
Trøndelag.211 This place is located strategically at the strait between Ørland
and Agdenes, i.e., the entrance to the Trondheimsfjord. Around this strait there
are many Iron Age burials. The three-ailed hall was found associated with
several other buildings during excavations in the summer of 1997. Two phases of
settlement were identified at this site. The first was dated to Pre-Roman Period
and displayed a farm consisting of three long-houses. House D was 27.9 m
long. It was distinguished from the other houses, since it was built of very large
posts, up to 50 cm in diameter, and it had a clay floor. It had several fireplaces
stretching over the whole length of the building and no stable for the animals
was found. Because of these features the house was interpreted as a hall, i.e.,
a building used for extraordinary purposes. All three houses (C, D and E) were
enclosed by a ditch and a fence. In the next phase, dated to Roman Iron Age
and Migration Period, it seems as if House A took over the functions of House
D from the previous phase. The houses in the second phase were also enclosed,
this time by a fence consisting of a row of parallel posts. At these houses too,
several fireplaces and cooking pits were found. One fireplace was dated to
the period AD 390–550, perhaps indicating when the farm was abandoned. The
long continuity of the farm and houses with certain functions indicates that
this place was once very special. North-east of Hovde is a place called Hov and
another place called Balsnes, perhaps ‘the isthmus of Baldr’.212 The former name
indicates that pagan cult took place in this area. Middle and Late Iron Age houses
have also been found at other places close to Hovde. The medieval church,
for instance, was erected at Viklem. About eighty post-holes were found there
and forty cooking pits and hearths, dating back to the period between
AD 400 and 800. The post-holes were remains of long-houses, but also a
pit-house.213 Beside the church there is also a Late Iron Age royal burial mound.
Most likely pre-Christian ceremonies took place at this site.
Interesting Middle and Late Iron Age long-houses have also been found at
other places in Trøndelag. Beside the three royal burial mounds at Bertnem
in Overhalla, Nord-Trøndelag, some type of long-houses were erected (see
fig. 27C). These houses are dated to Migration and Merovingian Period, but we
do not know much about them.214 Important Late Iron Age houses have also
been found at, for instance, Melhus in Gauldal, and at Skei in Steinkjer.215
The most interesting parallel to the ceremonial building in Mære is found at
Borg at Vestervågøy, i.e., the second largest of the Lofoten islands. At this place
five gold foil figures were found associated with a monumental hall building.
This place is also interesting in the context of the Lade Earls, since this fam-
ily is said to have originated from the north of Norway, more precisely from
Hálogaland. In the excavations at Borg in the 1980s and 1999–2002 a settlement
was found dated to c. AD 200–1000.216 The investigations were concentrated
on a house structure I:1, where a huge house called I:1b (64 m long and 7–8 m
wide) was erected during the Migration Period. Around AD 600 an even larger
house called I:1a was raised, 83 m long and 7.5–9 m wide. Both houses were
three-aisled wooden buildings and both comprised certain rooms which
were interpreted as a “hall-room”. A lot of finds were made at structure I:1. Some
finds were exclusive and may have indicated cultic actions, such as two pottery
jugs and the seventeen or eighteen glass beakers imported from the Continent
and the British Isles. Most interesting in the context of the Trøndelag sanctu-
aries is the five gold foil figures found in the younger “hall-room” (Room C)
of Borg I:1a, depicting the motif of a couple (fig. 8c). These objects seem to
be related to post-holes and the high-seat area. The assumed high-seat was
located in the north corner of the room, where a bronze vessel was also found,
as well as the gold pointer, the Tating ware jugs and most of the glasses. In this
hall-room (14 m long) there were also some unusual fireplace structures, which
may indicate that ritual activities took place there during the Early Viking Age.
It was argued that this room had two main functions; during most of the year
217 Elsewhere in Norway, Iron Age halls have also been found at Borg, Forsand in Rogaland, see
Løken 2001; and at Huseby, Tjølling in Vestfold. See Skre 2007a.
218 See M. Olsen 1926, 230f.; Røskaft 2003, 56; Sandnes & Stemshaug 2007, 225.
219 See e.g., T. Andersson 1986; Sandnes 1987; 1992; Brink 1996a; Vikstrand 2001, 253; de Vries
1977; SOL, 131f.
220 Sandnes & Stemshaug 2007, 225.
221 Vikstrand 2002, 132.
222 See Helleland 1996, 136f.; Sandnes & Stemshaug 2007, 225.
223 See Haraldsen 1998; Lidén 1999, 43; L. Larsson 2011, 201.
224 Røskaft 2003, 53–76, 210.
used during the entire Late Iron Age in this area. Originally it had the plain
meaning ‘farm’, but during the Viking Age it seems as if it gained the meaning
‘the farm par excellence’ or ‘the farm where the banqueting hall was located’.225
In my opinion, this term could also be interpreted ‘ceremonial building’.
The place-name Vang appears in Trøndelag. It has also been interpreted
as cultic. Magnus Olsen, for instance, has suggested that a special kind of
meadow (vangr) often was located beside the hof sanctuaries in Norway as
an equivalent to the kirkevangen, i.e., the meadow where the people gathered
beside the church for different kinds of activities.226 This interpretation has
more recently been accepted.227
Place-names, including the term haugr, have also been referred to as “cul-
tic” in Inn-Trøndelag. According to Magnus Olsen, they designated central
farms with ancient burial mounds, where the religious and political commu-
nities regularly gathered.228 It is uncertain, however, whether pre-Christian
ceremonial buildings were raised at these sites. During the Middle Ages
fylkiskirkjur were erected there.
the gable. Just inside the door stood the high-seat pillars with the so-
called holy nails fixed in them, and beyond that point the whole build-
ing was considered a sanctuary. Inside the main hof was a structure built
much like the choir in churches nowadays, and in the middle a raised
platform (stalli) like an altar. On this platform lay a solid ring weighing
twenty ounces, upon which people had to swear all their oaths. It was
the business of the hofgoði to wear this ring on his arm at every public
meeting. There was a sacrificial bowl on the platform (stalli) too, with a
sacrificial twig shaped like a priest’s aspergillum for the blood of animals
killed as offerings to the gods to be sprinkled from the bowl. This blood
was called hlaut. Inside the choir-like part of the building (afhús) the fig-
ures of gods were arranged in a circle right round the platform (stalli).
Every farmer had to pay tax to the hof. Another of their duties was to sup-
port the hofgoði in his mission, just as farmers nowadays have to support
their chieftains. It was the goði’s business to see to the hof and maintain it
properly at his own expense, as well as hold sacrificial feasts.230
It has been debated whether this description of Þórólfr’s hof should be con-
sidered reliable source for the history of religions.231 It seems as if some infor-
mation in this account does not carry an independent source value and may
be dependent on Snorri Sturluson’s description of ceremonial buildings in
Trøndelag in Hákonar saga góða.232 Several scholars have also argued that the
author (or editor) of the saga used concepts which most likely must be regarded
as late constructions. The sacrificial terms hlaut and hlautteinn in the sense of
230 Hann setti bœ mikinn við Hofsvág, er hann kallaði á Hofsstǫðum. Þar lét hann reisa hof, ok
var þat mikit hús; váru dyrr á hliðvegginum ok nær ǫðrum endanum; þar fyrir innan stóðu
ǫndvegissúlurnar, ok váru þar í naglar; þeir hétu reginnaglar; þar var allt friðarstaðr fyrir
innan. Innar af hofinu var hús í þá líking, sem nú er sǫnghús í kirkjum, ok stóð þar stalli á
miðju gólfinu sem altari, ok lá þar á hringr einn mótlauss, tvítøgeyringr, ok skyldi þar at
sverja eiða alla; þann hring skyldi hofgoði hafa á hendi sér til allra mannfunda. Á stallanum
skyldi ok standa hlautbolli, ok þar í hlautteinn sem stǫkkull væri, ok skyldi þar støkkva með
ór bollanum blóði því, er hlaut var kallat; þat var þess konar blóð, er svœfð váru þau kvikendi,
er goðunum var fórnat. Umhverfis stallann var goðunum skipat í afhúsinu. Til hofsins skyldu
allir menn tolla gjalda ok vera skyldir hofgoðanum til allra ferða, sem nú eru þingmenn
hǫfðingjum, en goði skyldi hofi upp halda af sjálfs síns kostnaði, svá at eigi rénaði, ok hafa
inni blótveizlur. Eyrbyggja saga 4, Ísl. Fornr. 4, 8f. Trans. Hermann Pálsson & P. Edwards
(slightly modified).
231 See e.g., Hultgård 1993, 228ff., 232ff.; 1996, 36ff.; Perkins 1994, 56f.; Jón Hnefill Aðalsteinsson
1998, 35ff.
232 See e.g., Düwel 1985, 10f.; Hultgård 1993, 228; Perkins 1994, 53.
‘sacrificial blood’ and ‘sacrificial twig’, for instance, cannot be found in Viking
Age terminology. In the pre-Christian vocabulary these two terms referred
to ‘lot’ or ‘lot twig’. They thus passed through a semantic development and
changed their meaning sometime during the Late Viking Age and Early Middle
Ages (see above and ch. 9 below).233
Nor has the description of the ceremonial building and its interior been
considered reliable. The author adds details in his description, which in a strik-
ing way resembled the structure and ritual objects of a church.234 According
to Olaf Olsen, for instance, the medieval author of Eyrbyggja saga had the
Christian church in mind when describing the hof at Þórsnes.235 When he
described the afhús, he added “much like the choir in churches nowadays”
(sem nú er sǫnghús í kirkjum). This comment, according to Olsen, proved that
the scribe actually was describing a church in this passage.
The criticism against Eyrbyggja saga’s description of Þórólfr’s ceremonial
building with all its details is partly sound. But it should also be stated that
the archaeological evidence of ceremonial buildings today is completely
different from the situation in the 1960s, when Olsen wrote his book. Since the
archaeological knowledge about ceremonial buildings has increased dramati-
cally during the last few decades, we must also reinterpret and re-evaluate the
testimony about such buildings appearing in the sagas. New archaeological
evidence deriving from several places in Scandinavia indicates, for instance,
that choir-like rooms or separate buildings connected to or situated beside
the feasting halls may have functioned as the sanctuary part of these houses
(see above). This could definitely be associated with the afhús mentioned in
Eyrbyggja saga.236 Archaeologists have also recently shown that a small cultic
house was erected at Hólmur in south-eastern Iceland (see below).
It should also be emphasized that the author of Eyrbyggja saga did not
invent the story of Þórólfr by himself. Most likely he proceeded from older tra-
ditions when writing his own text. Such traditions appear in Landnámabók
(probably compiled as early as the twelfth century), where the basic features
of this story may be found.237 Þórólfr’s ceremonial building is mentioned here
too: “There he [Þórólfr] erected his home and built a large ceremonial building
(hof mikit), and consecrated it to Þórr, and now the place is called Hofstaðir.”238
This indicates that some traditions about Þórólfr and his ceremonial building
existed in Iceland at least as early as the second half of the twelfth century, and
probably also earlier, since Þórólfr is mentioned as a landnámsmaðr in Ari’s
Íslendingabók (c. 1120–32).239
There are some descriptions of Icelandic hof buildings in the Old Norse
prose traditions, which may have inspired the author of Eyrbyggja saga when
describing Þórólfr’s hof. The most important text is probably the one found
in Úlfljótslǫg, which, for instance has been preserved in the Hauksbók redac-
tion of Landnámabók (the text is presented in ch. 5 below). This description
of the interior of the hof partly resembles the account on the hof at Þórsnes in
Eyrbyggja saga, where we also hear about a ring placed on an altar (stalli). The
connection between the goði and the hof also appears in these both texts. In
addition, the financial system of the sanctuaries (gefa toll til hofs) mentioned
in Úlfljótslǫg harmonizes with the information found in Eyrbyggja saga, where
we hear that “every farmer had to pay tax to the hof”. Olaf Olsen was, however,
sceptical of the historical value of Úlfljótslǫg as well. According to him, this
text is only a learned medieval construction of the pre-Christian Law, cult and
society.240 In my opinion, there are some elements connected to the descrip-
tion of the ceremonial building in this text and in Eyrbyggja saga which may be
based on old oral traditions. I will come back to them in the thematic discus-
sion below and evaluate some details in these texts, but I will show here that
archaeological evidence and place-names indicate that the term hof in Iceland
really may have designated a quite large building where public religious rituals
took place.
238 þar reisti hann bæ sinn ok gerði þar hof mikit ok helgaði Þór; þar heita nú Hofstaðir. See
Landnámabók S85, H73, Ísl. Fornr. 1, 125.
239 See Íslendingabók, Ísl. Fornr. 1, 10.
240 O. Olsen 1966. See also ch. 5 below.
Viking Age hall building was found in 1908 by the archaeologist Daniel Bruun
(with Finnur Jónsson).241 It measured 36 metres long and it was 8 metres wide.
Since this large building was located at the place called Hofstaðir ‘the hof place’
it was interpreted as a hof, i.e., a ‘ceremonial building’.242 It was argued that
this house resembled the large ceremonial buildings called hof in the Sagas of
Icelanders, e.g., Þórólfr’s hof mentioned in Eyrbyggja saga. Not only the name
but also the large size and the structure of the house indicated that it was a hof,
i.e., a “pagan temple”. The hall was divided into two parts. There was first a ban-
queting room, marked AB. There was also a separate room at the north end of
the hall (marked C). It was interpreted by Daniel Bruun and Finnur Jónsson as
a shrine or the inner sanctuary, where the images of deities were kept (afhús).
There were also two further rooms at the western side of the building, marked
D and E. In addition there was a strange pit or orbital ruin (marked G), outside
the hall to the south, full of ash and animal bones. It was interpreted as a place
where rubbish was placed after the ceremonial feasts in the “temple”.
For several decades there was general consensus in research that the
building at Hofstaðir should be interpreted as a pre-Christian Scandinavian
“temple”.243 In the middle of the twentieth century, however, some doubts
were expressed about this interpretation. It was argued that this settlement
differed little in form from other Iron Age dwelling houses and farmsteads in
Scandinavia.244 These doubts were even more emphasized by Olaf Olsen, when
he re-excavated the pit (G), to the south of the hall in 1965.245 He argued that
Hofstaðir was a quite ordinary chieftain farm. The hall there was sometimes
used as a place for ceremonial banquets, but mostly it was employed for other
more ordinary and profane activities. The pit (G) was interpreted as a cooking
pit, with a size of 4.60–5.40 metres. It was used for ritual cooking in connec-
tion with religious feasts and gatherings. Olsen rejected the idea that the room
at the north end of the hall (C) could be interpreted as an “inner sanctuary”
(afhús). He argued that the choir-like structure at the northern gable had a
door only on the outside, and thus could not be compared with the afhús men-
tioned in Eyrbyggja saga.
Results from the new excavations, conducted between 1991 and 2002,246
indicate that the adjacent room (C2), built in phase II (c. 980–1030 AD), also
had a connection directly into the hall,247 and thus could be interpreted as
some kind of an afhús. According to Gavin Lucas, the function of this room is
very unclear, because of the context of the finds.248 Only a polished bone pin
was retrieved there. This room/house was rebuilt during phase V (1300–1477)
called C1. Its function is likewise unknown, but its massive design indicates
something unusual.249
During the new excavation, a small “hall” (D1) was discovered in the south-
western part of the hall.250 This “hall” was built in phase II (c. 980–1030), and
probably had a direct passage into the major hall (AB). The interior of this
structure covers an area of 8 × 3.2 m and its roof was supported by paired posts.
The room also had a hearth, which suggests that people spent some time there.
It was interpreted as a domestic room.
The new excavation indicates that the large “bowed, aisled hall” (AB) was
erected in phase I, c. 940 and abandoned in phase II, around 1030.251 This hall
building, in Scandinavian style, was divided into three rooms. The northern
room had two great cooking pits, which had been cleaned out repeatedly. Most
likely food processing occurred in this room (see further ch. 6 below). The cen-
tral room seems to have been used for sleeping, eating, and other activities.
Perhaps some small craft activities took place at the southern end of this room,
since many of the beads were found in this area. The central hearth in this
room was used for warmth and cooking. Most likely the high-seat was located
opposite this hearth. This must have been the most important space for cer-
emonial feasting and could perhaps be described as the “hall-room”. The south-
ern room of the hall building may have been used as food storage.
The new archaeological analysis of the Hofstaðir bone material has gen-
erated some sensational results, which also have relevance for a cultic inter-
pretation of the place. A minimum of 23 cattle skulls were retrieved in two
clusters from outside the walls of the hall at structure A2 and D (the minor
hall).252 These skulls show evidence of specialized butchery and prolonged
246 On the new excavations, see e.g., Adolf Friðríksson & Orri Vésteinsson 1997; Orri
Vésteinsson 2001; 2007; Lucas & McGovern 2008; Lucas 2009.
247 See Lucas 2009, 124.
248 See Lucas 2009, 125ff., 388, 391.
249 See Lucas 2009, 155.
250 See Lucas 2009, 131ff., 388, 391.
251 See Lucas 2009, 62ff., 112ff., 373ff., 386ff.
252 See Lucas 2009, 236ff.
central location and the monumental size of the house (according to the
recent excavation 38 metres long) probably functioned as manifestations of
power.256 Perhaps the bloody slaughter, the ritual decapitation and display
of the cattle (bull) heads had a similar function. Thus we can interpret it as a
chieftain settlement which also had public cultic functions, as e.g., at Helgö
or Borg in Lofoten. Activities of an everyday character were also carried out
at Hofstaðir. The household economy was based on the production of piglets,
goat-milk products, and young sheep in preference to wool. It seems as if desir-
able food was produced there. Also forging and other crafts took place there.
Most likely, it was only in connection with the seasonal ritual feasts that the
hall was transformed into a sanctuary and perhaps also occasional dwelling
space for the people who gathered at Hofstaðir to take part in the religious
ceremonies.
Hofstaðir was probably an important chieftain site in the society of the
Mývatn region, but it was not the first farm settled there. Sites such as Sveigakot
and Hrísheimar were settled as early as the last quarter of the ninth century.
Hofstaðir’s significance only starts around 940. Perhaps the farm was estab-
lished there in order to control the flow of coastal resources into the Mývatn
region. The hall was erected high up on the homefield, rendering it visible
in the immediate landscape, thus expressing some kind of symbolic power.257
Perhaps this farm competed with other farms in the region, such as Reykjahlíð,
to be the most powerful chieftain farm there. It seems, however, as if Hofstaðir
failed in that competition. In one sense Hofstaðir may have been very impor-
tant during its existence, namely as a cultic site and a place of rich feasting.258
The settlement of Hofstaðir was abandoned in phase III (1030–1070), which
is marked by several acts of closure, such as the ritual depositions of skulls
and the sheep in A2 and D. A new settlement was established in the south-
western area of the levelled farm mound, where a medieval church (eleventh
century) was also built. This abandonment of the feasting hall should probably
be related to the conversion of Iceland.259
The context of the hall at Hofstaðir differs in some respects from the
halls in the Mälaren area and Norway. While the halls in the Mälaren area,
such as the one at Helgö, have a long structural continuity, the hall at Hofstaðir
256 Cf. Lucas 2009, 252. The normal house size in Iceland was 12 to 16 metres. See Lucas
2009, 376.
257 Lucas 2009, 402.
258 Lucas 2009, 404ff.
259 See Lucas 2009, 407f.
There are also archaeological indications of more specific and small pre-
Christian cultic houses in Iceland. At Hólmur close to Hornafjörður (Höfn)
in south-east Iceland, a specific cultic house (“blót house”) was discovered by
excavations in the summers of 1997 and 1999 close to a Viking Age burial situ-
ated on a mound or a hill (15 × 25 m. and approx. 2 m high).272 Hólmur seems
to have been a Viking Age farmstead and the cult house and the grave were
located on the small mound about 250 metres north-east of this settlement in
an exposed position. The house is small (2 × 2.16 metres) and has a NW–SW ori-
entation with its entrance facing north-west. It had two roof-supporting posts
and there was an oven in the south-east corner. The house was sunken and
must be considered as a typical pit-house. Several objects were found inside
the house, such as two pieces of whalebone vertebrae with cutting abrasions.
On the floor in front of the oven, many objects were found, for instance, a
worked disc-shaped piece of soft rock (round disc). In addition to this, char-
coal, cremated bones and fire-cracked stones were also frequently discovered.
All finds are quite common at settlements in Iceland, except the round stone
disc. Around the house there was an area containing a vast amount of fire-
cracked stones, which were related to ritual activities. Besides the grave, sev-
eral post-holes were also found outside the house. A couple of them seem to
have formed a kind of entrance or gate to the mound. Crude iron was placed
in the post-holes. Hearths and cooking pits were also found outside the
house. Certain objects were discovered in them, such as gaming pieces, loom
weights, burnt and unburnt bones and fire-cracked stones. The cult house was
established during the first phase of the landnám period, c. 870–880, and it
was used during the entire Viking Age.
Vikstrand does not rule out that hof may refer to some kind of ceremonial
building or a feasting hall in Iceland too.274
On the basis of the finds made at Hofstaðir, Mývatnssveit, Orri Vésteinsson
has recently argued that the hall there was built as an attempt to create a politi-
cal and religious centre based on the resistance to growing Christian influence
in Iceland.275 The pre-Christian cult there was maintained until 1030, although
the legal decision of conversion was taken in AD 1000 at the Alþingi. By means
of a topographical analysis, he argued that the farms with designations con-
taining the element hof in Iceland were never part of the original settlement
cluster, but rather should be seen as secondary formations. They were estab-
lished after the farming communities had been formed. The intention was to
create central places in the already existing settlements as resistance nodes
against Christianity. According to Orri, the word hof in place-names refers
to the feasting hall which was the arena for different types of socio-political
activities. The religious, pagan connotation of the term is, however, first seen
in the late tenth century, when these halls became the centres of resistance to
the new religion.
274 Vikstrand 2001; 2002, 132. Olaf Olsen also accepted that the hof names in Iceland may
have connoted religious aspects. However, this word should not be interpreted as ‘temple’,
i.e. a building or place intended exclusively for religious activities: “Ligeledes kunne det
påvises, at ordet hof ikke nødvendigvis må oversættes ved ‘temple’, men antagelig kan
opfattes som betegnelse for en gård, på hvilken der er foregået kulthandlinger, vel især
blótgilder, for en store gruppe mennesker end gardens egne beboer.” O. Olsen 1966, 192.
275 Orri Vésteinsson 2007
tional hall buildings and more specific cultic houses. The latter type is, how-
ever, less visible in the archaeological sources from these areas, compared to
the situation in the southern parts of Scandinavia. The find contexts of cer-
emonial buildings in the Mälaren region and Trøndelag are sometimes distin-
guished by exclusive objects such as gold foil figures and imported glass. In
connection with the cultic sites of these areas, monumental graves were some-
times erected (see ch. 12). Compared to this situation, the material context of
ceremonial buildings and sanctuaries in Iceland is poor. The archaeological
finds made at Hofstaðir nevertheless indicate that those who built this hall
either were chieftains or had ambitions to be powerful leaders in the district
where they lived.
In Part 2 of the present study, some contextual matters on the issue of the cer-
emonial buildings and ruler strategies in Late Iron Age Scandinavia have been
outlined, such as the sources and their value for historical reconstructions, as
well as the regional-contextual approach. In chapter 2 the social and political
structure and historical development in the three investigation regions (the
Mälaren area, Trøndelag and Iceland), were examined. It can be concluded
that a tendency towards centralization of power from the Migration Period
onwards in the Mälaren area could be discerned, with kings and earls at the top
of society. A similar centralizing tendency was also observable in Norway and
Trøndelag from the early Viking Age onwards. As a result of this power concen-
tration involving kings and earls, Norwegian magnates from the lower elite and
farmers emigrated from their homeland to Iceland. There the Icelandic Free
State was established, which had a decentralized power structure, without roy-
alty. In Iceland power was thus split among many local chieftains, who only
had limited possibilities to exploit their subjects. A concentration of power did
not appear there until the thirteenth century.
Chapter 3 has presented a case study where the religious strategies for polit-
ical authority and legitimacy of three ruling families were investigated in some
written sources. This case study included the royal family named Ynglingar
from the Mälaren region, the Earls of Lade from Trøndelag, and the chieftain
family named Þórsnesingar from Iceland, i.e., representatives from the three
regions addressed in the present monograph. The kings and earls from the
Mälaren region and Trøndelag claimed to be of divine origin. This strategy was
not seen among the chieftains in Iceland. These chieftains argued, however,
that they were the deity Þórr’s dearest friend. The conclusion was that the
centralizing tendencies which were observed in the societies of the Mälaren
region and Trøndelag had generated a type of monumental and bombastic
ruler strategy which was not as visible in the more decentralized Icelandic
chieftain society. This conclusion constitutes the fundamental hypothesis of
the entire monograph.
The hypothesis of the present monograph is that the political-structural dif-
ferences in the three regions must have had consequences for the formation of
ruler and power strategies associated with cultic places and ceremonial build-
ings. Religious strategies for rulership in Sweden and Norway (among earls and
kings) would therefore be different from those found in Iceland (among chief-
tains). As we shall see below in Part 4, the investigation of the source mate-
rial shows that this hypothesis can partially be confirmed, but perhaps not as
significantly as expected. The similarities between the three regions proved
to be striking, especially when it comes to ritual and symbolic strategies. We
therefore now turn to these similarities in Part 3.
The issue of cultic leaders in Late Iron Age Scandinavia has rarely been
discussed in previous research. Besides surveys in handbooks of ancient
Germanic and Scandinavian religion,5 and references in some philological and
onomastic studies,6 only a few scholars have exclusively investigated ancient
Scandinavian cultic leaders thoroughly.7 Previous discussion has been polar-
ized; two lines of interpretation can be discerned, represented by scholars
from two different fields of study: philology (onomastics) and the history of
religions. Some historians of religions, for instance Folke Ström, argue that
the ancient Scandinavians lacked a professional priesthood. According to
him, there were no priests who exclusively had the assignment to serve the
divine powers.8 The political ruler, the king or chieftain, made contact with
the deities at the public sanctuaries on behalf of the people at the sacrificial
feasts and in other rituals at the cultic sites. This type of religious leadership
has often been associated with the notion of sacral kingship. Historians of reli-
gions who investigate the entire Germanic area sometimes argue that priests
existed in some parts of it. Jan de Vries, for instance, states that in the southern
area profane and sacred leadership functions were separated. The situation in
the northern area seems to be different.
Bei den Südgermanen haben die Priester neben der weltlichen Obrigkeit
gestanden. [. . .] Die Quellen, die wir für Skandinavien besitzen, scheinen
darauf hinzuweisen, daß die Trennung der weltlichen und priesterlichen
Funktionen hier nicht, oder jedenfalls nur sehr spät, stattgefunden hat.9
Other scholars think that the Scandinavians had specialized priests. These are
mainly philologists, especially specialists in onomastics. Klaus von See argues
that the ON term goði refers to an exclusively priestly office. Only in Iceland, in
his opinion, where the historical situation was very special, did the goði office
develop into a leadership including several functions, such as law and other
“secular” aspects. Klaus von See also states that originally the Germanic people
strictly distinguished between religious and judicial aspects.10 This issue has
been debated throughout the twentieth century and beyond.
John Kousgård Sørensen states that a priestly class, in Danish “præstestand”,
existed in Late Iron Age Scandinavia. He focused on the term OScand ‑vé(r),
‑vi(r), ‑væ(r) (PN *wīhar), which he interpreted as ‘priest’. According to him,
compound nouns including this term reflected a differentiated hierarchical
priesthood.11 Other specialists on toponomastics have observed terms in the
place-name materials which could refer to exclusive religious specialists: e.g.,
ON goði, *vífill, and perhaps *lytir.
In general the specialists in toponymy apply designations such as “priests”
(“Priester”),12 “pagan priests”,13 “priesthood” (“Termini der Priesterschaft”),14
“priest office” (prästämbete)15 and “pagan vicarage” (hedniska prästgårdar)16 to
ancient Scandinavian contexts. One exception is Per Vikstrand, who uses the
concept of “designations of cultic leaders” (kultfunktionärsbeteckningar).17
The difference between historians of religions and philologists/specialists
in toponymy is probably due to the different evaluation of the source catego-
ries and different methodologies. Historians of religions rely mainly on his-
torical and narrative sources, while philologists put their trust in the linguistic
material and etymology.18 The difference of opinion is also due to confusion
of terminology. It seems as if the philologists have used the common category
“priest” without any definition or deeper reflection. Nor have historians of reli-
gions defined what they mean by “priest”. In this discipline, however, sugges-
tions as to how operational concepts such as “priest” and “priesthood” can be
conceived are sometimes encountered.19
In an article published in 1998 I criticized this terminology. On the basis
of classical phenomenological treatments, I proposed analytic definitions of
the categories “priest” and “priesthood” and tested them on the Scandinavian
materials.20 I arrived at the conclusion that general features of priests or
10 Von See 1964, 78f., 105–112. Cf. Maurer 1873; Heusler 1911.
11 Kousgård Sørensen 1989, 5–33.
12 Laur 2001, 223.
13 Brink 1996a, 266.
14 T. Andersson 1992a, 518.
15 Strid 1999, 103.
16 Hellberg 1976, 11; 1986a, 61ff.
17 Vikstrand 2001, 386, 427.
18 Cf. Vikstrand 2001, 396.
19 See e.g., Sabourin 1973; Widengren 1969.
20 For definitions of “priest” and “priesthood”, see e.g., Sabourin 1973; Widengren 1969;
Sundqvist 1998, 78f.
priesthood were vague in early Scandinavia. There is weak evidence for initia-
tions into or formal training for a religious office.21 The cultic officials do not
seem to constitute an organized and hierarchical form of institution, appearing
as an independent social stratum in society. Neither was there a priestly insti-
tution normalizing world view or ritual practice. Since the concept of “priest”
(from Greek πρεσβύτερος ‘the older’, πρέσβυς ‘old person’) was formed and
developed in a Christian context I argued that it is more correct to use more
neutral concepts, such as cultic leaders, cult performers or religious officials.22
The general categories to be used in the study of religion have been much
discussed. During the first half of the twentieth century, the phenomenolo-
gists of religion developed instrumental tools and operational concepts for the
discipline by means of a comparative method. By using a synchronic approach
they elucidated structures in the religious material which were supposed to
appear transhistorically and transculturally. In that manner certain categories
were elaborated, such as “myth”, “ritual”, “priest” and “temple”.
In recent decades the comparative perspective and the phenomenology of
religion have been much criticized.23 This criticism rejects universal theories
and concepts produced by the comparativists; the phenomenologists, it says,
have observed only the similarities and ignored the particularities and differ-
ences. The critics argue that the generalized categories are rarely based on
solid empirical material and that the old comparativists often proceeded from
a priori assumptions. When construing a category such as “temple” or “priest”,
for instance, the cognitive conception is strongly influenced by Christian and
western thinking. The use of such concepts in non-Christian contexts always
causes a risk of serious misinterpretation.24
In my opinion this criticism has been sound in parts. On the other hand,
I think that some common analytic concepts are needed to enable interdis-
ciplinary communication and transcultural comparisons of religious phe-
nomena, even if they must always to a certain extent be subjective. We must,
however, always keep in mind that the concepts applied do not always cor-
respond to the native categories. The distinction between etic and emic levels
elaborated in anthropology is very enlightening on that particular point.25 All
operative concepts applied must therefore always be well defined. We must
21 Traces of initiation may be seen in mythical traditions, however. See mainly Schjødt 2008.
Cf. Sundqvist 2009b; 2010a.
22 For a more detailed argumentation, see Sundqvist 1998, 76ff. and also 2003a; 2003b; 2007.
23 E.g., J. Z. Smith 2000, 23–44; 2004; Flood 1999; Gilhus & Mikaelsson 2003, 61ff., 79ff., 83ff.
24 Cf. Hewitt 1996, 16; Rüpke 1996, 241. If our tools (i.e. our terms) “do not do the job properly,
then they must be honed or replaced”. Jensen 2014, 7.
25 E.g., Lett 1996, 382f.
also avoid value-loaded terms that apparently have their background in west-
ern or Christian contexts or problematic evolutionistic discourses.
This criticism of concepts has also involved the notion of “priests”. Jörg
Rüpke, for instance, has formulated a critique of Joachim Wach’s typology of
ten categories of religious authority, including concepts such as priest, priest-
hood, prophet, shaman, healer, magician and diviner. According to Rüpke these
terms are taken from specific cultural contexts but are used transculturally.
He states that: “these types are defined by certain functions and performances
that sometimes are embarrassingly missing from the culture just analysed.”26
As an alternative Rüpke attempts to describe cultic leaders as agents of
control within systems of symbols. They may thus be defined as controllers.
These controllers may monopolize certain ritual activities. They may control
theological products and dogmas, as well as holy scriptures or oral traditions.
Writing may be a way to preserve verbal conceptualizations of symbol systems.
The controllers could therefore also monopolize the systems of documenta-
tion, i.e. writing systems.
In my view, however, concepts such as controller, theology and dogma
fit badly into the ancient Scandinavian context. The term “cultic leader” is
proposed instead. It refers to a person who was temporarily responsible for
certain public religious functions in society and at the cultic sites.27 The cul-
tic leader had other societal duties beside his or her religious tasks. He or she
also functioned as a general political leader. The term “religious specialist”
designates an exclusive religious office in the present study, i.e. when a more
intensified and permanent specialization had taken place.28 We will now turn
to the formation of cultic leadership at the sanctuaries in the three investi-
gation areas.
Different types of sources indicate that chieftains controlled the cultic sites
and the ceremonial buildings in the three investigation areas. These chieftains
were sometimes designated goðar (OSw *gudhar). This term (ON goði sg.) is
a derivation of the noun goð ‘god’, indicating that this denomination in some
33 This laconic information may have quite a high historical source value. See e.g., Dillmann
1997.
34 Landnámabók, Ísl. Fornr. 1, 307, note 12.
35 The Old Norse medieval traditions report that kings and earls appeared in Sparbyggvafylki,
during the Early Viking Age. Snorri mentions, for instance, an earl called Ketill jamti
(son of Ǫnundr jarl). He lived in Sparbyggvafylki, during the Early Viking Age. He fled to
Jämtland when King Eysteinn Upplendingakonungr conquered the land c. 780. Hákonar
saga góða 12, in Heimskringla I, Ísl. Fornr. 26, 164f.; Óláfs saga helga 137, in Heimskringla II,
Ísl. Fornr. 27, 241. Snorri also reports about a king of this fylki, who fell in battle just before
Haraldr Finehair’s appearance. Haralds saga ins hárfagra 7, in Heimskringla I, Ísl. Fornr.
26, 99.
36 See the discussion in e.g., Stenvik 1996; Røskaft 1997, 237; 2003, 138f.; Lidén 1999, 45.
37 It is possible that hofgoðar were important in the Late Viking Age as well. One manuscript
of Heimskringla mentions that during Guðbrandr’s life (c. 1000 AD) there was a hofgoði of
the Dalesmen who was called Þórðr ístrmagi. In the other manuscripts Þórðr is called
hǫfðingi. See Phillpotts 1912–1913, 271.
38 See e.g., Landnámabók SH7–8; H11; S289 H 250; S310 H270. For more complete documen-
tation, see Strömbäck 1928b; Birkeli 1932, 24ff.
39 It seems, however, as if Þórhaddr went to Iceland more or less voluntarily. The expres-
sion hann fýstist til Íslands “he wanted [to travel] to Iceland” indicates this. Jón Viðar
Sigurðsson et al. 2005, 96.
40 . . . at hofi því, er Þorbjǫrn . . . hafði varðveitt. Landnámabók S 368, Ísl. Fornr. 1, 368.
41 Brennu-Njáls saga 87–88, Ísl. Fornr. 12, 208–220. Óláfs saga Helga 112, Heimskringla II, Ísl.
Fornr. 27, 183–188; Flateyjarbók II, 189.
42 Farbregd 1986; Røskaft 2003, 66ff.; Nordeide 2011, 104ff.; 2012.
43 M. Olsen 1926, 218. Cf. Sandnes 1992, 15f.; Sandnes & Stemshaug 2007, 479.
44 Farbregd 1986.
charcoal and burned stone suggest that ritual cooking took place there, from
c. 100 BC to AD 1000. In connection with these pits post-holes were also dis-
covered which not were related to house foundations. Perhaps they could be
associated with ritual structures.
There are other hov-places in this area which may have been connected to
local cultic sites and also ceremonial buildings, such as Hov in Ålen, Hov in
Soknedalen, Hov in Orkdal, and Hove at Byneset.45 These farms all had a cen-
tral position in the local settlements and most likely they functioned as pub-
lic gathering places for the local judicial assembly and pre-Christian cult. It
seems as if local magnates owned these farms and from them they controlled
the social-political activities in the settlement districts during pre-Christian
periods.46 At several of these places early medieval churches were also erected.
Several of them may be regarded as private farm churches (hǿgendiskirkjur).
A religious and political power continuity may probably be discerned at
these sites.
It can be concluded that rulers of different dignity thus had sanctuaries
and ceremonial buildings erected on their own farms in Trøndelag during the
Viking Age. Some of these farms were also central places for the entire fylki
district. At Melhus, in Gauldœla fylki, for instance, many finds have been made
on the farm, mainly dating to the Late Iron Age. The finds indicate a high social
status, for instance an expensive sword and a buckle. Viking Age burial mounds
are also attested at this farm.47 During archaeological investigations at Melhus
and Skjerdingstad (just south of Melhus), several three-aisled buildings dating
from the Iron Age have been found. Some of them had fireplaces, and one had
three fireplaces, indicating that no animals were held there.48 The medieval
written sources report that ceremonial banquets were celebrated at Melhus.
It is mentioned, for instance, that Earl Hákon took part in banquets at this
farm.49 It is plausible that banqueting halls were erected at Melhus during the
Viking Age. It seems as if Snorri believed that Melhus was an important reli-
gious centre in tenth-century Trøndelag. He states, for instance, that Ásbjǫrn
of Meðalhús (= Melhus) was one of the chieftains, who organized public
and noblest families in Iceland who had the power position and the opportu-
nity to play a central role at the judicial courts and in public cult. Usually they
erected the hof buildings on their own farms.
Olsen’s argument is, in my opinion, well founded. People from the lower
strata of society in Iceland could probably never be “selected as goði” or
“appointed to a hof ” just because they were wise and righteous. Neither are
there any signs in the sources that people were initiated into the office of
goðorð after a long period of religious and judicial education. The cultic func-
tion and the title as goði was probably originally inherited within certain fami-
lies who belonged to the elite of Iceland.
70 On “the wise chieftain” and other qualifications of chieftains, see Jón Viðar Sigurðsson
1999, 84–150.
71 That cultic assignments were inherited in a family is known from many ancient cultures.
See Sabourin 1973, 102–135.
that Þórólfr occupied a cultic office, which also included the charge of the hof.
This office was passed on within his family to his sons and grandchildren.72
Both Landnámabók and Eyrbyggja saga state that several male members of his
family were titled goðar, i.e. they were regarded as some kind of cultic leaders.
Eyrbyggja saga 3, for instance, mentions that Þórólfr had a son called Hallsteinn,
who was born in Norway. Chapters 5 and 6 tell how he went to Iceland together
with Bjǫrn Ketilsson. Hallsteinn considered it a slur on his manhood that
he should have land granted him by his own father Þórólfr at Hofstaðir, so he
crossed over to the other side of Breiðafjǫrðr to a place called Hallsteinsnes at
Þorskarfjǫrðr, and staked his claim there. Later in the text Hallsteinn was called
goði af Hallsteinsnesi.73 Similar information is also recorded in Landnámabók,
where he is called Þorskafjarðargoði.74 This text says that he sacrificed to Þórr,
since he wished that the god could send him his high-seat posts. After some
days a big tree came ashore, which Hallsteinn used for high-seat posts.75 In my
opinion, it is possible to interpret the tradition thus: Hallsteinn was qualified
to occupy a goðorð since he was a son of the chieftain and cultic leader Þórólfr.
Even though it not explicitly mentioned, Eyrbyggja saga intimates that
Þórólfr’s other son, Þorsteinn þorskabítr Þórólfsson, was also a kind of a goði,
who took care of the hof at Hofstaðir after his father’s death. The context of
the narrative suggests that Þorsteinn lost the full custody over this sanctuary.
According to chapter 9, Þorsteinn had a great conflict with Þorgrímr Kjallaksson
and Ásgeirr á Eyri. Þorgrímr and Ásgeirr threatened to desecrate the sacred
field at Þórsnes, which led to an open battle between the Kjalleklingar and
the Þórsnesingar.76 There had been deaths on both sides and many wounded.
People decided to send for Þórði gelli, who at that time was a leading chieftain
at Breiðafjǫrðr. He was considered the likeliest man to make peace between
the fighting parties. Þórði decided that no compensation was to be paid for the
killings and woundings at Þórsnes. Since Þórsnes now was defiled by the spill-
ing of blood in enmity it was considered useless for holding assemblies, so a
new place had to be chosen for this purpose. In order to ensure friendship
between the fighting parties and establish reconciliation, Þórði declared that
72 Several scholars have regarded Þórólfr as a goði, see e.g., Wessén 1924, 170; Baetke 1942,
133f.; Strömbäck 1975, 41f. Cf. DuBois 1999, 65f.; Sundqvist 2007, 25ff.
73 Eyrbyggja saga 48, Ísl. Fornr. 4, 136.
74 Landnámabók M 25 S85, Ísl. Fornr. 1, 124ff. He is also mentioned by Ari in Íslendingabók,
Ísl. Fornr. 1, 10.
75 Landnámabók S123, H95, Ísl. Fornr. 1, 163f.
76 Eyrbyggja saga, Ísl. Fornr. 4, 14ff. This tradition also occurs in Landnámabók S85, Ísl.
Fornr. 1, 125f.
from then on Þorgrímr Kjallaksson was to bear half of the coast of maintain-
ing the hof sanctuary at Hofstaðir, and that he and Þorsteinn were to share the
sanctuary dues and support of the farmers equally between them.77 Þorgrímr
was also supposed to back Þorsteinn in all his lawsuits and safeguard the sanc-
tity of whatever place Þorsteinn might choose for the new assembly. Þorgrímr
was from then on designated Þorgrímr goði.78 The text clearly indicates that
Þorsteinn þorskabítr, before the conflict with the sons of Kjallakr, was the only
person who occupied the cultic office at Hofstaðir. He probably inherited this
office from his father Þórólfr, as Hallsteinn had done. After the conflict with
Þorgrímr, however, Þorsteinn was forced to share the chieftaincy (goðorð)
with him.
The interpretation that Þorsteinn acquired the chieftaincy (goðorð) and the
management of the hof at Hofstaðir by means of the inheritance from his
father harmonizes with the continuation of the saga, for it is mentioned that
Þorsteinn later built a great farm at Helgafell, where he erected a hof. At his
new farm, Þorsteinn and his wife had a son called Grímr. It seems as if Grímr
inherited his father’s cultic role and the charge of the sanctuary: “Þorsteinn
dedicated this boy to Þórr, calling him Þorgrímr, and said he should become a
hofgoði.”79
Eyrbyggja saga goes on to say that Þorgrímr Þorsteinsson, in his turn, had
a son called Snorri.80 He inherited his father’s farm and the sanctuary at
Helgafell. “He was now in charge of the hof and was thus called goði (Hann
varðveitti þá hof; var hann þá kallaðr Snorri goði).” We may note that the scribe
of the saga used the verb varðveita ‘be in charge of, take care of’ when express-
ing the relation between the cultic leader and his sanctuary.81 This expression
was also used when describing Þórólfr’s relation to his sanctuary in Mostr; i.e.
Þórólfr watched over the Þórr’s hof (varðveitti . . . Þórshof) located there.
The author of Eyrbyggja saga intimates that several members of Þórólfr’s
family for several generations occupied a religious office which included the
charge of hof buildings in the areas around Breiðafjǫrðr. It also seems as if
77 Þorgrímr Kjallaksson skyldi halda uppi hofinu at helmingi ok hafa hálfan hoftoll ok svá þing‑
menn at helmingi, . . . Eyrbyggja saga, Ísl. Fornr. 4, 17.
78 Var hann af því kallaðr Þorgrímr goði. Eyrbyggja saga, Ísl. Fornr. 4, 18.
79 Eyrbyggja saga, Ísl. Fornr. 4, 19.
80 Eyrbyggja saga, Ísl. Fornr. 4, 27.
81 Fritzner (1954) translates the verb varðveita ‘tage vare paa noget for at det ikke skal komme
til Skade’. He also suggests a translation of the specific expression varðveita hof ‘siges den
Mand der har det paa Gaarden, vedligeholder det og besørger Gudstjensten’.
the sons and grandsons of Þórólfr were devoted to the cult of Þórr.82 In my
opinion, the idea of making Þórólfr and his descendants into a kind of “priest
family” who cared for local sanctuaries at Breiðafjǫrðr was not a literary con-
struction made by the scribe. Most likely he built this idea on an older tra-
dition. In Landnámabók too we may note that several descendants of Þórólfr
were titled goði. Þórólfr’s son Hallsteinn was called Þorskafjarðingargoði, while
his grandchild was called Þorgrímr goði Þorsteinsson þorskabíts, and his great
grandchild Snorri goði Þorgrímsson.83
The sources report that the goðorð could also be acquired in other ways as
well. It could be shared between two persons, received as a gift or even pur-
chased. Hrafnkels saga Freysgoða, for instance, mentions that Þorkell leppr
Þjóstarsson had passed over his position of authority (mannaforráðr) to his
brother Þorgeirr when he went abroad. Before that he had been a goðorðs‑
maðr. The text indicates that he originally inherited the title goði from his
father. In the course of events, Þorgeirr offers to pass back the goðorð and the
position of authority to his brother Þorkell for a period, but then, he suggests,
they could share it equally between them. Þorkell, however, refuses to take this
offer from his brother, since he felt that Þorgeirr was the more accomplished
of the two for this assignment.84 Eyrbyggja saga likewise says that Þorgrímr
and Þorsteinn shared the goðorð at Þórsnes. We must not forget that these
statements appear in medieval texts. Most scholars accept, however, that the
goðorð could be handled in such way.85
The evidence from the saga literature thus suggests that chieftains who were
designated hofgoðar and goðar erected and were in charge of the hof buildings
in Iceland. Some of them seem to have occupied a cultic office in their old
homeland as well, for instance Þórhaddr and Þórólfr. They actually brought
some parts of the old sanctuary with them when they emigrated from Norway
to Iceland. The ceremonial buildings were usually built at the farms of the
chieftains. Some sagas indicate that the office of hofgoðar/goðar and the care
of the sanctuary were inherited within the family. Such an office could also be
acquired in other ways. It could, for instance, be received as a gift or purchased.
The sources indicate that the Icelandic goðar should not be seen as persons
82 Some of Þórólfr’s sons, like their father and grandfather, carried the name of the god as
a first element in their names, e.g., Þorsteinn and Þorgrímr. Whether this naming cus-
tom within goði families really reflects ancient conditions is much debated. See e.g.,
T. Andersson 1992a; Vikstrand 2009.
83 Landnámabók, S86, H74, M25. Ísl. Fornr. 1, 124–127.
84 See e.g., Hrafnkels saga Freysgoða, Ísl. Fornr. 11, 112ff.
85 See e.g., Byock 2001, 14.
there were regarded as this deity’s spouse (“ektefelle”).93 The weak evidence
can, in my opinion, neither refute nor corroborate this. But perhaps we can at
least relate Þuríðr hofgyðja to Freyr, since she was the sister of Þórðr Freysgoði,
whose family was called Freysgyðlingar ‘priestlings of Freyr’. In the region of
their home there is a place called Freysnes ‘Freyr’s Headland’.94
A lausavísa made by the skald Þorvaldr Koðránsson (tenth century) pro-
vides early and probably reliable evidence of an ON gyðja called Friðgerðr in
a cultic context in Iceland.95 According to Kristni saga, Þorvaldr arrived at her
farm in Hvammr in western Iceland together with a bishop. They preached
the faith while Friðgerðr was in the hof sanctuary sacrificing (enn Friðgerðr var
meðan í hofinu ok blótaði).96 Friðgerðr’s son Skeggi laughed at the missionaries.
Þorvaldr then uttered this verse:
93 See M. Olsen 1926, 247ff. Olsen referred to the late tradition about Gunnarr Helmingr
who played the role as Freyr when travelling around in Svetjud. He had a female compan-
ion who also appeared to be a “Frøis-prestinne”. Flateyjarbók I, 337ff. Cf. Ǫgmundar þáttr
dytts, Ísl. Fornr. 9, 112ff. The text does not state, however, whether this female was a gyðja
or not. She is just called Freys kona “Freyr’s wife”.
94 See E. O. G. Turville-Petre 1964, 261.
95 This stanza is preserved in Kristni saga and the Flateyjarbók version of Óláfs saga
Tryggvasonar.
96 Kristni saga 2, (ed). Kahle, 9. Cf. Kristni saga 2, trans. Grønlie, 36.
97 Skj. B1, 105; A1, 110. “I preached the precious faith,/ no man paid heed to me;/ we got scorn
from the sprinkler/—priest’s son—of blood-dipped branch./ And without any sense,/old
troll-wife against poet/—may God crush the priestess—/ shrilled at the heathen altar.”
Trans. Grønlie. Grønlie interprets hlautteins hreytir as “the sprinkler of blood-dipped
branch”. I follow Düwel’s (1985, 30f.) interpretation here “the one who cast lots”.
may not have been chosen by mere accident. Perhaps they refer to the activi-
ties that the skald really observed during his stay at Hvammr. The stanza also
says that Friðgerðr was shouting from the pagan “altar/platform” (gall of heið‑
num stalla). Even if the ritual context is not fully clear, it seems as if Friðgerðr
was performing sacrifices at the “altar/platform” (stallr/stalli), while Skeggi was
casting lots.98 The term stallr/stalli indicates that they were at a sanctuary, per-
haps inside a hof building. The prose in Kristni saga supports this interpreta-
tion. The version found in Flateyjarbók also states that Friðgerðr was sacrificing
during Þorvaldr’s visitation.99 The sources thus suggest that the gyðja and her
son performed both sacrifices and divination rituals at a ceremonial building,
while Þorvaldr prayed about Christ for the people in Hvammr.
98 For a more thorough examination of the stanza, see Düwel 1985, 30f.
99 . . . Þorvalldr talde tru firir henni en hon blotade j mote. Flateyjarbók I, 270.
100 See e.g., Hellberg 1986a; Elmevik 1966; 1990; 2003; Brink 1996a; Vikstrand 2001.
101 It has been argued that gyðjur also appeared in the Mälaren region. Some specialists on
onomastics have argued that this term (OSw *gydhia) is represented in the place-names
of this area. Lars Hellberg (1986a) has suggested that the name Gödåker in Tensta par-
ish, about 20 km north of Uppsala, could be interpreted as an Old Swedish *Gydhiuaker,
‘prästinnans åker’. He admits that this interpretation, from the linguistic point of view,
is not without problems when referring to the earliest attestations of the name (1492,
Gydakrom, Gedaker). These name forms also appear in quite late medieval documents
from c. 1500. Hellberg’s interpretation was, however supported by Lennart Moberg (2002),
who stated that the interpretation ‘den hedniska prästinnans åker’ was plausible. Several
distinguished specialists on onomastics have, however, rejected this interpretation. See
T. Andersson 1992a; Vikstrand 2001, 388f.
102 Hellberg 1986a, 61.
Alle was not only the ODa goþi of the Sølver, he was also an honour-worthy
thegn, i.e. a free man and successful ‘warrior, champion’ of high rank.112
(1) sunuR na | ut smiþ | a kata (2) uifrþaR . . . un . . . (3) kuþaskaki faþi
The first part of the sequence kuþaskaki in line 3 can be interpreted either
as the personal name Gudhi/Guþi in the genitive case or as a genitive form
113 Strid 1999, 101ff. See also Jacobsen & Moltke in DR 411: “kuþa kan opfattes . . . som akk.
af subst. goþi, en gode; i så fald kunde lakunen fx. udfyldes uia- el. frus-kuþa, hellig-
dommenes gode el. Frøsgode.” Brink (1999a, 431) accepts Strid’s interpretation. See also
Hellberg 1976, 9, note 14. Critically considered by Marold 2000, 276.
114 See Jansson 1987, 134; Düwel 2008.
115 Hellberg 1986b; Strid 1999, 102. Per Vikstrand suggested a new interpretation: ‘(alla) de fria
(öländska) männens vi’ or ’allmänt vi’. See Vikstrand 2001, 345; 2007, 206.
116 Strid 1999, 102. For another interpretation, see Vikstrand 2007, 206.
117 Hellberg 1979, 129. Cf. Brink 1999a, 427; Strid 1999, 104.
118 Von Friesen 1914.
119 My translation based on Ragnar Kinander, Sm 144, 292–296.
of a common OSw noun (appellation) *gudhi. This personal name is not well
documented.120 We may assume that the concept *gudhi here denotes a per-
son with religious leadership roles.121 Perhaps Skeggi’s father was a *gudhi or
he came from a family of *gudhar.122 Perhaps this family previously had mili-
tary functions at the hillfort (Godhastæinn), beside their religious assignments.
Whether Guða-Skeggi himself was a cultic leader is uncertain.
The picture of the eastern Scandinavian *gudhar/goþar extracted from runic
inscriptions thus harmonizes with the information about ON goðar in the
medieval Icelandic prose texts. In these sources the goðar seem to act in multi-
functional leadership roles as general chieftains. Most likely the cultic leaders
in Svetjud (and Denmark) called *gudhar (ODa goþar) should also be regarded
as general chieftains. It seems as if these chieftains could be related to local
or regional sanctuaries, where they also organized the public cultic activities.
Most of these sanctuaries were probably located at the chieftains’ farms, but
some could perhaps be regarded as communal cultic sites.
It thus seems as if some kind of chieftains and political leaders took care of
public sanctuaries in all three investigation regions. These chieftains were often
designated with titles indicating a cultic function such as the ON goði (OSw
*gudhi) and the female counterpart gyðja. My investigation indicates, how-
ever, that these chieftains in general should be considered as “cultic leaders”,
i.e. they had other societal duties beside their religious tasks.
It is entirely certain that the chieftains designated with the title goði (or the
gyðja) never monopolized the public cult in Scandinavia. According to Old
Norse narrative sources the king (konungr), and the earl ( jarl) in Svetjud and
Trøndelag could appear in similar religious roles and take care of sanctuaries.123
It seems as if the kings and earls of Svetjud and Norway were related to specific
chief sanctuaries at rulers’ farms, which had a regional or an inter-regional sig-
nificance for the Svear and the Þrœndir, for instance Uppsala and Lade. There
120 It may appear in a few inscriptions. Sm 96; Vg 187; U 579. Ingrid Sanness Johnsen states
that the proper nouns Góði and Guði, as well as the appellative *gudhi/goði are theoreti-
cally plausible. Johnsen 1968, 162.
121 von Friesen 1933.
122 Källström 2007, 243.
123 Phillpotts 1912–1913, 267.
was also, however, one “regional sanctuary” in Trøndelag which seems to have
been a “cultic centre of the community type” (if we use Vikstrand’s terms),124
namely Mære. The cult there seems to have been organized by a cooperative
of Þrœndir chieftains, at least during the second half of the tenth century. The
organization of cult and cultic leadership at the inter-regional and regional
sanctuaries of Uppsala, Lade and Mære, will be treated in chapter 14 in
Part 4. In what follows I will just argue that the kings and earls in Trøndelag and
Svetjud sometimes appeared as protectors and cultic leaders of the sanctuar-
ies. They arranged great ceremonial feasts there, sometimes with support from
other subordinated cultic leaders.
role is also ascribed to King Hákon the Good in Hákonarmál 18 (c. 960) (see
ch. 4 above).
5.3.2 In Svetjud
Skaldic poetry reports that the Viking Age kings of the Svear had much the same
duties as their Norwegian colleagues. In the ninth-century poem Ynglingatal,
King Yngvi of the Svear is called vǫrðr véstalls, which may be interpreted as
“the custodian of sanctuary altar”.126 The term vé has the meaning ‘cultic place,
sanctuary’.127 Most likely there were ritual structures and/or buildings at such
places, as indicated by the composition véstallr ‘sanctuary stand, altar’.128 The
word vǫrðr signifies ‘watchman, keeper, custodian, guardian’. A similar type of
epithet to vǫrðr vestalls also appears in connection with Norwegian rulers. Earl
Sigurðr, according to Sigurðardrápa, was called vés valdr, for instance. Runic
inscriptions from Sweden testify that such designations also really existed in
eastern Scandinavian contexts. In the monumental inscription of Rök from
Östergötland (Ög 136) (c. 800 AD), for example, the sequence sibi uiauari
appears. It may be interpreted as Sibbi vīaværi, i.e. “Sibbi, the custodian of the
sanctuaries”.129 Sibbi was not only a great ruler in the central settlement dis-
tricts of Östergötland; he was also closely connected to certain vi sanctuaries.
In a Viking Age runic inscription from Västergötland the name Vīurðr appears,
which may be interpreted as a byname or designation of a ruler, ON *vévǫrðr
m. ‘sanctuary custodian’.130 In several Swedish runic inscriptions the personal
name (or byname) Vīseti (ON Véseti) appears. It may be connected with the
appellation ON *véseti m. ‘the one who dwells at (or sits) or superintends a
sanctuary’.131 Such nomenclature probably originally designated rulers who
performed the double roles of political and cultic leader at the vi sites.
126 On this kenning, see Sundqvist 2002, 196ff.; 2007, 94.
127 The term vé is probably connected to the PG adjective *wīha- ‘holy’, preserved in Goth
weihs, OHG wīh ‘holy’, and in Germ. Weihnachten. Vikstrand 2001, 298.
128 The manuscripts diverge regarding the second element of this word. The manuscript
Kringla has vestallz ‘sacred stand’, which must be considered the most reasonable reading.
See e.g., Finnur Jónsson 1931 (1913–16), 629; Åkerlund 1939, 91; Wessén 1964, 24, 64; Bjarni
Aðalbjarnarsson, in Heimskringla I, Ísl. Fornr. 26, 42; Marold 1987, 66, 88. For another
opinion, see Adolf Noreen 1925, 231. In Old Norse the word stallr (m.) has the basic mean-
ing ‘stand’. In the compound word véstallr it acquires a sacral significance. Cf. Düwel 1985,
38; Hultgård 1993, 230ff. See also the poetic expression of heiðnum stalla “over the pagan
altar”. Skj. A1, 110; B1, 105.
129 For interpretations of the sequence uiauari, see e.g., Peterson 1994, 73.
130 Vg 73. See Peterson 2007, 257.
131 See Peterson 2007, 257.
(such as the terms *gudhi, *vivil and perhaps *lytir), however, indicates the
existence of religious specialists. How should this contradiction be explained?
Is it possible to give a coherent picture taking into consideration both perspec-
tives? Per Vikstrand has produced one suggestion. He argues that some kind of
religious specialists may have been protected by and worked under the “great
men” in the same way as other craftsman used to do.136
There is some support for this line of reasoning in available sources. It seems
as if the rituals performed at the great public feasts were very comprehensive
and complicated. Perhaps the rulers (such as kings and earls) needed some
kind of specialists for different aspects and separate rites during the public
sacrifice, although it was the ruler who had the overarching responsibility for
the public sacrificial feasts.There are, however, very few authentic attestations
of sacrificial rituals and the cultic performers are rarely mentioned (except the
rulers).137 In Hávamál, st. 144, we find some formulations alluding to the sacri-
ficial rituals:
Perhaps the verbs biðia, blóta, senda and sóa reflect different aspects of the
sacrificial ceremony. Klaus Düwel states regarding these terms: “Es wird sich
nicht um synonyme Opferausdrücke handeln, sondern jeder Begriff dürfte
eine bestimmte Art von Opferhandlung betonen”.139
It is probable that the sacrifice included ritual slaughter (ON sóa), per-
formed outdoors at an Old Norse hǫrgr or stallr, i.e. structures of stones and
wood. The Eddic poem Hyndlolióð st. 10 alludes to some customs performed
at the hǫrgr. The stones had been reddened by ox blood during the sacrifice
and turned into glass (nú er griót þat at gleri orðit) by repeated sacrificial fires
(see ch. 9 below). The verb sóa is usually translated ‘to sacrifice, to slaughter’.140
136 Vikstrand 2001, 396. Such relation could be compared with the Greek and Roman patron-
and-client-relationship. See e.g., Jackson 2014b.
137 Cf. Hultgård 1993. See further ch. 9 below.
138 “Do you know how to ask, do you know how to sacrifice, Do you know how to dispatch, do
you know how to slaughter?” Trans. C. Larrington.
139 Düwel 1970, 234. Cf. Hultgård 1993, 221f.
140 Fritzner 1954 and de Vries 1977, 528.
It appears twice in a ritual context in Hávamál st. 144–145 with this meaning.141
Also in Ynglingatal (st. 8 (5)) it gives clear religious-cultic associations, as it is
combined with the adjective árgjarn ‘eager for a good year’.142
Some parts of the victims were given (senda; cf. gefa) to the gods and per-
haps hung in a grove, an Old Norse lundr. When depicting the cultic feast at
Uppsala Adam of Bremen states in Book IV,27: “The sacrifice is of this nature:
of every living thing that is male, they offer nine heads, with the blood of which
it is customary to placate gods of this sort. The bodies they hang in the sacred
grove that adjoins the temple. Now this grove is so sacred in the eyes of the
heathen that each and every tree in it is believed divine because of the death
or putrefaction of the victims.” (see the Latin text in ch. 4 above). Adam’s infor-
mation about sacrificial rituals in holy groves seems to be reliable. Evidence
from ancient Scandinavia and other parts of Germanic area attests such cults.143
Some parts of the meat were cooked and consumed during the feast in
the hall as a communion meal, where also the libations were made. The story
referred to above, about Earl Sigurðr and Hákon the Good, from Hákonar saga
góða ch. 14–18, reflects probably such a ceremony. In the texts these feasts are
usually called Old Norse blótveizla (sg.). Perhaps the term blóta fits well in this
context.144
For this reason the complicated sacrificial rituals during the great ceremo-
nial feasts held at regional or inter-regional central places and royal farms most
probably needed some kind of specialists. In my opinion, however, these reli-
gious officials were not permanently employed as “priests”. The terms in topon-
ymy (e.g., *gudhi and *vivil) does not give sufficient empirical support for an
exclusive religious office in ancient Scandinavia. Most likely these officials—
who probably were recruited from chieftain families—had functions similar to
the bailiffs and the king’s and earl’s representatives at the royal demesnes. They
managed the crown lands and central places, and supported the “great men”
in different spheres of society, both in political and judicial matters, as well as
in religious duties. During the sacrificial feast, for instance, they had more spe-
cific religious functions and performed as cultic specialists, however, they were
all subordinated the ruler (i.e. the king or the earl) who was the ultimate leader
and organizer of the feast. These religious functions gave them a lot of social
prestige and they were during the feasts in such contexts designated with honor
titles, which originally (in most cases) were inherited within their families, such
141 In Hávamál st. 145 the skald states: Betra er óbeðit, enn sé ofblótið,/ ey sér til gildis giof;/
betra er ósent, enn sé ofsóit. In Hávamál st. 109 sóa plainly means ‘to slain or slaughter’.
142 See Sundqvist 2002, 248ff.
143 Hultgård 1996, 38f. and Sundqvist 2002, 128ff.
144 Düwel 1970, 226f. and 234; Sundqvist 2002, 249, note 61.
as goði, þulr and *vífill. These specialists may to some extent be regarded as
professional practitioners of the cult. The “great men” probably rewarded them
for these assignments, for instance, by means of precious gifts and protection
etc. This could perhaps be described as a patron-and-client relationship based
on a redistribution or an exchange system which was favorable for both sides,
i.e. “clientelism”.
It could be suggested thus that in certain aristocratic milieus the king or the
earl bore the ultimate responsibility for the sanctuaries and the sacrificial feasts
held at these sites. He may have functioned as a commission agent for the feasts
and he was certainly regarded as the custodian of the holy places (see ch. 8).
When performing the great sacrificial ceremonies at the gatherings, he some-
times needed support from other cultic leaders/specialists, who were subordi-
nated him during these rituals. Such cultic system and the co-operation between
chieftains at (inter-)regional sanctuaries will be treated further in ch. 14.
As noted above, sources suggest that political leaders on different social levels,
in all three investigation areas, managed and controlled the public sanctuar-
ies. It seems as if these cultic structures or buildings were often erected on
the farms of the rulers. This feature is probably present in all of Late Iron Age
Scandinavia. But what was the driving force behind the erection of such pub-
lic sanctuaries? One possible motive could be economic aspects and income
for the rulers, because it seems as if tribute was paid to the sanctuaries in all
three areas.
145 See the Latin text in ch. 4 above. Adam’s indignation has been interpreted as an indication
of the accuracy of his statements. Hultgård states: “Adams upprördhet antyder att hans
uppgifter är riktiga, även om man kan tolka denna möjlighet för de kristna att friköpa sig
Adam tells us that kings (i.e. chieftains) and people were obliged to visit and
send their gifts (dona transmittunt) to the “temple of Uppsala” during the
major calendrical sacrifices.146 Adam states that no one was granted exemp-
tion from these ritual feasts. This information may be compared with a tra-
dition reproduced by Snorri Sturluson in Óláfs saga helga ch. 77, where the
Uppsala sacrifices are likewise described: “At the time when heathendom still
prevailed in Svetjud, it was an old custom there that the main sacrifices were
held at Uppsala in the month of gói. Sacrifices were to be made at that time for
peace and victory for the king. People from all Svíaveldi were to resort there.”
Snorri uses the expression skyldi þá blóta; skyldu menn þangat sœkja; skyldi
þar . . . vera þing etc. The verb skylda should be translated as ‘to be obliged,
bound’.147 Both Adam and Snorri intimate that the assembly gathered at the
same time, in February–March, as the sacrificial feasts at Uppsala. It is not
unlikely that the ruler who was in charge of the sanctuary took tribute on this
occasion.148 Perhaps this ruler was identical with the Uppsala king in the Early
Viking Age, i.e. before the royalty converted to Christianity. In the eleventh
century it might have been a pagan and conservative bailiff who “cared” for
the “temple” in Uppsala. It is possible that the term Uppsalaauðr (OSw øþer)
‘Uppsala wealth’ was originally related to this tribute system. In Ynglinga saga
10, Snorri mentions that the Svear sent tribute to Uppsala, where the (euhemer-
ized) god Freyr received lands and cattle. He also states: “This was the origin of
Uppsalaauðr” (see ch. 12 below).
Old Norse sources referring to Trøndelag indicate that participation in the
pagan cultic feasts in the sanctuaries was compulsory. In Hákonar saga góða 14,
Snorri states concerning the blótveizlur in this region:
It was ancient custom that when sacrifice was to be made, all farmers
were to come to the hof and bring along with them the food they needed
while the feast lasted. At this feast all were to take part in the drinking
från de religiösa festerna på ett annat sätt. Det är enligt min mening snarare ett utslag av
en tolerant religionspolitik från sveahövdingarnas sida.” Hultgård 1997.
146 It should be noted that taxes and tributes in early medieval Denmark were always desig-
nated as “gifts”. See Poulsen 2012 and T. Zachrisson 2013.
147 Norrøn ordbok has “leggje på ein som skyldnad el. plikt, drive, tvinge, nøyde (e-n el.
e-m) . . . ; binde ein til noko; leggje noko på ein; . . .”. Heggstad et al. 1993, 390. Fritzner has
“drive, tilskynde, tvinge en til noget saa at han ikke vel kan unddrage eller undslaa sig der-
for, undlade at gjøre det, paalægge en noget som hans pligt eller skyldighed, undergive en
noget som hans skyldighed, som en byrde forpligtelse, der paahviler ham, forskylde noget,
gjøre eller vise sig fortjent dertil”. Fritzner 1954, Vol. III, 401f.
148 Cf. Wessén 1924, 196; Brink 1990b, 56.
of ale. Also all kinds of livestock were killed in connection with it, horses
also; . . . .
In the passage above the verb skylda also frequently appears: “when sacrifice
was to be made (þá er blót skyldi vera), all farmers were to come to the hof
(at allir bœndr skyldu þar koma) and bring along with them the food needed
while the feast lasted (ok flytja þannug fǫng sín, þau er þeir skyldu hafa, meðan
veizlan stóð). At this feast all were to take part in the drinking of ale (at veizlu
þeiri skyldu allir menn ǫl eiga).” Some of these expressions indicate that the
farmers were obliged to take part in the rituals at the hof. It is possible that
the public cults and ceremonial feasts performed by the Hlaðajarlar also
involved economic aspects. They may very well have been part of an ancient
system of redistribution. In exchange for “the ruler’s cult”, festivals, banquets
and feasts, the people paid tribute or offered service to the earls. The occa-
sion when Earl Sigurðr held his generous feast was something unusual and
opposite to custom (i.e. the tradion where the farmers paid for the cultic feast
themselves).
There are also archaeological indications of this economic system at other
cultic sites in the Mälaren region and Trøndelag. It has been argued that bread
played an important role at the ceremonial feasts of Helgö.149 Rotary querns
have been found close to the hall as well as corn placed in a round store-house
(Rutenberge). Moreover, at the stone setting at this site, bread was deposited,
probably as a gift to the gods. Close to the hall of Sanda too, wheat grains and
rotary querns have been discovered. Archaeologists have argued that the seed
that was ground there must have been brought in from somewhere else, since
arable land in this area was limited.150 Perhaps it came as tribute from the
farmers who intended to participate in the cultic feast of Sanda. Perhaps
the grain at Helgö was also collected as tribute in connection with a cult.151
Gold foil figures have been discovered at the ceremonial buildings of Mære,
Borg (Lofoten), Helgö, Husby in Glanshammar, Svintuna and the assumed
hall at Eskilstuna. Several scholars have interpreted them as “temple coins”.152
In order to participate in the religious feasts and rituals people would have
had to pay with these golden objects. In my opinion there are problems with
this interpretation. The fact that some of these objects were discovered in
post-holes indicates rather that they were sacrificed in connection with the
foundation of the house or that they had ideological implications (see ch. 11
below).153 Whether the foils should be regarded as “temple coins” is thus uncertain.
153 K. Lamm in J. P. Lamm 2004, 51. The posts have however been replaced several times in
both the case of Uppåkra and Helgö. Hence, the gold foils could have ended up in the
most recent of these rebuildings. These buildings are thus not constructed just once at
the same spot, but several times which complicates the matter.
154 Many Icelandic sources report that the farmers had to pay duties to the sanctuaries and
the hofsgoðar. Egils saga ch. 84, for instance, mentions Oddr, who was chieftain (hǫfðingi)
in Borgarfjǫrðr, in western Iceland. He was also hofsgoði and cared for the sanctuary. All
men paid him for that (. . . hann var hofsgoði ok réð fyrir hofi því, er allir menn guldu hoftoll
til. . . . Egils saga, Ísl. Fornr. 2, 293). See further Wessén 1924, 196ff.
155 O. Olsen 1966, 43ff. Cf. Males 2010, 158, 219–222.
156 Olsen (1966, 47) supported this statement with a reference to Grágás I, 1, 20f.
157 “Hovtolden kan—som så meget andet i stykket om Úlfljóts love—være et vidste havde
eksisteret før tiendens gennemførelse.” O. Olsen 1966, 48. Cf. Males 2010.
158 Nummus ab unoquoque mare uel foemina annuatim in huius simulacri cultum doni nomine
pendebatur. Eidem quoque spoliorum ac predarum pars tertia deputabatur, . . . . . . . Hanc
itaque statuam totius Sclavuie pensionibus cultam finitimi quoque reges absque sacrilegii
respectu donis prosequebantur. Quam inter ceteros etiam rex Danorum Sueno . . . Saxo
14.39.7–8. Archaeological evidence from Groß Raden (ninth century), Ralswiek and
Feldberg (eighth century), Wollin (tenth century) and Parchim (eleventh–thirteenth cen-
turies) supports the existence of “sacrosanct temples” and “cult halls”. Written sources
mention cults at Rethra and Szczecin. Slupecki 1993.
(c. 1122–32), where it is stated that Grímr geitskǫr explored the whole of Iceland
before the Assembly (alþingi) was established: “Each person gave a penny to
him, and he gave that money afterwards to the hof buildings.”159
It is almost impossible to find out whether the system of hoftollr existed in
pre-Christian Iceland, with compulsory fees regularly paid to the sanctuaries
and the hofgoði. There is no clear evidence for this in the more reliable written
sources. Since “sanctuary fees” seem to appear in other parts of Scandinavia,
as well as among the West Slavs, it is reasonable to believe that some kind of
tribute was paid to the hofgoðar in Iceland too. Perhaps the farmers brought
animals or meat to sanctuaries during the public feasts. Some parts were con-
sumed during the ceremonial meals, while other parts were given to the one
who organized the feast. The evidence from Hofstaðir in Iceland may partly
support such an assumption (see ch. 4 above).
It is quite possible that not only economic interests motivated the rulers to
erect sanctuaries at their farms and organize public cult there. It may have
been a matter of prestige and political power (see ch. 6 below), perhaps even
linked with the leaders’ property right to the ceremonial building and the sanc-
tuaries. Most likely this role of the political leader was also a strategy for gain-
ing political authority, empowerment and legitimacy. The rulers were believed
to have a specific relationship to the Other World; the kings and the earls were
sometimes regarded as descendants of the gods, while the chieftains stated
that they were “best friends” with the deities in cultic contexts (see ch. 3 above).
Because of these genealogical and cultic relations, the farmers and other com-
mon people in general accepted or even expected that the ruler should be the
one who took care of the sanctuaries, since he/she was the one who had these
unique links to the Other World. Therefore they were also willing to pay tribute
to the managers of the sanctuaries. The sources reflecting the transition period
in Svetjud and Norway indicate quite clearly that the people, who still were
pagans, expected and even required the first Christian kings to take part in the
traditional and calendrical sacrificial feasts and to perform sacrifices on their
behalf (see chapter 9 below).
There was actually a reciprocal relationship and exchange system between
the rulers and the farmers; the elite used the ceremonial feasts at the
159 En hónum fekk hverr maðr penning til á landi hér, en hann gaf fé þat síðan til hofa.
Íslendingabók, Ísl. Fornr. 1,7.
sanctuaries located on their farms as repayment for the tribute, while the
farmers needed the cult of the ruler—he who had a close relationship to
the gods—as a means of metaphysical protection and perhaps also as enter-
tainment, including skaldic performances. The cult at the political leaders’
farms was a means of consolidating social structures.
The rulers’ interest in public cult and sanctuaries could, on the other hand,
also be regarded as a sign of a society with little specialization and division
of labour.160 The Viking Age rulers appeared with a wide spectrum of leader-
ship functions. A more stratified society with specialized occupations did not
develop until the Middle Ages.161
It can be concluded that political leaders, on different social levels, in all three
investigation areas, Svetjud, Trøndelag and Iceland, could be regarded as a kind
of cultic leaders. In one or another sense these leaders had influence on or
even controlled the public sanctuaries and the cultic activities that took place
there. Often ceremonial buildings seem to have been erected on the farms of
these chieftains and rulers. One reason for organizing the common cult at a
public sanctuary could be economic interests. It seems as if tribute was paid
to the rulers who organized and performed the cultic feasts at the cultic sites.
This feature too appeared in sources from all three investigation areas. At the
central sanctuary of Uppsala and perhaps at Lade, the income could be quite
large, since all Svear and Þrœndir were according to these sources more or less
obliged to attend the feast at their respective sanctuary. But there could also
have been ideological motives behind the rulers’ interest in public cult and sanc-
tuaries. Since these political leaders were believed to have specific genealogi-
cal and/or cultic links to the Other World, common people accepted and even
expected that they were the ones who took care of the sanctuaries and the cult
performed there.
In Svetjud and Trøndelag it seems as if the earls and kings also constructed
and took care of inter-regional sanctuaries and also organized cultic feasts,
important for all Svear or all Þrœndir. These types of sanctuaries and feasts do
not occur in the Icelandic context. The issue of regional differences related to
cultic organization will be treated more thoroughly in chapter 14 below. Now
we will turn the structure of the cultic sites and ceremonial buildings and their
possible cosmic symbolism.
Ceremonial buildings and sanctuaries were thus not only a source of income
for the ruling elite in the three investigation areas. They were also regarded
as arenas where these groups could express political power. In this chapter I
will argue that the ritual structure of ceremonial buildings and sacred sites
in Svetjud, Trøndelag and Iceland sometimes displayed common features. In
all three regions there were certain symbolic-ritual elements or objects which
quite regularly appear in association with the ceremonial buildings. These ele-
ments played an important role during cultic activities, but were also strategies
for gaining political authority and legitimacy.
The ceremonial buildings and the sanctuaries must sometimes have been
impressive in an immediate sense for people approaching them. Certainly
they constituted the materiality of religious ideology and provided agency
for the political leaders who owned or controlled them. They represented
in a concrete way the presence of divine and political power in the human
world.1 Some of these ritual structures found at sanctuaries may also have been
regarded as symbols which reflected mythical or cosmic aspects, at least for
those who had access to religious and cultural knowledge.2 Most of these ritual
structures also occur in connection with “mythical buildings” and “mythical
landscapes” mentioned in the Old Norse traditions. In my opinion, there may
have been a micro-macrocosmic relationship between the real ceremonial
buildings and the “mythical buildings”. The real cultic place—the house and
the high-seat—was seen as a reduplication of the mythical landscapes. In that
sense the cultic place was also regarded as a threshold to the Other World, i.e.
a meeting place for humans and divine beings. The owner and user of such
place had enormous power since he or she appeared at a sacred place that was
related to the gods. The specific character of these sites was perhaps necessary
for creating dignity and power. These conditions may be seen as a symbolic-
ritual or cosmic strategy for gaining legitimacy, which may have been common
to all three areas.
3 See e.g., Eliade 1974 (1949), 367–387; 1987 (1957), 58ff.; 1991 (1949), 6ff. Eliade built his theory
on other scholars, such as Uno Holmberg [Harva] 1917; 1922; Edsman 1944.
4 Sedlamayr quoted in Eliade 1987 (1957), 61f.
In modern theoretical debate, Eliade’s ideas about “the sacred” have some-
times been regarded as problematic.5 He has been criticized for following an
extraordinary theological reasoning about the “religious man” (homo religio-
sus), which cannot be verified. He has also been attacked for his unconditional
universal perspectives and lack of consideration for specific cultural contexts.
When making comparisons he always looks for similarities and disregards dif-
ferences. A number of scholars also disagree with Eliade’s apolitical interpreta-
tion of cosmic myths.6 Some even argue that his theories about cosmic pillars
as centres of the universe are not based on sound empirical grounds.7 In any
case, most historians of religions agree that cultic sites often reflect a specific
symbolism which may include references to the cosmic-mythical world.8
During the last few years Eliade’s ideas have been applied to Scandinavian
conditions, not so much by historians of religions as by archaeologists and spe-
cialists in place-names, i.e. researchers who explore the ancient landscape.9
The archaeologist Lotte Hedeager, for instance, has argued in an innovative
article that the process by which foreign objects from the outside world were
acquired and transformed into “prestige objects” at Gudme, “the home of gods”
on Funen, Denmark, was embedded in a cosmological system.10 In order to
understand this cosmology, Hedeager searched for information in Old Norse
mythical traditions found in, for instance, The Poetic Edda and Snorra Edda
(c. 1220). According to Hedeager, the topographic model of Gudme was delib-
erately based on the paradigmatic structure of Ásgarðr as it is described in
these Norse texts. The large hall of Gudme was arranged as a counterpart to
Glaðsheimr (sic!), i.e. the mythical home of Óðinn.11 The small hall, located
south of the great hall, represented the mythical building Vingólf, i.e. the hall
of the goddesses. Beside the hall buildings at Gudme, there was also a work-
shop area. A similar structure may also be found at Ásgarðr, where a forge was
5 See e.g., Paden 1992, 84f.; 1996; McCutcheon 1997, 42ff., 74–100 and passim; Flood 1999, 5f.;
Knott 2014, 492.
6 See Smith 1987, 17ff.; McCutcheon 1997, 42ff.
7 Smith 1987, 1–23.
8 See e.g., Widengren 1969, 328ff., 339ff.; Brereton 1987; Smith 1987.
9 See e.g., Hedeager 2001; 2011; Brink 2001, 90; Vikstrand 2001, 26f.; Andrén 2004; 2006; 2014;
T. Zachrisson 2004a; 2004b; Jørgensen 2009, 332. But see also e.g., Drobin 1991; Drobin &
Keinänen 2001; Nordberg 2003; 2008; Warmid 2004; Sundqvist 2004a; 2007.
10 Hedeager 2001; 2011, 137–163. On the place-name, see Kousgård Sørensen 1985, 133ff.;
1992, 233. Critically considered by Brink 2011.
11 According to Grímnismál 8, Glaðsheimr is not the house but rather the place where Óðinn
lives. This place is the location of Valhǫll. See Simek 2006, 112. It is unclear whether Snorri
refers to the house or the place when mentioning Glaðsheimr in Gylfaginning 14.
located at Iðavǫllr. The lake situated in the western part of the Gudme settle-
ment, and the springs connecting Gudme lake with Gudbjerg to the west and
Galbjerg to the north, were related to the mythical wells called Urðr’s well
and Mímir’s well. Hedeager concluded her analysis by stating that Gudme was
“a replication of Asgard, its divine counterpart”.
Hedeager’s investigation is interesting in several senses, and her cross-
disciplinary approach of applying theories developed in the study of com-
parative religion to archaeological materials is very fruitful. She also points to
important aspects which were overlooked by Eliade, namely that the sanctu-
ary also had political and economic functions. Gudme was a site for the elite.
The cosmic symbolism at this site can therefore be related to political motifs
and be seen as part of the mechanism that creates authority. But there are
also some problems related to Hedeager’s attempt. My major objection is her
uncritical and too literary way of a reading of the Old Norse mythical texts, and
particularly her method of using these texts directly as sources for the Late
Roman Iron Age and Migration Period world view at Gudme.12 As far as I can
see, she does not evaluate the individual texts to which she refers. When she
construes Old Norse cosmology, for instance, she does not make a distinction
between the testimony contained in Snorri’s prose and the poetic traditions.
For Hedeager, Snorri’s words are an accurate reflection of pre-Christian myths.
She does not take into consideration the fact that Snorra Edda is a later literary
and Christian adaptation of older traditions. Snorri himself arranged his mate-
rial in accordance with literary models and rhetorical devices that were com-
mon among learned writers in the Middle Ages. For a historian of religions,
therefore, Snorri’s text is more unreliable as a source of ancient Scandinavian
mythology than the information contained in the Eddic lays and skaldic
poetry, which may have been composed in a Viking Age context. An example:
Hedeager argues that the temple called Vingólf, ‘the friendly house’, was situ-
ated close to Óðinn’s hall in the mythical topography. This information appears
only in Snorri’s text. In fact, Vingólf is never mentioned in the old poetic tradi-
tions. Historians of religions have argued that Snorri invented this name him-
self.13 Some scholars suggest that it is very young (“sehr jung”) and should be
read as Víngólf, ‘wine house’, i.e. another name for Valhǫll.14 It is thus possible
12 For a more thorough and extensive criticism of Hedeager’s theory, see Sundqvist 2011b.
13 Simek 2006, 363.
14 See Braune 1889. It has been argued that the name could be read and interpreted as
Vingólf ‘hall of friends’, i.e. the location of the warrior paradise, einherjar. Simek 2006, 363.
See also Finnur Jónsson (1890), who criticized Braune’s interpretation and thought that
the term was very old. He argued that it should be interpreted as ‘det venlige, hyggelige
that it was Snorri’s own idea that Vingólf was a separate house located some-
where near Valhǫll. His information on this specific place cannot be regarded
as reliable within a Viking Age cosmological context, and is even less reliable
within the Late Roman Iron Age and Migration Period world view.15
Nor does Hedeager go into the textual-criticism problems related to these
texts. When scrutinizing the different manuscripts of Snorra Edda, one can see
that the information on Vingólf varies. In the Snorra Edda manuscript known
as the Codex Regius, for instance, it is said that the goddesses’ building, Vingólf,
was a hǫrgr: Annan sal gerðu þeir, þat var hǫrgr er gyðjurnar áttu, ok var hann
allfagr. Þat hús kalla men Vingólf (Cod. Reg. Edda).16 However, if we refer to the
manuscript called Codex Upsaliensis, the hǫrgr in the same passage designates
a structure inside the hall (salr), not the building itself (Annan sal gerðu þeir er
hǫrgr var í . . ., Cod. Ups. Edda).17 It is thus unclear what type of building Snorri
had in mind when mentioning Vingólf, as the medieval witnesses diverge in
their descriptions.
There are many other complex issues connected with the reconstruction
of the ancient Scandinavian cosmology that Hedeager does not take into
consideration. According to Snorri, the ash tree Yggdrasill had three roots,
one of which went to Mímir’s well, one to Urðr’s well, and one to the well of
Hvergelmir, which was in the realm of the dead, Niflheimr (Hel).18 This state-
ment, i.e. that three roots extended to three wells, is not found in the Eddic
lays. Perhaps this reflects Snorri’s own systematization. Most historians of reli-
gions believe that there was only one tree and one well, both of which went
under different names.19 In addition, the dichotomy and dualism between
hus’ and that it referred to the goddesses’ temple. Kauffmann (1892) interpreted Vingólf as
‘the hall of lovers’, i.e. the place where the shield-maiden makes the deathless hero happy.
15 Snorri actually gives conflicting information about Vingólf. In Gylfaginning 3 Vingólf is an
alternative name for Gimlé (Gimlé eða Vingólf ), i.e. the heavenly place where good people
will live after Ragnarǫk (see Gylfaginning 3 and 5; cf. Vǫluspá 64). This information is not
compatible with the idea that Vingólf was the temple of the goddesses. See also Snorri’s
Edda (Ed) Faulkes 1988, 176.
16 Gylfaginning 14. (Ed) Faulkes 1988, 15.
17 Gylfaginning 13 in The Uppsala Edda. (Ed) Heimir Pálsson 2012, 24. In the Codex Upsaliensis
the name was Vinglóð, which makes the issue even more confusing.
18 Gylfaginning 15.
19 De Vries 1956–57, §583; F. Ström 1985, 97; Nordberg 2003, 171; Simek 2006, 205.
20 See e.g., Gurevich 1969; Meletinskij 1977; Hastrup 1985, 147.
21 See e.g., Clunies Ross 1994, 48–56; Brink 2004, 292–298; Ney 2006; Vikstrand 2006. See also
the critical remarks of Sebastian Cöllen 2011 and especially by Løkka 2010.
22 Sundqvist 2004a; 2007; 2011b; cf. Söderberg 2005; Jørgensen 2009.
23 Cf. Miller 2005; Henare et al. 2007; Engelke 2012.
buildings, i.e. the entire sanctuary. In each category I will discuss whether the
ritual structure of these buildings and places was constituted according to a
common model. I will also relate the results of these investigations to struc-
tures appearing in the mythical landscape. It will be argued that the structure
of the mythic-cosmic world was deliberately projected on to real sanctuaries
in the investigation areas. By projecting cosmic symbols and structures on the
human ceremonial buildings, these buildings were given form and meaning.24
This cosmic symbolism had several functions. It guaranteed the presence of
the divine powers at these structures. In a ruler-ideological perspective it was
also important to show that the political rulers who owned and controlled
these sacred places were closely related to the divinities and the mythic land-
scape. When entering his feasting hall at the sanctuary the ruler appeared on
a stage, which was surrounded by mythic-cosmic symbols, indicating that his
power came from a source beyond this world.
In the history of religions it has often been noticed that religion is often inte-
grated with political factors. The construction and location of ceremonial
buildings are therefore not only determined by religious motifs, but also by a
deliberate choice of localities, architecture and rituals, intended to symbolize,
authorize and legitimize political power. I will first concentrate on the locali-
ties and architectures of ceremonial buildings in the three investigation areas,
and look for possible references to and/or similarties with the cosmic-mythic
world.
24 It is a well-known fact that myths often function as a norm or prototype for human con-
duct and behaviour. See e.g., Malinowski 1948. Cf. Honko 1972; Segal 1998.
from the rest of the settlement.31 The halls at Helgö are likewise situated on
terraces and thus visible from afar (see ch. 4 above and fig. 12B below).
The dwellings of the political and religious elite were thus associated with
an elevated position in Svetjud. The majestic impression of these houses was
also reinforced by means of architecture and size. These houses were large,
had three aisles, and a roof carried by powerful posts erected in pairs. In Old
Uppsala the hall on the southern plateau was about 50 m long and 12 m wide.
One of the post-holes (G) was 1.47 m deep and at the surface some 50–90 cm in
diameter.32 This indicates that the house was not only long but also high. Two
of the post-holes (9 and 12) discovered by Sune Lindqvist under the present
church had a diameter of about 85 cm. No. 12 was as deep as 1.25 m.33 It has
been suggested that these holes may have belonged to a building, which thus
must have been high. One of the largest halls in the Mälaren region was found
in Lunda. It was 50 m long and in a topographic position which must have
overawed those who approached the building. One of the halls in Fornsigtuna
(the Garrison) was huge as well, some 40 m long and 10 m wide.
There are not so many detailed descriptions of the architecture of the cere-
monial buildings in the medieval written sources referring to Svetjud. However,
most of the textual evidence emphasizes that the pre-Christian buildings in
Uppsala were large and imposing. In Ynglinga saga 10, Snorri says, for instance:
“Freyr erected a huge hof at Uppsala and made his chief residence there.”34
That Uppsala had one or several large and impressive halls is mentioned sev-
eral times in Old Norse and medieval Latin texts. Ynglinga saga 36 says that
Ingjaldr built a hall when his father Ǫnundr died: “He had made ready a hall
in no wise smaller or less stately than the [hall] Uppsalr, which he called the
Hall of Seven Kings.”35 There he held a great funeral feast (erfi) for his father.
31 At many other sites in this region we may find possible hall buildings on platforms and
high positions in the landscape, see e.g., Granby-Hyppinge in Vallentuna, see Skre 2007a,
230; at Karvik in Bromma, see Ringsted 2005; Skre 2007a, 230; the cultic house at Alby in
Botkyrka parish, Södermanland, see Hjulström 2008; at Husby, Glanshammar, see Ekman
2000; at Vendel (Tuna) in Uppland, see Arrhenius 1998, 185; and at Runsa in Uppland,
Olausson (ed.) 2011 and Olausson 2014. The hall at Barksta, Västmanland, was also very
high. Karlenby 2005.
32 Hedlund 1993, 66.
33 Nordahl 1996, 47–50.
34 Freyr reisti at Uppsǫlum hof mikit ok setti þar hǫfuðstað sinn, . . . Ynglinga saga, in
Heimskringla I, Ísl. Fornr. 26, 23. Trans. Hollander, modified.
35 Hann lét búa sal einn, engum mun minna eða óvegligra en Uppsalr var, er hann kallaði
sjau konunga sal. Ynglinga saga, in Heimskringla I, Ísl. Fornr. 26, 66. Trans. Hollander,
modified.
The traditions about King Aðils mention a great hall or maybe a cultic building
at Uppsala. When describing Rolf’s visit to Aðils in Uppsala, Snorri used the
expression konungs hǫll,36 while Arngrímur Jónsson’s translation of Skjǫldunga
saga refers to a domus magnifica, “magnificent house”,37 and Saxo to Atisli
forte penates “Aðils’ strong house, hall”.38 Snorri also mentioned that Óðinn
“erected a large hof in Fornsigtuna and made sacrifices according to ancient
customs among the æsir”.39 The thirteenth-century authors thus believed that
great and impressive ceremonial buildings were erected in Svetjud in pre-
Christian periods. Archaeological evidence indicates that they surely were
right on that point.
The halls and cultic buildings in Svetjud were probably not only large, they
were also decorated. When Adam described the “temple” in Uppsala he stated
that it was “all fitted out with gold” (quod totum ex auro paratum est). Scholion
139 mentions that “a gold chain surrounds the temple” (catena aurea templum
circumdat). It hangs, shining, from the gables of the building, visible to all who
approach it. Adam’s description, however, may have been “enhanced” (not to
say distorted) by literary elements. Anders Hultgård has argued that the text
may have been affected by two classical literary figures, evidentia and ekphrasis
(Latin descriptio). Both sharpen a description. They would, for example, exag-
gerate the impression of a work of art, or of a building. Their purpose was to
make the narrative vivid and clear, and to encourage the reader’s imagination.
On the other hand, it is also conceivable that ceremonial buildings were in fact
decorated with gold and precious metals. There is archaeological evidence for
this: for instance, molten gold was discovered at one of the post-holes of the
Migration Period hall at Gudme. According to Frands Herschend, the posts at
Gudme might have been decorated with gold, but a fire had caused the gold
to melt and run down into the post-hole.40 The large Merovingian Period hall
on the southern plateau in Old Uppsala was also decorated with spirals and
perhaps heads of slaughtered animals.41
36 Ynglinga saga, in Heimskringla I, Ísl. Fornr. 26, 56–59. Snorra Edda, Skáldskaparmál ch. 44.
37 Skjǫldunga saga, Ísl. Fornr. 35, 30f.
38 Saxo 2.6.3.
39 [Óðinn] gerði þar mikit hof ok blót eptir siðvenju Ásanna. Ynglinga saga, in Heimskringla I,
Ísl. Fornr. 26, 16.
40 See Herschend in A.-S. Gräslund 1997, 108−109 and Petersen 1994, 36.
41 Personal communication from Per Frölund.
of Trondheim.46 The farm with the assumed cultic site at Værnes most likely
had a strategic location as well, between the Trondheimsfjord and the mouth
of Stjørdalselva, i.e. between important lines of communication.47 Similar
conditions may also be seen at the farm and the assumed cultic site of
Melhus. The modern name Melhus is a reduced form of an Old Norse name
Meðalhúsar ‘the middle farm’ (‘midtgarden’).48 It has been argued that this
name is secondary to a more original name of the farm. When the farm was
split up, the new name Meðalhúsar appeared, since it was situated between
two other farms. The name itself may thus indicate centrality in the settlement
district. The locations of the three fylki churches—in Alstadhaug, Sakshaug
and Haug—have all been assumed to have been pre-Christian cultic sites
before Christianity arrived in this region.49 Alstadhaug and Sakshaug espe-
cially have impressive situations in the landscape, high up on hills, and visible
from the sailing route far down in the Trondheimsfjord.
The hof places where ceremonial buildings were probably erected seem in
general to have had central locations in the settlement districts of Trøndelag.
They are situated in fertile plains at the bottoms of valleys where rivers run.50
For instance, the farm of Hov in Ålen has this central position by the river Gaula
in Gauldalen. It was placed in one of the fertile plains around the river. Hov in
Soknedalen also has this central position in the settlement district. It is located
in the upper parts of the river Sokna, where fertile plains surround it. Hove in
Meldal has a similar location at the bottom of the valley and right beside the
river Orkla. This farm belonged to the central settlement of the Orkladaføret.51
The farm Hov in Orkdal is strategically situated at the mouth of Orkla and just
beside the Trondheimsfjord. Finally, Hove in Selbu is located in the bog plains
by the lower parts of the river Nea, and close to the lake Selbusjøn, the natural
centre of the settlement.
Ceremonial buildings with impressive topographic situations are known
from Late Iron Age high-status settlements elsewhere in Norway. Borg in
Lofoten, for instance, has a great topographic location.52 The impressive main
building (I:1a) with the “hall room” (C), dating to the Viking Age, was situated
on the top of a low ridge, at the highest point of the site.53 At exactly the same
spot, the large Merovingian Period house (I:1b) had been erected earlier.
The tradition of building halls on plateaus cannot be seen in the archaeo-
logical sources from Trøndelag or northern Norway, however. Neither the cultic
house at Mære nor the Iron Age hall at Hovde was built in this way. In neither
case, however, was this necessary in order to express loftiness or an exposed
position, since the topographic conditions were otherwise so favourable. This
is also the case of Borg in Lofoten. As far as I know, only one hall found in
Norway was built on a plateau, namely the hall at Huseby, Tjølling, close to
Kaupang in southern Norway.54 This plateau was 36 m long and 13 m wide, and
had bowed long sides. It was situated on the crest of a rocky hillock and it was
widely visible.
In other parts of Norway, more impressive banqueting halls have been dis-
covered. The most interesting parallel to the ceremonial buildings of Trøndelag
is the hall building at Borg, Vestervågøy, since it also included gold foil figures.58
The Migration Period hall (I:1b) was 64 m long and 7–8 m wide, while the
Merovingian and Viking Period house (I:1a) was roughly 83 m long and 7.5–9
m wide. This latter hall is the most impressive multifunctional hall building
hitherto found in Scandinavia. Elsewhere in Norway, large Iron Age halls have
also been found, for instance at Huseby, Tjølling in Vestfold. It was 35 m long
and 7.9 to 11.7 m wide and located on a terrace.59
The written sources do not provided much guidance about the exterior char-
acter of the ceremonial buildings of Trøndelag. Snorri Sturluson comments
on the hof at Lade in, for instance, Hákonar saga góða 17, where he describes
it as a large banqueting hall where ceremonial feasts took place. Óláfs saga
Tryggvasonar (manuscript AM 310 4to), of Oddr munk, reports that Earl Hákon
had a great hof sanctuary (mikla hof) at this site, with plenty of divine images.60
Njáls saga 87 states that the greatest hof sanctuary in Norway was located at
Lade. If we combine the information found in Old Norse traditions, we get the
impression that the Viking Age hof of Lade was a large hall, intended for great
public gatherings and ceremonial feasts. Some sources cannot be regarded as
historically reliable concerning these matters. In Egils saga 11, for instance,
there is a description of a great, but temporary banqueting house, located on
the island of Torgar in Hálogaland, where Þórólfr Kveld-Úlfsson held a feast for
Haraldr Finehair:
There was a large barn (kornhlaða) that Þórólfr had fitted out with
benches so that the drink could be served there, because he did not have
a room large enough to accommodate that number of people [800 people
in total]. Shields were mounted all around the building. The king sat in
his high-seat, and when the upper and lower benches were both filled . . .61
58 See mainly Johanson & Munch 2003; Herschend & Mikkelsen 2003; Munch 2003a and
2003b.
59 Skre 2007a.
60 Óláfs saga Tryggvasonar eptir Odd munk Snorrason. Ísl. Fornr. 25, 193.
61 Egils saga, Ísl. Fornr. 2, 28f. Transl B. Scudder.
62 Meulengracht Sørensen 2003.
the king, by showing that he had more men than the king at his disposal in his
large (temporary) hall building. In this case the saga description of the feasting
house may not be used as a historical source, since it has a literary function.
The archaeological evidence from Borg nevertheless indicates that Viking Age
ceremonial buildings may have been quite large in Norway.
It seems as if the size and character of hall buildings in Trøndelag and else-
where in Norway varied a lot. The large hall in Borg, Lofoten, was most likely an
enormous status symbol. Whether such impressive ceremonial buildings also
existed in Trøndelag is uncertain, but the archaeological finds at Hovde, for
instance, indicate that a similar type of three-aisled hall buildings also existed
there in the Middle Iron Age. Even though this house was not as big as the one
in Borg, it must have been awesome for people passing through the strait at the
mouth of the Trondheimsfjord. The large post-holes of the Viking Age house at
Mære, and its location on the top of the hill, indicate that this house too must
have appeared splendid to those approaching it.
63 See Landnámabók S85, H73, Ísl. Fornr. 1, 125. See also above.
64 Kjalnesinga saga, Ísl. Fornr. 14, 7.
Iceland, since Vatnsdœla saga 15 states: “He built a great hof a hundred feet
long.” It was built at his homestead and “he called the farm Hof”.65
Whether the saga authors are reliable when describing the great size of the
hof sanctuaries is uncertain. As noted in chapter 4, there is not much archaeo-
logical evidence of large ceremonial buildings in Iceland. Apart from the hall
building (38 m in length) on the chieftain farm of Hofstaðir at Mývatnssveit,
there are few finds indicating large long-houses in Iceland. The normal halls/
long-houses in Iceland fall between 12 and 16 m. The hall at Aðalstræti, Reykjavík,
for instance, was about 16 m in length. Also very small is the late tenth-
century hall found at Sveigakot, Mývatn. This hall is a typical Scandinavian
Viking Age hall, but in miniature, with concave long walls, a three-aisled
design and a centrally placed hearth.66 The floor area was 7.3 m by 3.2 m. The
fact that Icelandic halls/long-houses in general are small should be taken into
consideration when assessing the house at Hofstaðir as a monumental or large
house.67 The internal area of Hofstaðir II was almost twice as big as the areas
in the houses of Aðalstræti, Grelutóttir, Granastaðir, Eiriksstaðir, Vatnsjǫrður,
Goðatættur II, Sveigakot and Hvítarhólt IX.68 The Late Iron Age halls in Svetjud
and Trøndelag are much grander than the hall buildings of Iceland.
The archaeological evidence indicates, at any rate, that the picture painted
by the medieval saga scribes of large halls and cultic buildings in the Viking
Age was not totally distorted. There are also some more realistic descriptions in
the sagas of the architecture and size of Viking Age feasting halls in Iceland. It
seems as if the medieval transmitters had sometimes witnessed such buildings
themselves. One such description is rendered in Fóstbræðra saga 13, which is
dated to the second half of the thirteenth century. The text tells us that Þorgeirr
Hávarsson was sent on a mission to Iceland by King Óláfr Haraldsson c. 1020,
in order to avenge one of the king’s retainers who had been hurt by a man
called Þórir. Þorgeirr arrived at Reykjahólar in western Iceland and engaged
a carpenter called Vélagr and together they built a feasting hall (skáli): “[they]
each built one side of the skáli. It was planked on the long walls, but did not
have any other plank-walls; those planks lasted right up to the time when the
first Bishop Magnús was at Skálholt.”69 This skáli was thus built c. 1020 and it
65 . . . hann reisti hof mikit hundrað fóta langt . . . Bœr sjá skal heita at Hofi. Vatnsdæla saga, Ísl.
Fornr. 8, 42.
66 Orri Vésteinsson 2009, 495.
67 Lucas 2009, 376f.
68 Lucas & McGovern 2008, 21 and Lucas 2009, 376.
69 . . . ok gerði sínum megin hvárr þeira skálann. Skálinn var um endilangt þilinn, en eigi ǫðrum
þiljum; þau þili héldusk allt til þess, er Magnús byskup var at staðnum í Skálaholti inn síðari.
Fóstbræðra saga. Ísl. Fornr. 6, 184. Trans. Meulengracht Sørensen 2003.
70 Hann er nítján alna ok fjogurra tuga alna langr; hann stóð enn, er Árni biskup inn síðari var
vígðr til Skálaholts. Fóstbræðra saga. Ísl. Fornr. 6, 184.
71 Meulengracht Sørensen 2003, 266.
72 Lucas 2009, 402.
73 See Lucas 2009, 236ff.
74 Orri Vésteinsson 2009, 500.
ethnicity.75 Sveigakot may very well have been controlled by either the rich
farm at Baldursheimur or the farm at Grænavatn.
It is interesting to note that the small pit-house (2 × 2.16 metres) at Hólmur
close to Hornafjörður (Höfn) in south-east Iceland, which has been interpreted
as cultic house (“blót house”), has an extraordinary location. It was situated in
an exposed position on a mound or a hill (15 × 25 m and about 2 m high) some
250 metres north-east of the Viking Age farm at Hólmur.76 Perhaps the location
of this house also had a symbolic significance.
ably alludes to the hof and the hǫrgr of Vǫluspá 7, when he mentions that the
gods decided to build a gold-glittering hof and next to it a hǫrgr called Vingólf:
It was their [the gods’] first act to build the hof that their thrones stand in,
twelve in addition to the throne that belongs to All-father [Óðinn]. This
house is the best that is built on earth and the biggest. Outside and inside
it seems like nothing but gold. This place is called Glaðsheimr. They built
another hall, this was the hǫrgr sanctuary that belonged to the goddesses,
and it was very beautiful. This building is called Vingólf.79
Most likely Snorri also had another poem in mind when describing the “best
and biggest hof on earth” built by the gods at Glaðsheimr, namely Grímnismál.
According to the list of godly residences in this Eddic lay, the gilded building,
which was situated at Glaðsheimr (‘bright home’ or ‘joy-home’),80 seems to be
identical with Valhǫll, i.e. the hall where the Æsir deity Óðinn gathered the
warriors slain in battle (einherjar) around him.81 In stanza 8 we may read thus:82
79 Var þat hit fyrsta þeira verk at gera hof þat er sæti þeira standa í, tólf ǫnnur en hásætit þat
er Alfǫðr á. Þat hús er bezt gert á jǫrðu ok mest. Allt er þat útan ok innan svá sem gull eitt. Í
þeim stað kalla menn Glaðsheim. Annan sal gerðu þeir, þat var hǫrgr er gyðjurnar áttu, ok
var hann allfagr. Þat hús kalla menn Vingólf.
80 Simek 2006, 112. Cf. Lindow 2001, 145.
81 The name Valhǫll should probably be interpreted as ‘the hall of the slain warriors’. It
has been argued that this name is late, even though it appears in tenth-century sources
such as Hákonarmál, Eiríksmál and Grímnismál. The word hǫll in the sense of ‘banquet-
ing hall’ was not used in name-giving in Scandinavia. Hǫll is probably a loan-word from
OE heall or OSax, OHG halla, which means ‘banqueting hall’. This does not entail that
the mythological conception behind the term Valhǫll must be a late construction. Most
likely the old expression for this conception was Óðins salr ‘Óðinn’s hall’, which appears in
Eiríksmál, for instance. It also appears in place-names, such as Onsala and Odensala (both
<Óðinssalir). Brink 1996a; 1999b; Nordberg 2003.
82 It should be noted that Snorri in Gylfaginning 2 describes Valhǫll as a “high hall” (háva
hǫll). Its roof was covered with gilded shields like tiles (var lagt gyltum skjǫldum svá sem
spánþak).
83 Glaðsheimr a fifth is called,/ there gold-bright Valhǫll/ rises peacefully, seen from afar; . . .
It is interesting to note that “the gold-bright Valhǫll” was raised peacefully and
“seen from afar”. It could be compared with, for instance, Adam of Bremen’s
information about the “temple” in Uppsala, which may be quoted here once
again: “A golden chain goes round the temple. It hangs over the gable of the
building and sends it glitter far off to those who approach, because the shrine
stands on level ground . . .” We have already noted that archaeological finds
indicate that ceremonial buildings, at least in Sweden and Norway, had an
elevated position in the cultural landscape, and thus could be seen from afar.
Archaeological evidence also supports the assertion that such buildings occa-
sionally were adorned with gold and other objects, like Valhǫll, for instance the
Migration Period hall at Gudme, and the Merovingian Period house at Uppsala.
In the mythical traditions there are also other hall buildings, such as
Breiðablik, Glitnir and Valaskjálf.84 Whether they should be seen as separate
mythic buildings or different designations of one cosmic hall is uncertain.
Breiðablik ‘the far-shining one’ and Glitnir ‘the shining one’ have names which
indicate an affinity to Valhǫll, which according to Grímnismál was considered
as ‘gold-bright’. The name Valaskjálf may also be related to Valhǫll. According
to Snorri, Óðinn owns this hall.85 The first element of the name could also be
associated with the son of Óðinn, Vali. It is also possible that these names refer
to different cosmic halls. In any case, some of these mythical names seem to
emphasize important aspects of ceremonial halls, namely, that they should be
decorated and visible from afar.
It seems as if the hall buildings and cult houses in Svetjud, Trøndelag and
sometimes in Iceland occupied a prominent position in the landscape. The
sovereignty of these buildings was reinforced by their magnificent size and
architecture. Central locality, elevated position, size and architecture were
probably associated with ideological aspects and a deliberate strategy of being
observed and associated with the divine world. Most of these features appear-
ing in relation to ceremonial buildings may be seen also in connection with
essential “mythical houses” in the divine world. Perhaps these recurrent fea-
tures and correspondences indicate an “ideal state”, the existence of a “com-
mon model”, and/or a cosmic symbolism of the cultic sites in the three regions
(see further in ch. 15 below). Without doubt the ceremonial buildings in all
three investigation areas were important agencies and/or symbols of power
for the elite. The “bombastic expressions”, the monumental sizes and splendid
84 See e.g., Grímnismál 6, 12, 15; Gylfaginning 16, 21, 31. See also Simek 2006, 44, 113, 346.
85 Gylfaginning 16.
The value of this text as a source for Viking Age conditions can obviously be
debated. It is inserted in a passage depicting how King Óláfr kyrri Haraldsson
86 There are two manuscripts of Fagrskinna. The manuscript *A burned in Copenhagen
1728, but was copied by Ásgeir Jónsson three times in AM 52 fol., AM 301 4to and AM 303
4to. *A is dated to the fourteenth century. Also manuscript *B burned in Copenhagen
1728, but was copied by Ásgeir Jónsson once (in OsloUB 371 fol.) and by Eyjólf Björnsson
twice (AM 51 fol. and AM 302 4to). One page of this manuscript is preserved NRA 51. The
manuscript *B is dated to c. 1250. See Fagrskinna, Ísl. Fornr. 29, lxi. On Fagrskinna, see
Kolbrún Haraldsdóttir 1994.
87 Þat var forneskjusiðr í Nóregi ok í Danmǫrk ok í Svíaveldi, þar sem váru konungsbú ok
veizlustofur, váru dyrr á hvárumtveggja enda stofunnar, en konungs hásæti var á miðjan
lengbekk, þann er vissi móti sólu. Sat þá dróttning á vinstri hǫnd konungi. Var þat þá kallat
ǫndvegi ok enn ágætasti sess út í frá hvárn tveggja veg til kvenna ok karla, er næst var ǫnd-
vegi, en hinn óvegligastr, er næstr var durum. Enn gǫfgasti maðr, sá er var gamall ok vitr, var
kallaðr konungs ráðgjafi, sem í þat mund var konungum títt at hafa með sér gamla spekinga
til þess at vita forna siðu ok dœmi sinna forellra. Þessi maðr skyldi sitja á annan langpall
gagnvert konungi, ok hét þat et óœðra ǫndvegi. Váru konur á hœgri hǫnd hónum, en á vinstri
hǫnd sátu karlar. Var þá títt hǫfðingjum at bera ǫl um eld ok drekka á ǫndvegismann sinn.
Þótti þat vera mest virðing at sitja fyrir konungs ádrykkju. Fagrskinna, Ísl. Fornr. 29, 300.
My trans.
(r. 1066–1093) moved the high-seat (hápall) from the middle of the long wall to
the middle of the “crossway stool” (á miðjan þverpall) at the gable.88 The dis-
tance of time between Óláfr kyrri’s reign and the scribe of Fagrskinna is, on the
other hand, just a little more than a hundred years. Consequently, it is possible
that Fagrskinna’s description of the hall is reliable. The scribe also quotes a
stanza contemporary with King Óláfr in support of his description. This stanza
was composed by Arnórr Þorðarson jarlaskáld in honour of Earl Þorfinnr, and
it describes the drinking rituals which were supposed to be observed by the
person sitting opposite the ruler.89 Since the content of the poem supports
parts of the prose account, the latter may arguably be considered a trustworthy
source for interiors of Late Viking Age halls in general.90
Most likely the description of the interior of ceremonial buildings in
Fagrskinna influenced other medieval authors and scribes writing about feast-
ing halls (veizlustofur). Snorri Sturluson, for instance, may have been inspired
by this description when presenting Óláfr Haraldsson’s hirðstofa ‘retainer hall’
at the royal residence (konungsgarðr) in Niðaróss.91 In this description, he says
that the hall had doors at both ends, and that the king’s high-seat (hásæti) was
situated in the middle of the hall. Opposite the king’s high-seat was the “lesser
high-seat” (ǫðru ǫndugi), where the king’s marshal sat, and next to him the
noble guests. On those occasions when ale was consumed, fires were lit in
the middle of the floor. All these statements resemble Fagrskinna’s description
of feasting halls. It is quite possible that Snorri partly borrowed from this text
when describing the interior of the feasting halls. Perhaps he also borrowed
from it when describing the hof buildings in Trøndelag.
88 Similar information appears in Óláfs saga kyrra 2, in Heimskringla III, Ísl. Fornr. 28, 204.
See Birkeli 1932; Steinsland 1991, 77.
89 Finnur Jónsson interpreted it thus: “Jeg tiltalte (drak til) krigeren,/ da jeg sad ligeoverfor
ham;/ fyrsten drak altid hver vinter/ fyrsten drak altid hver vinter/ det brave mandskab
til.” Hétk, þás (hvern vetr) sǫ́tum,/ hrafns verðgjafa, (jafnan/ lið drakk gramr á góðar)/ gag-
vart (skipa sagnir). See Skj. B1, 316.
90 The passage in Fagrskinna probably has a specific relevance to conditions in Trøndelag.
Before the passage about the banqueting halls (veizlustofur), a stanza made by Steinn
Herdísarson is quoted where Óláfr kyrri is called þengill Þrœnda ‘the king/chieftain of the
Þrœndir’. See Fagrskinna, Ísl. Fornr. 29, 299. See also Skj. B1, 282. In the prose immediately
after this stanza the increasing significance of Niðaróss is emphasized.
91 See Óláfs saga helga 57, in Heimskringla II, Ísl. Fornr. 27, 72.
as a place of honour, and intended for the owner of the hall: the king or
chieftain.92 The high-seat was probably elevated in some way, so that the ruler
actually looked down on the other people in the room. The other participants
seem to have been seated according to rank and gender at specific places in
the ceremonial building, but on a lower level than the hall owner. The terms
fótpallr ‘foot bench’ or fótskǫr ‘foot step’, which are associated with the high-
seat in some sources, indicate that this seat stood on some sort of dais or plat-
form.93 The elevated position of this seat is even marked in the first element
of the compound term hásæti; the adjective há- means ‘high’. According to the
medieval texts, the high-seat was a social and political fixture with ideological
implications. In addition to that it played a central role in religious and judicial
contexts. It was thus a symbol of tradition, power and authority. A person who
had the right to occupy such a seat was called hásætismaðr, ǫndvegishǫldr or
ǫndvegismaðr.
There has been debate in previous research concerning the placing of the
high-seat in the hall and whether hásæti and ǫndvegi were identical or not. I
will not go too far into that debate, just summarize it, and then make some
comments on it in the context of ceremonial buildings in the investigation
areas. Based on the information in Fagrskinna and other Old Norse texts, it was
argued early on that the pre-Christian high-seat (whether it was called hásæti
or ǫndvegi) was situated in the middle of the northern long bench.94 On the
opposite side of the high-seat was “the lesser seat” (óœðra ǫndvegi) intended for
the highest-ranking guest, or the most honourable person after the ruler of the
house. In the middle of the eleventh century this custom was changed when
the high-seat was moved to the eastern gable. Hjalmar Falk opposed some
of these ideas and stated that a distinction must be made between ǫndvegi
and hásæti. The former term was applied to the place of honour in farmers’
houses, while the latter term referred to the seat of rulers in royal halls.95 A
similar proposal was made by Emil Birkeli. According to him the concept and
92 See e.g., Birkeli 1932; Holmqvist 1962; Drescher & Hauck 1982; Steinsland 1991; Herschend
1997; 1998; H. Beck 2000; Sundqvist 2014. According to Old Norse prose texts, high-seats
(ǫndvegi) also appeared in the houses of minor chieftains and farmers. See Birkeli
1932, 29ff.
93 See e.g., Ynglinga saga 36, in Heimskringla I, Ísl. Fornr. 26, 66; Haralds saga in Hárfagra 8,
in Heimskringla I, Ísl. Fornr. 26, 100; Saga Magnúss góða 23, in Heimskringla III, Ísl. Fornr.
28, 37; Flateyjarbók I, 42.
94 Keyser 1868; cf. Gudmundsson 1894; Nikolaysen 1881–91.
95 Falk 1911–16.
phenomenon of hásæti was developed and emancipated from the old ǫndvegi.
It was also accommodated in a new and more aristocratic context.96 The old
term ǫndvegi referred originally to the seat of honour in farmers’ houses, which
was always situated in the north-eastern corner. In this context Birkeli also
discussed the etymology of the term ǫndvegi. Usually it has been interpreted as
a compound word with the preposition ǫnd ‘opposite’ (‘rett imot’) as the first
element and vegi (derived from vegr) ‘side, edge, margin’ (‘side, retning, kant’)
as the second element.97 The compound word should be translated as ‘the seat
opposite [the sun]’ (‘sætet rett imot [solen]’). In Birkeli’s opinion, however,
the first element ǫnd should be translated ‘soul’ and the second element vegi
(vegr) as ‘road’. The term ǫndvegi should thus be interpreted as ‘the road of the
souls’ (‘åndeveien’), i.e. the place where dead ancestors have their place in
the house. The ǫndvegi in the north-eastern corner of the house was thus origi-
nally a sacred sphere where the ancestor cult took place.98
Birkeli has gained weak support for his theory about the meaning of ǫndvegi.99
His ideas about the location of the high-seat have also been questioned.
Wilhelm Holmqvist stated, for instance, that Falk and Birkeli were leaning too
much on written evidence concerning this issue.100 According to Holmqvist,
the location of the high-seat must be related to building technology. In a hall
or a long-house, with an open fireplace in the centre of the room, it is natural
to place the high-seat in the middle of the long wall. During the Middle Ages
a new type of (smoke) oven was introduced in houses, which led to the corner
location of the high-seat. Frands Herschend is also critical of the reasoning in
previous research, and has proposed an alternative schematic development of
the position of the prominent seats in the house, hásæti or/and ǫndvegi.101 His
proposal was based on new empirical data gained from recent archaeological
finds in Iron Age halls.
It is quite possible that Falk and Birkeli were on the right track when they
argued that the aristocratic and royal hásæti was developed from the ǫndvegi
of farmers’ houses,102 but it is probably impossible to prove this theory from
existing sources. In the written sources from the thirteenth century, however,
103 Birkeli (1932, 6f.) stated, however, that Snorri only used the concept hásæti in royal con-
texts in his Heimskringla. Perhaps Snorri made a distinction between hásæti and ǫndvegi.
104 See e.g., Eyrbyggja saga 33 (Ísl. Fornr. 4, 91f.), where these two concepts refer to one and
the same seat.
105 Fritzner 1954.
106 See e.g., Ynglinga saga 36–37, in Heimskringla I, Ísl. Fornr. 26, 65ff., but also in more
historical accounts such as Fagrskinna 79, Ísl. Fornr. 29, 300 and Óláfr saga helga 79, in
Heimskringla II, Ísl. Fornr. 27, 113.
107 See Herschend 1995; 1998, 28.
108 See also K. Lamm 2004, 51; Lund Hansen 2011.
109 Holmquist Olausson 2001, 15; Holmquist Olausson & Kitzler Åhfeldt 2002, 16.
located in the middle of the long walls. Drinking goblets have been found at
these “seats”, and the hearth was located between them.110 It is possible that
the high-seat in the younger hall of Lunda (c. 450–650) was also close to the
hearth. The central post-holes which flanked it there were distinguished from
the other post-holes both in size and in the richness of the finds. They had a
diameter between 0.74 and 1.68 metres and were 0.46 and 0.78 metres deep.
In these holes a lot of objects were deposited, such as a bead, an amber-stone,
a comb and a fishing hook. The other post-holes of the building had very few
finds. The excavators suggest that the high-seat was in this area, close to the
northern wall. According to them, the finds deposited there indicate that some
kind of cult took place.111 The high-seat area has also been identified at the
Migration Period hall of Högom, Medelpad.112
High-seats may also be seen in the Late Iron Age iconography from eastern
Scandinavia. On one Gotlandic picture stone from Buttle Änge no. I the inte-
rior of a hall is depicted with two seats located opposite each other (see fig. 9a
and 9b). According to Frands Herschend, they must be interpreted as a “high-
settle” and a “lesser high-seat”. The right one is higher than the left seat, since it
is standing on a dais. It was interpreted as the “high-seat”.113 A “high-seat” and
a “lesser high-seat” are also evidenced on the stone coffin from Sanda, Gotland,
dated to the eleventh century (see fig. 10).114 Two persons are sitting opposite
each other in these seats, in some sort of a building with roof and walls. The
one who is sitting in the right seat, probably a “high-seat”, seems to be offered
a drinking vessel, or some other object, by a third person. Below this motif,
three men (or deities) are walking in a procession carrying different attributes.115
This motif emphasizes that the upper image may also concern some kind of
ritual or ceremony. This coffin is also decorated with a runic inscription at the
very top. This inscription (G 181) does not give any clues to interpret the images
below it, since it only says: Rōðvīsl auk Farbiorn auk Gunnbiorn “Rodvisl and
Farbjörn and Gunbjörn”.116
Figure 9a
Detail from the Gotlandic picture stone from Buttle Änge no. I.
After Herschend 1998, based on Lindqvist 1941–42)
Figure 9b
The picture stones from Buttle. Carl Gustaf
Rosenberg.
Figure 10
Stone coffin, Sanda, Gotland.
Photo: Bengt A Lundberg.
ATA, RIKSANTIKVARIEÄMBETET
226.
117 Cf. Olivecrona 1942; Hellström 1971, 221; Steinsland 1991, 66–86; Herschend 1997, 9, 49–59;
Sundqvist 2002, 268ff.; 2014.
regarded as some kind of cosmic pillars (axis mundi sg.) or an interface to the
cosmic world (see below). The ritual deposits of gold foil figures at the position
of the assumed high-seats in the halls of Helgö and Svintuna (Östergötland)
may be regarded as good evidence of such notions (see further ch. 12 below).
The religious significance of the high-seat is only slightly evidenced in the
written sources referring to eastern Scandinavia. Adam of Bremen, for instance,
describes the interior of the “temple” at Ubsola thus: Thor in medio solium
habeat triclinio “Thor occupies a seat in the middle of the dining-room.” Latin
solium connotes ‘high-seat, throne, insignia’.118 The prototype for this high-seat
must have been intended for earthly rulers. There might have been some kind
of micro-macro symbolism or ritual relationship between the real high-seats
in the halls of the Mälaren region and the divine high-seat of Þórr in the sanc-
tuary of Uppsala, as described by Adam.119 When the rulers entered the high-
seats they actually imitated the prototypical actions performed by the gods
in the mythical time. At the same time they were in contact with the divine
world when sitting in these seats (see ch. 7). It is also possible that the enter-
ing of the high-seat could be plainly seen as a performative ritual intended to
authorize and legitimize the ruler (see ch. 13 below). The ritual and ideological
dimension of the high-seats among the Svear is also attested in the Old Norse
sources. In these texts the high-seat plays an important role in connection with
drinking rituals and rites of passages performed by rulers (see further ch. 9 and
ch. 13 below).
The religious significance of high-seats in the Mälaren area may also be
attested in another sense. Among the objects in the tenth-century hoard from
Eketorp, Närke, was a silver miniature “high-seat” which had been used as a
pendant or amulet.120 This type of miniature is often interpreted as a repre-
sentation or attribute of the divinities. They also give us an image of how “real
high-seats” were shaped and decorated. In a Birka skeleton grave, no. 844, there
was a similar amulet to the one in Eketorp from the Viking Age (fig. 11).121 Nine
“Klotzstuhl-Amulette” have been discovered in Scandinavia.122 Two of them
were in the coffin graves nos. 632 and 968 at Birka and dated to the Viking
Age. One was also found at Ihre, Hellvi parish, Gotland. In addition Karl Hauck
118 Lat solium “der hohe erhobene Sitz, der Thron, . . . Regale”. Georges 1951 (1913), 2711. Habel
(1959, 371) has “Sitz, Stuhl, Bank, Tribüne, Katheder, Thron; Hochwarte; Sarg”.
119 See e.g., Nordberg 2003, 170ff.
120 Ekelund 1956; Stenberger 1971 (1964), 731; Drescher & Hauck 1982, 267–269.
121 Drescher & Hauck 1982, 267–269.
122 Drescher & Hauck 1982, 248–256.
Figure 11
A “high-seat” amulet (Sw. kubbstol) from Birka. Statens Historiska Museum.
The next evening King Hákon was compelled to take part in the ritual meal and
eat the horse-meat (hrossaslátr) and then enter the high-seat. Even though
Snorri’s account has been criticized by scholars, there is no reason to suspect
that he misunderstood the significance of the high-seat.
There are other Old Norse texts indicating the existence of high-seats in this
region. Landnámabók, for instance, mentions that the chieftain and hofgoði
Þórhaddr from Mære dismantled the hof and brought all the “temple soil” and
the high-seat posts (súlurnar) to Iceland (see ch. 5 below). These posts had
most likely been related to the high-seat of the old hof in Mære. A Viking Age
ceremonial building, perhaps a hof, has been detected in Mære, Trøndelag.
When Hans-Emil Lidén investigated this house, he discovered four post-holes,
which do not logically fit in relation to the rest of the building.128 Whether
these post-holes belonged to a building or stood within the building is uncer-
tain. They constituted a square measuring about one square metre. They were
located at the long wall, not far from the fireplace, and the gold foil figures
were found beside them. Inspired by information from Old Norse texts, such
as Landnámabók, Lidén first interpreted the four post-holes as remains of the
high-seat, or more precisely, the high-seat posts (ǫndvegissúlar). In a later
publication he believed that three of the four post-holes belonged to posts sup-
porting the wall.129
As far as I know, no other possible “high-seats” with gold foil figures have
been found in Late Iron Age houses in Trøndelag. The two gold foil figures
discovered at Hevne were probably located at the old farm, but no post-holes
or traces of houses were found there. At the hall in Hovde neither gold foil
figures nor remains of a high-seat were discovered. The closest parallel in
Norway to the finds at Mære is probably the structures and finds excavated at
Borg, Vestervågøy. In the Merovingian and Viking Age “hall-room” (Room C) at
this site (I:1a) five gold foil figures depicting the motif of a couple were found
(fig. 8C).130 These objects seem to be related to post-holes (actually one post-
hole) and an area interpreted as the place of the high-seat. The assumed high-
seat was located in the north corner of the room, where other finds were also
made which may indicate that ritual activities took place there during the
Early Viking Age. Another close parallel is the important hall-building with 29
gold foils found in Hov in Vingrom, Oppland. The gold foils were found in con-
nection with four post-holes, which may be interpreted as traces of a high-seat.
Fragments of glass vessels, weapons, seaxes, knives, pearls and a large amount
of strike-a-lights were also discovered there.131
The men wanted the bishop to run out [of the house], but he answered
that he did not need to escape his seat, since this was the will of God.
When the fire was over he sat there wholly unburned in his high-seat.136
These men probably believed that the high-seat opened the door to the Other
World. Perhaps they also wanted to die in full regalia (“in Amt und Weiden”) as
they sat in their high-seat.
131 Oral communication from Heid Gjøstein Resi, Kulturhistorisk museum, Oslo University,
via the reviewer of my manuscript.
132 See e.g., Eyrbyggja saga 13, Ísl. Fornr. 4, 23f.; Vatnsdæla saga 22, Ísl. Fornr. 8, 61.
133 Ásgrímr tók Njál af hestinum ok bar hann inn ok setti hann í hásæti. Brennu-Njáls saga 118,
Ísl. Fornr. 12, 296. Cf. Grettis saga 19, Ísl. Fornr. 7, 61–71. See also Birkeli 1932, 42f.
134 Vatnsdæla saga, Ísl. Fornr. 8, 61f.
135 Eyrbyggja saga 33, Ísl. Fornr. 4, 91f.
136 villdu menn at byskup hefdi þar vt farit en hann sagdi sier hæfdi ei at ryma sitt sæti þvi þetta
værj guds vilie. Sat hann vpp obrunen med aullu j sinu hasæti þa af var elldurinn. Gottskalks
Annaler 1148, Islandske Annaler, 321. Cf. Hellström 1971, 220f. My trans.
The term ǫndvegissúlar also intimates that they were connected to the high-
seat (ǫndvegi) in some sense.141 High-seat posts are mainly attested in written
sources referring to Norway and Iceland, but some archaeological evidence
suggests that they also occurred in Svetjud.
When studying the literary sources it is clear that the high-seat posts some-
times were considered as ritual objects which could be separated from the
ceremonial building and the high-seat. When travelling to Iceland from
Norway, for instance, these posts were brought on the ships and used in a type
of divination ritual or in order to get good luck.142 Landnámabók states that the
first landnámsmaðr Ingólfr and his foster-brother Hjǫrleifr left their home in
Norway. Before leaving, Ingólfr made a great sacrifice in order to find out about
his future, and to achieve luck with his enterprises.143 He was advised to take
off to Iceland.144 It is mentioned that Hjǫrleifr never sacrificed. Anyhow he fol-
lowed Ingólfr to Iceland. As soon as Ingólfr arrived at the coast of Iceland, he
cast his high-seat posts overboard, in order to get luck.145 Ingólfr also declared
that he would settle down where the posts came ashore.146 He sent out his
slaves to find the high-seat posts. But it was not until the third winter after
their arrival that they could be found at an isthmus in south-western Iceland.
Although the conditions for living were not so propitious on this spit of land,
Ingólfr decided to settle there since the high-seat posts and the supernatural
powers had pointed out this place to him.
According to Landnámabók, many men followed Ingólfr’s example and
settled in Iceland.147 They also brought their high-seat posts from the sanc-
tuaries in Norway and cast them overboard when approaching Iceland. For
instance, Loðmundr inn gamli,148 Hrollaugr Rǫgnvaldsson149 and Þórðr skeggi
Hrappsson followed this custom.150 People from Trøndelag also performed
141 For discussion, see Birkeli 1932, 23ff.; Holmqvist 1962, 292.
142 Cf. Strömbäck 1928b; Baetke 1942; Clunies Ross 1998, 142ff.
143 . . . at blóti miklu ok leitaði sér heilla um forlǫg sín . . . Landnámabók, SH 7, Ísl. Fornr. 1, 42.
144 Fréttin visaði Ingólfi til Íslands. Landnámabók, SH 7, Ísl. Fornr. 1, 42.
145 . . . skaut hann fyrir borð ǫndugissúlum sínum til heilla. . . Landnámabók, SH 8, Ísl. Fornr. 1,
42. ON heill (f. and n.) means ‘luck’ or ‘good omen’. Heggstad et al. (1993, 176): ‘lykke, (godt
el. vondt) varsel, teikn’ or ‘ting som gir lykke’.
146 During the taking of land Hjǫrleifr was killed by one of his slaves.
147 The story about Ingólfr functions as a paradigmatic example of good behaviour and may
be considered as a dǿmisaga or an exemplum. See Wellendorf 2009; 2010.
148 Landnámabók, S289 H250, Ísl. Fornr. 1, 302.
149 Landnámabók, S289 H270, Ísl. Fornr. 1, 316f.
150 Landnámabók, H 11, Ísl. Fornr. 1, 124f.
such rituals. Landnámabók mentions that, for instance, the hofgoði Þórhaddr
from Mære took down the hof and brought all the soil of the sanctuary and
the high-seat posts to Iceland (see ch. 5 above). Eyrbyggja saga 4 presents a
similar story (see ch. 3 and 4 above). It thus seems as if the theme of casting the
high-seat posts was developed into a topos in both Landnámabók and the saga
narratives about the landnám.151
The management of the high-seat posts and the ritual occupation of
land in the saga accounts occur in a clearly pre-Christian context. Most schol-
ars have also argued that these narratives reflect genuine pre-Christian cus-
toms and rituals from the landnám period.152
This opinion has recently been questioned by Jonas Wellendorf, because
there are a couple of narratives in the Hauksbók redaction of Landnámabók
with a content similar to the accounts of high-seat posts, but with a clearly
Christian context.153 According to Dag Strömbäck, these Christian stories
reflected a “cult action continuity”. He interpreted them as a sign of the great
significance and bearing capacity the pagan custom also had for the early
Christians.154 Wellendorf opposes this way of reasoning and suggests that the
accounts of high-seat posts might be secondary to the Christian accounts found
in Hauksbók. In order to support his argument, Wellendorf presents two medi-
eval ecclesiastic narratives, which may have functioned as models for these
151 See e.g., Wellendorf 2009, 1018f.; 2010. There are some traditions of the casting of high-seat
posts which do not turn up in Landnámabók, for instance in Laxdæla saga 5, Ísl. Fornr. 5,
9 and Kormáks saga 2, Ísl. Fornr. 8, 205.
152 Dag Strömbäck (1928b) interpreted this pagan ritual as a way of transporting the sacred
power and good fortune (heill) from the old land to the new. One of the purposes was to
make the landvættir, the ruling spirits of the land, well disposed to the human intruders
into their domain. Margaret Clunies Ross (1998, 150) opposed Strömbäck and stated that
“the rituals described in the sagas and in Landnámabók as having been carried out by the
first settlers in order to secure their land claims were directed, not at the landvættir as
the first owners of the land, but at other humans who might appropriate what individuals
wanted for themselves.” Emil Birkeli (1932, 24ff.) connected the high-seat and its posts
with the ancestor cult. See also Gunnell 2001.
153 According to Landnámabók (H 21, Ísl. Fornr. 1, 59ff.), Ásólfr alskik came to Iceland to
preach about Christ. When he died the congregation wanted to build a church at his
funeral place. Illugi, who was sent abroad to find the building material, was forced to cast
all church wood overboard with a prayer when returning to Iceland. Three days later the
wood came ashore at Kirkjunes close to Akranes, where they built the church. There is
one similar story related to the Christian man Ørlygr Hrappsson in Landnámabók H 15, Ísl.
Fornr. 1, 53ff.
154 “. . . denna hedniska seds livskraft och stora betydelse . . .”. Strömbäck 1928b.
accounts in Landnámabók, namely Ægidius vita and Declans vita.155 The cult
of Ægidius was popular in Norway during the Middle Ages. He was included
and celebrated in the liturgy of Niðaróss. “The Legend of Ægidius” was also
translated into Norse.156 It tells that Ægidius travelled to Rome and saw a cou-
ple of doors which were carved and decorated with pictures of the apostles.
He liked them very much and wished that he could take them home to his
monastery at Nîmes in Provence. The pope therefore gave them to him. When
taking these doors home to Provence, Ægidius cast them into the River
Tiber. People thought that he had gone mad. But when he arrived at his
monastery, people told him that two fine decorated doors had been found
at the banks. According to Wellendorf, this narrative has much in common
with the account of Þórólfr Mostrarskegg’s high-seat posts. In both stories the
numinous objects find the predestined place in a miraculous way. And in both
accounts these objects are decorated with “sacred beings”, the apostles and
Þórr respectively. In my opinion, there are also some important differences.
In the accounts of Landnámabók and Eyrbyggja saga, the ritual with the posts
was performed as a divination in order to find a place to settle in a new land.
In these narratives the high-seat posts also play a part in a consecration ritual,
where the intention was that the new and unoccupied land should be trans-
formed from chaos to cosmos.157 At the place where the high-seat posts came
ashore the landnámsmenn went with fire around the area which was supposed
to be occupied and settled.158 The purpose of these permormative rituals
was to symbolically transform the land into cosmos, and to create a sacred
area suitable for a settlement and particularly a sanctuary.159 In “The Legend of
Ægidius” the doors arrived at a place which was already cosmicized (speaking
in Eliade’s terms), settled and even regarded as sacred. Thus the very core of
this legend was quite different from the Icelandic traditions.
The “Legend of Declan” admittedly has some similarities to the accounts of the
high-seat posts. This text explains why the Irish monk and saint, Declan,
settled at Ardmore, near Waterford. Declan had a little magic bell that he had
received from God. Once he was out at sea this bell came sailing on a stone.
He followed it and decided that where the bell reached land he would settle
155 Bibliotheca hagiographica latina. Subsidia hagiographica, 93 and 2116. Quoted from
Wellendorf 2009.
156 The Norse text is quoted in Wellendorf 2009, 1021.
157 See Eliade 1987 (1949), 29ff.; 1991 (1949), 10ff.
158 See Strömbäck 1928b, 203ff.
159 It was also a way to secure the territory by a symbolic means of showing one’s dominance
over other men. Clunies Ross 1998, 150.
down and found a bishopric. It came ashore on a small island off the Irish
coast, where he also established his episcopal see. This story thus has a great
deal in common with the narratives about the high-seat posts. One problem
is, however, that Declan was not celebrated in the Liturgy of Niðaróss. Nor was
this story translated into Norse.
It should be mentioned that Wellendorf never intended to discuss the his-
toricity of the ritual of casting high-seat posts; he was mostly interested in
the literary motif. In any case, his result has consequences for historians of
religions, since it undermines the historical source value of Landnámabók. In
my opinion, there are nevertheless reasons to believe that the high-seat posts
played a significant role for the pre-Christian religion, but perhaps not in the
way that is sometimes described in the Old Norse traditions. The connection
between these pillars and gold foils attested by archaeology supports such an
assumption (see further below).
160 . . . á hlutnum markaðr Freyr af silfri . . . Vatnsdæla saga 9–10, Ísl. Fornr. 8, 26, 29f.
161 Vatnsdæla saga 15, Ísl. Fornr. 8, 42.
162 Þat var Freyr ok gǫrr af silfri. Landnámabók, S 179, Ísl. Fornr. 1, 218.
objects such as seventeen stone gaming pieces, a human tooth, and a cruci-
ble fragment were deposited in holes close to the central hearth. They may
have been placed there since these holes were used for the high-seat posts.
Also in the halls of Borg, Lofoten, gold foils (perhaps representing Freyr and
Gerðr) have been found in holes of the assumed high-seat posts, and at Lunda,
Södermanland, deposited objects were found in a similar context (see ch. 4
above). It thus seems as if the transmitter of the Ingimundr tradition must
have known that small sacred objects were placed in the soil around the high-
seat posts of the hof sanctuaries and other ceremonial buildings during pre-
Christian periods.163 The story about Ingimundr, recounted in Landnámabók
and Vatnsdæla saga, in combination with the archaeological evidence, indi-
cates that the high-seat posts really had a religious significance during the
Viking Age, perhaps when the ceremonial building and the high-seat were
consecrated (see ch. 8 below).
There are also other aspects related to the high-seat posts which indicate
that they played an important role in the pre-Christian religion and ruler-
ship. In Eyrbyggja saga 4 and Landnámabók (S85, H73) we read that images of
Þórr were carved on high-seat posts which Þórólfr brought from Norway (see
above). When he erected them in his hof at Hofstaðir, in Iceland, he fixed nails
into them which were called “gods-nails” (reginnaglar), according to Eyrbyggja
saga. The information about the reginnaglar has been regarded as genuine in
this text. The first element of this compound, regin ‘(pagan) gods’, indicates
that the word is old.164 The term reginnagli appears also in a Christian poem,
although in a rather obscure context.165 Two kennings for nails found in the
Þulur (AM 758 I, 4to) confirm that the compound indeed had a pagan religious
significance; regingaddi ‘gods-sting’ and veraldarnagli ‘world-nail’. The latter
term may indicate that such nails were hammered into the high-seat posts,
since these posts were conceived as axis mundi symbols. At this place men
could encounter the mythical world.
It has been argued that the gold foils found by archaeologists at post-holes
in halls were originally attached to the roof-supporting posts. Lars Larsson, for
163 Lidén 1969, 19. Most likely the hlutr mentioned in the saga(s) may be considered as some
kind of amulet, perhaps a gold foil figure. However, no gold foil figures have been found in
Iceland, see ch. 11 below.
164 Olrik 1910; Simek 2006, 262f.; Bertell 2003, 127.
165 Þás þú rekr/ fyr reginnagla/ bóka-máls/ bœnir þínar. “Hvor du fører dine bönner frem for
præsterne” Glælognskviða 10, Skj, B1, 301. Finnur Jónsson’s Lexicon Poeticum (1931, 460)
has ‘bókamál’, ‘bogsprogets gudanagel’, betegner vist ‘bogsprogets’, d.v.s. den gudelig læres
‘hovedstøtte’, dermed menes ‘helgener’.
instance, suggests that many of the foils found at the cultic house of Uppåkra
“had been fastened on the posts as decorations”.166 According to Birgit
Arrhenius, a similar situation may be seen at the hall IA at Helgö. In post-hole
49, situated in room 3 (i.e. the “hall room”), she argues that a “phallus post” was
placed. Arrhenius proposes that figural gold foils could have been used to deco-
rate this wooden pillar, rather as the foils were used at Uppåkra.167 Perhaps
the term “gods-nails” (reginnaglar) attested in the sagas should be related to the
gold foils in such contexts. It could might as well refer to other ritual objects.
The information in Eyrbyggja saga and Landnámabók that images of Þórr
were carved on high-seat posts actually has very few parallels in the sagas.168
One example may be found in Fóstbrœðra saga 23, where Gríma is said to have
owned a large chair, on the back of which a large image of Þórr was carved
(var skorinn Þórr).169 Whether this image was carved on high-seat posts is quite
uncertain, however.
The decoration of some early medieval episcopal thrones found in Sweden
suggests, however, that high-seat posts were decorated with mythical images.
Representations of powerful animals are, for instance, cut on the elbow-
rests of the episcopal throne (late twelfth century) that once stood in Suntak
church, Västergötland. A similar decorated episcopal throne (thirteenth cen-
tury) appears in the stave church of Heddal, Øst-Telemark. The wood carvings
from the thirteenth century stave church of Hegge in Norway may also give
us an impression of what the high-seat posts could have looked like.170 The
Grimstad figure from the Roman Iron Age may likewise convey an idea of
the appearance of the earliest high-seat posts.171
Perhaps archaeological evidence from the Mälaren region also indicates
that high-seat posts could be decorated. In the hall (house 52) discovered at the
ruler site of Lunda, Södermanland, three small cultic images have been found
(see ch. 7). Gunnar Andersson and Sara Fritsch suggest that these figures were
only miniature images of big cultic sculptures, shaped like posts and connected
to the high-seat of the hall.172 They argue that figures A and B give a “post-like”
impression. Inside the big hall at Lunda four extraordinary post-holes were
discovered. They have a different character from the other post-holes where
posts carrying the roof were placed, not so much in terms of size—all the roof-
carrying post-holes were big and deep—but because different types of objects
were ritually deposited in these holes, such as loom weights and objects made
of amber. No such finds were discovered in the other post-holes. According to
Andersson and Fritsch, the decorated cultic posts of the high-seat were placed
in these four particular post-holes.
Freyr, the son of Njǫrðr, had seated himself in Hliðskjálf and looked into
all the worlds. He looked into Jǫtunheimr and saw there a beautiful girl,
as she was walking from her father’s hall to the storehouse. From that he
caught great sickness of heart.176
173 This name is sometimes also written Hliðskiǫlf. See Steinsland 1991, 66ff.; Simek 2006, 152.
174 Gylfaginning 17.
175 Many scholars have felt that Skírnismál is quite a young Eddic lay, see Steinsland 1991, 30ff.
176 Freyr, sonr Niarðar, hafði sezc í Hliðsciálf oc sá um heima alla. Hann sá í Iotunheima, oc
sá þar mey fagra, þá er hon gecc frá scála fǫður sins til scemmo. Þar af fecc hann hugsóttir
miclar. Trans. Larrington.
It has been argued that this prose was an integral part of the poem from the
very beginning.177 This is quite plausible. When reproducing this mythical
account in Gylfaginning (manuscript R), Snorri made Hliðskjálf into a central
element in his narrative, for instance. It indicates that at least he regarded this
motif from Skírnismál as quite genuine, even if he added some details and
moral aspects to the episode:
It happened one day that Freyr had gone into Hliðskjálf and was look-
ing over all worlds, and when he looked to the north he saw on a certain
homestead a large and beautiful building, and to this building went a
woman, and when she lifted her arms and opened the door for herself,
light was shed from her arms over both sky and sea, and all worlds were
made bright by her. And his punishment for his great presumption in
having sat in that holy seat was that he went away full of grief. And when
he got home he said nothing, he neither slept nor drank; no one dared to
speak with him.178
It is interesting to note that Hliðskjálf is called “holy seat” (helga sæti) and
that it was intended for Óðinn only, i.e. the ruler of the gods. According to
Snorri, Freyr was punished for sitting there. This is partly expressed in another
passage of Gylfaginning, where it is also emphasized that Hliðskjálf makes
Óðinn into an all-seeing and all-knowing god. This place is thus in some sense
numinous and endows the one who is seated there with supernatural abilities,
such as the ability to see into other worlds. In chapter 17 Snorri writes thus:
There also is a great place called Valaskjálf. This place is Óðinn’s. The
gods built it and roofed it with pure silver, and it is there in this hall that
Hliðskjálf is, the high-seat of that name. And when All-father (Óðinn) sits
on that seat he can see over all worlds.179
180 Óðinn oc Frigg sáto í Hliðsciálfo oc sá um heima alla. On this tradition, see de Vries
1956–57, §398.
181 When describing Ásgarðr he also adds in Gylfaginning 9: “There is a place called Hliðskjálf,
and when Óðinn sat on the high-seat there, he saw over all worlds and every man’s activ-
ity and understood everything he saw.” Þar er einn staðr er Hliðskiálf heitir, ok þá er Óðinn
settisk þar í hásæti þá sá hann of alla heima ok hvers manns athæfi ok vissi alla hluti þá
er hann sá. Gylfaginning 9. It has been noted that Snorri is slightly inconsistent here,
when he says that it was the place, and not the high-seat, that was named Hliðskjálf. See
Steinsland 1991, 68; Simek 2006, 152.
182 Hallfreðr’s lausavísa is in Skj. B1, 158 and Skj. A1, 168, while Þórólfr’s stanza is in Skj. B1, 388.
183 Steinsland 1991, 69.
184 I am following Steinsland (1991, 69f.) in this overview.
185 According to Eric Elgqvist (1944), Hliðskjálf should be interpreted ‘the watch-tower of the
band (lið)’.
186 Holtsmark 1964, 41.
187 Holtsmark has gained supported from Gerd Wolfgang Weber 1972.
institution with roots in Late Iron Age society. Snorri did not need Christian
iconography to relate Óðinn to a high-seat.188 The high-seat is actually a very
natural position for the ruler of the gods to be seated when appearing in myth-
ical halls, such as Valaskjálf. There is evidence in Viking Age skaldic poetry
relating Óðinn to a specific high-seat. A mid-tenth-century lausavísa, made by
Kormákr Ǫgmundarson, says that a dead man will soon drink mead together
with the gods. This will take place in “the high-seat of Óðinn”:
myndak ǫl at Óðins
i ǫndvegi drekka . . .189
This is repeated with a similar expression some 150 years later in the skaldic
poem Krákumál 29, which says that Óðinn had sent dísir from Valhǫll to invite
the dead man to his home. The dead man says that he will gladly drink beer
in the high-seat together with the Æsir (glaðr skalk ǫl með ǫ́sum í ǫndvegi
drekka).190 According to Frands Herschend we may have archaeological support
for such a notion.191 He points out similarities between the dead ruler’s posi-
tion in grave no. 8 (c. AD 800) in Valsgärde, and the general locality of the high-
seat in the hall building. The dead ruler in the grave is in his symbolic high-seat
together with all his equipment, weapons, the cup and cauldron, the drinking-
horn, gaming table and counters, the fireplace and so on. A ceremonial feast
is being celebrated, in the same manner as he used to do during his life in the
hall. Perhaps this feast is taking place in Valhǫll, together with Óðinn.
The “watch-tower theory” seems unnecessary when relating Hliðskjálf to
a wider literary context about Óðinn and his high-seat. Nor does this theory
suit well when it is related to the role of Óðinn in the ancient Scandinavian
pantheon, i.e. his function as the ruler of the gods. The role as guard in the
watch-tower is better suited to one of the servants in Ásgarðr. The ruler himself
(i.e. Óðinn) is supposed to be seated in his high-seat in the hall.192
We may thus conclude that Hliðskjálf in the mythical traditions should most
likely be related to Óðinn’s high-seat. When entering this seat Óðinn acquired
a numinous ability and a cosmic vision. He was able to see all worlds in the
universe, and thus he was also informed about everything that was going on
in every place. It is possible that some mythical ideas about the high-seat were
projected on the real high-seats of the investigation areas. These ideas may
have enforced the religious and ideological character of these seats, and made
them into a place where the real and divine worlds were joined. When the
rulers entered their high-seats they were enriched with certain supernatural
gifts, which perhaps were necessary when leading the cultic acts in the hall.
The association of Óðinn with the high-seat may also have emphasized these
rulers’ close relation to this sovereign god.
Fires were to be lighted in the middle of the hof floor, and kettles hung
over them. The sacrificial beaker was to be borne around the fire, and he
who made the feast and was chieftain, was to bless the beaker as well as
all the sacrificial meat. Óðinn’s toast was to be drunk first—that was for
193 See Óláfs saga kyrra 2: Þat var siðr forn í Nóregi, at konungs hásæti var á miðjum langpalli.
Var ǫl um eld borit. Heimskringla III, Ísl. Fornr. 28, 204. Cf. Óláfs saga helga 57: Ef gǫfgir
menn kómu til konungs, var þeim vel skipat. Við elda skyldi þá ǫl drekka. Heimskringla II, Ísl.
Fornr. 27, 72.
victory and power to the king—then Njǫrðr’s and Freyr’s, for good har-
vest and peace.
It thus seems as if Snorri regarded the ambulatory ritual around the hearth as
an old pre-Christian custom, which was integrated with the sacrificial meals
and cult performed for the pagan gods. In this account, the sacrificial meat was
cooked on the fire of the hof.
Archaeological evidence from the three investigation areas indicates that
Snorri’s accounts are not totally distorted on this point. It seems as if hearths
constituted an important element in the ceremonial buildings found there.
They often appear in contexts of finds, which suggest that they had symbolic
and religious functions beside the more practical aspects, such as giving
heat and light.194 Most likely they played a central role during the ceremonial
feasting (see ch. 9). Their locations, close to the high-seat, indicate that they also
had important ideological implications. In the ceremonial building discovered
at Helgö, for instance, a central hearth (no. 43) was discovered. It was located
in the hall-room, but not exactly in the middle of it. It seems as if it had a dis-
located position, slightly closer to the middle pair of roof-carrying posts, and
close to the position of the assumed high-seat in the south-western corner of
the room.195 Gold foil figures were found around the fireplace, as well as knives,
loom weights and whetstones.196 As noted above, gold foils may be related to
religious rituals, myths and ruler strategies. Also the fact that the assumed
high-seat was situated close to the fireplace arouses similar associations.
At Vallhagar on Gotland, the hearth was located in between the places of the
two high-seats, in the middle of the hall (house 16). This position indicates
that it played an important ceremonial role for the rulers at this site.197 The
context of the finds also indicates ritual activities. Drinking vessels have
been found close to it. Perhaps some kind of rituals were performed there,
similar to the ceremonies described in the written sources. At several halls in
the Mälaren area, hearths appear close to the places interpreted as high-
seats. At the Garrison in Birka, for instance, the hearth was situated in the
area where most high-status objects were deposited, i.e. the site of the assumed
194 One of Herschend’s (1993, 182) five criteria for halls is that “the hearth was not used for
cooking or for handicraft”. In my opinion, at least ritual cooking may sometimes have
taken place at the hearths inside the ceremonial buildings. But most likely such cooking
often also took place outdoors at cooking pits etc.
195 Herschend 1995, 224.
196 See Herschend 1995, 1998, 28ff.; K. Lamm 2004, 51ff.
197 Herschend 1998, 25f.
high-seat. A similar structure is also evident at the large hall of Lunda. It should
be noted that outdoor cooking pits and hearths appear at several halls in the
Mälaren area, for instance, at Lunda and Helgö. At such pits ritual cooking may
have thus taken place outdoors, and not at the hearth inside the ceremonial
building.
Hearths have also been found inside the assumed ceremonial buildings
found in Trøndelag and northern Norway. Ritual cooking took place, for
instance, at the Viking Age ceremonial building at Mære. Quantities of animal
bones and some characteristic cooking stones indicate this.198 Most interest-
ing in this context is the fireplace structure of the Late Iron Age hall in Borg,
Lofoten. The central fireplace was probably in daily use for cooking and for
heating and light.199 In hall-room C there were also two fireplaces. One of
them was an oblong pit and one was round. Both of them were filled with
fire-cracked stones and pure charcoal, but none of them had ashes or burnt
bones. This indicates a difference in use compared to the central fireplace.
Gerd Stamsø Munch suggests that the reason why no ashes and burnt bones
were found in the long hearth is connected to its use for cooking the slaugh-
tered animal at feasts. “The fire would have been heated intensely before the
meat was put into place, covered with red-hot stones and finally with sods,
the effect being to bake the meat slowly in fire which did not produce ashes
and burnt bones. The long-fire was probably even cleaned afterwards—this
seems particularly important if it was reserved for the preparation of very spe-
cial meals.”200 She also suggests that something similar may have been the case
for the round hearth, and concludes that both of them may have been used for
cultic activities. The round hearth is located very close to the corner, where the
high-seat was assumed to be, and where the gold foil figures and other prestige
objects were detected. This fact suggests that the hearths in room C may have
played a religious-ideological function for the ruler of Borg. Together with the
other ritual objects in room C (the high-seat, the gold foils etc.) it legitimized
the chieftain’s political as well as his religious power. During the religious feasts
it may have played a specific ritual and symbolic role, which emphasized the
central position of the ruler. Certain ceremonies, such as ambulatory rituals,
may have been performed at the round hearth in room C, during the ritual
drinking, as suggested by the written evidence.
Hearths have also been found in the ceremonial buildings of Iceland. Several
fireplaces were found in the Viking Age hall (AB) of Hofstaðir, Mývatnssveit.201
In the northern room there were two great cooking pits which had been
cleaned out repeatedly. Most likely food processing occurred in this room. The
majority of the carbonized cereal grain came from this room, which supports
this interpretation. Outside the north-western entrance fire-cracked rocks/
stones were found, which may have derived from the use of these cooking pits.
There was also a hearth in the central room of the building.202 During phase II
this hearth (2776) consisted of a rectangular stone kerb with a stone base and
had a size of 0.7 × 1.2 m. It was located in the areas where the high-seat may
have been situated and where several objects were deposited. This hearth was
used for heat and light, but may also have had religious and ideological func-
tions during the feasts. No ashes were found in situ during the new excavation,
but Daniel Bruun reports charcoal and bones in and around it, indicating that
it was used for cooking. Whether it was used for ritual or everyday cooking is
hard to tell, however. In any case, it was in this area the ceremonial feasts may
have taken place. The hearths in structures A2 and D (the minor hall) from
phase II are also interesting in this context. When the hall was abandoned,
cattle skulls were placed in two clusters outside the walls of these structures.
In A2 there was a hearth (3039) consisting of a sub-rectangular cut (0.6 × 1 m),
lined with stones and mixed with ashes.203 There was also a rectangular bar-
rel pit (3100), 1.64 m in diameter, and 0.9 m deep with vertical sides and a flat
base. Both suggest that this room was used for food preparation and storage.
But the barrel had been removed at some point and the pit backfilled with
metalworking waste.204 This action may have been symbolic in some sense.
The excavators suggest that the hearth and the pit were used for preparation
of food, perhaps in a ritual context, because the floor contained an extremely
high proportion of burnt bone. The cattle skulls and a slaughtered sheep car-
cass related to this room support this interpretation. There was also a hearth in
the small hall (D). Hardly any bones were found there, however. It thus appears
unlikely that preparation or cooking food occurred at that hearth. In any case,
the cattle skulls suggest that this room may have had a special function.
Hearths have also been found in other halls in Iceland, for instance at
Aðalstræti, Granastaðir, Grelutóttir, Eiriksstaðir, Goðatættur II, Vatnsfjörður
and Havitárholt IX.205 An oven and hearths were also discovered at the more
specific and small cultic house (“blót house”) at Hólmur close to Hornafjörður
In front of Þórr [the image of the god] was a platform, which was very
nicely decorated and covered with iron on the top. On this place there
had to be a fire, which could never be put out. It was called the conse-
crated fire.208
This passage has been debated. Some scholars have argued that the informa-
tion about the fire on top of the platform derives from the imagination of the
Útgarða-Loki.217 Even if this tale is young, the very core of it may consist of an
old riddle.218
These mythic traditions does not justify assuming a cult of fire or the exis-
tence of an Scandinavia fire god in the Late Iron Age. It is not impossible,
however, that the fire/hearth in the ceremonial building was consecrated and
considered sacred during the religious feasts in the three investigation areas.
Perhaps this was evident when the sacrificial meals were being prepared.
According to the Old Norse mythical sources, it seems as if the hearth and
the kettle in Valhǫll played an essential role. Grímnismál and Gylfaginning
report that the slain warriors (einherjar) were served with meat from the boar
Sæhrímnir, which was prepared by the cook Andhrímnir ‘the one exposed
to soot’ in the cauldron Eldhrímnir ‘the one covered in soot from the fire’. In
Grímnismál 18 it says thus:
The descriptions of Valhǫll and other mythical buildings resemble in general the
image we have of the real cultic buildings and feasting halls in the investiga-
tion areas.221 Like the real ceremonial buildings, Valhǫll was decorated and vis-
ible from afar to those approaching it. With no doubt the image of the mythical
hall originally was based on real buildings. Inside the mythical hall there were
seats for the gods, and in addition there was a high-seat which belonged to
Óðinn. When entering this seat Óðinn acquired a numinous ability and a cos-
mic vision. Most likely the real high-seats were also regarded as symbolic cen-
tres and interfaces between the real world and the cosmic world. During the
sacrifices and the religious feasts, the ruler and/or house-owner therefore had
to enter the high-seat, in order to create a relation between this world and
the divine world. In both the human feasting halls and the mythical halls the
hearth seems to have been an essential place, where different types of rituals
were performed in connection with the ceremonial meals.
221 As will be noted below, the martial dimension of the eschatological Valhǫll mythology
was probably more relevant for warrior lords with warrior bands appearing in Svetjud and
Norway. The institution of warrior lords probably did not exist in Iceland. See ch. 15 below.
Near this temple stands a very large tree with widespreading branches,
always green, winter and summer. [. . .] There is also a spring at which the
pagans are accustomed to make their sacrifices.
Eddic lays also describe how the tree extended its branches wide.225 Like the
tree in Uppsala, the cosmic tree in the mythic traditions seems to stand beside
an important building. According Grímnismál 25–26 there was a tree called
Læraðr situated beside Valhǫll (á hǫllo Heriafǫðrs).226 Sometimes the mythic
traditions emphasize that a well was situated beside Hár’s [Óðinn’s] hall, i.e.
Valhǫll. In Hávamál 111 we hear this from Óðinn:
The similarities between the description of Uppsala and the mythical tradi-
tions are striking, and have led to debate in previous research. Some schol-
ars have argued that the “temple”, the tree and the well in Adam’s account
of Uppsala are actually descriptions of a mythical landscape. These scholars
feel that Adam misunderstood the mythical traditions surrounding Valhǫll,
Yggdrasill and the well of Urðr and confused it with reality. Adam’s text there-
fore cannot be accepted as a trustworthy source on Viking Age sanctuaries,
they say.228 In my opinion, these connections may lead to another conclusion.
In what follows, I will argue that the mythical references in Adam’s description
render a cultic reality. The temple, the tree and the well may have been deliber-
ately arranged as a correspondences of notions about the mythical landscape.
This is not unique to Uppsala, as it can be seen at cultic places in other parts
of Scandinavia and on the Continent as well. In my opinion, these cosmic ref-
erences to Uppsala were significant for the rulers. Uppsala was not only the
site of the famous temple, it was also a political and economic centre. The rul-
ers of Uppsala used mythical traditions about the cosmos and the divine world
in order to gain legitimacy and power. Such a mythical or cosmic strategy may
also be seen at other cultic sites in the investigation areas of the present study,
as well as in all Germanic Europe.
229 A document from the Council of Tours in 567 describes how converted people who still
perform pagan rituals should be treated. Some of them worshipped mountains, trees and
wells. Concilium Turonense, MGH, Legum, Sectio III, Concilia, Tomus I, 133. In a letter
from Gregory I to Queen Brunhilde 597 it is stated that the cult of tree does not exist any
longer. Gregorii I. Papae, MGH, Epistolarum, Tomi II, Pars III, 7. See also Clemen, Fontes,
1928, 30. A Langobardic law, compiled during King Liutprand’s reign in 727, stipulates the
fines for those who worshipped trees and wells. Liutprandi Leges, Anni XV, Cap. 84. Leges
Langobardorum 643–866, F. Beyerle (ed.) 1962 (1947), 139. See also Clemen, Fontes, 1928,
38. At Concilium Germanicarum, led by Bonifatius in the year 743, the grounds for the
Carolingian mission strategies were worked out. For instance, regulations against pagan
customs were formulated. Some of these regulations were presented in the text Indiculus
superstitionum et paganiarum. In this text we read in ch. 6 De sacris silvarum, quae nimi-
das vocant “About forest sanctuaries, which they call Nimidas” and ch. 11 De fontibus sacri-
ficiorum “About sacrificial springs”. MGH, Legum Sectio II, Capitularia Regum Francorum,
Tomus I, 222f. See also Clemen, Fontes, 1928, 42f. In Capitulatio de partibus Saxoniae (769)
the cult of wells and trees was forbidden. Sacrifices at groves in a heathen fashion and the
custom of making meals in honour of the demons were also rejected. MGH, Legum, Sectio
II, Capitularia Regum Francorum, Tomus I, 69. See also Boudriot 1928, 34f., 38ff.; Homann
1976, 107ff.; Nilsson 1992, 27ff.
230 In ecclesiastic regulations formulated during the reign of King Edgar 959–975 it is stated
that every priest (preosta) must encourage Christianity and fight against heathen practice
(hæðendōm) and condemn cult at wells, trees and stones (forbeode wil-weorðunga . . . tre-
owum & on stānum). ALIE 1840, 248.
They [the Saxons] worship green trees and wells. They also worship a
large tree trunk, erected under the naked sky; in native language they call
it Irminsul, which in Latin means world-pillar, since it supports every-
thing [in the world].233
Rudolf’s text is not a direct source, since it is built on older accounts and there
seem to be secondary elements in it, such as the interpretation of the name
Irminsul.234 However, his information that Irminsul was a big tree-trunk or pil-
lar seems to be reliable, since the name Irmin-sul (Ermin-sul) means ‘the large
pillar’.235 It is thus possible that the pillar or tree trunk Irminsul was an impor-
tant cultic object at the chief sanctuary of the Saxons. Most likely it referred to
some cosmic symbolism.236
Cultic trees may be seen in Uppsala, not only in Adam’s text.237 According
to the U-version of Hervarar saga, a sacrificial tree, called blóttré, stood at the
assembly place of the Svear, perhaps in Uppsala (see the text in ch. 9 below).
231 About tree cult and cosmic pillars in Eurasia, see Holmberg (Harva) 1922; Drobin &
Keinenen 2001.
232 See in particular Palm 1948. Cf. Springer 2000; Drobin & Keinänen 2001; Sundqvist 2004a;
2007; Tolley 2009.
233 Frondosis arboribus fontibusque venerationem exhibebant. Truncum quoque ligni non
parvae magnitudinis in altum erectum sub divo colebant, patria eum lingua Irminsul
appellantes, quod Latine dicitur universalis columna, quasi sustinens omnia. Translatio
S. Alexandri, MGH, Scriptorum, Tomus II, 676. See also Clemen, Fontes, 1928, 60f. My trans.
234 Palm 1948, 86ff.; see however Springer 2000. Widukind, who was a monk in Corvey,
records in his chronicle Rerum gestarum Saxonicarum (c. 967) a cultic image made as a
pillar which was devoted to the god Irmin: . . . effigie columpnarum . . . Hirmin . . . dicitur,
(ed.) Lohmann & Hirsch 1935, 20f., see also note 3 there. The information that Irmin was
a god or forefather is considered uncertain. See e.g., Palm 1948, 89ff.; Simek 2006, 175f.;
Maier 2000b; Drobin & Keinänen 2001, 140f.
235 Cf. ON jǫrmun-, OE eormen- ‘large’, ‘enormous’, ‘elevated’ ‘wide’; in jǫrmungandr, ‘the big
[enormous] stave’, i.e. the Miðgarðsormr; in OE eormengrund ‘wide world’; and in a heiti
of Óðinn, Jǫrmunr. See e.g., Fritzner 1954, Vol. II, 244; Clark Hall 1916, 93; Palm 1948, 92f.;
Maier 2000b; Drobin & Keinänen 2001, 141. It is unlikely that Irminsul should be inter-
preted as ‘pillar of the god Irmin’. See Simek 2006, 175.
236 See e.g., Simek 2006, 175f.; Maier 2000b.
237 See Palm 1942, 101f.; 1948, 98ff.
When the Svear dethroned King Ingi Steinkelsson (r. 1080–1110), who refused to
perform sacrifices on their behalf, his brother-in-law Sveinn stepped forward
and told the Svear that he would perform sacrifice on their behalf if they would
give him the kingdom. All assented to that proposal. “A horse was led to the
meeting place, dismembered and distributed for eating, and the sacrificial tree
(blóttré) was reddened with the blood.”238 This text is usually regarded as an
uncertain source on pre-Christian religion, but the information about a cultic
tree harmonizes with Adam’s text.
There is also archaeological support for cultic trees in Western Sweden.
During excavations under the church of Frösön in Jämtland, numerous ani-
mal bones were discovered at a mouldered birch stump. Sixty per cent of the
bones belonged to wild animals, mostly bears, and 40 per cent were domestic
animals. The bones and the stump were dated to the tenth century.239 It should
be noted that the church was called Hoffs kirkio in a document from 1408. Still
today, a place south-east of the church is called Hov. It has been argued that
the Hov names in Jämtland refer to buildings where cult practices were car-
ried out.240 Thus, we cannot rule out the possibility that we have a structure
at Frösön similar that in Uppsala; a cultic house situated close to a cultic tree.
Many scholars have associated the ritual trees with the “guardian trees” (in
Swedish called vårdträd and in Norway tuntre) which appear mainly in late
folklore. Such trees were often situated on farms.241 A “guardian tree” could
be of different species, for instance, oak, lime, ash or maple. These trees were
related to the farm and the family living there and they concerned the fate of
the family members. Such trees could protect the people living on the farm,
against fire and other dangers. Rituals were performed at the tree, for instance
during childbirth. These customs were probably old and they remind us about
the great oak, barnstokkr ‘the child trunk’, which was placed in the middle
of the Vǫlsunga hall, according to Vǫlsunga saga.242 In this case too we may
find the combination of a feasting hall and a sacred tree, closely related to each
238 Hervarar saga. (Ed) Jón Helgason, 160. My trans. Cf. Hervarar saga, (ed.) E. O. G. Turville-
Petre, 71.
239 Iregren 1989; Magnell & Iregren 2010. See also Näsström 2001a; 2002, 113.
240 Vikstrand 1993.
241 Mannhardt 1904–1905 (1874–75) I, 51–54; Palm 1948, 56ff.; F. Ström 1961; Tillhagen 1995,
35ff., 61ff.; Steinsland & Meulengracht Sørensen 1998, 30f.; Vikstrand 2001, 279f.
242 Vǫlsunga saga, 6f. According to Hrólfs saga kraka, a tree was situated in the middle of
King Aðils’ hall in Uppsala. See Nordberg 2003, 164, note 1. Cf. Ninck 1935, 340. See also
Gunnell’s (2001) interesting cosmic interpretation.
other. This structure harmonizes well with examples presented above, even if
it appears in a source that is usually considered to be of low value as evidence.
As mentioned above, cultic trees may not always have been real trees, as at
Frösön. According to Anders Andrén, the triangular stone settings in Swedish
called treuddar ‘tricorns’ (or ‘tree settings’) which have been found at several
places in Sweden and Norway should be interpreted as symbolic expressions of
the world-tree Yggdrasill, with its three roots.243 They are mainly dated to the
Late Iron Age and they appear in different find contexts. One Viking Age tricorn
is situated just outside the cultic hall building at Helgö (see fig. 12a and 12b).244
It contained ritual depositions, such as crucibles, bread, arrowheads, pottery
and an Arabic coin dated to 819. Under the stone setting, in the middle, a sin-
gle post-hole was found, dating to the Merovingian Period. It too had remains
of ritual depositions. Perhaps a cosmic pillar of the Irminsul type stood there
before the stone setting was built. Torun Zachrisson has argued that the clay
found around the post-hole represented the mythical-cosmic white mud called
hvíta-aurr, which according to Vǫluspá 19 was poured (ausinn) on Yggdrasill.245
It is thus possible that both Helgö and Frösön have a ritual pattern reminding
us of Uppsala: beside the cultic house there is a sacred or cosmic tree.
Tricorns also appear in Trøndelag. At Skei, Steinkjer, for instance, several tri-
corns are situated close to a Late Iron Age burial field. Iron Age houses have
been discovered there as well as cooking pits.246 An interesting large tricorn
also appears at the assumed cultic site of Lunderhaugen in Brønny municipal-
ity, beside an Iron Age burial mound (35 metres in diameter). In the middle of
this structure four big stones were placed and several deposited objects were
found there. One of the stones was interpreted as a phallus symbol. This place
was used from 500 BC until at least AD 785.247 Andrén’s interpretation of tri-
corns as symbols of the cosmic trees has been applied to stone settings in other
parts of Norway, more precisely in Rogaland. Bjørn Myhre has argued that the
stone settings at the farms of Ullandhaug and Hauge-Tu, for instance, could be
interpreted as cosmic symbols, representing the mythical world-tree.248 The
tricorn at Hauge-Tu was situated close to a place where gold foils and a house
243 Andrén 2004, 406; 2014, 27ff. Tricorns have mainly been found in southern and central
Sweden and in southern Norway, see Myhre 2005.
244 T. Zachrisson 2004a; 2004b.
245 T. Zachrisson 2004a; 2004b.
246 Smedstad 1993.
247 Beverfjord & Binns 1994.
248 Myhre 2005.
Figure 12a One Viking Age tricorn outside the cultic hall building at Helgö (p. 242).
Photo: Anders Andrén.
255 Oral communication with the archaeologist Frölund 2013. See also: http://arkeologiup-
plandsmuseet.wordpress.com/2013/10/11/gamla-uppsala-nytt-ljus-over-gammal-offer-
brunn-2/. The well was first investigated by Martin Olsson (1912) and it appeared to be
lined with wood. The wood was dated in the 1950s by means of dendrochronology; as the
date obtained was 1659, it was rejected as a possible sacrificial well. Hult de Geer 1950;
Alkarp 1993, 43. The recent investigation indicates that it must have been in use as a sac-
rificial well at least in the 1100s (in a Christian context).
256 Damell 1980. Cf. Duczko 1993b, 21.
257 In Lindqvist 1936, 82.
258 Hultgård 1997, 28.
259 The origin of these ordeals is much debated. See Nilsson 1991.
260 See e.g., Hagberg 1967–1977; Ilkjær 1990; 2000; Fabech 1994. Traces of ship parts have also
been found in bogs and lakes in the Mälaren region, for instance at Örsmossen, Tensta
parish, Lake Söderbysjön, Nacka parish, and Rickebasta, Alsike parish. These finds were
probably traces of rituals, perhaps sacrifices. G. Larsson 2007, 238ff.
Smedstorp parish, both in Scania, and at a spring (a tributary of the Ätran) situ-
ated at Hjuleberg farm, Abild parish, Halland,261 but also at many other places
in Continental Europe.262 Cult at springs is also documented from Uppland.
At Gödåker, in Tensta parish, for instance, there is a spring beside a burial field
with finds, dating back to the Roman Iron Age. In this spring several sacrificial
objects were detected.263
Charlotte Fabech has argued that sacrifices in water and wetlands existed in
Scandinavia during the Stone, Bronze and Early Iron Ages, but disappeared at
the end of the Migration Period.264 Torun Zachrisson, on the other hand, has
stated that Fabech’s theory cannot be applied to the Mälaren area. Deposits in
water continued there after the Migration Period.265 Lotte Hedeager has shown
that ritual deposits in wetland also existed in the Viking Age at several places
in Scandinavia.266 Most interesting is the little lake at Tissø, Zealand.267 The
name Tissø has been interpreted as ‘the lake dedicated to Týr’ or perhaps
‘the lake dedicated to the god’.268 In this lake, several Late Iron Age finds have
been made which have been interpreted as sacrificial objects, such as weap-
ons and other artefacts. At the settlement beside the lake, a large ceremonial
hall was raised. Palisaded enclosures were situated close to the south-western
part of the hall. They enclosed a small structure that was interpreted as a cul-
tic house. A similar pattern is also in evidence at the Late Iron Age farm of
Järrestad, Scania, with the hall, an enclosed area with a small cultic building,
and beside this hall a wetland area with sacrificial wells and springs. Ritual
deposits have been found in the wetland area.269
A similar ritual structure may also be found at Late Iron Age central places
in the Mälaren region. At Tuna in Vendel, a feasting hall was situated on a plat-
form, just beside the famous boat graves. North of Tuna there is a farm name
Torsmyra. Place-name specialists have interpreted it as ‘the bog dedicated to
the god Þórr’.270 Perhaps wetland sacrifices were performed there. As far as
spring of the gods’.278 Whether this name, and some of the other names pre-
sented above, really reflects pre-Christian cultic conditions is, however, some-
what uncertain.279 Svavar Sigmundsson states that it would be dubious to
assume “a direct link between heathen worship and the many natural features
all around the country with placenames incorporating goð and blót. Some of
the goð names could equally be derived from the masculine noun goði (chief-
tain), and the blót names are so widespread that the assumption of any ancient
derivation is called into question.”280
Restrictions against cultic actions connected to water appear in some
sources referring to the transition period in Norway. According to “Viðbætir
við Óláfs sǫgu hins helga”, for instance, which is included in Flateyjarbók,
King Óláfr Haraldsson abolished the worship of rocks, woods, water, and trees
when doing mission work in Norway,281 while Gulathings-Lov kristinréttur hinn
nýi reports that it was forbidden to believe in gnomes and perform rituals at
rapids.282 In the Norwegian medieval laws there are otherwise few regulations
against cultic activities performed at sacred fens or springs. The lack of infor-
mation about this may be a consequence of the possibility that this custom
was taken over by the medieval Church.283 It should be noted, however, that
there are scholars who think that the cult at sacred springs was introduced by
the Church in Scandinavia, after the conversion there.284 This type of cult was
related to the worship of local saints in Europe. When the cult of local saints
was taken to Scandinavia, the spring cult was also introduced. There are some
saintly springs in Scandinavia showing some curious features which may indi-
cate that they were objects of cult even before conversion and the introduction
of Christianity in this area. For instance, the custom of making food offerings
in a spring dedicated to St Óláfr may indicate traces of cult continuity.285 Some
278 On the first element Goða- in place-names, see Ólafur Lárusson 1942, 76.
279 See Magnús Már Lárusson 1965, 59; Näsström 2001a; 2002, 120. Cf. Svavar Sigmundsson
(1992) sceptical attitude to the goð-names.
280 Svavar Sigmundsson 1992, 254.
281 . . . aull blot braut hann nidr ok aull god . . . ok mork onnur blotskapar skrimsl bædi hamra ok
hǫrga skoga votn ok tre ok oll onnr blot bædi meiri ok minni. Flateyjarbók III, 246. Cf. Åke
V. Ström 1975, 215.
282 . . . trua a landvættir at se j lundum æða haugum æða forsom . . . NGL 2, 308. See also Nilsson
1992, 35.
283 See Näsström 2001a; 2002, 120ff.
284 See Arwidsson 1965 and Näsström 2001a; 2002, 120ff.; cf. Döring 2003.
285 This appears in information from 1777 at Olofskällan in Sankt Olof’s parish, in Skåne. See
Arwidsson 1965, 55.
of these “Christian springs” appear in the context of folk religiosity, which may
also indicate that these cults built on ancient traditions.286
Perhaps the combination of a hall building and a sacred lake could be seen
in southern Norway. Close to the hall at Huseby, close to Kaupang in the middle
of the Tjølling area, there is a lake called Vittersen. Stefan Brink has argued that
the first element in the Norwegian name contains the word vítr, véttr, vettr, or
vættr, referring to a ‘supernatural being, spirit or god’.287 The name thus refers
to a sacred lake, which was located not far from hall found at Huseby (see
ch. 4 above).
286 Näsström 2001a; 2002, 120ff.; 2003. For Norwegian evidence of sacred wells and springs, see
Bø 1965.
287 Brink 2007, 56.
288 See ch. 4 above.
289 La Farge & Tucker (1992, 80f.) explain ON garðr thus: ‘yard’, ‘enclosure’ and ‘an enclosed
space’. Cf. Heggstad et al. 1993, 139; Vikstrand 2006; Løkka 2010; Cöllen 2011, 221.
290 Gylfaginning 8. Faulkes (1988, 171) interpreted Miðgarðr as ‘the rampart surrounding
the world of men and protecting it from giants’. Miðgarðr also detones the enclosed
area where men lived, i.e. ‘dwelling place in the middle’ (Simek 2006, 214), ‘the home of
humans’ or ‘central enclosure’ (Lindow 2001, 228f.).
291 Gylfaginning 9.
292 Jansson 1987, 138.
It is thus not impossible to imagine that the religious topography and cult
at Uppsala and other cultic sites in the investigation areas, at least for certain
people, (i.e. those who had access to religious and cultural knowledge), were
seen as a reduplication of the cosmic world. Perhaps real places and symbols
in Uppsala represented the world-tree Yggdrasill and the mythical well(s). The
ceremonial building in Uppsala and the activities that took place there might
very well have reflected notions about the life in Valhǫll. Like Ásgarðr, Uppsala
was at least for a period probably enclosed by a symbolic fence. All these ritual
attributes in Uppsala were probably markers indicating an axis mundi, the cen-
tre of the Universe. But perhaps it is an exaggeration to say that the setting
of Uppsala as a whole was a reduplication of Ásgarðr. It was rather a place
with features reflecting a widespread cosmological concept, including several
mythical notions and ideas.
In the previous chapter I argued that several elements related to the ceremo-
nial buildings in the three investigation areas may reflect a mythical symbol-
ism. So too do the cultic images, which, according to the written sources, were
placed inside these buildings. Anthropomorphic figures giving form to and/
or representing gods or mythical beings have been made in almost all cultures
of the world since the Stone Age. They also appear very early in Germanic
contexts.1 In the history of religions, a distinction has sometimes been made
between the German concepts of Idole “idol” (from Lat idolum) and Götterbild
“image of gods”.2 The former term refers to a symbolic representation of an
abstract and multivalent religious content, while the latter term, “image of
gods”, designates a figurative representation of a specific deity, which was
used in cultic contexts.3 In this study I will apply the concept “cultic images”,
which here refers to both anthropomorphic figures representing specific gods
(i.e. “Götterbilder”) and more abstract expressions of divine power. Since this
study also takes the materiality turn into consideration, it must be emphazised
that the cultic images were not always perceived as symbols or representations
of the divinity. For some people these images were plainly the way the gods
were being present, how they were defined in the first place, and how they
could be recognized.4
In the native language the concepts tréguð and skurðgoð were applied to
such objects, but also terms such as trémaðr, líkneski (probably a translation
loan of Latin simulacrum) as well as stafr and hlutr. The authors who described
the ancient Scandinavian religion in Latin used concepts such as idolum ‘idol’
and simulacrum ‘image, idol’. I will present some cultic images which mainly
appear in the context of ceremonial buildings in the three investigation areas.
These images may be seen in different sources, both literary and archaeologi-
cal materials. Some of them were miniatures and made of metal, while others
were made of wood.
According to Adam of Bremen, the Svear honoured the statues of the three
gods in the “Uppsala temple”. Thor (Þórr), occupied a throne in the middle
of the banqueting room while Wodan (Óðinn) and Fricco (Freyr) had places
on either side (see ch. 4 above). It is reasonable to suggest that Adam used
the Latin expressions statuas trium deorum, idolum and simulacrum to refer to
cultic images made of wood.5 In other contexts of his book these terms seem
to refer to wood images. In one passage Adam says that an “idol” (idolum) of
Thor (Þórr) was placed at the thing place of the Svear.6 Wolfred, from England,
smashed it with an axe, but was killed by the pagans for that deed. This cultic
image was most likely made of wood, since Wolfred used an axe when destroy-
ing it. Adam also mentioned that a well-known “idol of Fricco (Freyr)” (Fricconis
simulacrum) was located in Västergötland, not far from the church of Skara.
When Bishop Egino visited this place he smashed it.7 It seems as if this statue
too was made out of wood, since it “was broken into pieces” (in frusta concidit).
Uncertain Old Norse sources refer to pagan cult figures made of wood among
the Svear. According to the Þáttr af Ǫgmundi dytts ok Gunnari helminga a cultic
image (skur(ð)goð, líkneski) of Freyr is said to have been in a sanctuary among
the Svear.8 In Hauks þáttr hábrókar an image of the deity Lytir was brought
to the hall of Uppsala.9 Even if such sources may not be trusted, the Old Norse
concepts skurðgoð and tréguð indicate the existence of images made of wood
among pre-Christian Nordic people.10
Adam’s statement about the statues of the three gods Þórr, Óðinn and Freyr
in the Uppsala sanctuary should not be dismissed as pure fantasy. Place-names
such as Odenslund(a), Torslunda and Fröslunda support a cult of these gods in
pre-Christian Uppland (see ch. 4 above). Archaeological finds may also sup-
port the existence of cult images in pre-Christian Svetjud. A small Viking Age
bronze statue discovered in Rällinge, Lunda parish in Södermanland, has a
phallic shape and might represent Freyr (see fig. 13). The phallic shape resem-
5 These terms are used by Adam for the cultic images at the Uppsala sanctuary, see e.g.,
Adam II,58; IV,26–27. Hultgård suggests that the concepts of idolum ‘idol’ and simula-
crum ‘image, idol’ belonged to Adam’s polemic and anti-pagan terminology. See Hultgård
1997, 22.
6 . . . ydolum gentis nomine Thor stans in concilio paganorum . . . Adam II,62.
7 . . . [Egino] Ibi [sc. In Scaramensi] etiam opinatissimum Fricconis simulacrum in frusta con-
cidit. Adam IV,9. Adam also states that the bishops Adalvard and Egino travelled around
in the communities of the Götar and smashed pagan “idols”. Adam IV,30.
8 Flateyjarbók I, 337–339. Cf. Ǫgmundar þáttr dytt, Ísl. Fornr. 9, 112ff.
9 Flateyjarbók I, 579f.
10 See e.g., de Vries 1956–57, §270. Johan Fritzner (1954), for instance, translated the term
tréguð as ‘Gudebillede af Træ’.
Olof Sundqvist - 978-90-04-30748-3
Downloaded from Brill.com04/16/2019 08:51:54PM
via McGill University
The Relation Between Rulers and Cultic Images 267
Figure 13
A small Viking Age bronze statue discovered in Rällinge, Lunda
parish in Södermanland. Staten Historiska Museum.
bles Adam’s description of Freyr in the “temple”. It has even been proposed
that the small statue from Rällinge was shaped as miniature image of the big
Freyr statue located in the “Uppsala temple”.11 It should be noted, however, that
statues of Freyr were probably placed at other sanctuaries as well, for instance
the well-known “idol of Freyr” at Skara, which perhaps was situated outdoors.
Thus, images of the deity Freyr might have been quite common in eastern
Scandinavia (see also on Lunda below).
The walrus-ivory figure (miniature) from Lund in southern Sweden (height
4.7 cm) (from around the second half of the tenth century to the twelfth cen-
tury) has also been related to Adam’s description of the cultic images in “the
Uppsala temple”.12 According to Ivar Lindquist this figure represents Þórr.13
This figure has, for instance, the same characteristic features as Thor (Þórr) in
Adam’s account, i.e. the deity from Lund is seated on “log-chair” (Sw kubbstol)
similar to the way Þórr is seated on a throne in the “Uppsala temple”. Lindquist
also compared the Lund figurine with a large chair described in Fóstbrœðra
saga. An image of Þórr was carved on this chair.14 This chair was placed á
stufugólf mitt. This statement harmonizes with Adam’s account: “Thor (Þórr),
occupies a throne in the middle of the banquet room” (Thor in medio solium
habeat triclinio). Lindquist also argued that the Lund figurine resembled a
portrait, usually interpreted as Þórr, depicted on a runic store at Norra Åby,
Södermanland (Sö 86). The archaic circular eyes, for instance, are very similar
in both images. Finally he also argued that the ring-and-dot pattern on the
back of the Lund figurine, shaped like a T, represented Þórr’s hammer (com-
pare with the image of Þórr in Fóstbrœðra saga 23). These arguments, if taken
together, may indicate that the Lund figurine represents Þórr.15 The Lund statu-
ette may thus give a suggestion of how an image of Þórr may have been shaped.
Medieval laws also indicate that cultic images existed in Svetjud during the
transition period. In the Law of Uppland (1296) we have the following expres-
sion: “Nobody shall worship pagan idols (?) (affguþum) or believe in groves
and stones.”16 The OSw term affguþ most likely refers to an image represent-
ing a pagan god in this context, since it is related here to other cultic objects,
such as the sacred groves and the stones. We must further assume that these
practices and customs were still occurring when the laws were written down
at the end of thirteenth century.17 Perhaps similar prohibitions on venerating
pagan images are mentioned in the Guta Law: haita a . . . haþin guþ “invoke
heathen gods (i.e. images)”. There is also evidence of outdoor cultic images in
other reliable written sources, which may refer to customs used among eastern
Scandinavian people. When describing the Rūs people (perhaps Svear) on the
Volga Ibn Fadlān mentioned that after anchoring their ships, they made sac-
rifices at a tall piece of wood set up in the ground. “This piece of wood has a
face like the face of a man and is surrounded by small figurines behind which
are long pieces of wood set up in the ground.”18 These cultic images seem to
be made out of tall pieces of wood, perhaps carved slightly to look like a man
with a face. But they could also be regarded as aniconic symbols of deities, i.e.
non-figurative images of gods, such as the Early Iron Age “Pfahlgötter”, which
have been discovered at for instance Oberdorla.19 It has been suggested that
the main image in Ibn Fadlān’s account represents Óðinn, “the leader of the
tribe of deities known as Æsir”.20 It has also been argued that this figure may
represent Freyr, since he was “particularly associated with the Swedes”.21 In
my opinion, it is almost impossible to identify these representations of gods.
Anyhow, the passage produces strong evidence that the Rūs used cultic images
made out of wood during the Viking Age when sacrificing.
Place-names in eastern Scandinavia, including those with the Old Swedish
term staver ‘stave’ (Old Latvian and Lithuanian stãbas, stabs ‘pillar’ or ‘cultic
image’), may refer to a type of outdoor cult image or object. Staver appears in
theophoric place-names, e.g., Nälsta (Nærthastaff, 1354) in Spånga, Uppland,
‘the staff or cultic image connected with the divinity *Niærdh’.22 The row of
erected posts, both to the south and north of Uppsala, could perhaps be related
to the Old Swedish term staver and cultic images, since ritual depositions were
made in connection with some of them.23 These posts may, on the other hand,
have had other functions (see ch. 6 above).
There are in general very few cultic images made of wood preserved from
Late Iron Age Scandinavia. One such image was however found on the periph-
ery of the Mälaren area. Near Sundborn in the province of Dalarna a very sim-
ple figure was dug up in a peat bog. The modelling of this figure is very far from
the description that Adam gives of the images, according to him, were placed
in the Uppsala sanctuary. This figure is 122 cm long and it is a virtually non-
carved product, recently dated to the Middle Ages.24 “A few strokes of an axe
sufficed to indicate a head (with eyes and a mouth), shoulders, breasts (?), and
knees.”25 Perhaps it should be related to the cult of Sami deities, since Sami
lived in this area. There is also a quite exceptional figure from Nordmyra,
Njutånger in Medelpad, also found in a peat bog. This figure, probably dating
from the first half of the first millennium, has a “teardrop-shaped head resting
on a very long neck that ends in a sphere which in turn rests on a hemisphere”.26
This figure may very well have been connected to a structure in a building,
or to a vehicle; the hemisphere at the bottom of the figure indicates this.
Representations of the gods may also be found on two dimensional images.
The divine triad appearing in “Adam’s temple” may, for instance, be visible on
the Skog tapestry, Hälsingland, and the picture stone of Sanda, Gotland (see
fig. 10 above).27
External evidence thus corroborates that Adam’s description of the three
images of gods in the Uppsala sanctuary may be based on sound information.
According to Adam, these statues were erected in the triclinium, ‘the banquet
room’ of the “temple”.28 The terminology used by Adam (libatio and libare)
indicates that drinking rituals were performed during the ritual banquets
there, for instance, when honouring the three gods represented in the images
(see ch. 9 below). It is most likely that the rulers of Uppsala had some specific
relation to these deities during such rituals. According to Ynglingatal, several
of kings of the Ynglinga family were regarded as Freyr’s or simply the deity’s
offspring (Freys ǫ́ttungr, Freys afspringr, týss ǫ́ttungr). Perhaps the images in
the Uppsala sanctuary thus constituted ideological expressions in the public
space, such as the close (genealogical) relationship between the ruler family
and the divine world. It is also possible that the cult of Óðinn, Þórr and Freyr
in this sanctuary was a deliberate strategy used by the rulers in order to gather
different cultic groups and communities in Uppsala during the cultic feast, i.e.
groups who worshipped these particular gods (see further ch. 14).
related to Freyr, although other deities are also possible.31 The most character-
istic feature of them is their phallic nature. Since we have the important infor-
mation from Adam of Bremen that the image of Freyr in “the Uppsala temple”
was phallic (cuius [Fricco’s] etiam simulacrum fingunt cum ingenti priapo) it
seems most plausible to interpret these figures as representations of the deity
Freyr.32 Other sources indicate that the cult of Freyr was spread in this area
31 These phallic images have been interpreted as representations of the god Freyr by Anders
Hultgård, Gunnar Andersson and Sara Fritsch (Hultgård 2003a; G. Andersson et al. 2004,
136ff.). Andersson and Fritsch also present alternative interpretations. Since figures A and
C have visible belts they might represent the god Þórr. The belts on these images may
refer to Þórr’s megingiǫrð ‘belt of power’, mentioned in the Old Norse mythical traditions.
Gylfaginning 20; Skáldskaparmál 18. The weakness of this interpretation is that these
images do not carry Þórr’s most central and important attribute, the hammer Mjǫllnir.
Therefore, it is in my opinion not convincing to interpret these images as Þórr. As an alter-
native to the god Freyr, Andersson and Fritsch suggest that the images represent the god
Ullr. Ullr’s name has often been related to Gothic wulþus ‘glory’ or ‘lustre’. In Wulfila’s Bible
this term is used to translate Greek δόξα ‘glory, splendour, grandeur’. According to this
interpretation Ullr’s name must be interpreted ‘the one who transmits lustre, the splen-
did one’. Since some of the figures at Lunda were gilded it would therefore be natural to
connect these figures to Ullr. In my opinion, the etymology of the name is too uncertain
to make this interpretation convincing. Many alternative interpretations of the name
have also been proposed. For a short overview of the etymology of the name Ullr, see
Vikstrand 2001, 166.
32 That a phallic cult existed in Viking Age Scandinavia could also be attested by means of
the short narrative called Vǫlsa þáttr, Flateyjarbók, where ceremonies related to a horse
penis are mentioned (see ch. 9 below). See also Hultgård 2003a.
during the Late Iron Age. Many place-names in this area, for instance, have
Frö- as the first element (see ch. 14).33
It has been argued that the three figures at Lunda actually were miniature
images of big cultic sculptures, shaped like posts and connected to the high-
seat of the hall.34 A possible parallel to this is the assumed phallic pillar in the
hall on Helgö ‘the holy island’. One post-hole there had a very striking construc-
tion. Birgit Arrhenius describes it thus: “At the edge of the hole to the west, two
large saddle querns were placed beside each other on a filling of gravel and
clay mixed with dark soil. Both stones were turned upside-down, i.e. with the
depressed grinding surface facing downwards and the vaulted underside facing
upwards . . . In this way the post together with the two stones would have had a
shape very close to a Roman phallus, to be seen e.g., on phallus amulets . . . My
proposal here is that there was a large phallus post positioned almost in the
centre of the hall.”35 In addition Arrhenius suggests that some of the gold foils
found in this hall could have decorated this pillar. If Arrhenius’ interpretation
is correct, we may see a cult of Freyr at Helgö, since the motifs of the foils (see
above and below) and the phallic post could both be related to this deity.
The Old Norse texts often report that cultic images were also placed in the hof
sanctuaries of Trøndelag and elsewhere in Norway. These texts also indicate
that such images were made out of wood. In several cases we hear that rul-
ers, such as Earl Hákon, had such images in their sanctuaries. It thus seems as
if they were important for establishing a religious-political position. Many of
these texts are, however, related to the Christianization of Norway, particularly
to the mission actions of King Óláfr Tryggvason and King Óláfr Haraldsson.
These stories are often related to certain topoi which were common in the
hagiography and legendary sagas of the saints, and they must therefore be
treated with great care if used as historical sources. In what follows I will pre
sent some of these traditions, and discuss their value in the context of rulership.
saga Tryggvasonar 59, for instance, he describes how King Óláfr Tryggvason
dismantled Earl Hákon’s hof at Lade, and removed all the property and all
the decorations from it and from the image(s) of the god(s).36 When Snorri
described the interior of the same hof sanctuary in Lade in Hákonar saga góða
14–17, he did not mention any cultic images. But indirectly he intimated that
several gods were worshipped there since he stated: “They toasted Óðinn for
victory and dominion (til sigrs ok ríkis) and Njǫrðr and Freyr for good crops and
peace (til árs ok friðar).” In this case it was Earl Sigurðr who owned and ruled
over the sanctuary at Lade. Some Old Norse texts, which may be older than
Heimskringla, produce a similar image. According to one of the manuscripts
of Óláfs saga Tryggvasonar eptir Odd munk Snorrason (AM 310 4to) (c. 1190), for
instance, Earl Hákon had a hof sanctuary at Lade, which he had dedicated to
several deities.37
There are some younger traditions mentioning that King Óláfr desecrated
the sanctuary including the cultic images in Lade. Óláfs saga Tryggvasonar en
mesta 150 (c. 1300), for instance, states thus:
After these things King Óláfr sailed up the Frith to Lade where he com-
manded the hof to be broken up (lét brjóta ofan hofit). He seized all the
valuables that were in it, together with the vestments of the idols (alt
skraut af goðonum), and he took from the door the large gold ring (tók
gullhríng mikinn or hofshurðinni) that Earl Hákon had caused to be made.
The hof and idols were ordered to be burnt together with the deities
(brenna alt samman hofit ok guðin).38
36 . . . lætr brjóta ofan hofit ok taka allt fé ok allt skraut ór hofinu ok af goðinu/goðunum. Óláfs
saga Tryggvasonar 59, in Heimskringla I, Ísl. Fornr. 26, 308. The reading af goðinu appears
in manuscript Kringla, while the manuscripts of Frisianum and Jǫfraskinna say af goði-
num “from the gods”.
37 Ok þat hit mikla hof er þar stóð hafði hann eignat ótalligum guðum. Óláfs saga Tryggvasonar
eptir Odd munk Snorrason, Ísl. Fornr. 25, 193.
38 Óláfs saga Tryggvasonar en mesta I, Fornmanna sögur 1, 302. My trans.
King Óláfr now entered the hof, accompanied by a few men and some
of the farmers. And when the king came to where the gods were, he
found Þórr sitting there as the most honoured of the gods, adorned with
gold and silver. King Óláfr lifted up the gold-adorned rod he held in his
hands and struck Þórr, so he fell from the podium (stalli). Then the king’s
men ran up and shoved all the gods from their podiums (stallar).39
When it comes to Þórr there are striking parallels between Adam’ description of
the “Uppsala temple” and Snorri’s account of the Mære sanctuary. Thor (Þórr)
is, according to Adam, potentissimus eorum, “the most powerful of them”, while
Snorri depicts this deity as “the most honoured of all the gods” ([Þórr] var mest
tígnaðr af ǫllum goðum).40 It is quite clear that Snorri built this story in Óláfs
saga Tryggvasonar on older traditions. According to Theodoric the Monk’s
Historia de antiquitate regum Norwagiensium (XI) (c. 1180), for instance, Earl
Hákon had gathered images in a house dedicated to the pagan deities at Mære.
These images were burnt together with the house by King Óláfr Tryggvason.
Snorri’s most important source for this passage was probably Oddr Snorrason’s
Saga Óláfs Tryggvasonar (c. 1190) originally written in Latin. The Latin version
is lost, but an Old Norse translation of it is preserved. Oddr depicted the image
of Þórr inside the hof in Mære thus: “Þórr was in the middle of the sanctuary
and he was highly thought of.”41 In this version too, King Óláfr insulted the
pagan farmers by smashing their cultic images.
In one version of this story found in Óláfs saga Tryggvasonar (Flateyjarbók)
(1390) the author describes the image of Þórr in the sanctuary of Mære as sit-
ting in the seat of a carriage: Þor sat j midiu . . . sa var umbunadr Þors, at hann
sat j kerru.42 That the image of Þórr sometimes included a carriage is also men-
tioned in Njáls saga 88. In this text the outlaw Víga-Hrappr entered the sanctu-
ary (hof ) owned by Earl Hákon and the hersir Guðbrandr in Guðbrandsdalr:
“Then he saw Þórr’s carriage and he took from him [Þórr] another ring” (Þá sér
39 Óláfr konungr gengr nú í hofit ok fáir menn með honum ok nǫkkurir af bóndum. En er
konungr kom þar, sem goðin váru, þá sat þar Þórr ok var mest tígnaðr af ǫllum goðum, búinn
með gulli ok silfri. Óláfr konungr hóf upp refði gullbúit, er hann hafði í hendi, ok laust Þór, svá
at hann fell af stallinum. Síðan hljópu at konungsmenn ok skýfðu ofan ǫllum goðum af stǫl-
lunum. Óláfs saga Tryggvasonar 69, in Heimskringla I, Ísl. Fornr. 26, 317. Trans. Hollander.
40 Adam IV, 26; Heimskringla I, Ísl. Fornr. 26, 317. Cf. Dillmann 1997, 67.
41 Ok Þórr var í miðju hofinu ok hafði mest yfirlát. Óláfs saga Tryggvasonar eptir Odd munk
Snorrason, Ísl. Fornr. 25, 280. See also Meulengracht Sørensen 1991b, 239ff.; Steinsland
2000, 114ff.
42 Flateyjarbók I, 320.
hann kerru Þórs ok tekr af honum annan hring).43 This implies an image of Þórr
sitting in a carriage in the sanctuary. This idea is probably based on old tradi-
tions. We may see that one of the mythological tags used by the tenth-century
skald Kormákr Ǫgmundarson in his Sigurðardrápa (c. 960) was “Þórr is sitting
in his carriage” (sitr Þórr í reiðu).44 This poem may be related to Trøndelag
since it was composed in honour of Earl Sigurðr of Lade.
Narratives about cultic images in tenth-century Trøndelag are also repeated
in other late prose traditions. These traditions include some unique narra-
tives which may not be found in e.g., Heimskringla. In some instances it is
mentioned that cultic images of Freyr were placed in the hof sanctuaries in
Trøndelag. According to Óláfs saga Tryggvasonar in Flateyjarbók, for instance,
one image of Freyr (skurgoð þat er gert er eftir Frey . . . likneski Freyss) is said to
have appeared in Þrandheim, i.e. Trøndelag.45 King Óláfr Tryggvason had set
out to desecrate it. When he landed near the hof, he saw a stud of horses and
was told that they belonged to Freyr. The King mounted the stallion as if to
insult the god, while his men took the mares. They rode in triumph to the hof,
where Óláfr derisively seized the image. He took it down from its stand (hio
nidr godin af stǫllunum),46 put it under his arm, closed the sanctuary (hof) and
rode away from the place. A similar tradition is also mentioned in Óláfs saga
Tryggvasonar by Oddr munk Snorrason. In this text too, the king desecrated
the image of Freyr at a hof in Trøndelag: hljóp konungr í hofit ok braut þar í
sundr Frey.47
Several sources report that Earl Hákon of Lade worshipped Þorgerðr
Hǫlgabrúðr. Images of this mythical being were placed in sanctuaries.48 Some
of them seem to have been located in Trøndelag. In Flateyjarbók’s version of
Óláfs saga Tryggvasonar, the statue of Þorgerðr is mentioned.49 It is stated that
Þorgerðr had been Earl Hákon’s best friend and that she was worshipped in
Trøndelag. A sanctuary of Þorgerðr is also mentioned in the Færeyinga þáttr
(preserved in Flateyjarbók) and in Færeyinga saga 23.50 It seems as if it was
located somewhere in Trøndelag. Njáls saga 88 mentions that Earl Hákon and
the chieftain Guðbrandr had a great hof in Guðbrandsdalr, where they kept a
triad of cult images representing the deities (or mythical beings) Þórr, Þorgerðr
and Irpa.
“. . . he has a hammer in his hand and is of great size and hollow inside,
and he stands on a kind of pedestal when he is outside. There is a profu-
sion of gold and silver upon him. He receives four loaves of bread every
day and also fresh meat.” . . . Then they saw a great crowd of farmers come
up to the place of meeting who carried between them a big figure of a
á stólinn fyrir hana; . . . See Færeyinga saga, Ísl. Fornr. 25, 50f. See also a similar version in
Færeyinga þáttr, Flateyjarbók I, 144f. We may see a similar description of Þorgerðr’s sanc-
tuary in Jómsvíkingadrápa 32.
51 Óláfs saga Tryggvasonar en mesta II, Fornmanna sögur 2, 153, 162f.
52 E.g., Óláfs saga helga, Heimskringla II, Ísl. Fornr. 27, 183–190; Flateyjarbók II, 188–192.
53 Óláfs saga helga 112, in Heimskringla II, Ísl. Fornr. 27, 184.
man all glistening with gold and silver. And when the farmers already at
the place of meeting saw it, they all sprang up and bowed down before
this idol. Then it was set down in the middle of the assembly.54
The king was not impressed, however. He had the image struck so it fell in
pieces, and “out jumped mice as big as cats, and adders, and snakes”.55 This
phenomenon is recurrent in the hagiography and the mission literature, the
idea that the devil and his animals have their dwelling inside the images of
pagan gods.56 The story also ends with the conversion of the recalcitrant farm-
ers and St Óláfr’s admonitions that they must not decorate logs and stones
with gold and jewelleries.
54 . . . ok hefir hann hamar í hendi ok mikill vexti ok holr innan ok gǫrr undir honum sem hjallr
sé, ok stendr hann þar á ofan, er hann er úti. Eigi skortir hann gull ok silfr á sér. Fjórir hleifar
brauðs eru honum færðir hvern dag ok þar slátr við. . . . Þá sá þeir mikinn fjǫlða búanda fara
til þings ok báru í milli sín mannlíkan mikit, glæst allt með gulli ok silfri. En er þat sá bændr,
þeir er á þinginu váru, þá hljópu þeir allir upp ok lutu því skrímsli. Síðan var þat sett á miðjan
þingvǫll. Óláfs saga helga 112–113, in Heimskringla II, Ísl. Fornr. 27, 187f. Trans. Hollander.
55 . . . ok hljópu þar út mýss, svá stórar sem kettir væri, ok eðlur ok ormar. Óláfs saga helga, in
Heimskringla II, Ísl. Fornr. 27, 189.
56 See Näsström 2001a; 2002, 200; cf. Lassen 2011, 215ff.
57 Cf. Ǫgmundar þáttr dytt (Ísl. Fornr. 9, 112ff.), where the statue of Freyr is animated.
58 This text has also been preserved independently. See Perkins 2001, 27ff.
59 See Flateyjarbók I, 291ff. Cf. Óláfs saga Tryggvasonar en mesta I, Fornmanna sögur 1, 294f.
On this tradition see Perkins 2001, 27ff.
the island. At the end of this story the pagan god lost a contest with King Óláfr
Tryggvason and Rauðr converted to the new faith.
This story cannot be used as a source for pre-Christian religion and rulers.
However, it seems as if it was partly based on older traditions.60 Rauðr is prob-
ably mentioned in Rauðs þáttr in ramma, as preserved in Snorri’s Óláfs saga
Tryggvasonar 78–80. According to Snorri, Rauðr was much given to making
sacrifices and he was a great magician (blótmaðr mikill ok mjǫk fjǫlkunnigr).61
He also put up stiff resistance to King Óláfr’s mission. Snorri’s version states
that Rauðr lived at Goðey (or Goðeyjar in the plural), which is located at Sálpti
in Hálogaland. This place can be identified with modern Godøy in Saltfjorden,
near Bodø. There are no outstanding archaeological finds on this island, but
the spit of land immediately to the north-east that goes under the name
Goðeynes had a considerable Iron Age settlement. One burial mound there was
20 metres in diameter, while another was 35 m in length and had the shape of
a ship.62 The name Godøy could very well be interpreted as ‘the island of the
gods’.63 If this interpretation is accepted we have a gen. pl. goða- ‘the gods’ as
the first element, where the last vowel has fallen because of its hiatus position.
Per Vikstrand does not rule out the possibility that the first element of Goðey
could be a gen. sg. of an ON goði ‘cultic leader’.64 If so, the name could be inter-
preted as ‘the cultic leader’s island’/‘the religious specialist’s island’. Hence, in
my opinion, the very core of this story may be related to some oral traditions,
mentioning that a chieftain of Goðey or the Goðeyjar resisted Christianity.
Perhaps this chieftain also was regarded as a strong advocate of paganism.
There are some stories about a chieftain called Hróaldr in Oddr’s Saga Óláfs
Tryggvasonar, which probably were Snorri’s sources, and which may indicate
the existence of an old tradition.65 Whether this tradition originally included
ideas about cultic images is uncertain. It seems as if the images in the late tra-
ditions had something of a literary function and illustrate the demonic aspect
of the pre-Christian religion. But even if the narrative in Rǫgnvalds þáttr ok
Rauðs has a literary and Christian tendency, the terminology and expressions
may be pre-Christian, such as magna með miklum blótskap ‘to strengthen (with
supernatural power) by means of much sacrifice and/or pagan worship’.66 It
7.3.4 Staff and Podium: The Evidence from the Norwegian Laws
Even if the Old Norse sources describing cultic images in sanctuaries in gen-
eral are late and partly must be regarded as uncertain, we cannot rule out the
possibility that images of deities really existed in Trøndelag and elsewhere in
Norway during the Viking Period. In the Norwegian medieval laws we have
certain restrictions and prohibitions connected to the use of certain cultic
practices, which sometimes seem to include cultic images. The most extensive
source describing such prohibitions may be found in Den ældre Eidsivathings-
Christenret, which was used in south-eastern Norway. This law rules that
nobody may have in their house such things as belong to paganism:
The staff (ON stafr) and the podium (ON stalli/stallr) may very well refer to the
custom of having a cultic image placed on a sacrificial altar/platform in
the house. This information actually harmonizes very well with Snorri’s descrip-
tion in Óláfs saga Tryggvasonar 59 of the interior of the hof in Mære, which
says that Þórr sat on a stalli. Several other traditions connected to sanctuaries
in Norway also mention the existence of ritual structures called stalli/stallr.
The word stallr or stalli (m.) is translated by Johan Fritzner as ‘Stillads hvor-
paa Afgudbilleder ere oppstillede, Alter’. In the medieval prose it thus refers
to the pre-Christian sacrificial altar, as well as to the podium or platform for
the cultic images. The basic meaning of the word must be ‘a stand upon which
something could be placed’.70 Since we have the compound véstallr ‘sacred
stand’ in the early poetic language we may assume that some type of altars or
ritual platforms really existed in the Viking Age.71 Also the kenning referring to
Óðinn, stalla vinr ‘the friend of the altars/podiums’ in Egill Skalla-Grímsson’s
Berudrápa (c. 970) indicates the same.72 That the term was associated early on
with some kind of altar may be supported by Fjǫlsvinnsmál 40, where we have
the expression “to sacrifice at place of the altar” (blóta á stallhelgom stað). In
this case the word *stallheilagr is an adjective with the meaning ‘altar-holy’ or
‘hallowed by an altar’.73
In the law the staff (ON stafr) seems to be related to the podium. Fritzner
translates stafr (m.) as ‘noget som benyttedes ved afgudernes dyrkelse’.74 Here
we must make a comparison with the information given by Ibn Fadlān about
the tall pieces of wood set up in the ground, which may represent aniconic
images of gods. The stafr ‘staff’ in the house may very well be an aniconic repre-
sentation of the deity. The word staff may also be related to rods in enclosures
around sacred sites and assembly places.75 This idea is based on information
appearing in Den ældre Frostathings-Lov in relation to vébǫnd ‘sacred bonds’.76
This was a kind of band attached to the rods of such enclosure. It is uncertain
whether these rods played any specific role in public cult. Since the word stafr
appears in connection with the word stalli/stallr in the Den ældre Eidsivathings-
Christenret, the interpretation ‘cultic image’ seems more plausible, at least in
this context.
Perhaps something similar to the stafr described in the written sources can
be seen in the archaeological materials. In the assumed small hof found at
Ranheim in Trøndelag, there were four post-holes arranged as a square. They
had smaller diameters than the post-holes used in the outer walls and they had
not been dug to the same depth. According to Preben Rønne, they did not play
any part in supporting the roof. He suggests that these holes originally were
posts or staves which also included images of gods.77
Den ældre Eidsivathings-Christenret also reports on “blót”: matblot. eða lær-
blot gort i mannzliki. af læiri eða af dæigi . . . The terms matblót and leirblót
refer to human-shaped images made of dough and clay. The exact function
of these images is not mentioned in the law. It has been suggested that they
were connected to magical practices.78 In my opinion, they might equally as
well be related to the religious sphere. These figures may very well represent
deities and/or mark the presence of them in the ritual room. The terms matblót
and leirblót indicate that they were used in the worship of gods, perhaps dur-
ing sacrificial rituals, because the verb blóta means ‘to sacrifice, or to worship
through sacrifice’. Perhaps such images were placed on the stalli/stallr.
Most likely other Norwegian laws also indicate the existence of cultic
images.79 Den ældre Gulathings-Lov 29 mentions that “sacrifices are also banned
for we are not permitted to worship any heathen god, or [any] mound, or [in
any] heathen sanctuaries (hǫrgr)”.80 Perhaps the implicit meaning of the first
part of this regulation is that sacrifices to an image representing a pagan deity
were forbidden. A similar formulation is repeated in Kong Sverrers Christenret,
which is directly derived from Den ældre Gulathings-Lov. In this law a prohibi-
tion against sacrifices to vættir ‘female creatures of supernatural kind’ is added.81
It is possible that the vættir were also represented by images, and thus such
figures implicitly are indicated in the law. Also in Den ældre Frostathings-Lov,
sacrifices to pagan gods were prohibited.82 Den ældre Eidsivathings-Christenret
outlawed the keeping of any objects connected to paganism,83 while Den ældre
Borgarthings-Christenret prohibited the population from putting their trust in
any “objects (beings) that they sacrificed to (blótskap)”.84 The term blótskap
has several meanings: ‘avgudsdyrkning, blotning’, ‘noko ein blotar til, avgud’,
‘avgudsoffer’.85 In the context of the law this concept may thus include the
meaning cultic images.
7.4 Cultic Images in the Sagas of Icelanders and the Image from
Eyrarland
Several Sagas of Icelanders report that cultic images were erected in the
hof buildings of chieftains’ farms in Viking Age Iceland. Eyrbyggja saga 4
mentions that such images of gods (goðar) were arranged in a circle right
round the platform (stalli) inside choir-like part (afhús) of the chieftain Þórólfr
Mostrarskegg’s sanctuary (hof ) at Hofstaðir in western Iceland. Þórólfr had
also high-seat posts in his new hof and Þórr was carved on one of them. It thus
seems as if he was surrounded with images of gods when taking part in the
ceremonies of his hof sanctuary. Þórr also seems to have been an important
god for the chieftain at Þórsnes.89 In Kjalnesinga saga 2 we hear that Þórr was
placed in the middle of the hof of the chieftain (goði) Þorgrímr (þar stóð Þórr
í miðju) at Kjalarnes, and other images of gods were situated on either side of
him. In front of Þórr there was a platform (stalli/stallr) with a consecrated fire
(vígðan eldr). These reports of hof interiors from Iceland are probably to some
extent dependent on Snorri’s descriptions of hof buildings in Trøndelag. There
is one difference, however. The images in the Icelandic hof buildings are not
placed upon the platform, but beside or around it.
Other sagas mention that cultic images were placed inside sanctuaries in
Iceland. In Hrafnkell Freysgoði’s sanctuary (hof, goðahús) at Aðalból in east-
ern Iceland, there were several statues of gods (goðar).90 Most likely one of
them represented Freyr. In Harðar saga ok Hólmverja 19 a sanctuary with an
image of Þorgerðr is mentioned. This sanctuary was erected at Bláskógar, and
it was called hof Þorgerðar Hǫlgabruðar. Hence it seems as if one deity played
a central role in the Icelandic ceremonial buildings. This deity may have had a
specific relation to the owner of the building.91
According to the very young Fljótsdæla saga 26, images of several gods and
goddesses were erected in the hof of Bersi inn spaki. They were placed in some
sort of high-seats situated on a platform. In the “lesser high-seat” Freyr and Þórr
were seated,92 while Freyja and Frigg were seated in the high-seat.93 According
to this account, the deities are alive and they are challenged by Helgi, who
was a Christian man. Helgi knocks the deities from their platforms down on
to the floor.94 This late story is rather fantastic and follows a common pattern
in the clerical literature. In my opinion, it cannot be relied upon when discuss-
ing pre-Christian ceremonial buildings in general. But the information that
the deities are sitting in their high-seats may build on old traditions. As noted
above, small Viking Age figures, found at several places in Scandinavia, depict
anthropomorphic miniature figures sitting on thrones.
The information in the Sagas of Icelanders produces some evidence of the
custom of having cultic images in the ceremonial buildings of the chieftains
in Iceland. Usually one deity seems to be the major god/goddess of the sanctuary,
89 According to Landnámabók (S85, H73, Ísl. Fornr. 1, 125), Þórólfr’s hof building was dedi-
cated to Þórr (helgaði Þór). It is also mentioned that he had an image of Þórr on his high-
seat post(s).
90 Hrafnkels saga freysgoða 2, 6, Ísl. Fornr. 11, 99, 124.
91 See also Víga-Glúms saga, Ísl. Fornr. 9.
92 Í ǫndvegi á hinn óæðra bekk sátu þeir í samsæti Freyr ok Þór. Fljótsdæla saga, 11, 295.
93 Þar sátu þær Frigg ok Freyja. Fljótsdæla saga, 11, 295.
94 . . . steypir þeim af stǫllum ok á golf fram. Fljótsdæla saga, 11, 295.
7.4.1 The Cultic Image from Eyrarland and other Images Discovered
by Archaeologists
Archaeological evidence may support the existence of cultic images in pre-
Christian Iceland. The best-known object is probably the eleventh-century
figurine from Eyrarland at Akureyri in northern Iceland, mentioned above
(see fig. 15). This figure is 6.7 cm high and cast in bronze. It depicts a man with
a moustache and beard, sitting on a chair and wearing a conical hat. The figure
is holding an object between his legs, which is often interpreted as a hammer
(or a cross). Many scholars have suggested that this image represents Þórr with
his hammer. This interpretation has, however, been debated.95 Richard Perkins
has made a very interesting interpretation of the Eyrarland statuette, where
he argues that Þórr here is blowing in his beard.96 By means of this magical-
religious action he raises wind. The idea of Þórr as a wind-raiser may be found
in the text Rǫgnvalds þáttr ok Rauðs, which has been preserved independently,
but is also incorporated in Óláfs saga Tryggvasonar en mesta. In this saga Þórr
raises wind by “sounding the voice of his beard” (þeyta skeggroddina/ skeg-
graustina). When King Óláfr is approaching the sanctuary of Rauðr, with his
sailing ships, Þórr performs this ritual act and raises a strong head wind for
the king and his companions. The purpose was to prevent the Christian king’s
mission in the area.
According to Perkins, images of Þórr “sounding the voice of his beard” were
common among Viking Age seafarers. Small Viking Age anthropomorphic
figures holding and/or blowing in their beards may attest this phenomenon,
such as the small image made of walrus ivory found in Lund, an amber statuette
from Feddet, Roholte parish, in Zealand, a bronze figure from Chernaia Mogila,
Chernigov, Ukraine, and a figurine made of whalebone from Baldursheimur,
Suður-Þingeyjarsýsla, Iceland.97
In my opinion, it is possible that the figure from Eyrarland depicts the god
in an important ritual role as he is sitting in his high-seat. Perhaps he is using
his hammer to consecrate the victim, perhaps he is sounding the voice of his
beard. It is possible, however, that this statuette and the other small images
were miniatures of large wooden statues of gods. These statues may have been
95 See e.g., Drescher & Hauck 1982, 263ff.; Perkins 2001, 82ff.
96 Perkins 2001.
97 Cf. Lindquist 1963; Perkins 2001, 53ff. 135f.
placed in the hall buildings and the hof sanctuaries of Viking Age Iceland as
described in the Sagas of Icelanders. Perhaps some of these statuettes were
seated in high-seats, as suggested by Fljótsdæla saga and Adam of Bremen. As
far as I know, however, no real anthropomorphic wooden images have been
found at hall buildings discovered in Iceland. Some abstract symbols of deities
have, on the other hand, been discovered in such contexts. A silver pendant,
with incised decoration in the form of a possible cross and a Þórr’s hammer,
was found at Hofstaðir in Mývatnssveit.98 It was discovered in the hall room
(AB) and must be dated to the first phase I (940–980 AD). There is also an
iron hook, which may be interpreted as a Þórr’s hammer, found in area A/C at
Hofstaðir.99
(blóta Frey). It is stated that Þorgrímr decorated the hall with tapestries for this
occasion. We cannot rule out that these tapestries had mythical motifs.103
Laxdæla saga is the only text to mention carvings on the walls of the feasting
hall. Some doors belonging to the Norwegian stave churches may give an idea
of how such carvings of pagan sanctuaries were shaped, for instance, the one
found at Hylestad church in Setesdal, Norway. It probably refers to a scene from
the traditions of Sigurðr fáfnisbani. Perhaps tapestries with mythic decorations
also covered the walls of such buildings as suggested by Gísla saga. The tapestry
from the Oseberg ship, for instance, may have occurred in such contexts. Some
motifs on it have been related to mythical and ritual themes, such as the hanging
scene.104 The Viking Age or early medieval tapestries from Skog in Hälsingland
and Överhogdal in Härjedalen could also be related to pre-Christian mythic
motifs.105 Austfararvísur mentions that húsbúnaði ‘wall-hangings (?)’ hang in the
hall of St Óláfr.
The existence of cultic images found in the Late Iron Age halls of the three
investigation regions suggests that the rulers wanted to remind the cult com-
munities in a concrete way that the political authority had a specific relation
to the gods and the mythical world. As noted above, the kings and earls in
Svetjud and Trøndelag claimed that they were descended from the gods, while
the chieftains in Iceland regarded themselves as “best friends” with a specific
deity. The former relationship should, in my opinion, be regarded as genealogi-
cal, while the latter relationship rather should be classified as cultic. This close
relationship, whether it was genealogical or cultic, was nevertheless empha-
sized when the authority was seen together with the cultic images on the pub-
lic scene in the sanctuary.
As noted above, some of the images represented the gods as they were
seated in the high-seat. When Adam described the statue of Thor (Þórr), for
instance, this god was placed in a high-seat (solium) in the “Uppsala temple”,
and so too were Þórr, Freyr, Freyja and Frigg in Bersi’s hof sanctuary, accord-
ing to Fljótsdæla saga. Small statuettes and miniatures discovered in Sweden,
103 Þorgrímr ætlaði at hafa haustboð at vetrnóttum ok fagna vetri ok blóta Frey ok býðr þangat
Berki bróður sínum ok Eyjólfi Þórðarsyni ok mǫrgu ǫðru stórmenni. . . . Þá er þeir Þorgrímr
bjuggusk um ok skyldu tjalda húsin . . . Gísla saga Súrssonar 15, Ísl. Fornr. 6, 50.
104 See Steinsland 2005, 297.
105 Anjou 1935 and A.-S. Gräslund 2001, 101ff.
Denmark and Iceland corroborate that the cultic images of the real sanctuaries
really were shaped in that fashion. When the gods were seated in the high-seat
they actually appeared in prototypical ritual roles, which had to be imitated by
the earthly rulers.106 As noted in chapter 6, Óðinn was endowed with specific
abilities when entering his high-seat Hliðskjálf. He could, for instance, see all
worlds in the universe. Some of these abilities were necessary for him when
exercising his power. When taking part in the ceremonial feast in the hall dur-
ing the sacrificial feasts, the earthly political leader had to be seated in the
high-seat in a similar way to the deity, especially when drinking the ceremonial
cups and eating the sacrificial food in the hall or the cultic house (see ch. 4
and 6 above, and ch. 9 below). Perhaps he was also endowed with numinous
abilities and cosmic visions when seated there. Most likely the political leader
also represented the god in this position. Thus when entering the high-seat
he achieved the ultimate prerequisite for making the proper decisions when
exercising power.
The cultic images may also have played another ideological role during the
sacrificial feasts. The ruler was the one who performed some essential parts of
the rituals when sacrificing to the gods on the behalf of the people. Sources
report, for instance, that he was also involved in drinking ceremonies which
were performed in honour of the gods (see ch. 4 above and ch. 9 below). These
deities were represented and marked present by the images at the sanctuaries.
The images constituted communication channels to the Other World when
people needed divine advice or changes of state, for instance, in nature, or
in relations between persons and groups in society. In his role as sacrificer,
the ruler was the one who communicated with the deities (see ch. 9 below),
who were concretely present or represented by the images in the public cultic
places. By means of these figures the executor of the sacrifices could create a
relationship or communication with the divine world. The rulers thus appeared
in a considerable and crucial ritual role in front of the participants during the
sacrificial ceremonies when interacting with these images. This public role of
the ruler was of course crucial when establishing power structures.
It seems as if images of deities and the other mythical objects which sur-
rounded the rulers in the hall buildings or cultic houses in all three areas also
had another ideological function. The sanctuary included (at least for some
people with cultural knowledge) a cosmic symbolism and reflected the mythical
world. This symbolism was even emphasized when the cultic images were also
106 The immaterial prototypical notion that the deity was seated in the high-seat derived, of
course, in the first place from the materiality of religion, i.e. the situation when the ruler
was placed in his seat.
placed there on or around the platform or carved on the high-seat posts. When
entering the ceremonial building, the king, the earl or the chieftain appeared on
this symbolic scene, where his close relationship to the gods was emphasized.107
Very much as Clifford Geertz has described “the theatre state” in nineteenth-
century Bali,108 the political leaders of Svetjud, Trøndelag and Iceland were
situated in a dramaturgical performance on a stage surrounded with ritual
props and objects representing mythical places, objects and beings. To the par-
ticipants in the halls or the cultic houses it must have seemed that the political
leader was in some way part of the divine world, when he appeared on such
a scene with references representing the mythical world. This experience was
very important for the rulers in these areas.
107 On dramaturgical aspects of the ritual space in ceremonial buildings, see Gunnell 1995;
2001.
108 See Geertz 1980; 1993 (1973), 334ff. See also the example of the Roman imperial cult in Asia
Minor, as described by Sr. F. Price 1994. The emperor there was located between human
and divine in a way which helped the Greek cities to accept their subjection to the exter-
nal power of Rome.
In Old Norse texts religious phenomena are sometimes associated with the
adjective heilagr ‘holy’, ‘holy, inviolable’ (cf. helga ‘sanctify, make holy, conse-
crate’, helgi ‘holiness, sanctity’; see also Proto-Germanic *hailagaz).1 A place
which was regarded as heilagr belonged to or was intended for the deities and
therefore had to be treated with reverence. At such places certain ritual restric-
tions and taboos2 were to be observed. Therefore they sometimes were demar-
cated from the profane space and enclosed with symbolic walls or bonds. The
rules of the sanctuaries were also important for the ruler strategies. The sacred
site, which was controlled or even owned by the ruler, could not be violated. It
was the ruler’s assignment to sanctify, protect, restore and perhaps also to ritu-
ally end these places. In skaldic poems rulers are therefore called vǫrðr véstalls
“the protector of the sacred stand” or vés valdr “the ruler of the sacred place”.
If these customs or regulations of the sacred site were disturbed, not only the
sanctity of the place was offended, but also the power, values, and the “world”
of ruler was violated. It was therefore of the utmost importance for a ruler
or a chieftain to protect and perhaps periodically re-establish the sanctity of
the ceremonial buildings and the sanctuaries. This type of performative and
1 O N heilagr is translated ‘heilag, heilag, ukrenkjeleg’ in Heggstad et al. 1993, 176. The basic
meaning of this adjective is actually disputed. Most likely heilagr contains a sacred compo-
nent of meaning, but its semantics do not exactly correspond with the analytic concept of
holy which has been developed in Liberal Protestant discourses by e.g. Nathan Söderblom
(1913; 1914) and Rudolf Otto (1917). On the problems with the meaning of ON heilagr, see e.g.,
Baetke 1942; Vikstrand 2001, 226ff.; Nordberg 2010; 2011. See also the verb helga ‘helge, gjere
heilag, vie, gjere ukrenkjeleg, fredlyse’ and the noun helgi ‘heilagskap, det å vere heilg el.
ukrenkjeleg’, Heggstad et al. 1993, 180.
2 The concept of “taboo” refers to ritual prohibitions or restrictions associated with phenom-
ena considered sacred and/or dangerous. Places, objects, persons, times, actions and words
may be permanently or temporary circumscribed by such rules. The term taboo derives from
Polynesian tapu (Tongan tabu; Hawaiian kapu) ‘marked thoroughly’, ‘that which is forbidden’.
Cf. Marett 1909; Söderblom 1913; 1914; Lehmann 1930; Radcliffe-Brown 1979 (1939); Steiner
1956; Douglas 1966; van Baaren 1972; Widengren 1969; Wagner 1987; Schmidt 1998. On the
concept of “taboo” in Germanic contexts, see Sundqvist 2005d.
ritual strategies may be seen in all three investigation areas, and on different
social levels, i.e. among kings, earls and chieftains.
In Old Norse the act of sanctifying land for oneself is called at helga sér land.3
By this act the claimed land gained sanctity and was granted by divine author-
ity. It seems as if it also was possible to sanctify a place or a building and then
dedicate it to a certain deity (see below).4 The purpose of these performative
rituals was primarily to claim land rights, but also (in a religious sense) to sepa-
rate the sacred area from the profane sphere. The sanctified land or place had
to be freed from evil powers and dedicated to the deity. Some places in the new
land also had to be sites or areas where men could encounter the mythical
world during cult. Therefore certain ritual restrictions and taboos had to be
observed there.5 These performative rituals executed by the chieftain in public
were also intended to show that from now on it was his rules/restrictions the
group or community must follow.
3 Strömbäck 1928b; Baetke 1942, 106ff., 153f.; Clunies Ross 1998, 149ff.
4 Consecration and dedication rituals related to sacred places are known from many cultures.
See Widengren 1969; Eliade 1974 (1949); 1987 (1957); 1991 (1949); Smith 1987. The medieval
Church likewise had rites for dedicating churches. Descriptions of such rituals exist in, for
instance, the Romano-Germanic Pontifical called the “Mainz Pontifical” (PRG 33 and PRG 40),
composed c. 950–962 in Mainz. A ritual close to this Roman-Germanic rite of dedication
of churches was probably also used in Scandinavia during the Early Middle Ages. See e.g.
P. Ström 1997.
5 According to Mircea Eliade (1987 (1957), 32), the consecration of a territory or a place should
be seen as a repetition of the cosmogony; “to organize a space is to repeat the paradigmatic
work of the god.” Possibly the renovations of sanctuaries in Viking Age Scandinavia could be
seen as repetition of cosmogony. See below.
6 See e.g. Strömbäck 1928b; Clunies Ross 1998, 122–157.
7 Dillmann 1997, 71, note 43. Critically considered by Sandnes 1987, 151.
8 In Landnámabók too, the ritual restrictions connected to the mountain Helgafell are men-
tioned. Landnámabók S 85, H 73, M 25, Ísl. Fornr. 1, 124f.
Most of these rituals and restrictions are also mentioned in the narrative about
Þórólfr preserved in Landnámabók. This text explicitly states that Þórólfr built
and dedicated the new hof to Þórr (gerði þar hof mikit ok helgaði Þór).9 By
performing these rituals the local group was informed that they must follow
Þórólfr’s rules.
In Eyrbyggja saga (but not in Landnámabók) it is recorded that Þórólfr
carried fire round the land he claimed. This ritual was probably also per-
formed in order to sanctify his new homeland. It is possible that this land was
related in one or another sense to the hof building. This may be supported by
other Old Norse texts. Landnámabók, for instance, tells of the goði and new set-
tler called Jǫrundr, who erected a hof at Svertingsstaðir, to the west of Fljót in
north-western Iceland. “There was an unclaimed piece of land east of Fljót,
between Krossár and Jǫldusteinn, and Jǫrund carried fire around it and dedi-
cated it to the hof sanctuary.”10 Perhaps other rituals were performed in such
contexts as well, such as swidden agriculture, the erection of wooden poles
to stake one’s claim, shooting an arrow over the land, ritually depositing sym-
bolic and highly personal objects at the boundaries of the land etc.11 Margaret
Clunies Ross argues that some of these rituals were associated with various
gods in the Norse pantheon: “Óðinn is often represented as dedicating enemy
armies to himself by shooting an arrow over them, while Þórr is associated
with the protection of the household and family property through high-seat
posts and other stave-like objects in various Old Norse myths and sagas.”12 It
should be noted that the claimed lands and places were often dedicated to
specific deities. In Landnámabók we read, for instance, that Þórólfr promised
to dedicate his entire land claim to Þórr (at helga Þór allt landnám sitt) and call
it after him.13
The information that the chieftains proclaimed ritual restrictions at sancti-
fied land and places is probably based on old traditions.14 In the Old Norse
sources, ritual restrictions and taboos are often also related to shrines called
vé ‘sacred place, sanctuary’.15 To kill a man in a sacred space (víg í véum) was
strongly forbidden and punished with curses and outlawry. Some of these reg-
ulations refer to Norwegian contexts. Landnámabók mentions that one of King
Haraldr’s escorts, Vali enn sterki, killed on sanctified ground and therefore was
outlawed (hann vá víg í véum ok varð útlægr).16 A person who killed on sanc-
tified ground was sometimes metaphorically described as a “wolf”.17 In Egils
saga 49, Queen Gunnhildr’s brother, Eyvindr, was considered a “wolf” after
killing at a vé sanctuary: “Because Eyvindr had committed murder at a sacred
place he was declared a defiler [actually a wolf (vargr)] and had to go into
outlawry at once.”18 This episode took place at a hof sanctuary at Gaular, north
of Sogn, in Norway. King Eiríkr also participated in the sacrificial feast there.
During the night a quarrel broke out that ended in assault. Eyvindr stabbed
Þorvaldr with his short-sword. The other men leaped to their feet “but none
13 Landnámabók S 85, H 73, Ísl. Fornr. 1, 124. See also Landnámabók M 8, Ísl. Fornr. 1, 344f.:
Ásbjǫrn helgaði landnám sitt Þór ok kallaði Þórsmǫrk.
14 Anglo-Saxon and Frankish sources, as well as texts from the Carolingian Period, indicate
that sacred places and objects in Continental Europe and England were connected to
certain restrictions during the Early Middle Ages. Lex Frisionum (800 AD), for instance,
reports how persons guilty of sacrilege against cultic sites and sacred objects were treated
by the pagan community. If someone destroyed a sanctuary and carried away the sacred
objects, he was taken to the shore of the sea which would be covered by the flood. There
he had his ears cut off and was sacrificed to the gods of the temple he had desecrated
(Additio XI, De honere templorum). In Section 5,1 it is stated that the person who violated
the temple (qui fanum effregerit) could be killed without wergild. See Clemen, Fontes,
1928, 59. Even if these passages appear in a Christian law they have been regarded as
authentic and old. They obviously have a background in pre-Christian traditions. Cf.
Nilsson 1992; Maier 2003, 117. For more examples, see Sundqvist 2005d.
15 Cf. PG *wīhaz; Old Saxon wīh-dag ‘holy day’, Old English wīg-bed ‘altar’ (literary ‘holy
bed’), or German Weihnacht ‘Christmas’ (literally ‘holy night’), and Goth weihen ‘to conse-
crate’ (ON vígja ‘consecrate, hallow, bless’). See Baetke 1942; de Vries 1956–57, §§240–244
(critically); Heggstad et al. 1993, 498; Vikstrand 2001, 298; Sundqvist 2005d, 251f. On the
word vé/vi in place-names, see e.g. T. Andersson 1992a; 1992b; 1992c; Vikstrand 2001.
16 Landnámabók. S 72, H 60, Ísl. Fornr. 1, 102f. See also S 376, H 331, Ísl. Fornr. 1, 374.
17 F. Ström 1942, 73ff., 128ff.; Näsström 2001a, 94; 2002, 138.
18 Eyvindr hafði vegit í véum, ok var hann vargr orðinn, ok varð hann þegar brott at fara. Ísl.
Fornr. 2, 125.
of them was armed inside [the building], because hof sanctity (hofshelgi) pre-
vailed there”.19 It thus seems as if the area inside the hof was also hedged with
taboo rules. It is interesting to note that this holy space in this text is desig-
nated vé ‘sacred place’.
Most likely it was a certain type of killing that was forbidden on the sancti-
fied grounds. The story of Þórhaddr (mentioned above) says that nothing was
allowed to be killed in the area where Mærina-helgi ruled, except the animals
on the farm. Most likely it was also permitted to kill consecrated sacrificial ani-
mals there. Manslaughter, on the other hand, was forbidden. After the fight
between the Þórsnesingar and the Kjallakssynir at the sacred field of Þórsnes,
the sanctified grounds were considered to be defiled by the spilling of blood in
enmity (þá var vǫllrinn óheilagr af heiptarblóði), according to Landnámabók. It
was then decided that the assembly should be moved to the eastern part of the
headland. This became a very sacred place too.20 The first element of the com-
pound heiptarblóð, heipt refers to ‘fiendskap, hat, ovund, misunning’.21 Thus
this pollution of the sacred place consisted of the spilling of blood in enmity, a
phenomenon caused by hatred and violence between men.
19 . . . en menn váru allir vápnlausir inni, því at þar var hofshelgi . . . Egils saga, Ísl. Fornr. 2, 125.
Cf. Vatnsdæla saga 17, Ísl. Fornr. 8, 48: Eigi er þat siðr at bera vápn í hofit, ok muntu verða
fyrir goða reiði, ok er slíkt ófœrt, nema bœtr komi fram; Landnámabók (S 179, Ísl. Fornr. 1,
220): Hrafn . . . hafði sverð gott; þat bar hann í hof; því tók Ingimundr af honum sverðit.
20 Þá var þat ráð tekit at fœra brutt þaðan þingit ok inn í nesit, þar sem nú er; var þar þá helgis-
taðr mikill . . . Landnámabók S 85, H 73, Ísl. Fornr. 1, 126.
21 Heggstad et al. 1993.
22 See particularly Carlie 2004.
they were even placed in the holes of the assumed high-seat posts. This may
be attested, for instance, at Borg in Lofoten. Two of the foils found there had
obviously been placed there intentionally. They lay under a stone, deep down
in the last phase of the post-hole. Gerd Stamsø Munch interpreted this thus: “It
is suggested that the plaques were deposited by the post at dedication of the
room for cult activities at Borg, a ceremony so important that it was necessary
that the plaques were still present when the post was subsequently replaced.”23
Similar finds have been made around the central post-holes at the ceremonial
houses of Mære, and at Hofstaðir (Mývatnssveit) (ch. 6 above and 13 below).
Perhaps these objects should be seen as traits of a consecration of the sanctu-
ary and particularly the high-seat, which was regarded as the ritual centre of
such a building. Similar rituals may have been performed in the Mälaren area.
Gold foils found in the post-holes of the hall of Helgö may attest to such ritu-
als (see ch. 11 below). Weapon deposits at the aristocratic hall of Birka may be
evidence of the same.24 Other ritual objects discovered in the post-holes of the
hall of Lunda may have played a similar ritual role (ch. 6 above).
Gotland was first discovered by a man named Þieluar. At that time the
island was so bewitched that it sank by day and rose up at night. That
man, however, was the first that brought fire to the island, and afterwards
it never sank again.26
It seems here as if the chaotic land was cosmicized (speaking in Eliade’s terms)
by means of a “fire ritual”, which perhaps also is mentioned in Eyrbyggja saga
(see above). This ritual act was performed by Þieluar, the mythical founder of
Gotland.27
Guta saga also indicates that ritual restrictions were observed at sacred sites
in Gotland. The text mentions that a sacrifice was performed at Vi (perhaps a
sacred place situated close to present-day Visby). Later, however, Botair built
a church there. The people of the island wanted to burn it down. Then Botair
went up on the top of the church and said: “If you want to burn it, you will
have to burn me along with this church.” Botair was supported by Likkair from
Stenkyrka. He said: “Do not persist in burning the man or his church, since
it stands at Vi below the cliffs.” As a result of this, the church was allowed to
stand unburnt. The name Vi probably refers to a pre-Christian sacred place
designated with the Old Gutnish word vi (cf. OSw vi, ON vé), i.e. a place where
no violence was allowed.
The idea that sacred sites (ON vé, OSw vi) in eastern Scandinavia were cir-
cumscribed with certain regulations and rules may also be attested in older
sources. According to the Oklunda runic inscription (N288) (ninth century), in
north-eastern Östergötland, protection against violence and asylum could be
found at one such vi site:
Gunnarr cut these runes. And he fled from penalty. Sought this sanctu-
ary (vi) . . .
Vi-Finn carved this (?).28
This inscription has been much debated.29 Scholars have often used the
Oklunda inscription combined with Swedish provincial laws in a r etrospective
27 On the possible hallowing with fire ritual in Guta saga and the problems related to such
interpretation, see Peel 1999, xx.
28 S.B.F. Jansson 1987, 37; von Friesen 1933, 152.
29 Cf. e.g. von Friesen 1933, 152f.; Salberger 1980; Ruthström 1988; Nilsson 1991; Peterson
1993; Lönnqvist & Widmark 1997; T. Zachrisson 1998, 47f.; Vikstrand 2001, 323ff.; Gustavson
2003.
The sources thus report that the sanctuaries in Svetjud, Iceland, Trøndelag and
elsewhere in Norway had ritual spaces which were considered as protected
30 Cf. Frense 1982, 92; Ruthström 1988, 68f., 73f.; Lönnqvist & Widmark 1997, 154f.
31 Cf. Sjöholm 1988, 18–32, 50f.; Nilsson 1989, 24–27; 1991, 487.
32 “Slutligen måste ändå frågan ställas, hur pass förkristen ristningen är till sitt innehåll, och
om inte det förfarande som den omvittnar till dels åtminstone kan ha influerats av konti-
nentalt och/kyrkligt rättstänknde.” Nilsson 1991, 488.
33 Nilsson 1991, 474–484.
34 According to Gutalagen, Fardhem, Tingstäde and Atlingbo were considered as asy-
lum churches. Similar regulations appear in the Östgötalagen (D VII:1, E III:2). See also
Gutalagen, Af mandrapi. (Ed) Pipping, 15–18: þa riþi þengat sum hann bandu will hafa
dragit. Cf. Ruthström 1988; Holmbäck & Wessén 1979 (1933–46), Vol. 4, 212f., 255f.
35 Cf. Franzén 1982, 113; Vikstrand 2001, 325.
36 Since uifin lacked the mark of the nominative it has been argued that it must be consid-
ered as an accusative object. See e.g. Salberger 1980; Ruthström 1988, 74. Peterson (1993,
33–40) argued in favour of the traditional reading and interpretation of the sequence uifin
þitta faþi, i.e. with uifin (RSw Vī-Finn) as a nominative form.
37 Flateyjarbók I, 337. See also Ǫgmundar þáttr dytts. Ísl. Fornr. 9, 111ff.
ground. These areas most likely included the ceremonial buildings, but also
ritual objects, structures and places located outdoors beside these buildings.
It seems as if such areas sometimes were enclosed in certain ways. The term
ON hofsgarðr ‘sanctuary-yard’ may indicate how such ritually restricted yards
were conceived. The room inside the ceremonial building and the sanctuary-
yard were thus regarded as sanctified and protected space and they could not
be violated. At such areas the hofshelgi prevailed. Most likely sacred meadows
could also be enclosed in some way during the religious and judicial gather-
ings. There are some traditions from Trøndelag related to judicial customs,
which also may be valid for attitudes related to sanctuaries.
It is old law that a bailiff from each of the various shires shall set up
enclosures [actually “sacred bonds”] here at the thingstead (scolo gera
vébönd her á þingvelli). And the enclosure shall have sufficient space so
that all who are appointed to the law court shall have ample room to sit
within it.39
The enclosures were intended to demarcate the area the so called ‘law court’
or ‘lay assessors’ were supposed to sit within. In more ancient periods this
space was probably conceived as a vé ‘sacred place’. The first element (vé)
in the term vébǫnd indicates this. In the ritual space inside the enclosures,
the þinghelgi ‘assembly sanctity’ or friðr ‘peace’ (cf. also grið40) thus pre-
vailed. The “sanctity” or “peace” of the assembly place (vé) thus could not
be violated. It seems as if this space was quite extensive, at least during the
Early Middle Ages. Den ældre Frostathings-Lov rules that forty men shall be
named from each “shire” ( fylki) of “Inner Trondheim” and sixty men from
each “shire” of “Outer Trondheim”. All these “lay assessors” were supposed to
38 Heggstad et al. (1993) explains it as “band som gjerder inn ein fredheilag stad, særleg om
inngjerdinga omkring lagretta på tinget”. Most likely sanctuaries had some type of enclo-
sures, equivalent to the phenomenon designated by the ON term vébǫnd.
39 Den ældre Frostathings-Lov, NGL 1, 127. Trans. Larson.
40 Cleasby & Vigfusson (1957 (1874), 215) explains grið as thus: ‘in pl., metaph. a truce, peace,
pardon; friðr is the general word, grið the special, deriving its name from being limited in
time or space (asylum)’. On þinghelgi, see de Vries 1956-57, §§ 148 and 242.
fit in this sanctified space within the vébǫnd. The assembly that gathered on
this ground could thus be interpreted as a social microcosmos, i.e. there were
men representing each “shire” ( fylki) of both Inn-Trøndelag and Ut-Trøndelag.
Together they symbolically represented the community which stood under the
Frostuþingslǫg or Þrǫndalǫg, i.e. ‘the law of Þrœndir’.41
There are other sources describing the nature of “court circles” in Norway
during the transition period. In Egils saga, for instance, there is a detailed
description of one at Gulathing, in Sogn, south-western Norway:
There was a flat plain where the court was established, and hazel staffs
were set in a circle, with a cord around them. This was known as “sacred
bonds” (vébǫnd). Within the circle sat the court, twelve men from the
Firðafylki, twelve from Sygnafylki and twelve from Hǫrðafylki.42
The spatial structure of the court circle (the “vé area”; cf. the term dómhringr)
was situated on smooth or levelled (sléttr) ground. It thus seems as if this
place was prepared by human hand, i.e. people had worked to remove all irreg-
ularities and impediments that may have been naturally present there. On this
ground twelve men representing each “shire” ( fylki) gathered. Together they
symbolically represented the community which was related to the Gulathing.43
In order to separate the sacred space of the court from the profane sphere,
staffs made of hazel were used at the Gulathing. This wood had strongly rit-
ual and apotropaic associations in Norway as well as among other Germanic
people.44 On these staffs the “sacred bonds” (vébǫnd) were connected. We hear
from other Germanic peoples that the place of the judge was marked off in a
similar fashion. The Gothic term staua ‘staff’, for instance, denotes ‘court’ and
‘judge’. The Old High German poem Muspilli (c. 870) describes the court
(mahal-stat) as kimarchôt ‘enclosed’,45 and in the Frankish law text Lex Ribuaria
(c. 630) we read that an oath must be sworn “vor des Königs Gerichtshügel
41 The assembly place of Frostathing was located at Logtu, on the Frosta peninsula in the
central part of the Trondheimsfjord. Røskaft 2003, 110.
42 En þar er dómrinn var settr, var vǫllr sléttr ok settar niðr heslistengr í vǫllinn í hring, en lǫgð
um útan snœri umhverfis; váru þat kǫlluð vébǫnd; en fyrir innan í hringinum sátu dómendr,
tólf ór Firðafylki ok tólf ór Sygnafylki, tólf ór Hǫrðafylki . . . Egils saga, Ísl. Fornr. 2, 154. Trans.
Scudder.
43 Lincoln 1994, 62.
44 Lincoln 1994, 62.
45 De Vries 1956–57, § 265.
im Kreise und im Haselnußgehege, das ist innerhal der Zweige”.46 The sacred
place within the enclosures could perhaps be associated with the term vangr
‘(sacred) plain, field’, which sometimes was related to the hof sanctuaries in
Trøndelag and Norway. Perhaps the place-name Fåvang (Fǫðvangr) in Ringebu,
Gudbrandsdalen, indicates this. It has been suggested that the first element
*fǫð- may be related to Gothic faþa ‘to make boundaries’. Fǫðvangr may thus be
interpreted as ‘the enclosed plain’.47 Other suggestions have been made too.48
46 Si quis pro hereditate vel pro ingenuitate certare coeperit post malo ordine, cum 6 in ecclesia
coniuret, cum 12 ad stappulum regis in circulo et in hasla, hoc est in ramo, cum verborum
contemplatione coniurare studeat. Lex Ribuari 69.5. Text and trans. (ed.) Eckhardt, 186f.
See also Clemen, Fontes, 1928, 31.
47 Olsen 1915, 137ff.
48 See e.g. Sandnes & Stemshaug 2007, 163.
49 The term temenos actually refers to the ancient Greek temple areas. The word τέμενος is
translated in Liddell and Scott (1888) as ‘a piece of land cut off’ . . . ‘a piece of land dedi-
cated to a god, the sacred precincts’.
50 Beronius Jörpeland et al. 2013; Sundqvist 2013, 89f. See also www.ArkeologiGamla
Uppsala.se, last accessed 1 May 2014, and oral information from Jonas Wikborg. On
processions in general, see Hultgård 2001a.
51 See T. Zachrisson 2010, 84.
52 T. Zachrisson 2004b, 151f., 162; 2010, 84.
cultic site of Lilla Ullevi in Uppland. Post-holes, cooking-pits and hearths indi-
cate this.53 At the hall of Husby, Glanshammar, dated to the Migration and
Merovingian Periods, there was similarly a demarcation or palisade.54 A gold
foil figure was discovered on this site. Together with the wall it supports the
idea that this hall and place may have had a religious function. The Iron Age
hall at Hovde, Ørland, in Sør-Trøndelag, was enclosed by ditches and fences.55
During the second phase there was a fence consisting of a row of parallel posts
fencing the house. Several fireplaces and cooking pits were also found associ-
ated with this area. Perhaps the fence here also enclosed a ritual area. In south-
ern Scandinavia we likewise find ritual enclosures around cultic buildings and
sacred sites. At Tissø, Zealand, a large ceremonial hall was erected in the Late
Iron Age.56 Palisaded enclosures were found at the south-western part of the
hall; they enclosed a small structure that was interpreted as a cultic house.
A similar pattern is also in evidence at the Late Iron Age farm of Järrestad,
Scania, with the hall, an enclosed area with a small cultic building, and beside
this hall.57
It thus seems as if the enclosed and the sacred places in the investigation areas
were protected with “peace” ( friðr or grið), which could not be violated. The
“peace” at the Frostathing was extended temporarily to a wider space than only
the assembly place during the thing-days. Violation of this “peace” was strongly
forbidden. Den ældre Frostathings-Lov reports thus:
Now it shall be told what peace those men shall enjoy who are chosen to
attend the Frostathing. Those who are appointed [to attend] and all oth-
ers who go to attend the thing shall be in each other’s peace (scal í griðum
við annan) while on the journey to the Frostathing and until they have
returned to their homes. And if any one shall violate this peace (gengir á
þau grið), he has forfeited his land, his chattels, and his right to peace
among men (oc friði sínum við alla men); and let him never return to the
kingdom.63
Violation of the grið or friðr ‘peace’ (cf. þinghelgi ‘assembly-sanctity’) was thus
punished with outlawry and considered a capital crime in medieval Trøndelag
and elsewhere in Norway. There is a passage in Egils saga 56 where King Eiríkr
and Queen Gunnhildr committed this crime. The saga reports that Egill trav-
elled from his home at Borg in Iceland to the Gulathing in Norway (c. 934), in
order to collect his legal share of the inheritance after Bjǫrn hǫlðr Brynjólfsson,
his wife Ásgerðr’s father. He had received reports that his brother-in-law Berg-
Ǫnundr, with support from King Eiríkr and Queen Gunnhildr, had seized all
the lands that Bjǫrn had owned. Berg-Ǫnundr was married to Bjǫrn’s other
daughter, Gunnhildr Bjarnadóttir. This journey was not without danger, since
Egill had previously been condemned to outlawry by King Eiríkr. During the
judicial process at Gula, Egill and his brother-in-law Berg-Ǫnundr were stand-
ing in the court-circle, inside the sacred bonds (vébǫnd), disputing about the
inheritance. The court also sat in this circle (see above). When things were
turning out in favour of Egill, Queen Gunnhildr commanded her brother
Askmaðr and the royal retainers to attack the court in order to prevent injus-
tice. The men ran to the court and cut the ropes where “the sanctuary” had
been staked out, broke the hazel poles and drove the court away, since no one
was carrying arms there.
The violation of the sacred space seems to be the turning point of this
narrative.64 According to the saga, which was composed during the first half of
the thirteenth century, this evil deed caused a rebellion against King Eiríkr and
Queen Gunnhildr in the shires of Vík and Trøndelag. Soon the king together
with his wife and their children had to flee the land. Even if this episode is liter-
ary and fictitious, the narrative reveals an attitude towards crimes committed
against sacred space, which may have existed in both the Viking Age and the
Early Middle Ages. This attitude also appears in a stanza quoted in the saga,
which is believed by most scholars to have been composed by the historic Egill
himself around 935. Here Egill cries out his disgust against King Eiríkr’s [and
Queen Gunnhildr’s] crime:
In this stanza Egill invokes the gods and asks them to avenge the deeds of the
king. King Eiríkr is here called “he who destroys sanctuaries” (þann er vé gran-
dar). This expression probably refers to the violation of his retainers against
the sacred bonds (vébǫnd) at the Gulathing, but it could also refer to any evil
deed the king had done against the pagan sanctuaries in Norway. The expres-
sion granda vé means plainly ‘to harm or injure a sanctuary’. In any case, the
verse and the prose together give strong indications of how sacred space at
both sanctuaries and thing-places was conceived during the transition period.
The sacred bonds (vébǫnd) thus separated the sacred space from its profane
surroundings at the thing-place in both Frosta and Gula, and probably at
many other “sacred” assembly places in Scandinavia. All those who entered
such areas had to leave their weapons behind, and the people standing inside
the bonds were treated as inviolable. The crime committed by King Eiríkr and
Queen Gunnhildr in violating this space was thus very serious in the eyes of
contemporaries. They acted in a way diametrically opposed to what a just and
decent ruler was expected to do. As noted above, a just ruler in Trøndelag,
according to one Viking Age poem, was called vés valdr ‘the ruler (protector)
of the vé sanctuary’. Similar conditions probably existed in other places too.
According to the Viking Age poem Ynglingatal, for instance, one pagan Svea
65 “Thus may the gods repay him (Eiríkr) for the seizure of my property; may the powers
drive the prince from his lands; may Óðinn and the gods be wrathful. Freyr and Njǫrðr
make the tyrant flee from his estates; may the god of the land hate the oppressor of prin-
cess, who destroys temples”. Text and trans. E. O. G. Turville-Petre 1976.
ruler was called vǫrðr véstalls ‘the custodian of the sacred stand (cultic site)’.
The chieftain or local ruler Sibbi, mentioned in the Rök inscription, was con-
sidered as vīaværi ‘the custodian of the vi sanctuaries’ (see ch. 5). The sacred
space was thus supposed to be protected by the chieftain, earl or the king,
and it could not be violated. The chieftains Þórhaddr of Mære and Þórólfr of
Mostr even proclaimed ritual rules which had to be obeyed when entering the
sacred space in Iceland (see above). They thus cared for and protected their
sanctuaries, assembly places and sacred lands with taboos.66 Landnámabók
records that Þorgrímr Kjallaksson and his brother-in-law threatened to des-
ecrate the holy assembly place at Þórsnes by discharging their bodily needs
there. The Þórsnesingar (Þórólfr’s sons) could not tolerate such desecration
of this holy place. It ended with a great fight where several men were killed.67
The major point in the episode of Egils saga is that King Eiríkr’s and Queen
Gunnhildr’s violation of the consecrated sphere led to a process where they
both de-authorized their own power by breaking the law.68 They acted in the
opposite way to a just ruler, who must protect the court-circle, since it was
regarded as a sacred space. The rule and order that prevailed at such places
must have been in the interests of the ruling power. The sacred space of both
sanctuaries and thing-places constituted the most important arenas where the
rulers could manifest their own power, by showing their authority, wealth and
glory in public and in ceremonial circumstances. The ritual restrictions
and taboos were therefore actually in favour of the ruler at such gatherings.
When he appeared on the public scene he did not need to feel physically
threatened by enemies. No weapons were allowed to be carried at such places.
In Egils saga 25, for instance, we read that “it is a custom here to meet the king
unarmed”.69 In this passage King Haraldr was attending a ceremonial feast
(var á veizlu) at Voss in Hǫrðaland, which took place in a hall (stofa), probably
a ritually protected place.
In the warlike society of Viking Age Norway it must have felt safe for the
leaders to inhabit places where ritual and religious restrictions prevailed, i.e.
66 Eyrbyggja saga 3 reports that Hrólfr (Þórólfr) watched over a Þórr’s hof
(varðveitti . . . Þórshof ) in Mostr, before he emigrated to Norway. Eyrbyggja saga, Ísl.
Fornr. 4, 6.
67 Þat þolðu eigi Þórsnesingar, er þeir vildu saurga svá helgan vǫll. Landnámabók S 85, H 73,
Ísl. Fornr. 1, 126.
68 King Eiríkr is even called a lǫgbrigðir by Egill in one stanza quoted in the saga. Egils saga,
Ísl. Fornr. 2, 165.
69 Þat mun hér vera siðr, at menn gangi vápnlausir fyrir konung, . . . Egils saga, Ísl. Fornr. 2, 63.
the sacred places. If someone attacked the ruler with weapons at such a place,
he or she not only offended this person and his office, but also the religious
and judicial rules which prevailed at such space. These regulations were estab-
lished and sanctioned by old custom (siðr). Indeed, any such violation had to
be considered a capital crime, since it was directed against the society and
perhaps even the cosmos itself. If the ruler was attacked when attending the
sanctified grounds, for instance in the ceremonial building or at the sacred
meadow, the anger of the mythical powers was also aroused. Such a violation
could lead to disturbances in the cosmos, because the sanctuary was regarded
as the mythic landscape on earth. Therefore the taboos and ritual restrictions
at cultic sites in Trøndelag could be seen as a strategy connected to the rul-
ers and political power, not only in this region, but also elsewhere in Norway,
Iceland and Svetjud.
ment saying that a hall should be put in order for the thane of the slain
Danish king Hnæf and his retainers. Clearly one cannot be a leader with-
out a hall, however temporary. . . . sagas like the fight at Finnsburg have
their obvious counterparts in the archaeological remains of the Late Iron
Age. It seems fair to propose the hypothesis that the struggle for power
among the leading families was to a certain degree a matter of fighting
each other with the purpose of destroying each other’s hall. Smashing
rather than plundering was the keynote of this kind of political rather
than economic warfare.70
The destroying of halls has even been described as ritualized warfare. The ene-
mies did not want to steal or plunder, as Herschend noted, even though there
may have been many expensive objects in these houses.71 It is quite possible
that the act of destroying or burning down the hall referred to a performative
ritual action of breaking down the entire world or cosmos of the enemy chief-
tain and challenging his rules and authority. Perhaps the episodes mentioning
the desecration of sacred sites in the Sagas of Icelanders could be interpreted
in a similar way. King Eiríkr’s and Queen Gunnhildr’s attack against the sacred
area of the Gula assembly could be seen as symbolic ritual act against the
power of the local and regional authorities, who were about to give Egill victory
at the trial. A similar purpose may have been behind Þorgrímr’s defiance at the
holy assembly of Þórsnes, with the intention to breake down the authority of
the Þórsnesingar.
In the texts referring to the conversion period it is quite clear that the
destruction of ceremonial buildings in Svetjud, Trøndelag and Iceland was a
symbolic and strategic action intended to exterminate the values and world
views of the enemies. Snorri Sturluson reports in Heimskringla how hof sanctu-
aries in late tenth-century Trøndelag were destroyed by King Óláfr Tryggvason
when he embarked on his mission work. The king’s enemies were the fylki
chieftains who wanted to preserve the old customs. In Óláfs saga Tryggvasonar,
for instance, Snorri describes how King Óláfr Tryggvason dismantled the hof of
Lade and removed all the property and all the decorations from it and from the
cultic image(s) (af goðinu/af goðinum).72 He took a large ring from the door of
the hof, whereupon he had the building burned down.73 This story relates that
not only the hof was desecrated but also other ritual objects related to it, such as
the decorations, the images, and the door ring. It also says that the farmers were
very upset about the king’s treatment of the sanctuaries and sent war arrows
to every fylki. King Óláfr maintained the same strategy when he attended the
sacrificial feast at the hof of Mære, in the inner reaches of Trøndelag, where he
struck the statue of Þórr off the podium and killed the leader of the farmers,
Járn-Skeggi.74 The desecration of the sanctuary and the death of their chieftain
made the farmers weak, and they finally accepted Christianity.
Similar conversion stories appear frequently in the Old Norse texts. When
Christian kings, such as King Óláfr Tryggvason, arrived at pagan cultic sites, the
hof buildings were burned down (brenna hof ), and the hǫrgr sanctuaries were
broken (brjóta hǫrga). According to the version of Óláfs saga Tryggvasonar
in Flateyjarbók, King Óláfr “had the hof buildings burnt down and the hǫrgr
sanctuaries broken and had churches erected instead”.75 He also destroyed the
cultic images.76 By such symbolic and performative actions, the missionar-
ies hoped that the pagans would lose their old beliefs and instead turn to the
Christian God.
Latin sources likewise record that the Christian missionaries wanted to
burn down the pagan sanctuaries.77 Adam of Bremen mentions that Adalvard
73 Síðan lét Óláfr konungr brenna hofit. Heimskringla I, Ísl. Fornr. 26, 308.
74 Óláfs saga Tryggvasonar 69, in Heimskringla I, Ísl. Fornr. 26, 317. Cf. Oddr’s Saga Óláfs
Tryggvasonar, Ísl. Fornr. 25, 279f.
75 . . . þa let hann brenna hof en briota hǫrga ok ræisa j stadinn kirkjur . . . Flateyjarbók I, 285.
76 . . . let ok eyda oll skurgod edr brenna edr briota . . . Flateyjarbók I, 285. This passage is
actually supported with a skaldic stanza dating back to the twelfth century and com-
posed by Hallar-Steinn. In his poem Rekstefja King Óláfr is praised since he condemned
the many hǫrgr sanctuaries and had cultic houses burned ( firna mǫrg ok hǫrga/ brenna
blóthús láta). He also ordered the pagan cultic images to be destroyed (bað hann heiðin
goð meiða). Skj. B1, 527; A1, 545.
77 Several sources indicate that during the conversion period the Church deliberately
offended traditional customs all over Europe. See Maier 2003, 116f.; critically considered
by v. Padberg 1995. Bede (Hist. I, 30–32; II, 15; III, 30), for instance, reports that the pagan
Anglo-Saxons had sanctuaries which were violated by the Christians. According to him
(Hist. II, 13), the former pagan cult leader (primus pontificum) Coifi (seventh century)
desecrated a sanctuary called Godmunddingaham, east of York, and broke several tra-
ditional rules in order to display his new belief: “For it was not before lawful for a priest
of the sacrifices (pontificem sacrorum) either to wear harness or to ride on other than
mare. Girded therefore with a sword about his loins he took a spear in his hand, and
mounting the king’s war horse set forth against the idols. Which sight when men saw
they thought he had been mad. But for all that he stayed not to profane the temple, as
the Younger came to Svetjud, avid to preach the Gospel and in a short time
lead the people in Sigtuna and Uppland to the Christian faith. He and Bishop
Egino agreed that they should go to the pagan temple of Uppsala and offer
Christ some fruit of their labours there, for they would willingly undergo every
kind of torture for the sake of destroying that house which was the seat of
pagan superstition. For if it was torn down, or preferably burned, the conver-
sion of all the people might follow (see ch. 4 above).78
Adam also mentioned the violation of the pagan images elsewhere in
Sweden. Wolfred, from England, for instance, smashed an image with an axe,
but was killed by the pagans for that deed,79 while Bishop Egino smashed the
well-known “idol of Fricco (Freyr)”, which was located in Västergötland, not far
from the church of Skara.80
As mentioned above, it was not only the pagan ceremonial buildings which
were destroyed, but also other cultic structures and objects appearing at the
sanctuaries were torn down. King Óláfr Haraldsson abolished the worship of
rocks, woods, water, and trees when doing mission work in Norway.81 It is inter-
esting to note that missionaries in Continental Europe cut down trees and cos-
mic posts, which were objects of pagan worship.82 These objects may very well
have been expressions or symbols of the pagan world view and cosmos. The act
soon as he approached near unto it, casting thereupon the spear which he held in his
hand; and much rejoicing for the acknowledgement of the worship of the true God he
commanded the company, which was there with him, to set fire to and destroy the tem-
ple together with all its courts.” This text was written more than hundred years after the
event and it seems to be imbued with a Christian tendency. Cf. Kuhn 1978, 234f. In any
case, the restrictions associated with sacred sites may build on authentic information. Cf.
Å. V. Ström 1975, 272; v. Padberg 1995, 167ff.; Simek 2003, 237.
78 Adam IV, 30. These plans were never put into action, however. When the pious King
Stenkil (Steinkell) heard about them, he kept the bishops from any such undertaking, and
declared that they would immediately be punished with death and he would be driven
from the kingdom. Probably everyone in Svetjud would also relapse into paganism.
79 Adam II, 60.
80 Adam IV, 9. (Latin text is quoted in 7.2 above). Adam also states that the bishops Adalvard
and Egino travelled around the communities of the Götar and smashed pagan “idols”.
Adam IV, 30.
81 Viðbætir við Óláfs sǫgu hins helga, Flateyjarbók III, 246.
82 There is evidence indicating that the Church deliberately desecrated holy trees devoted
to the pagan gods. Willibald described in Vita Bonifatii (c. 768) how the holy oak of
Jupiter (. . . appellatur robor Iobis . . .) in Geismar, Hessen, was cut down by the Christians
(Clemen, Fontes, 1928:45). Jupiter was here probably an interpretatio romana for Donar
(Þórr) (e.g. Simek 2006, 63f.; von Padberg 1995, 148; critically considered by Palm 1948,
49–74). The motive was probably to show that the pagan god Jupiter/Donar was without
of cutting down these trees symbolically implied that the pagan “world” was
destroyed. Perhaps also the authority of the ruler, who was supposed to protect
the sanctuary, also was broken in connection to such hostile actions.
The idea that the conservative Earls of Lade wanted to display themselves as
custodians of the pagan cultic sites appears in several sources. Besides the ken-
ning and designation of Earl Sigurðr as vés valdr there are references indicating
that his son Earl Hákon also sought to acquire a similar image. According to
the sources he seems to have been a devoted pagan worshipper, as his father
and grandfather had been. Both Fagrskinna and Heimskringla return to this
theme several times. It is quite clear from Snorri’s Haralds saga Gráfeldar that
the pagan cult in Norway was destroyed by the Christian sons of Eiríkr blóðøx.83
Eiríkr and his wife Gunnhildr converted to Christianity while living in exile in
England. Haraldr Gráfeldr Eiríksson and his brother therefore broke down the
pagan sanctuaries and sacrificial cult when they returned to Norway. Snorri
writes thus: “But when they [the sons of Eiríkr] entered upon the government
in Norway they had no success in converting the inhabitants, and all they
accomplished was to destroy the heathen fanes and to break up the sacrifices
(þá brutu þeir niðr hof ok spilltu blótum), and this brought them much enmity.
There came bad seasons in their time . . .”84 During Hákon Hlaðajarl’s reign,
however, heathendom was restored. He commanded that all people through-
out his dominions were to maintain the hof sanctuaries and the sacrifices in
them (at menn skyldu halda upp hofum ok blótum). Hákon is therefore praised
in the skaldic poem Vellekla 15–16 by Einarr Helgason thus:
power and that nothing happened when his tree was cut down. For the pagan population
this act must have been wretched and execrable.
83
Haraldr Saga Gráfeldar, in Heimskringla I, Ísl. Fornr. 26, 198–224. Cf. Fagrskinna, Ísl. Fornr.
29, 95–103, 112.
84
Haraldr Saga Gráfeldar, in Heimskringla I, Ísl. Fornr. 26, 203.
85 (15) “The wise man [Hákon] at once allowed the men of Þórr to uphold the plundered hof
lands and shrines of the gods well known to the people; and then the god of the fence of
the spears (i.e. the warrior) carried the wolf of slaughter (i.e. the sword) over the moun-
tains and all the sea—the gods guide that man [Hákon].”
(16) “And the sons of the Æsir (i.e. the gods), needful to men, return to the sacrifices; the
mighty tender of the red board of the meeting of Hlǫkk (i.e. Hákon) wins fame by such a
thing. Now the soil flourishes as before—again the destroyer of the wealth of the spear-
bridge [Hákon] allows the merry messengers of the gods to inhabit the sanctuaries.” Text
and trans. E. O. G. Turville-Petre 1976, 60f. modified. This poem was reproduced in sev-
eral manuscripts. Cf. Skj. A1, 122–131. See also Finnur Jónssons’s reconstruction in Skj. B1,
117–124 and SkP I, 280–329.
86 St. 15’s second helming does not occur in Fagrskinna.
87 Hákon var ríkr ok tók at efla blót með meiri freku en fyrr hafði verit. Þá batnaði brátt árferð ok
kom aptr korn ok síld, grøri jǫrðin með blómi, . . . . Fagrskinna, Ísl. Fornr. 29, 111. Snorri makes
similar comments on these stanzas: . . ., þá bauð hann þat um ríki sitt allt, at menn skyldu
halda upp hofum ok blótum, ok var svá gǫrt. . . . Inn fyrsta vetr, er Hákon réð fyrir landi, þá
gekk síld upp um allt land, ok áðr um haustit hafði korn vaxit, hvar sem sáit hafði verit. Óláfs
Saga Tryggvasonar, in Heimskringla I, Ísl. Fornr. 26, 241–243.
88 See F. Ström 1981, 450.
89 Sér hón upp koma ǫðro sinni/ iorð ór ægi iðiagrœna; falla forsar, flýgr ǫrn yfir,/ sá er á fialli
fisca veiðir. . . . Muno osánir acrar vaxa, / bǫls mun allz batna, Baldr mun koma. Vǫluspá
59 and 62. Cf. F. Ström 1981, 451. Helmut de Boor (1930) argued that the terminology of
Vǫluspá resembled the vocabulary of the poems devoted to Earl Hákon. Perhaps Vǫluspá
had some connection to the circle of skalds connected to the tenth-century earls of Lade.
Cf. Schier 1981, 415ff.; Steinsland & Meulengracht Sørensen 1999, 36ff.
re-establishing the pagan cult and restoring the sanctuaries which had been
damaged by the Christians, Earl Hákon transformed the misery in Norway to
a mythical scene identified as the “perfect new world” after Ragnarök, where
“soil flourishes as before” and “the gods inhabit the sanctuaries”. Earl Hákon
is seen in this context as a saviour and a creator of prosperity and wealth.
These mythical allusions are clearly used to create authority and legitimacy. As
long as the ruler maintains the sacrificial cult and protects the cultic sites and
the sanctified grounds, the divine powers and ideal condition of the mythi-
cal world may be manifested among humans. Earl Hákon’s restoration of the
pagan cultic sites may therefore be seen as a repetition of the cosmogony.
Age hall building at Hovde, Ørland, in Sør-Trøndelag, was also rebuilt sev-
eral times.98 There are also several layers of settlements at the cultic house of
Mære.99 The hall of Borg in Lofoten had at least two phases. This is also the
case with the Viking Age hall of the chieftain farm at Hofstaðir, Mývatnssveit,
in north-eastern Iceland.100
Perhaps the periodic renovation of ceremonial buildings and sanctuar-
ies could be seen as a repetition of the cosmogony. As the world and cosmic
process must be renewed by means of calendrical sacrifices, the cultic sites
must also be restored and sanctified periodically. The responsibility for this
renewal and consecration lay with the ruler. When the ruler is called vés valdr
or vǫrðr véstalls this implies that he not only was seen as the one who must
take care of and protect the sanctuary, but that he also periodically had to
reconstruct and consecrate it. By re-establishing and restoring the old sanc-
tuaries the ruler transformed the earthly fane into a mythical stage identified
as the “perfect new world”, where “soil flourishes as before” and “the gods
inhabit the sanctuaries” as described in Vellekla.
to the assumed hof were pulled out and all of the wood removed. The assumed
hǫrgr was carefully covered with stones and clay. The entire cultic site was
then entirely covered with earth. These actions took place during the tran-
sition period in the tenth century.104 As noted above, some rituals were also
performed at Hofstaðir when the great hall (AB) was dismantled in phase III
(1030–1070) (see ch. 4 above). Several acts of closure were performed, such
as the ritual depositions of the skulls. In Uppsala the large hall building was
likewise ritually ended at the beginning of the ninth century. After it burned
down the spiral decorations were intentionally placed in the post-holes (see
ch. 4 above). Most likely the ruler of these places played a central role in these
rituals. Since he was designated vés valdr or vǫrðr véstalls he also took charge of
ending them ritually. With these performative rituals he may have signalled to
the local or regional group as well as the competing magnates that he no longer
had any claims to authority at this place or area.
Sources thus indicate that the rulers consecrated lands and places with a com-
plex of rituals which were performed in public. In connection with that, they
proclaimed the ritual rules of these sacred lands and places. By means of these
performative rituals they demonstrated that it was their rules and restrictions
the group or community must follow. The rulers were also the custodians of the
sacred places. If the ruler (at least the king and the earl) was successful with
these ritual assignments he could create a good relationship with the divine
powers. This relation also had a cosmic dimension, in that the land he ruled
over could be transformed into a “paradise-like world”. The ruler could also rit-
ually end or close a cultic site, by removing certain important objects from the
place, such as the high-seat posts. He could also re-establish and consecrate
sanctuaries which had been desecrated. These ritual roles were performed in
public and rendered the ruler designations such as vǫrðr véstalls. In my opinion
this ritual strategy was applied to chieftains, earls and kings in the investiga-
tion areas. In the next chapter we will turn to other types of rituals performed
by the rulers when appearing in sacred space and ceremonial buildings.
9.1 Hákonar saga góða and the Ceremonial Feasts at Lade and Mære
1 See e.g. Radcliffe-Brown 1952, 153–55; cf. Platvoet 1995; Bell 1997, 128ff.; Rappaport 2000.
2 See Bell 1997, 27–29.
3 See e.g. Herschend 1993; 1997; 1998; Brink 1996a; 1999b; Enright 1996; Lönnroth 1997; Näsström
2001a; 2002; Nordberg 2003; Sundqvist 2002; 2006; 2013 and 2014.
King Hákon the Good and his father King Haraldr. Snorri depicts in general
terms the interior of a hof, the objects and actions there, for instance, the ritual
slaughter of animals and the treatment of the sacrificial blood. He also tells us
that the beaker ( full) was to be carried around the fire, and that the chieftain
who gave the feast was to bless the beaker (signa fullit) as well as all the sacrifi-
cial food (ok allan blótmatinn). When the toasts were drunk to the gods, ritual
formulae were proclaimed, such as til árs ok friðar.
This description of ceremonial feasts in Hákonar saga góða has been much
debated (see ch. 4 above). The sharp criticism of Snorri’s text has partly been
justified, but in some senses it failed. Recently scholars have modified it and
produced, in my opinion, a sounder view of Snorri’s account, where both possi-
bilities and problems of source criticism are noted (see ch 4 above). In my opin-
ion, several elements in Snorri’s text are supported by other reliable sources
and may thus be genuine. In what follows I will discuss some details in this
description and try to estimate what parts of it may build on more ancient tra-
ditions. I will first present those elements in Snorri’s description which I con-
sider inauthentic as regards pre-Christian rituals. Than I will point out aspects
I believe to be authentic. My purpose here is to illuminate these notions with
sources from the investigation areas. These comparisons are intended to show
that these elements existed in all these areas.
At this feast all were to take part in the drinking of ale. Also all kinds of
livestock were killed in connection with it, horses also; and all the blood
from them was called “hlaut”, and “hlautbolli”, the vessel holding that
blood; and “hlautteinar”, the sacrificial twigs.
In his book Das Opferfest von Lade, Klaus Düwel analysed the semantic devel-
opment of the words hlaut, hlautteinar and hlautbolli used in Snorri’s text.
Düwel argued that in most cases Snorri either misunderstood these concepts
or mixed them up with Christian ideas with no basis in pre-Christian culture.4
Düwel is no doubt right in his criticism. It is very doubtful that the term
hlaut designated ‘sacrificial blood’ in pre-Christian language and that hlautte
inn and hlautbolli meant ‘sacrificial twig’ and ‘sacrificial bowl’. The m
eanings
of these terms are not attested in skaldic poetry or the Eddic lays. Most
likely hlaut originally referred to a ‘lot’ used in divination rituals.5 Snorri is
perhaps the first one to use the term hlaut in the sense of ‘sacrificial blood’.
Most likely the scribes of Eyrbyggja saga and Kjalnesinga saga took over this
meaning when applying this term in their descriptions of hof buildings.
When describing the “pagan use” of the sacrificial twigs (hlautteinar) Snorri
writes: “These were fashioned like sprinklers, and with them were to be smeared
all over with blood the ‘stallar ’ [altars] and also the walls of the hof, within
and without, and likewise the men present were to be sprinkled with blood.”
It is quite possible that Snorri was influenced by ecclesiastical ceremonies in
this description.6 In the medieval church holy water was sprinkled over the
congregation. Perhaps examples in the Old Testament may also have played
some role in this context, since it is said there that Moses sprinkled blood on
his people.7 An indication of Christian impact on this text may also be seen in
the comparison that Snorri made between the hlautteinar and the stǫklar ‘holy
water twigs’. The Old Norse term stǫkkull (sg.) designates the Latin aspergillum,
an object that was often used in the holy water rituals of the medieval Church.
Hence the information about the use of the hlautteinar in the contexts of hof
buildings is not completely reliable (see however below on the blood rituals).8
According to Hákonar saga góða 14, the chieftain who held the feast was
to bless the beaker as well as all the sacrificial food (. . . signa fullit ok allan
blótmatinn). Scholars have been suspicious about this ritual. Ernst Walter, for
instance, convincingly argued that Old Norse signa was a loan-word from the
Christian concept, Latin signare (signo) ‘to make a sign (cross-sign)’ (see ch.
4).9 Whether the loan-word signa indicates that the content of the text is late
and not built on pre-Christian notions is uncertain, however, for this term
appears in Eddic and skaldic poetry, as well as in an eleventh-century runic
5 Hultgård (1996, 53) referred this term to the Gothic hlauts (a masculine a-stem) which repro-
duces the Greek concept κλῆρος (klēros) ‘lot, part’. It seems as if it is this meaning which is
evident in the first element of the concept hlautviðr ‘lot-twig’ in Vǫluspá 63, for instance.
6 See Hultgård 1993, 230; 1996, 45. See also Simek (2006, 151) who comments on the term hlaut
teinn thus: “the concept would appear to have its origins more in the imagination of the
Christian authors of the High Middle Ages, who created a heathen counterpart to the sprin-
kler (asperges) of Christian liturgy.”
7 See Exodus 24:4–9. Cf. Düwel 1985, 34ff.
8 A more original use of the term hlautteinn is attested in a lausavísa (c. AD 980) by Þorvaldr
Koðransson, where it has the meaning ‘lot-bough, lot-twig’. Skj. B1, 105. Cf. Skj. A1, 110. The
term hlautteinn should probably be associated with the concepts blótspánn or hlautviðr, i.e.
the ‘twig(s)’ used in divination rituals.
9 Walter 1966.
Calendrical rituals and feasts exist in most cultures around the world.14 They
give “socially meaningful definitions to the passage of time, creating an ever
renewing cycle of days, months, and years . . . Calendrical rituals occur periodi-
cally and predictably, accompanying seasonal changes in light, weather, agri-
cultural work, and other social activities.”15
9.2.1 Trøndelag
The feasts in Trøndelag depicted in Hákonar saga góða 13–18 could be defined
as calendrical. They were performed “at the beginning of winter” (um haus
tit at vetri (vetrnóttum)) and at Yule (ON jól).16 This information seems to be
built on an old and firm oral tradition.17 Such feasts may be attested in writ-
ten sources from all three investigation areas. Equivalents to the concept jól
( júl) (neutr. pl.) (‘Yule-feast’, ‘Gastmahl, Festmahl’; cf. the kenning Hugins jól
‘Fest, Trinkgelage des Raben’) can also be seen in other Germanic areas, such as
Gothic fruma Jiuleis, and Old English gēol (neutr.), gēola (mask.), giūli (mask.),
and iūla (mask.).18 According to Snorri, there seem to have been three calen-
drical religious feasts celebrated in Trøndelag (and Svetjud) during the pre-
Christian era. These cultic feasts coincided temporally with the quarters of
the year.19 The feasts at the winter-nights (vetrnætr) took place in October, the
midwinter sacrifices (miðsvetrarblót) in January, and the feast to greet the sum-
mer ( fagna sumri) took place in April. Perhaps there was also a Midsummer
feast in July (miðsumarsblót).20 In Snorri’s account it seems as if rich farmers
in northern Norway also arranged sacrificial feasts at their farms three times
every year. In Óláfs saga Helga ch. 117 he writes about Sigurðr Þórisson, who
lived at Ǫmd at Þrándarnes (Trones):21
and the third at the beginning of summer. And when he had adopted
Christianity he still persisted in this way of holding feasts. So he had a
great banquet in autumn for friends, then a jól feast in winter to which he
invited a great many people; and at Easter too he had numerous guests.22
Earlier in Óláfs saga Helga 107 Snorri wrote about the calendrical sacrifi-
cial feasts which were still performed on the winter nights (i.e. October) in
Inn-Trøndelag during St Óláfr’s reign. The Christian king was informed that
all toasts were then brought to the Æsir deities, and they followed old hea-
then custom. Cattle and horses had been slaughtered and the pedestals (of
images) reddened with their blood, and sacrifices were performed in order
to improve the harvest.23 Later, during the winter, the king was told that the
people from the inner reaches of the Trondheimsfjord had assembled in great
numbers at Mære and that there were big sacrifices made at midwinter for
peace and a good season (blótuðu þeir þá til friðar ok vetrarfars góðs).24 The king
was upset about this and arranged a meeting with the chieftain of the farmers
called Ǫlvir and accused the Þrœndir of arranging midwinter sacrifices (miðs
vetrarblót). Ǫlvir told the king that the farmers were not guilty of that. They
just celebrated the jól feasts, including social drinking.25 After Easter the king
despatched a man called Þoraldi to a meeting. He was the king’s steward on
his estate at Haugr. He told the king that the people living in the interior parts
of Trøndelag were pure heathen in their belief, even though a few there were
baptized:
22
Hann var því vanr, meðan heiðni var, at hafa þrjú blót hvern vetr, eitt at vetrnóttum, annat
at miðjum vetri, þriðja at sumri. En er hann tók við kristni, þá helt hann þó teknum hætti
um veizlur. Hafði hann þá um haustit vinaboð mikit ok enn jólaboð um vetrinn ok bauð þá
enn til sín mǫrgum mǫnnum, þriðju veizlu hafði hann um páska ok hafði þá ok fjǫlmennt. in
Heimskringla II, Ísl. Fornr. 27, 194. Trans. Hollander.
23
Þat haust váru sǫgð Óláfi konungi þau tíðendi innan ór Þrándheimi, at bœndr hefði þar
haft veizlur fjǫlmennar at vetrnóttum. Váru þar drykkjur miklar. Var konungi svá sagt, at
þar væri minni ǫll signuð Ásum at fornum sið. Þat fylgði ok þeiri sǫgn, at þar væri drepit
naut ok hross ok roðnir stallar af blóði ok framit blót ok veittr sá formáli, at þat skyldi vera til
árbótar. Heimskringla II, Ísl. Fornr. 27, 177.
24
Heimskringla II, Ísl. Fornr. 27, 178.
25
’Hǫfðum vér,’ segir hann, ’jólaboð ok víða í heruðum samdrykkjur. Ætla bœndr eigi svá
hneppt til jólaveizlu sér, at eigi verði stór afhlaup, ok drukku menn þat, herra, lengi síðan.
Er á Mærini mikill hǫfuðstaðr ok hús stór, en byggð mikil umhverfis. Þykkir mǫnnum þar til
gleði gott at drekka mǫrgum saman.’ Heimskringla II, Ísl. Fornr. 27, 179.
The king took then his ship and sailed up to Mære where he had Ǫlvir
and many other men killed. He converted the people to Christianity, placed
priests there, and finally erected and consecrated the church there.
There are other and more ancient sources indicating that the calendrical
festivals mentioned by Snorri existed in Norway during the transition period.
Ágrip, for instance, indirectly supports the existence of four pre-Christian
calendrical feasts. It describes how Óláfr Tryggvason substituted ritual drinking
for pagan sacrifices and libation rituals at Christmas and Easter, St John’s Mass
(summer) and at St Michael’s Mass (fall).27 A praise poem called Haraldskvæði,
composed by Þorbiǫrn Hornklofi about AD 900, gives strong evidence of the
calendrical Yule feast (ON jól) in southern Norway (see below).
26 En þat er siðr þeira at hafa blót á haust ok fagna þá vetri, annat at miðjum vetri, en it þriðja
at sumri, þá fagna þeir sumri. Heimskringla II, Ísl. Fornr. 27, 180. Trans. Hollander.
27 . . . felldi blót ok blótdrykkjur ok lét í stað koma í vild við lýðinn hátíðadrykkjur jól ok páskar,
Jóansmessu mungát ok haustǫl at Míkjálsmessu. Ágrip, Ísl. Fornr. 29, 22.
28 Þá skyldi blóta í móti vetri til árs, en at miðjum vetri blóta til gróðrar, it þriðja at sumri, þat
var sigrblót. Heimskringla I, Ísl. Fornr. 26, 20.
29 de Vries 1956–57, §§305–314; cf. Nordberg 2006b, 76–78.
30 Scholion 141 says that the sacrifice was at the vernal equinox (hoc sacrificium fit circa
aequinoctium vernale), i.e. the beginning of the spring planting. This sentence is uncer-
tain, as it is only attested in manuscript A3, dated to 1434. Hultgård 1997, 30.
31 O ÖD I, 321.
was celebrated at the first full moon after the first new moon following the w
inter
solstice, while the Disthingen took place at the third full moon.32 Since these
feasts took place during the winter, some of the rituals were probably performed
inside halls and cultic buildings, where the rulers also took part in the ceremo-
nies. In Austrfararvísur a calendrical feast called álfablót was celebrated in west-
ern Svetjud (see ch. 8 above and below).
Most likely calendrical feasts also took place in Iceland during the Viking
Age. The sacrificial feasts at the winter-nights (vetrnætr) are well-attested in the
sources. According to the Sagas of Icelanders, they took place in the autumn,
probably October. Eyrbyggja saga 37, for instance, mentions that Snorri goði
gave a great feast at his home in western Iceland (Snæfellsnes) and invited
a lot of friends during the autumn, at the winter nights (at vetrnóttum). The
guests were served with ale and there was drinking.33 Some sagas report that
the object of sacrifice during these feasts was Freyr. In Gísla saga Súrssonar 15,
for instance, we may read thus:
The events of this saga take place in north-western Iceland. A sacrificial feast
at the winter-nights is also mentioned in chapter 10 of Gísla saga. It seems as
if chieftains and magnates prepared and perhaps paid the cost of these feasts:
In those days it was the custom to celebrate the coming of winter by hold-
ing feasts and a “winter-night sacrifice”. Gísli no longer sacrificed after he
left Viborg, but he still held feasts and showed the same magnanimity as
before.35
36 Þar var veizla búin at vetrnóttum ok gǫrt dísablót . . . Víga-Glúms saga 6, Ísl. Fornr. 9, 17.
37 See de Vries 1956–57, §311; Näsström 2001a; 2002, 167.
38 Egils saga, Ísl. Fornr. 2, 105–111.
39 Flateyjarbók I, 420. Cf. F. Ström 1954.
40 . . . nú mun Ljót, móðir hans, blóta í mót sumri, sem hon er vǫn at þeira sið . . . Vatnsdæla
saga, Ísl. Fornr. 8, 67. Trans. Wawn.
41 See mainly Lucas & McGovern 2008.
42 Lucas 2009, 404ff.
43 Magnell & Iregren 2010.
Hákonar saga góða 13–18 tells us that Sigurðr Hlaðajarl was “the most ardent
heathen worshipper” (var inn mesti blótmaðr), as Hákon his father had been.
We are also told that Sigurðr maintained all sacrificial feasts there in Trøndelag
on the king’s behalf. When the Christian king Hákon the Good appeared in
Trøndelag and did not want to take part in the traditional sacrifices with the
farmers, they became very upset. At the Frostaþing they asked him to partici-
pate in the rituals. Later in this story the king was more or less forced to take
part in the sacrificial meals at the sanctuaries of Lade and Mære and dedicate
toasts to the native deities worshipped there. It thus seems as if the Viking
Age rulers of Trøndelag not only organized the cult (see ch. 5); they were also
expected to perform the sacrifice on behalf of the people during the public
religious feasts. Perhaps they consecrated the beaker and the food, for instance.
Similar information is attested in older texts, too.44 Ágrip (1190 AD), for
instance, tells us that the farmers of Trøndelag (Þrœndir) in Mære demanded
that King Hákon the Good “sacrifice as other kings used to do or we throw you
out of the country, if you do not follow us in this thing”.45 Fagrskinna, written
sometime between Ágrip and Heimskringla, also reproduces this tradition in
short words. At the assembly in Mære “the Þrœndir gave him [King Hákon] two
things to choose between. Either he should sacrifice as kings used to do before
and in such a way maintain old law/custom for good crops and peace; or else
they would throw him out of the country”, and “because of this threat and out
of love for his friends the king yielded to their demands and sacrificed”.46
The three prose texts in Fagrskinna, Ágrip and Heimskringla thus suggest
that the people expected the king to take part in the religious feast. According
to Ágrip and Heimskringla he had to eat the sacrificial meat (i.e. horse liver) in
order to become a legitimate king. By means of these rituals he ratified ancient
law. According to Fagrskinna and Heimskringla, he had power over good crops
and peace (ár ok friðr) for the people. It thus seems as if he was a necessary
ritual link to the gods during these feasts, according to these texts.47
44 On this theme, see particularly Jón Hnefill Aðalsteinsson 1992; 1998, 57ff.
45 . . . báðu hann blóta sem aðra konunga í Nóregi, “ella rekum vér þik af ríki, nema þú gerir
nekkvern hlut í samþykki eptir oss.” Ágrip, Ísl. Fornr. 29, 8.
46 . . . ok á því þingi gørðu Þrœndir konunginum tvá kosti, at hann skyldi blóta eptir vanða
enna fyrri konunga ok fylla svá en fornu lǫg til árs ok friðar, elligar mundu þeir reka hann af
ríkinu, ef hann vildi <eigi>) . . . . . . . . ., þá gørði hann eptir bœn þeira ok blótaði. Fagrskinna,
Ísl. Fornr. 29, 80.
47 See Jón Hnefill Aðalsteinsson 1992; 1998.
9.3.1 In Svetjud
Snorri’s ideas that the pagan rulers in Norway were expected to perform ritu-
als and be involved in public cult were thus not taken out of thin air. He had
several sources for these notions (see ch. 4 above). These ideas are actually
attested in other parts of Late Viking Age Scandinavia and in other types of
sources. Scholion 140 of Adam of Bremen’s text, for instance, says:
When not long ago the most Christian king of the Svear, Anunder, would
not offer the demons the prescribed sacrifice of the people, he is said, on
being deposed, to have departed “from the presence of the council, rejoic-
ing” that he had been “accounted worthy to suffer reproach for the name
of Jesus.”49
According to the context it seems as if these events took place at the public
feast in Uppsala. It should be noted that this scholion appears in the oldest
manuscript of Adam’s text, namely A2, which usually is dated to about 1100
(see ch. 4 above). It was thus made only 25 years after Adam wrote his auto-
graph on contemporary conditions in Uppsala.
It is likely that conditions at the sanctuary of Uppsala changed during the
early decades of the eleventh century. After King Óláfr sœnski Eiríksson con-
verted to Christianity there was probably no king of the Svear who performed
the central rituals there. The traditional order was thus disturbed. This may
be the reason why kings are so absent in Adam’s account of the Uppsala cult.
This scholion indicates that the people who gathered in Uppsala during the
eleventh century still expected the ruler to perform the major sacrifices on
their behalf. At such great sacrificial feasts in Uppsala, however, there must
have been several cultic leaders involved in the different ritual actions (see ch.
14 below). Most likely the king was seen as a kind of leader of these specialists
and perhaps a necessary symbolic link to the gods.
King Anunder was thus driven from the public feast and assembly at
Uppsala, since he would not perform as a cultic leader at the sacrifices. This
image agrees well with some information given in the U-version of Hervarar
saga (c. 1300), which says that the Svear removed King Ingi Steinkelsson from
office, since he would not make sacrifices on their behalf at the assembly
which probably was held at Uppsala:
At the assembly that the Svear held with King Ingi, they gave him two
choices. Either he would maintain the ancient law or he had to give up
the kingdom. King Ingi said that he was not ready to give up that faith,
which was the right one. The Svear shouted and threw stones at him and
drove him off the assembly place. Sveinn, the king’s brother-in-law,
remained at the thing. He told the Svear that he would make sacrifice on
their behalf if they would give him the kingdom. All assented to that pro-
posal. Sveinn was taken as king over all Svetjud. A horse was led to the
meeting place, dismembered and distributed for eating, and the sacrifi-
cial tree was reddened with the blood. All Svear thus abandoned
Christianity, re-introduced the sacrifice and drove Ingi off and he trav-
elled to Västergötland. Blótsveinn was king over the Svear for three
winters.50
Because of the change of style, most scholars accept that the chronicle of the
Svea kings has been added to the original version of Hervarar saga.51 It may
be part of a lost historical source moulded from legendary and mythical ele-
ments. Some parts of this tradition are also preserved in Orkneyinga saga,
which tells of the Christian King Ingi of Svetjud who fought against p
aganism.
50 Á þingi nokkuru, er Svíar áttu við Inga konung, gerðu þeir honum tvá kosti, hvárt hann vildi
heldr halda við þá forn lǫg eða láta af konungdómi. Þá mælti Ingi konungr ok kveðst eigi
mundu kasta þeiri trú, sem rétt væri. Þá æptu Svíar ok þrǫngdu honum með grjóti ok ráku
hann af lǫgþinginu. Sveinn, mágr konungs, var eftir á þinginu. Hann bauð Svíum at efla blót
fyrir þeim, ef þeir gæfi honum konungdóm. Því játa þeir allir; var Sveinn þá til konungs tekinn
yfir alla Sviþjóð. Þá var fram leitt hross eitt á þingit ok hǫggvit í sundr ok skipt til áts, en roðit
blóðinu blóttré. Kǫstuðu þá allir Svíar kristni, ok hófust blót, en þeir ráku Inga konung á brott,
ok fór hann í vestra Gautland. Blót-Sveinn var þrjá vetr konungr yfir Svíum. Hervarar saga.
(Ed) E. O. G. Turville-Petre, 70f. Cf. (Ed) Jón Helgason, 160. My trans.
51 For instance Schück 1910, 14–15; Reifegerste 1989, 196–200.
The farmers, however, chose another king, called Blótsveinn (the brother
of the Queen) who retained sacrifices.52 This tradition may have been trans-
ferred to Iceland by Markus Skeggjason, who was probably Ingi’s skald before
being made Lawman at Þingvellir in 1084.53 Blótsveinn’s existence has been
questioned and his story has been regarded as a medieval exemplum, that is, a
homiletic story.54 Yet Blótsveinn (in Old Swedish Blodhsven) is also mentioned
independently of Icelandic traditions in The Legend of St Eskil (Legenda sancti
Eskilli). Here, the sacrifices are said to occur in Strängnäs.55
It thus seems as if the Late Viking Age cult community in Svetjud expected
the king to perform or at least participate in the common sacrifices at the
Uppsala sanctuary. In my opinion, there is no reason to doubt this informa-
tion. Perhaps the sacrifice was only valid when the king took part in the rituals.
A plausible interpretation is that the king was the most important mediator
between the human and the divine world during the sacrifices at this sanctu-
ary. His ritual role in the cult was thus necessary for the community, and per se
it also legitimated his high position in other social contexts.
Medieval written sources thus support the idea that pagan rulers were
involved in the sacrificial feasts in eastern Scandinavia, which included cer-
emonial meals and drinking rituals. This idea may also be indicated in other
types of sources, sources which are undoubtedly pre-Christian. Some runic
stone monuments of the Blekinge group may give information about politi-
cal leaders in cultic functions. They were carved not before AD 550 and per-
haps not after AD 750.56 Stentoften (DR 357), Gummarp (DR 358) and Istaby
(DR 359), have a proper noun Haþuwulfr. On Stentoften and Istaby there
are also two persons called Hariwulfr and Heruwulfr. The Björketorp stone
(DR 360) can be associated with this group, although it lacks the name Haþu-
wulfr. It has almost exactly the same curse as Stentoften. As the name
Haþuwulfr appears on three stones in this area scholars believe that he was a
local ruler in Blekinge some time around AD 600.57
52 Kom því svá, at bœndr tóku sér annan konung, þann er þá helt til blóta, Sveinn, bróðir drott
ningar, ok var kallaðr Blót-Sveinn. Orkneyinga saga, Ísl. Fornr. 34, 89–90. Snorri was also
acquainted with this tradition and mentioned Blót-Sveinn. See Heimskringla III, Ísl. Fornr.
28, 263.
53 This theory was first suggested by Henrik Schück 1910, 17. Cf. Reifegerste 1989, 198f.;
P. H. Sawyer 1991, 37; Foote 1993, 24.
54 Lönnroth 1996, 150f.
55 S RS II: 1, 391–399.
56 On datings of these inscriptions see Krause 1966, 203ff., Moltke 1985, 137–147; Williams
2001.
57 E.g. Krause 1966, 203–214; Williams 2001; Düwel 2008, 21f.
With nine bucks, with nine stallions Haþuwulfr gave good growth.59
Haþuwulfr performed a sacrifice where he offered nine bucks and nine stal-
lions. In that way he gave the people a good crop. Perhaps he also arranged a
great sacrificial feast at his hall building or at an outdoor sanctuary, with plenty
of meat for all the people.
Santesson’s interpretation has many advantages compared to previous
attempts, and many scholars accept it.60 Her interpretation can be supported
on several grounds, also from the perspective of history of religions. For
instance, the number nine corresponds to the sacrifices at Uppsala and Lejre
mentioned by Adam and Thietmar.61 As in Uppsala, Lejre and Lade, the runic
inscription indicates that horses were sacrificed in Blekinge too.62 It should
also be noted that the Proto-Nordic word jāra in this inscription (here marked
with the j-rune, i.e. an ideograph (Begriffsrune)) is probably equivalent to the
Old Norse ár that appears in Snorri’s text in connection with sacrifices, mean-
ing ‘good crops’ (see below).63 Stentoften thus supports the idea that a ruler
performed a sacrifice, or at least that he commissioned it.
This runic inscription indicates that persons belonging to the ruling elite
officiated in the public cults and performed ceremonial acts. The changes in
sacrificial custom that have been pointed out by archaeologists may reflect
changes in social organization,64 from tribal chieftains to “petty kinglets” and
finally more permanent ruling families. The family of Haþuwulfr seems to be
a dynasty, according to the naming customs.65 The new ruling elite used reli-
gious ceremonies when manifesting their power and sovereignty. They prob-
ably institutionalized public cults which they tied to themselves and their
residences.
9.3.2 Iceland
Sources report that the Viking Age chieftains in Iceland, i.e. the goðar, also had
central ceremonial functions during public meetings, such as law assemblies
and cultic feasts. According to the Hauksbók redaction of Landnámabók, the
chieftain (goði) was to wear a ritual ring on his arm at such meetings, espe-
cially when he himself was at the head of the affairs (the text is paraphrased
in ch. 5). This ring probably had an important religious significance since it
otherwise was placed on the ritual platform in every chief hof. The text also
reports that the ring was first reddened in the blood of a neat which the chief-
tain himself had sacrificed at the meeting. The connection between chieftains
(goðar) and sacrifices may be indicated by skaldic poetry dating back to the
tenth century. In a lausavísa (8), composed by Hallfreðr vandræðaskáld, Óðinn
is called hrafnblóts goði ‘the goði of the raven sacrifice’.66
Eyrbyggja saga 3–4 also suggests that the chieftain Þórólfr Mostrarskegg in
Iceland (and Norway) performed central roles at the hof sanctuary and exe-
cuted parts of the sacrificial rituals (see ch. 3 and 4 above). Similar information
is reported about the chieftain called Þorgrímr goði in Kjalnesinga saga 2. He is
described as blótmaðr mikill ‘a great sacrificer’, who had a hof sanctuary built
on his farm at Kjalarnes in south-eastern Iceland.67 Most likely the informa-
tion in Kjalnesinga saga and Eyrbyggja saga is to some extent dependent on
Hákonar saga góða 14, when describing the chieftain and his role as a sacrifi-
cer. They do not add much new information in relation to them. On the other
hand, Landnámabók, which probably is older than Snorri’s text, also describes
Þórólfr as a blótmaðr mikill ‘a great sacrificer’.
Other Icelandic traditions similarly report that chieftains (goðar) were
involved in the sacrificial rituals at the public sanctuaries. Gísla saga Súrssonar 15
mentions that Þorgrímr Þorsteinsson goði arranged a feast in his drinking hall at
the end of autumn to celebrate the coming of the winter-nights. During this feast
he also sacrificed to Freyr (blóta Frey).68 It has been suggested that this is one of
the oldest Sagas of Icelanders.69 Thus we may assume that the information about
the cultic role of the chieftain was based on a quite reliable oral tradition, since it
also harmonizes with information from other sources.
Snorri’s information that King Hákon was supposed to perform some central
rituals during the sacrificial feasts in Viking Age Trøndelag has many paral-
lels in sources referring to the ritual functions of rulers in all three inves-
tigation areas. They were thus not only organizers of cultic feasts, but also
cultic performers during the sacrifices. Most likely they were regarded as a
necessary ritual link between the human and divine world, and mediators in
giving blessings to the people, such as good crops and peace. The close genea-
logical and/or cultic relation to the Other World made them suitable for this
position and function.
Snorri states in Hákonar saga góða that sacrificial blood (hlaut) was smeared
on the platforms (stallar) and on the walls of the hof. As noted above, the term
hlaut is dubious in this context. But this does not mean that the ritual itself
needs to be suspect. In Óláfs saga Helga 107, Snorri presents similar informa-
tion without mentioning the term hlaut. He states that sacrificial feasts were
still performed in Inn-Trøndelag during St Óláfr’s reign and “that cattle and
horses had been slaughtered and the platforms reddened”.70
Klaus Düwel was very critical of these passages in Snorri’s text.71 He argued
that the ritual actions connected with the sacrifices in these texts had no
Germanic origin, that is, that Snorri had no pre-Christian sources for these
68 Gísla saga Súrssonar 15, Ísl. Fornr. 6, 50. The text is quoted above.
69 Jónas Kristjánsson 1988, 258.
70 . . . at þar væri drepit naut ok hross ok roðnir stallar af blóði ok framit blót ok veittr sá formáli,
at þat skyldi vera til árbótar. Heimskringla II, Ísl. Fornr. 27, 177.
71 Düwel 1985, 32ff.
rituals. Rather, when describing how blood was sprinkled on the altars in
Trøndelag, Snorri modelled his account on ancient Hebrew rituals described
in Exodus. In Snorri’s time, parts of the Old Testament had been translated into
Old Norse in a text called Stjórn. In this text there was a passage about the Old
Jewish sacrifice in chapter 100 (i.e. Exodus. 24):
Moyses tok halft fornarblodit ok hellti þui i kerit. En halft blod hellti hann
yfir alltarann.
According to Düwel this text clearly resembled the passage about the sacri-
ficial blood in Hákonar saga góða 14. Düwel argued, further, that during the
Early Middle Ages, the sacrificial rituals of the Old Testament were considered
pagan, and pagan religion was thought to be the same in all places; this led
Snorri to use this material in his account of the ancient rituals in the Trøndelag.
It is possible that Snorri took some expressions from the description in
Stjórn when writing about the sacrifices in Trøndelag. It seems, however, as
if the word stallr/stalli (m.) is a genuine term in pre-Christian Scandinavian
sacrificial contexts (see ch. 7 above). The compound véstallr ‘sacred stand’, for
instance, appears in the early poetic language.
9.4.1 Pouring Sacrificial Blood on Ritual Objects and “The Temple Soil”
Snorri’s statement in Hákonar saga góða that the sacrificial blood from the ani-
mals was smeared on the sacred objects, called stallar, may thus not be totally
distorted. His statement that the walls of the hof were smeared with such blood
may likewise be reliable. Independent sources indicate that sacred objects and
cultic structures were reddened with sacrificial blood in western Scandinavia.
In Hyndlolióð 10 the goddess Freyja exclaimed thus about one of her adorants:
“He’s made an altar (hǫrgr) for me, faced with stone. Now that stone has turned
to glass. He’s reddened the new altar with ox blood.” Admittedly Hyndlolióð is
attested in a late manuscript from the second half of the fourteenth century,
and usually the poem is regarded as young and dated to the twelfth century.72
Yet the content indicates that the poet had good information about ancient
customs and rituals. It seems quite unlikely that a medieval Christian skald
would have created from his own imagination the idea that the stones of the
hǫrgr had been reddened by the blood from the ox. These stones had probably
acquired a glaze as an effect of the many sacrificial fires on the blood flowing
72 This manuscript is called Flateyjarbók Gks 1005, fol. (F). For a recent treatment, see Cöllen
2011, 192–247.
over them.73 It is possible that some kind of burnt offering took place there,
where the victim was sent as a gift to the gods.
Similar information to that in Hyndlolióð also appears in other texts.
According to the U-version of Hervarar saga, a holy tree called blóttré, situ-
ated in Svetjud, was reddened with the blood from a sacrificial horse (see
above). Whether the word blóttré in this text really refers to a tree is somewhat
uncertain. One possibility is that this term designates the wood of the ritual
platform or altar (ON véstallr, stallr, stalli), which was reddened with the sacri-
ficial blood.74 According to Ynglinga saga 15, the Svear smeared the stallar in
Uppsala with sacrificial blood. This text reports that they sacrificed their king
called Dómaldi for better crops (blóta til árs), and that they attacked and killed
him and reddened the ritual platforms (rjóða stalla) with his blood.75 This
account has no historical background, but it is built on the ninth-century poem
Ynglingatal 8 (5), which definitely renders a pagan world view. The poem reports
that the Svear carried swords against their own ruler and stained the earth red
(at fold ruðu) with his blood. They sacrificed him, eager for crops (árgjǫrn . . . of
sóa skyldi). No stalli/stallr is mentioned in this stanza. The poem indicates, how-
ever, that the earth of the place was reddened with sacrificial blood.
A passage referring to a possible sacrifice made to the álfar in Kormáks
saga also points in the same direction. The injured hero Þorvarðr was recom-
mended to pour the blood of a bull on to one of the hills/mounds (rjóða blóð
graðungsins á hólinn) inhabited by álfar and to prepare a meal for them from
the meat of the bull in order to heal his injury.76
As we noted in ch. 5, several chieftains who dismantled their hof buildings
in Norway when travelling to Iceland brought some of the “temple-soil” to the
new land. One man who did this was the hofgoði, Þórhaddr the Old, according
to Landnámabók. According to Eyrbyggja saga 4, Þórólfr also brought “temple-
soil” from Norway to Iceland, i.e. the soil from under the ritual platform or
altar (svá moldina undan stallanum). The information from the written sources
indicates that the soil below the ritual platform/altar (sg. stalli/stallr) was holy
since the sacrificial blood had poured down on it.77 The written sources also
report that ritual objects which were placed on the stallar were reddened with
blood during the sacrificial rituals, for instance the “oath-ring”.78
In several non-Christian religions, the custom of pouring blood on sacrifi-
cial altars is common, for instance in Greek religion, as well as in the Jewish
burnt offering.79 In the Greek sacrifice the blood was collected in containers
and then poured or sprinkled on the altar, on the hearth or in sacrifice pits.
Small animals were cut up on the altar itself.80 Also interesting is the Roman
sacrificial practice whereby blood reddened the altar in front of the temple.81
Lucretius, for instance, gives a vivid picture of “altars washed with the stream-
ing blood of beasts”.82 Some of the blood was caught in containers. Virgil men-
tions that when Aeneas’ companions sacrificed to Hecate, they “put knives to
the animals’ throats and catch the warm blood in vessels”.83 There is even evi-
dence indicating that the priest tasted the blood. The animal was then butch-
ered, cooked and sometimes eaten by the worshippers.
buildings (see ch. 11 below). Perhaps the earth where they were found was once regarded
as sacred “temple soil”. Cf. Gunnell 2001.
78 According to the Hauksbók redaction of Landnámabók, the oath-ring was placed
on the platform. This ring was reddened in the sacrificial blood of a neat by the goði.
Landnámabók, H 268, Ísl. Fornr. 1, 313.
79 Burkert 1983; Hultgård 1996, 45. On the blood at sacrifices, see especially Ekroth 2002,
242–276.
80 Burkert 1983. Fine descriptions of Greek sacrifices and sacrifices can be find in Hesiod
(Theogony, v. 535–57) and especially in Homer (Odyssey I, Book III, 432–63, 70–72), where
blood also plays a role. According to Gunnel Ekroth (2005), only some of the blood was
sprinkled on the altar at the thysia sacrifices, while most of it was consumed as food by
the worshippers. Unlike Hebrew religion, there is no evidence for blood taboo or prohibi-
tion for eating blood in Greek religion.
81 See e.g. Ogilvie 1994, 41–52; Beard 2002, 36ff.; Näsström 2005.
82 Lucretius, V, 1200.
83 Vergil, Aeneid VI, 248–249. On Roman sacrifices, see also Cato: De agri cultura 83, 132, 134,
139–41.
Figure 16a
Gotlandic picture stone from Lärbro (Hammars I).
Photo: Bengt A Lundberg. ATA,
Riksantikvarieämbetet.
Figure 16b Detail of the picture stone from Lärbro (Hammars I).
Photo: Olof Sundqvist.
84 “Das Gestell, das auf dem Bildstein abgebildet ist, könnte dem kultischen Gegenstand
stalli oder stallr nahekommen, den Snorri und der Verfasser der Eyrbyggjasaga erwäh-
nen.” Hultgård 1993, 234f.
cultic function than being the place where the sacrificial objects were pre-
sented and given to the deity. This ritual platform seems anyhow to be located
at a cultic site out-doors, beside a tree, where a victim is hanging.
One puzzling thing in Snorri’s account of the sacrificial ritual at Lade is that
the sacred platforms are referred to in the plural, stallar. Whether these stallar
were located inside or outside the hof is not mentioned. In both Landnámabók
and Eyrbyggja saga the stalli/stallr, in the singular, seems to be inside the hof.
Archaeological evidence indicates that altars or ritual platforms could be
placed either inside a house or outdoors. Inside the small cultic house (5 × 7 m)
in Borg, Östergötland, for instance, there was a foundation built up with strong
and flat stones.85 It was located along the eastern wall, at the end of the path,
which split the room in two halves. Just to the south of the foundation two amu-
let rings were discovered. The foundation has been interpreted as a véstallr.86
The traces of a rectangular platform inside the hall building V A at Helgö,
mentioned by Arrhenius, could perhaps also be interpreted as reflections of
a véstallr. In my opinion, such an interpretation is better than the proposal
seiðhjallr (see ch. 4 above). An outdoor ritual structure or platform (perhaps
equivalent to a véstallr or hǫrgr) might be evidenced at Lilla Ullevi in Uppland.87
Uppsala. The text is usually regarded as having a low source value, as it is from a
late fourteenth-century manuscript and contains fictional elements, but at the
same time it has aspects that may be very old. The story tells how King Haraldr
Finehair had sent two men, Haukr and Vígharðr, to Bjarmaland, in order to
bring furs to the king. They were blessed on their journey by the “witch” Heiðr.
King Eiríkr, Haraldr’s enemy, discovered this and sent two retainers, Bjǫrn and
Salgarðr, to follow Haraldr’s men. Before they went on with their journey, they
needed support from the divine powers. Eiríkr held a feast (væizla) at Uppsala.
It has been suggested that Lytir is another name for Freyr.99 An appellation
*lytir may occur in some Swedish place-names, e.g. Lytisbærg, at Östra Husby,
Östergötland, Lytislunda, in Österrekarne, Södermanland and Litslunda,
in Lillhärad, Västmanland.100 It is derived from OSw luter, loter ‘lot which is
used in divination’ (cf. ON hlaut) or a Proto-Nordic *hluti-wīhar “Losdeuter,
Wahrsager, Opferpriester”.101 The appellative may thus refer to a cultic leader/
specialist who performed divination rituals.
In the account of the god Lytir, sacrifices were first made to him. When
Lytir indicated that he was present, the deity’s wagon was pulled, perhaps in
a procession, to the hall by draught horses (eykr sg.). Carriage and horses may
have ritual implications. It is said that the carriage became heavy when the
god was in it. Auguries were performed before it—King Eiríkr asked questions
of the god (i.e. his image) after he had given him the libation sacrifice in the
hall. The story may therefore reflect the practice of augury combined with
different types of sacrifices.
Thus it is possible that divination rituals were performed by the rulers when
making sacrifices at the platforms (stallar) of the sanctuaries. These rites were
probably performed in order to obtain divine guidance in crisis situations and
when something dangerous had to be done. The action to be taken by the ruler
thus acquired divine legitimacy.
The most likely reconstruction from the forensics of the skulls requires at
least a two-person team, one of whom struck the animal between the
eyes (effectively killing it and certainly stunning it into momentary
immobility) while the second swung a fairly broad-bladed axe at the neck
or base of the skull for a beheading stroke. There is a definite concentra-
tion of beheading blows from the right side towards the left of the
a nimals’ skulls. If the axe-wielder was right handed, this means this per-
son was probably standing to the right of the sacrificial animal. If the
team got their timing right, the beheading stroke would produce a blood
fountain as the animal’s heart would still be beating.104
According to Lucas and McGovern, this is a ritual and violent practice intended
to dramatize the death.
This interpretation could be supported with sources referring to the other
parts of the investigations areas. For it seems as if decapitation of the victims
(including bloody elements) was a widespread ritual practice at sacrifices in
several parts of Scandinavia. Adam of Bremen, for instance, suggests that it
was only the heads of the sacrificial animals (and humans) that were sacrificed
to the gods at the temple of Uppsala. He writes thus: “The sacrifice is of this
nature: of every living thing that is male, they offer nine heads.”105 The term
capita should probably be understood in this context in its concrete sense as
‘heads’.106 Perhaps it was only these parts that were given to the gods, while the
meat was consumed during the ceremonial feast. This idea could also be sup-
ported by means of archaeological evidence from Jämtland. Under the medi-
eval church at Frösön ‘the island dedicated to the god Freyr’, Viking Age animal
bones were found around a birch stump.107 These bones were interpreted as
remains from animal sacrifices. Most of the bones belonged to the heads of
the animals. Perhaps these heads were given to the gods at a sacred tree or a
holy grove. The ceremonial meal may have taken place in the Viking Age cer-
emonial building beside this outdoor cultic site, indicated by the place-name
Hov. A similar situation may be found at Borg, Östergötland. At this site a small
house was discovered, which was interpreted as a cultic building. Outside this
building 75 kilos of unburned bones were discovered, together with 98 amulet
rings.108 The bones, which belonged to horses, dogs and pigs, were interpreted
as remains of sacrifices. Most of the bones consisted of skulls and jaws from
animals. It is thus not impossible to assume that the custom of offering heads
of animals was widespread in Scandinavia. This evidence supports the inter-
pretation that the skulls from Hofstaðir could also be seen as sacrificial gifts to
the gods, while the meat was consumed during ceremonial feasts celebrated in
the chieftain’s hall (see below). It also indicates that sacrifices involved bloody
rituals as described in the written sources above.
Perhaps the blood in these rituals also had the performative function to
change the situation for the participators, i.e. to construe new conditions
and a new reality. Inspired by René Girard’s theory of sacrifice as a form of
scapegoating,113 Lucas and McGovern suggest that the main reason why the
rituals at Hofstaðir were performed with such unnecessary violence and
bloodshed was expiation and the removal of tensions and conflicts within
the community. The extreme violent action was thus a way to draw away from
interpersonal conflicts and channel it out in other ways. The violent bloody
ritual changed conditions for the community, especially in periods of social
stress. Since, the archaeological records from Hofstaðir show that these ritu-
als were season based and performed during the summer, I suggest that this
calendric ritual rather may be seen as a repetition of the cosmogony made
in order to strengthen the cosmic process. According to Snorri’s version in
Gylfaginning of the creation myth, the slaying of the primordial giant Ymir
made by the divine brothers Óðinn, Vili and Vé in order to create cosmos,
involved a bloody slaughter.114 The blood of Ymir killed all giants (i.e. the evil
forces) except Bergelmir, by means of drowning. The blood was also used when
making the seas and the lakes. This mythic action may have been repeated
in the sacrificial contexts as suggested by Bruce Lincoln.115 During the ritual
performance the sacrificial blood alluded to Ymir’s blood. The bloodshed
implied a definite change of condition for the participators; from now and on
the world was renewed and the seasonal changes were guaranteed as well as
the crops and general welfare etc. For many people, who did not have access to
the religious and cultural knowledge, this symbolic meaning of the blood was
not perceived. Perhaps the treatment of blood was only a signal of the sacred
act as such. These dramatic acts were executed by representatives of the local
elite, who probably gained much prestige by performing these rituals. In con-
nection to these sacrificial rituals the rulers also arranged feasts in their halls,
including ceremonial meals as described in Hákonar saga góða. According to
this and some other traditions the custom of eating horse meat seems to be
crucial there.
116 See Jón Hnefill Aðalsteinsson 1992; 1998, 57ff. On the medieval Christian attitudes con-
cerning the consumption of horse meat and prohibitions against these customs, see
Sanmark 2004, 216ff.
117 Ágrip, Ísl. Fornr. 29, 8. The text is quoted above.
118 Fagrskinna, Ísl. Fornr. 29, 80. The text is quoted above.
119 Heimskringla II, Ísl. Fornr. 27, 177.
120 See Sanmark 2004, 222.
121 GuL 20. NGL 1, 11f. The prohibition on eating horse flesh in GuL appeared in a chapter
that forbade the eating of meat during the seasonal fast. Anders Hultgård (1996, 42) has
argued that the fact that horse meat was forbidden at all times of year suggests that horses
previously were consumed in ritual contexts.
The phrase “to leave for a heathen land” ( fare a land hæiðit) is significant. It
also appears in the outlawry clauses regarding those who refused to baptize a
child, neglected to pay tithe, or performed illegal marriages and divorces. All
these practices were thus not in keeping with Christian customs.125
Archaeology supports the idea that pre-Christian ceremonial feasts, includ-
ing meals with meat, took place in Trøndelag at cultic sites with halls and cultic
houses. Sometimes we may assume that rulers were involved in these rituals. At
the Late Iron Age houses at Mære, where the gold foil plaques were also found,
other finds were made which may be associated with pre-Christian ceremonial
feasts in aristocratic settings. For instance, pieces of pottery and glass were
discovered as well as quantities of animal bones (including horse bones) and
some characteristic cooking stones mixed with charcoal. The cooking stones
and burnt debris seem to have been connected to the latter Viking Age build-
ing, which Lidén interpreted as a pre-Christian hof.126 About the same time as
the gold foils were deposited at this building, it is highly likely that ceremonial
meals were also prepared and celebrated there.
At the chieftain farm of Hove in Åsen, Ut-Trøndelag, there is a small ridge
called Hovshaugen, where several finds were made indicating that pre-Chris-
tian cultic meals were celebrated there over a long period.127 Several pits with
charcoal, burned stones, and animal bones (sheep and goats) were found
there. It was suggested that these pits were used for ritual cooking and per-
haps sacrifices. Perhaps these meals sometimes took place at some buildings
of the farm Hove. This cultic site was used for a very long time, from c. 100 BC
to AD 1000. Also at the central farm Melhus, in Gauldœla fylki, we find traces
of ritual meals in the context of rulers and settlements. A great many cooking
pits were discovered at the churchyard of Melhus, with dates stretching over a
long period. The oldest pits may be dated to Early Iron Age, but most of them
were dated to the period between 900 and 1300. Their function is debated.
Bones were found in some of the pits, for instance remains of pigs but also of
other domesticated animals. The archaeologist Anne Stalsberg suggests that
many of these structures are remains from the building of the stone church.
The pre-Christian remains situated on the plain below the yard, however, may
reflect an outdoor gathering place of both political and religious significance.128
Ceremonial feasts at aristocratic halls may be attested in northern Norway.
According to Gerd Stamsø Munch, the hearths in the hall-room C (I:1a) of Borg,
Lofoten, were used for ritual baking of the slaughtered animal meat (see ch. 6
above). These meals were probably an important element in the construction
of authority at the ceremonial building of Borg.
this information is reliable, since Ari said that he personally retrieved it from
“old intelligent people”. These people may have lived with one foot in paganism
and one in Christianity.130
The investigation at Hofstaðir indicates that the ritual slaughter of cattle
had a seasonal basis (see above). The evidence suggests recurring ritual activi-
ties over a period of years, and not a single mass killing event.131 During these
calendrical feasts a lot of meat was consumed in the hall. Perhaps the richness
of these feasts was an expression of power claims. Lucas describes this phe-
nomenon as “competitive feasts”, “occasions where the intention is to display
or show off the success of a household or community”.132 The aim may have
been both to strengthen the solidarity within the group and to acquire sta-
tus. The skulls hung out in the open air on the walls/roof may be regarded as
“trophy displays” from these grandiose feasts. Such feasts could be seen as a
strategy for rulership in Viking Age Iceland
Hákonar saga góða. The composition of this saga is usually dated to sometime
between 1220 and 1275.136 Viking Age burial customs were performed in
Gotland until the thirteenth century. The archaeologist Lena Thunmark-Nylén
states: “Not until about the time of the Guta Law (c. 1220), when the last pagan
customs were strictly forbidden and criminalized, do we lose contact with the
remaining non-Christians.”137 This statement hints that the information in
the Guta Law and the Guta saga may build on reliable traditions.
9.5.4 Horse Sacrifices and Ritual Meals in the Royal Milieus at Uppsala
Adam of Bremen intimates that sacrificial meals were also celebrated at
Uppsala during the Viking Age. The term commessationes (sg. commesatio)
probably refers to such meals.138 It is also interesting to note that Adam men-
tions that horses were sacrificed in Uppsala. When describing the sacrificial
grove he states that “even dogs and horses hang there with men” (Ibi etiam
canes et equi pendent cum hominibus).139 Even if the information about human
sacrifices is regarded as an exaggeration by Adam, made in order to demonize
the old religion in Uppsala,140 the idea about horse sacrifices may be true. It
harmonizes well with the information we have about horse sacrifices at Mære
mentioned in Ágrip and Heimskringla. Perhaps horse meat was also consumed
during the ceremonial feasts at Uppsala, as in Lade and Mære. Several other
sources report that horses were used as sacrificial animals in Svetjud, espe-
cially in aristocratic contexts. The U-version of the Hervarar saga has already
been mentioned. It narrates that Blótsveinn had to participate in a horse sac-
rifice and eat horse meat when he accepted the kingship of Svear. The excava-
tions at Uppsala and its surroundings have also uncovered traces of ceremonial
practices involving horses: cremated fragments of horses were discovered,
among other animal bones, in Uppsala’s Eastern and Western Mounds (from
the early and second parts of the sixth century respectively).141 A horse has
been found south of Prästgården, located together with boat grave 1, dated to
the ninth century.142 A lot of bones, connected with the Viking Age settlement
area at Norra gärdet, just north of Old Uppsala church, were also discovered.
Some were horse bones, but bones from cattle, sheep, pigs, and even frogs were
also found. The archaeological and osteological analyses support the hypoth-
esis that some of these bones were house offerings: in House 1 (post-hole A32),
for instance, a dog jaw bone, a horse jaw bone, and other bones from a horse
were found and interpreted as foundation sacrifices.143 Deposited horse bones
have also been found in association with the row of posts (see above). During
the excavations in the summer 2012 a lot of sacrificial bones were also found
in the eastern part of Old Uppsala (see ch. 4 above). Sacrifices of horses, pos-
sibly from the Late Iron Age, have been discovered at several places in east-
ern Scandinavia, some of which appear to reflect aristocratic milieus.144 Horse
bones found in a water-hole outside the ring fort at Eketorp in Öland, have
been associated with offerings connected to communal meals.145 At this site
there was also a ceremonial building, dating to a period between the fifth and
the seventh centuries.146 A similar cult including ceremonial meals probably
existed at Skedemosse, in Öland where horse bones, gold hoards, and other dis-
coveries have been made and dated to the Middle and Late Roman Iron Age.147
thick and clearly stratified. The layer included all kinds of deposited objects,
such as surgical instruments, crucibles, bread, arrowheads, whetstones, birch-
bark containers, ice nails, pottery, flint and an Arabic silver coin. Sherds from a
number of vessels were also found at the south end of the setting, beside a bare
rock. The sherds included an amphora of Rhineland origin. Different types of
saddle and rotary querns were also found there, as well as baking ovens and
bread. They were also related to pre-Christian sacrificial rituals. On the stony
ledge some amulets were discovered as well, such as miniature spears, Þórr’s-
hammer pendants, fire-steel pendants, miniature shears and amulet rings,
usually considered as symbols of ancient Scandinavian deities.
In my opinion, the sacrificial rituals performed at the stony ledge must be
related to the ritual activities that took place in the hall (I A), such as the com-
munal meals and banquets. The sacrificial rituals performed at Helgö were
probably very complex in nature, and were performed at several places. The
ritual slaughter and the handing over of victims to the gods may have taken
place outdoors at the rock and the ledge, while the communal meal was cel-
ebrated in the hall. Without doubt it was the hall owner, i.e. the ruler of Helgö,
who played a central role in the rituals that took place both inside and outside
the hall building. Not only the drinking vessels but also the great amount of
animal bones (calf, pig and hare), and all the finds under the stone setting (such
as the bread) may reflect the complicated rituals which were performed at this
sacred site during the calendrical feasts and on other occasions. In connection
with these ceremonies there may have been a team of cultic leaders taking
part in the preparation of the sacrificial feast. Most likely the ruler or chieftain
was the supervisor of these actions. His leading ritual role was emphasized
during the ceremonial feast when he sat in the high-seat inside the hall. Helgö
thus constitutes an extraordinarily good example of sacrifices, banquets and
sacrificial meals performed under the surveillance of the ruler, in the context
of a ceremonial building.
9.5.6 Lunda
At the large Iron Age hall building at Lunda there are also finds which may be
associated with drinking rituals and ceremonial meals in an elite milieu, such
as broken glass originating from Frankish beakers. As at Helgö and Birka, the
Lunda hall had a hearth which may have played a certain role in ambulatory
rituals and ceremonies placing the hall owner (i.e. the ruler) in a central posi-
tion, close to the assumed high-seat. Most likely the ruler of Lunda sat in this
high-seat when celebrating the ceremonial feast. The high-seat in this hall also
emphasized the central position of the ruler during the rituals.
As at Helgö, ritual activities took place both inside and outside the hall at
Lunda. In the yard north of the hall there were several cooking pits, most likely
intended for cooking ritual meals. In one huge cooking pit (3 × 2 m in area
and 1.1 m deep) unburned bones of pigs were found. The archaeologist Eva
Skyllberg relates these bones to the god Freyr.151 In my opinion, she might be
right. In the Old Norse mythical traditions it seems as if Freyr had a special
relation to the pigs. One of his attributes was the pig called Gullinbursti.152 It
was also close to the pit with pig bones that the three phallic images were
discovered. Most likely they represent Freyr. Both pigs and Freyr may also be
related in different senses to the Yule feast ( jól). That Freyr was connected to
the Yule feast is well-established in research.153 Economic arguments may also
relate pigs to this winter feast. Pigs rummage in the earth during the autumn,
looking for beechnuts, truffles, roots and other things. When there is frost in
the ground it is time to slaughter them. This is the reason why the traditional
swine slaughter took place in late autumn or early winter in Sweden. In my
opinion, there are thus good reasons to believe that religious feasts were cel-
ebrated at the cultic hall of Lunda, perhaps during the winters. The great hall
building, the exclusive ceremonial glass and the consumption of pig indicate
that such feasts were controlled by the ruling power. Most likely these meals
were a way to construct authority for the political power at Lunda.
In Hákonar saga góða, Snorri describes the pre-Christian drinking rites ( full,
minni) connected with the sacrificial feasts at the hof of Lade. In my opinion,
there is no doubt that such rituals were crucial during the pre-Christian reli-
gious meetings in the ceremonial buildings of Svetjud, Trøndelag and Iceland.
That certain religious terms connected to such rituals, such as minni, may not
be traced back to pre-Christian times is not a sufficient argument for reject-
ing the idea that the phenomenon itself is ancient. Sources of different types
clearly indicate that ceremonial drinking and meals were important in the
aristocratic halls and cultic buildings in all three investigation areas.
9.6.1 Norway
The compound word blótdrykkja—‘drinking bout at pagan sacrifice’154—seems
to refer to ancient drinking rituals in Norway. It is attested in Ágrip, which
describes how Óláfr Tryggvason abolished the old custom and replaced it with
Christian drinking ceremonies at Christmas and Easter, St John’s Mass, and
St Michael’s Mass.155 The expression drekka jól—‘celebrate Yule with drinking-
ceremonies’—appears in royal contexts in a praise poem to Haraldr Finehair,
Haraldskvæði, composed by Þorbjǫrn hornklofi around 900. The poem reads:
This dance may ultimately derive from cult.160 Anders Hultgård argues in the
same vein when he suggests that leikr perhaps referred to plays and rituals
connected to the pre-Christian Yule.161 Þorbjǫrn’s stanza could thus be used
as reliable evidence, in the context of the practices of rulers, of pre-Christian
libation rites.
In a mythologized account in Hákonarmál stanza 16, Hákon is invited to
Valhǫll to drink beer with the gods (þigg þú at ǫ́ sum ǫl), reflecting again the
relationship between rulers and ritual drinking.162 Further, ceremonial beer
drinking at feasts is attested in saga texts referring to pre-Christian Norway.
Egils saga narrates the visit of King Eiríkr blóðøx and his wife Gunnhildr to
Atløy for the dísablót.163 The saga describes it as an outstanding banquet with
much drinking inside the house (veizla in bezta ok drykkja mikil inni í stofunni).
When Egill and his men are finally allowed to enter the house (stofa), Ǫlvir,
perhaps a cultic leader,164 is allowed to sit beside the king and the queen in the
high-seat. Beer is carried out to drink memory toasts (minni) and for each toast
a horn has to be emptied.165 Egill is angry with Bárðr, who at first forbade him
and his men to take part in the sacrificial feast. The episode is supported by a
stanza composed by Egill, dated to the beginning of the tenth century:
Sǫgðuð, sverriflagða,
sumbleklu, herkumbla
– því telk – brjótr, þars blétuð
– bragðvísan þik – dísir;
leynduð alls til illa
ókunna hjǫrrunna,
ilt hafið bragð of brugðit,
Bárøðr, hugar fári.166
This saga text, combined with the tenth-century lausavísa, provides us with
reliable evidence of pre-Christian ceremonial drinking connected with the
dísablót in Norway. These drinking customs took place in a building called
stofa, perhaps a banqueting hall at the royal farm.
In this quotation the Latin verb libo (libare) must be interpreted in its more
precise meaning, i.e. ‘to perform a libation sacrifice to the deity’.168 It thus
seems as if the worship of the gods in the banqueting room (triclinium) was
intimately associated with the drinking rituals, perhaps with the ruler sitting
in his high-seat. Also in connection with the “unseemly” songs that were sung
at these feasts, “libations” (libatio sg.) were performed: “For the rest, the trivial
songs that they use in this sort of libation rites are manifold and shocking,
167 Hásæti Hjǫrvarðs konungs var búit gagnvart hásæti Granmars konungs . . . Heimskringla I,
Ísl. Fornr. 26, 68.
168 For this interpretation, see Dillmann 1997, 66.
so it is best to keep quite about them.” It seems likely that these ritual songs
had sexual connotations, since Adam disapproved of them so much. Perhaps
the “unseemly songs” and the libation sacrifices in the banqueting room of
Uppsala were related to a fertility cult.169
Drinking customs in ceremonial buildings may be attested by archaeologi-
cal finds in the Mälaren area. Large amounts of glass have been discovered at
house group 2 at Helgö, for instance.170 No less than 96 percent of the 1,600
glass fragments found on the island were discovered at this house group.
Interestingly, there was a particular concentration in the ceremonial hall I A
at Foundation I, more specifically at the place interpreted as the “high-seat”.
The glass was clearly high-status and comprised, for instance, fragments of
filigree glass claw and cone beakers. It has been estimated that these glass frag-
ments belonged to at least 50 beakers and bowls of different types dated to the
period from the Roman Iron Age to the Viking Age. It cannot be excluded that
these beakers were used for ceremonial purposes, for instance in connection
with ritual banquets.171 The warrior hall discovered in Birka yielded finds simi-
lar to those in the hall at Helgö. In the western part of that house broken glass
from exclusive beakers was found. Most likely the high-seat was located at this
place, where the king sat during his visits to Birka (see ch. 6 above).
Prohibitions in normative Christian texts from eastern Scandinavia directed
against pagan customs and ideas may also give some support to the idea that
libations were essential at pre-Christian sacrificial feasts. In the medieval Guta
Law (in the section “Af blotan”) we read that invocations with food and drinking
were prohibited when they did not follow Christian customs.172 These pagan
customs may also be related to some information in the medieval Guta saga.
The Old Gutnish term mungat ‘ale, beer, feast’ refers to a specially-brewed ale
for times of celebration, such as religious feasts. Perhaps it was also drunk on
169 There is more evidence of this in Gotland, where smaller assemblies held lesser sacrifice
with cattle, food and drink. Guta saga, (ed.) Peel, 4f. In this text these ceremonies are
described with the Old Gutnish noun blotan (f.) ‘sacrifice’ (cf. ON blót ‘pagan worship,
sacrifice’). The one who performs such rituals is called sauþnautr ‘person with whom one
boils meat, comrade-in-sacrifice, boiling-companion’ (cf. siauþa ‘cook, boil’; cf. Gothic
sauþs ‘sacrifice’).
170 Lund Hansen 2011.
171 Holmqvist 1980, 56–65; T. Zachrisson 2004a; 2004b; 2010; Lund Hansen 2011.
172 Þet ier nu þy nest et blotir iru man num mier firi buþin Oc fyrnsca all þaim sum haiþnu
fylgir Engin ma haita a huathci a hult eþa hauga. eþa haþin guþ. huatki a vi eþa staf
garþa . . . nequara þa miþ mati eþa miþ dryckiu senni sum ai fylgir cristnum siþi. . . .
Gutalagen, af blotan, (ed.) Pipping: 6−7. On these texts, see T. Blomkvist 2002. On religious
drinking rituals, see Drobin 1991; Nordberg 2003, 175ff.
Gotland in connection with the ritual dedications to the deities, for a good
harvest, for the dead and so on, as described in Hákonar saga góða.
Sources narrate that ritual drinking also occurred in Iceland during the
calendrical feasts. Such rituals were often arranged by the local chieftains. Gísla
saga Súrssonar 15, for instance, reports that Þorgrímr decided to hold a feast at
the end of autumn to celebrate the coming of the winter-nights (at vetrnóttum
ok fagna vetri). During this feast sacrifices to Freyr were made (blóta Frey) and
there was a lot of drink to be served (drykkja skyldi vera . . .). At the same time
Gísli held a great feast at his farm, with copious drinking.
173 Bændr segja, at þeir vilja, at konungr blóti til árs þeim ok friðar, svá sem faðir hans gerði.
Heimskringla I, Ísl. Fornr. 26, 170.
174 . . . skyldi blóta eptir vanða enna fyrri konunga ok fylla svá en fornu lǫg til árs ok friðar. Ísl.
Fornr. 29, 80.
175 See Grønbech 1997 (1912).
176 Wessén 1924, 177ff. See also Baetke 1942, 3–39; F. Ström 1954; 1985; de Vries 1956–57, §§ 149,
184, 292, 461.
This is the beginning of our Law, we shall turn to the east and pray to the
Holy Christ [that he may procure] good crops and peace.185
In the Frostathings-Lov, fast days are dedicated “for prosperity and peace and
for good health to all men”.186 This instance produces solid evidence that the
formula was used in Trøndelag during the transition period. Other evidence
from the transition period in Norway indicates that ár ok friðr was ancient and
taken over by the Church. Glælognskviða, composed by Þórarinn loftunga in
the first decades of the eleventh century, says of St Óláfr:
hann of getr
af goði sjǫlfum
ár ok frið
ǫllum mǫnnum.187
The ancient expression has most probably been set in a new Christian con-
text by the skald. The priest and skald, Sigvatr Þórðarson, designates King Óláfr
jǫfurr ársæll in the poem Knútsdrápa 3.188 This stanza was quoted in Ágrip 31.
The scribe of Ágrip noted that when King Sveinn and his mother Alfífa took
over the reign of Trondheim, famine and misery prevailed: aldregi var ár á
þeira dǫgum.189
The expression ár ok friðr also appears in medieval sources that derive from
eastern Scandinavia: Gutalagen, which was barely influenced by Old Norse
tradition, says: “he [God] grants us prosperity and peace, victory and health”
(hann unni os ar oc friþ sigr oc hailsu)—and it is probable that Christianity in
Gotland adopted the pre-Christian formula.190
The expression ár ok friðr was thus probably known in all three investigation
areas.191 Most likely it has a pre-Christian origin but during the transition
period was taken over by the Church. According to Hákonar saga góða, this
185 Þat er upphaf laga varra at ver scolom luta austr oc biðia til hins helga Crist árs oc friðar.
NGL 1, 3.
186 . . . til árs oc friðar oc til heillsu allum monnum . . . NGL 1, 141.
187 “He, by himself, receives from God, good year and peace for all men.” Skj B1, 301. My trans.
This is actually the earliest attestation of the formula in the Old Norse traditions.
188 Skj B1, 232.
189 Ágrip, Ísl. Fornr. 29, 31.
190 Gutalagen,(ed.) Pipping, 3.
191 It is superfluous to mention that the formula ár ok friðr was known in medieval Iceland.
All references to it in the Old Norse sources suggest this.
formula was recited in connection with the ritual toasts during the calendrical
feasts in the ceremonial building. In this text it appears that the king himself
had to perform some of the drinking rituals and perhaps also recite this for-
mula. The king had to sacrifice in order to produce “good harvest and peace”
for his people. It seems as if the concept of ár ‘good year’s crop’ is often associ-
ated with rulers, at least in Sweden and in Norway.192 Sources suggest that the
pagan ruler, by means of sacrifices, was able to produce such good conditions.
The formula was thus an important ritual means for creating authority in cer-
emonial buildings.
The sources thus indicate that rulers held sacrificial feasts including ritual
drinking in the halls in all three investigation areas. The sources also suggest
that the rulers during these banquets had the opportunity to demonstrate
their wealth and to display their generosity in a concrete sense. According
to Hákonar saga góða 14, for instance, Sigurðr Hlaðajarl is described as very
generous to the farmers, since he paid all the expenses for the cultic feast by
himself. Snorri supports his statements here with a stanza taken from Kormákr
Ǫgmundarson’s Sigurðardrápa (960 AD). In this stanza it seems as if the feast
was held at a sanctuary (vé), which was supervised by Earl Sigurðr. It is intro-
duced thus: “You do not need to bring with you/ neither basket nor tankard/ to
the generous man.”
In the Sagas of Icelanders we see a similar attitude, where chieftains invited
their subjects to the calendrical ceremonial feasts in order to show their
generosity and splendour. Gísla saga Súrssonar 15, for instance, reports that
Þorgrímr decided to hold a feast at the end of autumn to celebrate the coming
of the winter-nights. During this feast sacrifices were made to Freyr and the
hall was decorated with tapestries, while the floor was strewn with rushes. The
text mentions that Þorgrímr invited his brother Bǫrkr and Eyjólfr Þorðarson to
this feast, but also “many other men of distinction”. There were one hundred
and twenty men in all and a lot of drink was served.
It thus seems as if wealth and generosity were key concepts for the Viking
Age rulers when constructing authority. In the Sagas of Icelanders, individual
Norwegian-Icelandic chieftains are therefore often described as auðmaðr ‘a
rich man’ (cf. auðigr maðr), rausnarmaðr ‘a generous man’ or stórlátrmaðr
‘a generous man’.193 Distributing gifts and holding great banquets were thus
important means for rulers to tie subjects to themselves all over Scandinavia.
The notions of this gift-giving system are expressed in the Eddic lay Hávamál
41–42, for instance, where we read:194
The historian Jón Viðar Sigurðsson has studied this system in relation to chief-
tains in Iceland. He argues that gifts were used to establish or renew ties of
friendship or loyalty. It was a reciprocal redistribution system. One could only
accept a gift if one was able to give another gift in return. Otherwise one did
not have the right to own or use the gift. Jón Viðar concludes: “Because of
the strong obligation to reciprocate, gifts were a good instrument by which the
various types of chieftain could bind supporters to them.”196
In Iceland the gifts consisted mainly of weapons, horses, oxen, jewellery,
clothes and fabrics.197 The feasts in the hall, however, served virtually the same
purpose as the gifts. In many cases the chieftains held great feasts and invited
193 Laxdæla saga, for instance, says about the chieftain Ingjaldr Sauðeyjargoði: “He was a
wealthy man who liked to throw his weight around” (. . . hann var auðigr maðr ok mikill
fyrir sér). Laxdæla saga 14, Ísl. Fornr. 5, 28. Þórólfr Mostrarskegg, according to Eyrbyggja
saga, ran his farm in fine style (as a rich and generous man) and had a large number of
men with him (Þórólfr gerðisk rausnarmaðr mikill í búi ok hafði fjǫlmennt með sér, . . .).
Eyrbyggja saga 4, Ísl. Fornr. 4, 10. Þorsteinn Ingimundarson from Hof was generous to his
neighbours with the goods from his estates (Þorsteinn frá Hofi var stórlátr af búi sínu við
heraðsmenn; . . .) Vatnsdæla saga 31, Ísl. Fornr. 8, 84.
194 The classical study of gift-giving systems in archaic societies was written by Marcel Mauss
(1997). Mauss introduced his study by quoting Hávamál. For such a system in Iceland,
see Jón Viðar Sigurðsson 1999. This system could perhaps also be related to the so called
patron-and-client-relationship (often called clientelism) mentioned in ch. 5 above, and
more known from the classical areas, see e.g. Jackson 2014b.
195 “With weapons and gifts friends should gladden/ one another, that is most obvious;/
mutual givers and receivers are friends for longest,/ if the friendship is going to work
at all./ To his friends a man should be a friend/ and repay gifts with gifts; . . .” Trans.
Larrington.
196 Jón Viðar Sigurðsson 1999, 91; 2010, 25ff.; 2011, 77f.
197 Jón Viðar Sigurðsson 1999, 92; 2010, 25ff.; 2011, 77f.
a lot of people. At the end of these feasts they sometimes distributed gifts to
the participants, in order to enhance their reputation. Thus both feasts and
gifts were means to construct authority and bind people to certain interests of
the rulers. Such gift-giving could also be practised as a kind of payment to loyal
subjects and could be performed in ritual forms in the hall. Egils saga 55 nar-
rates that after helping King Aðalsteinn to defeat King Óláfr (rauði), Earl Aðils
and the Scots in England, Egill Skalla-Grímsson came to the hall of the king
when he was sitting and drinking, and celebrating his victory. The king took a
gold ring from his arm, put it on the point of his sword, went to the floor and
stretched the sword over the fire to Egill. This episode is supported by a stanza
made by Egill himself (see further ch. 11). In the Anglo-Saxon poem Beowulf,
the rulers distributed rich gifts to their retainers in the hall during the feasts.
Sometimes the lord was therefore called OE bēag-gifa (-gyfa) “ring-giver”. He
is also called OE manna mildust “the most generous of men”.198 Occasionally
the noble females also took part in these gift-giving ceremonies (see below).
Among the Franks too, the rulers were expected to distribute gifts from the
booty. This is attested by an account recorded by Gregory of Tours. When
Clovis was still a pagan he plundered churches, and all the objects from the
booty were later distributed among his retainers in Soissons.199
These references indicate old systems of redistribution in several parts
of ancient Europe, where the ruler tied loyal people to himself by means of
gifts. There are no equivalent sources from eastern Scandinavia describing
the gifts and drinking ceremonies in the hall as exhaustively as in Egils saga
and Beowulf. However, skaldic poems, runic inscriptions and archaeological
finds may give some indications. In Ynglingatal the Svea king Vanlandi is called
menglǫtuðr ‘spendthrift of jewellery’ or ‘generous person’, while King Agni is
called gœðingr ‘magnate’ or ‘a man with a lot of property’. It is possible that
such epithets were ascribed to rich rulers who had manifested their wealth in
their halls and generously distributed gifts to their retainers. It must be men-
tioned, though, that Ynglingatal was composed by a Norwegian skald. These
attributes of kings may thus have nothing to do with eastern Scandinavian
conditions. However, runic inscriptions from eastern Scandinavia also indi-
cate that rulers had to be generous and care for their retainers. According
to Santesson’s interpretation of the seventh-century runic inscription from
Stentoften, Haþuwulfr arranged a great sacrificial feast, perhaps with plenty
of meat for all the people (see above). This event was extraordinary in several
senses and was therefore carved into the runic monument. Similar attitudes
also appear in Viking Age inscriptions from the Mälaren area.200 The Turinge
stone inscription (Sö 338) tells us that the chieftain Thorstein and his brother
“treated their retainers well” (heldu sina huskarla vel). According to this
inscription Thorstein was a landmaðr, i.e. some kind of chieftain who also had
a lot of landed property.201 In the inscription on the Väppeby stone (U 703),
at Veckholm, the dead man is praised for being generous with food and elo-
quent (mandr matar goðr ok malsrisinn). Most likely he was eloquent when
he appeared in his hall. Similar information appears in the inscription on the
Gådi stone (U 739) where Holmbjǫrn praises himself for being “generous with
food and eloquent”. Qualities of this type are also praised on the runic stone
from Hagstugan (Sö 130) where the magnate Domare is honoured for being
mildan orða ok matar goðan.
The sources thus indicate that the rulers of the Mälaren region—chieftains
or kings—during the banquets had the opportunity to demonstrate their gen-
erosity and to create alliances and friendship, by giving their men food and
gifts, probably in ritualized forms in the hall. Evidence of this system may also
be seen in a late Viking Age runic inscription on a bone discovered in the block
called Trädgårdsmästaren in Sigtuna:
The king is generous with food. He is the richest [man]. He is full of favour.
This inscription may have been carved by coincidence on a bone when Mari
took part in a banquet organized by the king in one of the halls discovered in
200 See Jansson 1985, 63f., 131ff.; 1987, 158ff., 127; Herschend 1994, 187ff.
201 Palm 2004, 239.
202 Gustavson et al. 1992, 166. My translation based on Gustavson.
Sigtuna.203 Since the bone is dated to around 1100 it is possible that the term
kunungr refers to King Ingi the Elder. It thus seems as if Christian kings and
magnates also held great banquets in halls during the transition period. The
bone from Sigtuna is an indication of such events.
In Hákonar saga góða no females appear as cultic leaders during the sacrifi-
cial feasts. Other Germanic sources often state that noble women took part
in the ceremonial banquets performed in halls.204 The most informative text
on such ceremonies in the context of halls is the Old English poem Beowulf,
written sometimes between eighth and eleventh century.205 According to this
poem King Hroðgar owned a hall (OE heall or sele) called Heorot.206 In this
hall the retainers and other guests were seated according to social rank on
the long benches. It was the assignment of Queen Wealhtheow to pour mead
into the participants’ cups.207 This too was done in ritual forms and according
to rank. By means of the formalized drinking ceremonies Wealhtheow creates
a social balance between the king and the retainers in the hall. She is therefore
called freoðuwebbe ‘weaver of peace’.208 In Beowulf it appears that the queen, in
addition to her role in the cup-offering ritual, also gave gifts and advice to the
retainers in more or less ceremonial forms in the hall. For instance, she gave
Beowulf a necklace, a coat of mail, and much advice.209 These ritual gifts may
have been part of an old redistribution system. By means of the gifts, the queen
and king won loyal men, manpower and raw materials. Beowulf thus reflects
ceremonies in the warrior hall, which may be connected to the ruler strategies
for gaining legitimacy. This text should not, perhaps, be treated as a historical
source, but it probably reflects a mentality which was common among aristo-
crats and comitatus retinues in the Late Iron Age society of northern Europe
and in the Early Middle Ages (see further ch. 15 below). Similar ceremonies
are also described in medieval Old Norse texts. Orkneyinga saga 86 (dated to
c. 1200) mentions that Earl Rǫgnvaldr of Orkney sailed to France and came
to a seaport which was named Nerbón (Narbonne). At this place a noble woman
called Queen Ermingerðr was ruling over people. She and her kinfolk invited
the earl and the best of his retainers to a splendid banquet (til veizlu virðili
grar). “There was lavish entertainment at the feast, and nothing that might be
done in his honour was denied the earl. The earl was sitting feasting one day
when the queen came into the hall escorted by a group of ladies and carrying
a serving-bowl. She was in her finest clothes, with her hair falling loose as is
customary with virgins, and a golden tiara upon her forehead. She served the
earl, while her companions began to entertain them with music.”210 This scene,
205 On the dating of Beowulf, see e.g. Klaeber 1950, cvii–cxiii.; Wrenn & Bolton 1996; Mitchell
& Robinson 1998.
206 Beowulf 1015; 1640ff. The text is edited by Klaeber 1950. On Beowulf and its connection to
Late Iron Age Scandinavia, see Bo Gräslund’s (2015 MS) interesting forthcoming study.
207 Beowulf 612ff. On this text, see e.g. Enright 1988; 1996. On her role as gift-giver, see Beowulf
1215ff.
208 Beowulf 1942. See e.g. Enright 1988; 1996, 21f. See also Klaeber 1950, 335.
209 Beowulf 1188–1231.
210 Ok er þeir kómu til veizlunnar, var þar inn bezti fagnaðr ok engi hlutr til sparaðr, sá at jarl
sins sómi væri þá meiri en áðr. Þat var einn dag, er jarl sat at veizlunni, at drottning gekk
inn í hǫllina ok margar konur með henni; hon hafði borðker í hendi af gulli. Hon var klædd
inum beztum klæðum, hafði laust hárit, sem meyjum er títt at hafa, ok hafði lagt gullhlað um
enni sér. Hon skenkti jarli, en meyjarnar léku fyrir þeim. Orkneyinga saga, Ísl. Fornr. 34, 210.
The text is based on the vellum Flateyjarbók codex written around 1390. Trans. Hermann
Pálsson & P. Edwards.
211 See also Orkneyinga saga, ch. 52, Ísl. Fornr. 34, 112, and Sundqvist 2007, 60ff. In this context
a quotation from Archbishop Hincmar of Reims in the ninth century must also be men-
tioned. In his De Ordine Palatii ch. 5 we read thus: “Für die rechte Ausstattung des Hofes
und besonderes für die Jahresgaben der Vasallen—abgesehen von der Versorgung mit
Speise und Trank und dem Unterhalt der Pferde—waren vornehmlich die Königin und
unter ihr der Kämmerer zuständig.” Quoted from A.-S. Gräslund 2006, 164.
212 See Fritzner 1954.
213 It thus seems as if the society was not only ruled by men. Women played roles as lead-
ers at several levels in society. See e.g. Jesch 1996, 28ff., 63ff., 193ff. Admittedly men con-
trolled the public space during the Viking Age, but in certain spheres women may have
performed as leaders, e.g. the public cult. There are also women who transgressed gender
boundaries in the sources. On the position of women in pre-Christian society, see e.g.,
Jesch 1996; B. Sawyer 2000; Kristoffersen 2000, 127ff.; A.-S. Gräslund 2001, 65–89; Bagge
2005, 21ff.; Ney 2004; Quinn 2005; Hultgård 2005; Sundqvist 2005a; 2007; 2015.
After this the Vǫlsi was passed from hand to hand, and everyone who received
it uttered a verse, often obscene, and always accompanied with the puzzling
refrain: þiggi Mǫrnir/þetta blæti. When King Óláfr came to the farm he threw
the Vǫlsi on the floor to the dog, and finally he succeeded in converting the
people living there.
The credibility of this text has been discussed.216 The narrative is put in a
frame which concerns St Óláfr and his efforts to Christianize the Norwegians.
The description of the old religion may thus have been deliberately made
bizarre. But there are some elements in the story which seem to be very ancient.
For instance, the magically charged words “linen” and “onion” (linalaukar) are
carved on a bone knife found in a female grave discovered at Fløksand, Norway,
and dated to the Late Roman Iron Age. This inscription has been associated
with the ritual treatment of the horse penis by the housewife in Vǫlsa þáttr.217
There are other elements in this text which may reflect ancient customs. When
the king throws the Vǫlsi on the floor, the housewife performs an interesting
ritual. As she is chanting a verse, she also heaves herself above the frame and
ridge of a door (hefi mik of hjarra ok of hurðása) to see if she can rescue the
sacred sacrificial object (ef ek borgit fæ blætinu helga) (i.e. the Vǫlsi). According
to Ibn Fadlān (tenth century), a possible Scandinavian slave-girl performs a
similar ritual during a chieftain burial. She had voluntarily accepted to fol-
low her master to the Other World. Before she was ritually killed, according
to Ibn Fadlān, she did this: “they brought the slave-girl to a thing that they had
constructed, like a door-frame. She placed her feet on the hands of the men
and was raised above that door-frame. She said something and they brought
her down . . . I quizzed the interpreter about her actions and he said, “The first
time they lifted her, she said, ‘Behold, I see my [dead] father and my [dead]
mother.’ ”218 By means of this ritual she had the ability to see into the Other
World, where her parents were already dwelling. There are thus probably
traces of old genuine folklore and rites in this story, which may have survived
until the conversion period.219
E. O. G. Turville-Petre suggested that Vǫlsi and Mǫrnir (also called vingull
and beytill) are one and the same, namely the emblem of Freyr.220 The phallus
was not only the emblem, however, but also the embodiment of the god. Vǫlsi-
Mǫrnir was thus, according to Turville-Petre, both the sacrificial victim and
the recipient of the sacrifice. He relates this idea of sacrifice to Hávamál 138
and the words of Óðinn: “wounded with a spear and given myself to myself”.
Since Mǫrnir in other poetic contexts refers to ‘sword’, i.e. a natural phallic
symbol and perhaps a representation of Freyr, this interpretation is plausible.
It should be noted, however, that Mǫrnir is female plural of Mǫrn, and that
the name could refer to the partner(s) of the fertility god, namely the fertility
goddess(es) or the giantess(es).221 This interpretation would also make sense
of the phrase: þiggi Mǫrnir/þetta blæti. The ritual in Vǫlsa þáttr thus symbol-
ized a hieros gamos, where Vǫlsi represented Freyr, and Mǫrnir his female cov-
enanters. In cultic contexts Mǫrnir could be represented by the housewife and
other female participants.
The good farmer Holmgautr had this [stone] raised in memory of his wife
Oðindisa. A better housewife will never come to Hassmyra to run the
farm. Rauð-Ballir carved these runes. She was a good sister to Sigmund.224
222 It should be noted, however, that the ON húsfreyja and RSw hīfrøyja have different first
elements hús ‘house’ and OSw hi or hæ ‘family’. Both words (húsfreyja and hīfrøyja) are
nevertheless usually translated ‘housewife’.
223 See e.g., Jesch 1994.
224 Text and interpretation may be found in Vs 24, Jansson 1964, 69–76. Trans. to English in
Jesch 1996, 65.
The information that the housewife “runs” or “rules over” the farm [the village
or settlement] (byi [boi] raðr) is important.225 It indicates that Oðindisa (Ōðin-
Dīsa) had a leading position in a more public sphere, perhaps also outside the
walls of the building. The Runic Swedish verb rāða may be interpreted as ‘rule
over’.226 The text also says that Oðindisa’s husband was “a good farmer” (boandi
goðr). This expression is intended to show that Oðindisa was married into a
good family. The information that she was a sister of Sigmund also intended
to indicate her high position in the local community and that she originally
derived from an influential family.
There is slight evidence indicating that the housewife of Hassmyra was
involved in cultic matters. The compound name Oðindisa is only attested in
this inscription. In a formal sense the iō-stem dis should not appear with a
weak inflection in this case.227 According to the onomastic specialists Roland
Otterbjörk and Thorsten Andersson this may have a simple explanation.228 The
female name was probably originally Disa.229 The name of the god Oðin- has
been added to her secondarily, as a byname prefix.230 The name Oðindisa indi-
cates that she had a specific relation to the god Óðinn. Perhaps it even suggests
that in her youth she appeared in ritual functions in the cult of Óðinn. At least
she was faithful to the old Æsir gods during the conversion period.
Old Norse sources indicate that other housewives in pagan Svetjud were
involved in religious rituals at sanctuaries, especially during the calendrical
feasts. According to Austrfararvísur (c. 1020),231 quoted by Snorri Sturluson in
225 The runologist Thorgunn Snædal writes to me in a e-mail dated 20 November 2006 that the
u-rune in this sequence actually not is stung, as Jansson states. She proposes that the pas-
sage should be read bui raþr and normalized to boi ræðr, i.e. that Oðindisa ruled over the
farm, the animals, the people, the stores, the crew etc. Cf. ON búsforráð ‘management of
household affairs’ Cleasby & Vigfusson 1957 (1874), 86. On ON bú, see Fritzner 1954, Vol. I,
204ff. See also Peterson (1994, 8), however, who states that the sequence could be read byi
and interpreted as a dative of a RSw byr ‘gård, by’.
226 Peterson 1994, 43.
227 Peterson 1981, 149.
228 Otterbjörk 1983, 110; T. Andersson 1992a, 512; 1993, 45f.; Peterson 2007, 171. I would like to
thank the runologist Jan Axelson for these references.
229 OSw dis means ‘woman, mythical female being’; cf. ON dís.
230 According to Thorsten Andersson, Oðindisa belongs to a group of “Theophore
Personennamen”: “Da die schwache Form des Namens auffallend ist, erscheint die
Erklärung ansprechend, daß es sich hier um den Namen Disa mit dem Götternamen
als sekundärem Beinamenpräfix handle.” T. Andersson 1992c, 512; 1993, 45. Cf. Vikstrand
2009.
231 Skj. B1, 220–225. For an English translation, see Page 1995, 48ff.
his Óláfs saga helga 91 (c. 1230),232 the Christian skald Sigvatr Þorðarson, with
some companions, made a journey to Svetjud at the beginning of the elev-
enth century. One evening they came to a farm called Hof. Whether Hof in this
case refers to a place-name is uncertain. It is possible that it should be inter-
preted as an appellation hof, i.e. a building where cultic activities took place.
The poem says “the door was shut” (hurð vas aptr). It also says that Sigvatr and
his fellows were sent away, since the people there declared that it was “holy”
(heilagr) there. The text mentions that they went to another farm. There was a
housewife (húsfreyja) in the doorway: “No farther can you enter, you wretch!”,
she said, “we are heathen people, and I fear the wrath of Óðinn” (hræðumk ek
við Óðins reiði). The poem also mentions that “she was holding álfablót there in
the house”, i.e. she was preparing sacrifices to the mythical beings called álfar.
According to some scholars, the calendrical feast called álfablót was identical
with the pre-Christian Yule (ON jól).233 It seems as if the Yule feast had some
connections to the cult of Óðinn.234 The Æsir-god is called Jólnir (or Jóli) in
Þórsdrápa st. 12 and in Ágrip ch. 1.235 Also the poetic kenning Hugins jól ‘the
(Yule) feast of Huginn’, must in one or another way be related to Óðinn or more
exactly to his raven. During such feasts ceremonial drinking took place inside
the ceremonial buildings. Most likely, however, álfablót, was a religious feast
celebrated during the autumn, separate from the Yule feast, which took place
at midwinter or the winter solstice. It is possible that housewives had leading
cultic roles at such feasts and also gained social prestige for this.
Egill and his men arrived on the island, Gyða arranged an expensive banquet,
with plenty of drinking (var þar drykkja mikil ok dýrlig veizla). During the eve-
ning Gyða approached Egill. It is in precisely this context Gyða is designated
húsfreyja. A couple of days later she arranged a new banquet at her farm. Egill
and his men went into the hall room (í stofu) and started drinking. There were
many people and a great feast (var þar þá mikil veizla).
Egils saga thus reports that a housewife could be involved in public ban-
quets at halls in Norway. This housewife appeared as the arranger of the feast,
and perhaps she also performed as some kind of ceremonial leader when she
approached Egill and was designated húsfreyja. It is very hard to estimate the
source value of Egils saga when it comes to rituals taking place in the Viking
Age. The prose must be considered as indirect secondary materials, but some
of the stanzas on which the prose partly is based may have been composed by
Egill himself, sometime during the first half of the tenth century. They could
thus be considered as direct or contemporary sources. There are several stanzas
said to be made by Egill in chapter 64, and one mentions also the boy Friðgeirr.
No stanzas support, however, the scene where Friðgeirr’s mother invited Egill
to the ceremonial feast and the ritual drinking. It is thus uncertain whether the
information about the ritual feast was based on an old oral tradition.
237 Laxdæla saga, Ísl. Fornr. 5, 6–15. In Íslendingabók and Landnámabók she is called Auðr in
djúpúðga. On this tradition, see also Herschend 1998, 137ff.; Hedeager 2011, 118ff.
238 Íslendingabók, Ísl. Fornr. 1, 6, 26; Landnámabók, S 12–14, 94–110, H 12–14, 81–84, Ísl. Fornr. 1,
50f., 136ff.
239 Íslendingabók, Ísl. Fornr. 1, 26.
240 Clunies Ross (1998, 147ff.) emphasizes that certain rituals in these contexts could be spe-
cifically for females.
Unnr then entered the hall, followed by a large group of people. When
the hall was filled, everyone was impressed by the magnificence of the
feast.242
Unnr was very old when the wedding feast was held, and therefore she
announced that her inheritance should be handed over to her grandson Óláfr
when she was dead. During the feast Unnr died as well and the feast then con-
tinued, in commemoration of both Óláfr’s marriage and Unnr’s death. Óláfr
now owned the farm at Hvammr and its property. When the feast concluded
he gave generous parting gifts to his most respected guests.243
It is quite obvious that Unnr was the one who previously distributed gifts at
banquets at Hvammr. In the version of Auðr’s (Unnr’s) death in Landnámabók
it is clear that she also had this role.244 In this text no wedding feast is men-
tioned. It just says that she invited her kinsmen to an expensive feast (bjó dýr
liga veizlu). When the feast had been going for three nights she distributed gifts
to her friends (valði hon gjafir vinum sínum) and gave them good advice (réð
þeim helræði). She said that the feast must continue for three more nights, but
that this feast was going to be her “inheritance feast” (erfi).
241
Laxdæla saga, Ísl. Fornr. 5, 9; Landnámabók, S 97, Ísl. Fornr. 1, 139.
242
Síðan gekk Unnr inn í skála ok sveit mikil með henni. Ok er skálinn var alskipaðr, fannsk mǫn
num mikit um, hversu veizla sú var skǫrulig. Laxdæla saga, Ísl. Fornr. 5, 12.
243
. . . gefr Óláfr stórmannligar gjafar þeim mǫnnum, er þar váru mest virðir, áðr á brott fóru.
Laxdæla saga, Ísl. Fornr. 5, 13.
244
Landnámabók, S 110, Ísl. Fornr. 1, 146f.
According to the sources, Unnr thus acted as a male magnate during the cer-
emonial feasts in the hall buildings. She arranged and invited people to these
feasts, she entered the hall in some kind of ritual procession, and also took
part in the drinking ceremonies. Unnr also distributed gifts to loyal people and
friends during these assemblies. Unnr was thus some kind of leader in the local
community, and therefore she also appeared in these roles at public gatherings.
dressed in fine clothing made of materials of the highest quality. She had a rich
set of jewellery and a knife with a silver-decorated sheath. In addition she had
two oval brooches and one equal-armed brooch. Most interesting, in the light
of the present study, was the small pendant of bronze. It was made in the
shape of a woman holding a beaker in front of her. This type of pendant is usu-
ally called “valkyria” and they have been found at several places in the Mälaren
Valley. They indicate a ceremonial role of females, i.e. the important function
of pouring wine and beer for the guests during the ceremonial meals. It is
quite possible that the noble lady in the boat grave had this ritual function in
the Early Viking Age cult of Uppsala. The Merovingian Period hall of Uppsala
was situated just a few hundred metres from this grave. Ceremonial banquets
probably took place in this hall. Most likely another hall was erected in the
vicinity during the Viking Age, where the lady in the boat grave perhaps played
a role during the ceremonial feasts.
At Valsgärde, about 2–3 kilometres to the north of Old Uppsala, several
rich Viking Age burials of females were found. They included some expensive
glass and metal vessels. Perhaps some of these vessels were used at ceremonial
banquets.253 A Late Iron Age hall building was situated just beside the burial
field at Valsgärde.254 It is quite possible that the women buried in these graves
had a similar role to that of Wealhtheow in Beowulf during the ceremonial ban-
quets. A chamber grave (Bj 523) in Birka, dated to the tenth century, may lead
us to a similar conclusion. In this grave a female was buried with two drinking
horns with silver mounts.255 At Birka too we have evidence of a Viking Age hall
where ceremonial banquets took place.
Grave 85 at Valsgärde contained the cremated remains of an old woman, one
horse, two dogs and a cock as well as a pig jaw. In her grave there was also a staff
with a bone animal head. This animal-headed staff reminds us of the animal
heads at Oseberg in Norway, where two females were buried. In the Oseberg
grave, several Viking Age beast sculptures were found. It has been argued that
they had a religious function.256 It is quite possible that the animal-headed
staff in Valsgärde 85 also had a religious function.257 Most likely the lady there
was a powerful woman. Probably she also played important roles in the public
cult of Valsgärde. Anne-Sofie Gräslund writes thus: “Concerning the woman in
Vgde 85, in all probability a husfru, my impression is that her religious power
was strong and thereby, of course, also her social status.”258 Gräslund supports
her assumption with the animal-headed staff, but also with the bones of the
various animals found there. They may have had a symbolic meaning in rela-
tion to the burial.
Female ceremonial leaders may also be found at Tuna in Badelunda,
Västmanland.259 At this place rich grave goods were found in a Roman Iron
Age chamber grave, such as a necklet, two arm-rings and one finger-ring made
out of gold. There were also two bronze vessels of Hemmoor type and glass.
In this chamber grave a noble woman was buried. Perhaps the glass and the
vessels were used for serving wine in ritual contexts. At the same place eight
Merovingian Period and Viking Age boat graves were discovered, all of them
containing females. Close to the burial field there is a place called Närlunda.260
The first element in this name is a genitive form of the name of the deity
*Njärd. Is it possible that the females in Tuna were related to Närlunda and
that they were the cultic leaders in the public worship of *Njärd?
Archaeological evidence from Norway indicates that rich females may
have been involved in cult and used ritual strategies to gain legitimacy during
the Late Iron Age. Besides the rich Oseberg ship grave, the great mounds at the
farms of Hauge and Tu in Rogaland may also be related to aristocratic ladies
who performed cultic functions during religious assemblies.261
When the authors described the inner parts of the hof sanctuaries in the Sagas
of Icelanders, they often commented on the “oath-rings”, which were placed
on the altars (stalli, stallr sg.).1 In these texts such rings were often called stal‑
lahringr, hringr or baugr. Certain rituals were performed in connection with
them, which seem to be related to the sacrifice, but also to judicial matters.
Oaths, for instance, were sworn at them. Such oaths were called baugeiðr or
lǫgeiðr. Sources indicate that these rings also had other connotations for the
Early Icelanders. It seems as if they sometimes were seen as symbols of honour
and dignity related to the chieftains and cultic leaders called goðar.2 When
appearing in public functions at sacrifices or judicial assemblies, the goði had
to carry the ring in his hand or on his arm.
In opposition to some previous researchers,3 I will argue that the testimony
from the saga authors regarding these ritual rings is built on old oral traditions.
For there are clear indications in more direct sources that such rings served this
cultic and judicial purpose. I will also argue that this context of ritual rings was
not restricted to Iceland, but can be seen in sources deriving from other parts
of the investigation areas, particularly Svetjud. In this area too we can see ritual
rings, which seem to appear in the context of pre-Christian sanctuaries and
ceremonial buildings. In my opinion these rings were not only important reli-
gious and judicial objects there, but also essential media or symbols of status
and sovereignty. I will start with Forsa, Hälsingland, where a very interesting
iron ring with a runic inscription has been preserved. Much of the discussion
in this chapter will be concentrated on this ring, since, in my opinion, it is one
1 The concept of “oath-ring” derives originally from C. J. Thomsen [& Petersen] 1886. Thomsen
designated all types of Bronze Age rings as “oath-rings”. We do not know what kind of func-
tion these Bronze Age rings actually had. In the present study the term “oath-ring” refers
to such rings as had a religious or judicial significance at assembly places and sanctuaries
during the Late Iron Age. On this concept, see also Beck 1986, 538ff.; Jacob-Friesen 1986; Brink
1996b, 46; Habbe 2005, 134–144.
2 In some contexts and for certain people these rings provided agency to the carrier in a more
direct way, without being perceived as symbols. Cf. e.g., Miller 2005 and Engelke 2012.
3 See e.g., O. Olsen 1966; Kabell 1975.
The ring from Forsa in Hälsingland, Sweden, is forged and decorated with
runes (see fig. 17a). It is about 43 cm in diameter, and has for a long time been
in the possession of the church in Forsa. When it was first observed in schol-
arly contexts in the eighteenth century it was nailed to the door between the
porch and the interior of the old church which was demolished in 1840. Today
the ring is kept in the Forsa congregation archives. For a long time there was
a close consensus among scholars that the runic inscription on this ring was a
medieval clerical legal enactment from the twelfth century. Recently, however,
it has been argued that the ring should not be looked upon as a clerical
record.4 It has been stated that the content of the inscription is a Viking Age
enactment, more precisely from the ninth century.5 The function of the ring has
also been reinterpreted. In previous research it was stated that the ring should
be seen as a “church-door ring” (Sw kyrkdörrsring). Today scholars argue that
the Forsa ring was an “oath-ring”, i.e. a pre-Christian ring with ritual or symbolic
value. It was used by the people of Hälsingland when swearing oaths during
the common judicial assembly. This ring was kept at an important cultic site
in Hälsingland.
In what follows I will discuss the dating and the function of the ring. I will
argue that those scholars who believe that this ring is old and had a ritual or
symbolic function in a pre-Christian religious-judicial context are on the right
track. In connection with this I will discuss the shape of the ring.
4 Liestøl 1979; Ruthström 1990; Brink 1996b; 2010; Widmark 1999; Källström 2007; 2010a;
Sundqvist 2007; Williams 2008.
5 Brink 1996b. Liestøl (1979) holds the door open for the possibility that the ring should be
dated to the tenth century.
6 E.g., Bugge 1877; von Friesen 1933; Johnsen 1968.
Figure 17a
The ring from Forsa in Hälsingland,
Sweden.
Photo: Bengt A Lundberg. ATA,
Riksantikvarieämbetet.
Figure 17b
Detail from the ring from Forsa.
Photo: Bengt A Lundberg. ATA,
Riksantikvarieämbetet.
text of the inscription, which has often been quoted. It was based on Sophus
Bugge’s reading from 1877.7 Gerhard Hafström’s interpretation in the article
about the Forsa ring in Kulturhistoriskt lexikon för nordisk medeltid (1959) has
been normative and is very close to Wessén’s interpretation:
En tvegill oxe och två öre till biskopsstaven för första mässfallet;
två oxar och fyra öre till biskopsstaven för andra mässfallet;
The most important argument for those who stated that the inscription had a
medieval-clerical background was provided by the two Runic Swedish terms
lærðir and staf, which appeared in Wessén’s normalized version of the inscrip-
tion. The first word was interpreted as ‘a learned person’, i.e. ‘a priest’, while the
other term was a metonymic expression for ‘the bishop’ or the ‘the office of
the bishop’. The “bishop-staff/rod” was a symbol of the office of the bishop
during the Middle Ages.9 The ring should thus be related to a medieval context
and concerned a clerical-judicial record. Some scholars felt that the inscription
was about fines for omissions inside the church, or as Hafström interpreted it,
about fines when no service was held (Sw mässfall). These fines were paid to
the bishop. The enactment on the ring thus reflected a phase when the Church
was organized in Hälsingland.
Recently some scholars have argued that the medieval dating of the inscrip-
tion is disputable, specifically if the forms of the runes and the language of the
inscription are taken into consideration.10 The runes appearing on the ring are
of the type called “short-twig runes” or “stub runes” (Sw kortkvistrunor).11 They
usually occur in older runic inscriptions, such as the one on the Rök stone,
Östergötland (Ög 136) and the Tjängvide stone from Alskog, Gotland (G 110),
which are both dated to the Early Viking Age.12 Most runic inscriptions with
“short-twig runes” are from the ninth or the tenth centuries. In this context
the Forsa ring has been explained as an odd phenomenon, an unusual late
inscription with “short-twig runes”. The form of the runes is thus an obstacle
to those who date the Forsa ring to the twelfth century. A similar problem per-
tains to some word forms in the inscription. The verb fāði ‘cut, painted’, for
instance, and the particle svāð ‘so that, if’ are archaic terms, which normally
13 See further Holmbäck & Wessén 1979 (1933–46) III, lvif. See also Peterson 1994, 11, 67. It
should be mentioned that the verb RSw fā appears in other runic inscription from elev-
enth-century Hälsingland; Hs 14 Malsta; Hs 21 Jättendals kyrka.
14 Wessén 1940, lvi; Liestøl 1979. On the problems with the sound value of the os-rune and its
use as a criterion for dating runic inscriptions, see Williams 1990.
15 Brink 1996b.
16 See Liestøl 1979.
17 Svað liuðir æigi at liuðretti. Liestøl 1979, 17ff.
18 Liestøl 1979, 22.
Figure 18
The door ring from Delsbo church, Hälsingland, Sweden.
Photo: Unknown. ATA, Riksantikvarieämbetet.
1990. He argued that it must be read <a>t vis kil[t]an. This could be normal-
ized and interpreted as at vis gildan “for the restoration of a vi [cultic site] in
a valid state” (“för att återställa vi [kultplatsen] i gillt skick”).19 The enactment
thus concerned the treatment of the pre-Christian cultic site called RSw vī
‘sanctuary’. He interpreted the word gildan as a verbal noun ‘restoration in a
valid state’ (“återställande i gillt skick”). Ruthström’s new interpretation was
important in several respects. It made it possible to show what type of crime
the scheme of fines in the inscription concerned, namely the mismanagement
of a cultic site.20 By means of this interpretation it was also possible to omit
the problematic term “tvegill” which previous researchers had identified in the
inscription. This term is not attested in the Scandinavian languages.
With support from Liestøl’s and Ruthström’s new interpretations, Stefan
Brink argued in an article from 1996 that the inscription on the Forsa ring must
go back to the ninth century.21 According to Brink, it concerned a pre-Christian
legal enactment for the attendance of a cultic site. The designation of cultic
sites OSw vi (ON vé) was most likely well known in Hälsingland during the
pre-Christian period. There are at least five place-names from this province
where this element appears.22 At such a place this ring could have had its real
Sitz im Leben. Brink interpreted the entire inscription in this way:
: uksatuiskilanaukauratuâstafatfurstalaki :
uksatuâaukaurafiurataþrulaki :
: inatþriþialakiuksafiuraukauratastaf :
aukaltaikuiuarRif anhafskakiritfuriR
: suaþliuþiRakuatliuþritisuauasintfuraukhalkat :
inþaRkirþusikþitanunrâtarstaþum :
aukufakRâhiurtstaþum :
inuibiurnfaþi :
One ox and two aura [in fine] [to?] staf [or] aura staf [in fine] for the
restoration of a cult site (vi) in a valid state for the first time; two oxen and
four aura for the second time; but for the third time four oxen and eight
aura; and all property in suspension, if he does not make right. That the
people are entitled to demand, according to the law of the people that
was decreed and ratified before. But they made [the ring, the statement
or?], Anund from Tåsta and Ofeg from Hjortsta. But Vibjörn carved.23
Brink felt that the term liuðrettr ‘the law of the people’ referred to the law of
the province of Hälsingland, which during this period only consisted of some
common customs, legal rules and enactments detached from their context, as
we may see in the inscription on the Forsa ring. He argued further that the ring
had its real function at the provincial assembly (Sw landsting) at Kungshögen
in Hög parish, which probably was the political and religious central place of
Hälsingland during the Viking Age and the Middle Ages.24 At this place this
ring was probably a symbol of power and perhaps an “oath-ring” comparable
with the ones seen in the Old Norse sources. People thus swore their holy oaths
on this ring at the assembly. He also argued that it was Anundr and Ofegh who
made this ring and formulated the legal enactment inscribed on this object.
They were some kind of judicial and cultic leaders.
One problem remains when interpreting the inscriptions as pre-Christian,
namely the term staf (perhaps a dative, without any sign of the case inflection
‑i). Previously the runologists interpreted it as “to the bishop”. The term Old
Swedish staf (staffr, staver; cf. ON stafr) is present in the pre-Christian vocabu-
lary and designates there a cultic object located at the cultic site. It appears, for
instance, in the place-name Nälsta (< *Niærdharstaver) ‘the staff of the deity
*Niærdh’. Such cultic staffs are also attested in the reliable written sources.
Ibn Fadlān, for instance, mentions the Rūs (probably Svear), who sacrificed at
some kind of staffs or posts which were placed in the ground (see ch. 7 above).
Brink suggests that the staf in the inscription on the Forsa ring may be inter-
preted in a metonymic sense, as referring to the one who used the staff as a
sign of dignity, i.e. the chieftain or the leader responsible for the cultic site.25
There are, however, no other examples in the runic inscriptions or Old Norse
poetry where RSw stafr or ON stafr in a figurative sense refers to a leader of
society or cult. The meaning of staf in this inscription is quite enigmatic. It can-
not be excluded, however, that it should be related to a ritual object appearing
at the cultic site.26
Forsa ring was nailed to the door of Hög Church before it was transferred to the Forsa
Church. Brink 1996b, 40. See also Hafström 1959.
25 Brink 1996b, 37. Ruthström (1990, 49ff.) argues that staf refers to the fence which was set
up around the sanctuary.
26 On a cultic stav in place-names, see Vikstrand 2001, 292ff.
27 Widmark (1999, 123) emphasizes that both vocabulary and syntax are archaic in this
inscription and could very well be dated to the ninth century. She suggests that the three
vertical lines which introduce the inscription could be interpreted as i-runes, with a
magic intention. It could also support a pre-Christian background for the inscription and
the ring. Magnus Källström (2010a) has rejected this interpretation and argued that these
signs only tell where the reading should start. See also Bugge 1877, 7.
before in this discussion. In what follows I will present some of them, which
concern the shape and appearance of the Forsa ring.
The Forsa ring has at one side a three-branched figure, which at first glance
may be taken as a Christian cross or a lily (see fig. 17b). But it could also be inter-
preted in other ways. There are two runic stones in Södermanland with icono-
graphic motifs which remind us of the figure on the Forsa ring. On the Åby
stone (Sö 86) and the Stenkvista stone (Sö 111) from the settlement of Rekarne
(Rekarnebygden) there are large hammers depicted in the middle above the
runic coil (see fig. 5). According to the historian of religions Anders Hultgård,
these motifs may directly be associated with the pre-Christian religion and the
hammer of the deity Þórr, called Mjǫllnir.28 The motifs on the runic stones dif-
fer to some degree from the figure on the Forsa ring. The images of the stones
are more T-shaped than three-branched. Other sources from the investigation
areas, however, indicate that Þórr’s hammers were sometimes three-branched
at the top of the handle. The assumed Þórr figure from Eyrarland, Iceland, for
instance, holds a three-branched object in his hand which is usually inter-
preted as Þórr’s hammer (see fig. 15 above).29 The hammer of the figure on the
Altuna stone, Uppland, also seems to be three-branched (see fig. 6). There is
great consensus among researchers that this figure represents the god Þórr. On
this stone he is struggling against the Miðgarðsormr. This motif is known from
literary sources.30 Similar three-branched Þórr’s hammer symbols appear on
a ring from Hilda, Innvik, Norway, a Þórr’s hammer made of silver from
Läby, Uppland, one from Mickels, Gotland, and intimated on several silver
pendants with Þórr’s hammers (see fig. 19a and 19b).31 In the Mälaren area,
particularly, small Þórr’s hammers made of iron are common, of similar form,
sometimes placed on larger rings, known as Þórr’s hammer rings.
When the Forsa ring is observed together with the assumed hammer as a
whole, a very characteristic form emerges. It remind us undeniably of the fire-
steel-shaped pendants which have been found at several places in Scandinavia,
for instance in Borg, Östergötland (see fig. 20).32 These pendants, according to
the archaeologists, have a form symbolism which belongs to the Viking Age.33
They often appear together with the miniatures representing Þórr’s hammer.
28 Hultgård 1992, 89ff. On the hammer as a symbolic expression of ancient Scandinavian
religion, see A.-S. Gräslund 1983–84; 1985; Hultgård 1999.
29 But see also Perkins 2001.
30 See e.g., Hymiskvida and Gylfaginning 48. See also Meulengracht Sørensen 1986.
31 For a thorough survey of the Þórr’s hammer symbols, see Perkins 2001, 82–149; A.-S.
Gräslund 2001, 55ff.
32 Liestøl 1979, 21ff.
33 On the dating of these objects, see Jonsson 2006.
Figure 19a
Þórr’s hammer made of silver from Laby, Uppland. SHM
Figure 19b
Þórr’s hammer from Mickels, Nar, Gotland. SHM
Figure 20
Fire-steel-shaped pendants from Borg, Östergötland. SHM
The relationship between Þórr, the hammer and the fire-steel is in my opinion
quite natural: Þórr used his hammer Mjǫllnir when making thunder and fire.34
According to the Old Norse traditions, Þórr’s hammer had also other func-
tions. It was used, for instance, when consecrating or protecting various phe-
nomena and objects. In one Eddic poem the hammer consecrates a wedding
couple; it was placed in the lap of the bride.35 According to Gylfaginning, the
hammer was also used at the funeral pyre of Baldr: “Then Þórr stood by and con-
secrated the pyre with Mjǫllnir.”36 The “may Þórr consecrate” formula appears
in several Viking Age runic inscriptions.37 On the Glavendrup stone (DR 209)
from Funen (tenth century), for instance, it says thus: “may Þórr c onsecrate
these runes.”38 Another version of this formula may be seen on the Jursta stone
(Sö 140) from Södermanland: “may Þórr protect [this monument].”39 On this
stone a symbol of the hammer is also depicted. When the symbol of the ham-
mer appears on the Forsa ring and the runic stones from Södermanland, they
probably have the function of protecting the holy objects (the rings and the
monuments) as well as the runic inscriptions.
It is actually no big surprise that a Þórr’s hammer symbol appears associ-
ated with a ring. There is often a relationship between Þórr and “oath-rings” in
several Sagas of Icelanders.40 It thus seems as if Þórr was related to the swear-
ing of oaths, in a similar way to the god Ullr.41 With support from Irish annals
it has been stated that a Viking king in Dublin had a ring which he called
“Þórr’s ring” ( fail Tomair). It was captured by King Maelseachlainn in the
year 994.42
In my opinion, the fire-steel shape and the Þórr’s hammer image indicate
that the Forsa ring had a Viking Age background and that it played an essen-
tial role in the cult of Þórr. There are additional arguments from the history of
religions which may corroborate that the Forsa ring was a pre-Christian cultic
object. In what follows, I will try to shed new light on the religious and ideo-
logical function of the Forsa ring by focusing on the context where this ring
and other similar ritual rings may have appeared. I will particularly focus on
the ritual space where the rings may have appeared.
It was Stefan Brink who first suggested that the Forsa ring originally was an
ancient “oath-ring” which was kept at the general assembly and cultic site
of Hög in Hälsingland.43 When making this interpretation he supported his
argumentation with information from the Sagas of Icelanders. In these sagas a
certain type of sanctuaries are described, called hof. In connection with them
there were officials known as goðar. During the sacrificial feasts these goðar
appeared as some kind of cultic leaders. One of the sacred attributes used by
these officials was a ring (stallahringr, hringr or baugr), which the goði either
carried in his hand or wore on his arm during the religious and judicial meet-
ings. Otherwise this ring was kept on the platform or altar (stalli, stallr) of the
hof sanctuary. The ring was sometimes used during the sacrificial rituals. It was
reddened in the blood of the animal which had been sacrificed at the sanctu-
ary. This ring was also used in judicial contexts. Oaths, for instance, were sworn
on this ring. They were called baugeiðr or lǫgeiðr.
The most important description of “oath-rings” may be found in the
Úlfljótslǫg (the text is paraphrased in ch. 5 above). According to this text, a
ring weighing two ounces or more had to lie on the ‘stalli’ in every chief’s hof
building, and every chieftain (goði) had to wear this ring on his arm at all the
assemblies he headed, having first reddened it in the blood of a neat which he
himself had sacrificed there. Every man who was there to transact any busi-
ness, as by law provided by the Court, first had to take an oath on that ring and
name for the purpose two or more witness in evidence. He was to say:
I take oath upon the ring, a lawful one so help me Freyr and Njǫrðr and
the Almighty God, to this end that I shall in this case prosecute or defend
or bear witness or give award or pronounce doom according to what I
know to be most right and most true and most lawful, and that I will deal
lawfully with all such matters in law as I have to deal with while I am at
this assembly.44
altar or platform in the middle of the room, called stalli. On this platform lay
an open ring,45 weighing twenty ounces,46 upon which people had to swear all
their oaths. It was also the business of the hofgoði to wear this ring on his arm
or hold it in his hand at every public meeting. Similar information appears in
Kjalnesinga saga 2, which mentions that a ring made of silver was placed on
the platform in Þorgrímr goði’s hof sanctuary at Kjalarnes. When there was an
assembly the hofgoði had this ring in his hand. People also swore oaths on this
ring at all litigations.47
Víga-Glúms saga reports that the chieftain (goði) Glúmr was supposed to
swear an oath upon a ring, affirming that he was innocent of the manslaughter
of Þorvaldr krókr.48 This had to be done at three hof sanctuaries in Eyjafjǫrðr.
When the men were assembling at Djúpadalr, six men entered the hof. Then
the saga says in general terms that whoever had to take the “hof-oath” placed
his hand on the silver ring, which was stained red with the blood of the cat-
tle sacrificed, and which was supposed to weigh not less than three ounces.49
Then Glúmr spoke and made his oath:
I name Ásgrimr to bear witness, and Gizurr in the second place to bear
witness that I take the “hof-oath”, on the ring, and I say it to the Æsir-god.
When Þorvaldr krókr got his death-blow, I was far away and did not take
45 That some symbolic (oath-)rings had to be open is attested elsewhere. A ring was dis-
covered at the small cultic house in Järrestad, in eastern Scania, which is located beside
a multifunctional hall (Söderberg 2005). This ring had a diameter of 9.5 cm. It was forged
and made of round-iron and it had a secondary made break into it. This feature could be
associated with the Icelandic “oath-rings” which could not be closed and which had to
be carried by the goðar during sacrificial ceremonies. It could also be associated with the
ritual restrictions which the Roman priest of Jupiter, the flamen dialis, had to follow. He
was not allowed to carry closed rings. Aulis Gellius, Noctes Atticae X, 15, 7. In this context
I must also refer to the interesting Pietroassa ring found in Romania and usually dated
to the fourth century. Those parts of the ring which have been preserved have a runic
inscription. According to the runologist Helmer Gustavson (2004), it says “Goternas odal
är helig”. In the nineteenth century this ring was stolen from the museum of Bucharest
and cut into pieces. There is a drawing of it from 1867 indicating that it was an open ring,
exactly like the Icelandic “oath-rings”.
46 . . . ok lá þar á hringr einn mótlauss, tvítøgeyringr . . . One ounce is approximately 27 grams.
The ring thus weighed around half a kilo.
47 Á þeim stalli skyldi liggja hringr mikill af silfri gerr; hann skyldi hofgoði hafa á hendi til allra
mannfunda; þar at skyldu allir menn eiða sverja um kennslamál ǫll. Kjalnesinga saga, Ísl.
Fornr. 14, 7.
48 Víga-Glúms saga 25, Ísl. Fornr. 9, 85f.
49 Sá maðr, er hofseið skyldi vinna, tók silfrbaug í hǫnd sér, þann er roðinn var í nauts blóði, þess
er til blóta væri haft, ok skyldi eigi minna standa en þrjá aura. Ísl. Fornr. 9, 85f.
part in that action, when the point and edge was reddened with blood.
Now let those men who are skilled in such matters, and who stand by,
look to my oath.50
The same oath was also sworn at the hof buildings of Gnúpafell and Þverá.
Since Glúmr committed perjury, he was later forced to leave his homeland at
Þverá.
Also interesting is the information in Droplaugarsona saga 6, which says
that the magnate (mikill maðr) Sveinungr concluded an oath at the “altar-
ring” together with two men, who probably were witnesses: Vann þá Sveinungr
eið at stallahring ok tveir menn með honum, . . .51 The compound expression
stallahringr indicates that the “oath-ring” had a special relationship to the
stalli/stallr, i.e. the platform or altar, which was situated in the hof sanctuary.
The expression vann eið at stallahring also appears in Eyrbyggja saga 16 in con-
nection with a goði.52
These descriptions in the Sagas of Icelanders have been debated. The Danish
archaeologist Olaf Olsen argued that they do not say anything at all about pre-
Christian religion and ancient customs in Iceland.53 Neither was Úlfljótslǫg,
according to him, based on ancient tradition. The oath formula, for instance,
with the expression “the Almighty God” (hinn almáttki áss), gave a Christian
impression.54 Aage Kabell argued in the same vein as Olsen. He rejected the
historical value of the descriptions of the oath-rings in the sagas and claimed
that they were made up by imaginative medieval scribes.55 Kabell also felt that
the other texts referring to oath-rings in some way were corrupt.
This distrustful attitude towards the historical source value of the Sagas
of Icelanders and Landnámabók has been a “leitmotiv” in twentieth-century
saga research. In principle all information about pre-Christian religion has
been rejected by some scholars. In my opinion, this critical attitude has been
pushed too far,56 because there is information in the more direct sources
which may confirm some of the information in the sagas. There are, for
50 . . . ek nefni Ásgrím í vætti, annan Gizur í þat vætti, at ek vinn hofseið at baugi, ok segi ek þat
Æsi, at ek vark at þar ok vák at þar ok rauðk at þar odd ok egg, er Þorvaldr krókr fekk bana.
Líti nú á eið, þeir er spekimenn eru ok við eru staddir.
51 Droplaugarsona saga ch. 6, Ísl. Fornr. 11, 153.
52 Eyrbyggja saga ch. 16, Ísl. Fornr. 4, 30.
53 O. Olsen 1966.
54 O. Olsen 1966, 49.
55 Kabell 1975, 33.
56 Several scholars have objected to the hyper-critical position in research, see e.g., Brink
1996b, 49f.; Jón Hnefill Aðalsteinsson 1998; Jón Viðar Sigurðsson 1999, 35ff.; 2011.
instance, some passages in the Eddic poetry where the custom of swearing
oaths on rings may be confirmed. Hávamál 110 mentions a baugeiðr ‘ring-
oath’, and Atlaqviða in grœnlenzka 30 states that oaths were sworn on Ullr’s
ring (eiða opt um svarða . . . oc at hringi Ullar). There is also a note in the
Anglo-Saxon Chronicle dated to 876 which is of great interest in this con-
text. It states that western Scandinavian Vikings swore oaths on a sacred
ring and promised that they would not continue to ravage in the realm of
King Alfred.57 This source must be regarded as historically reliable.
According to some reliable Old Norse texts, oaths were also sworn on other
sacred objects, such as the edge of the shields, but also on the side of a ship
and the back of a horse. The Eddic poem Vǫlundarkviða st. 33, for instance,
states thus:
57 Her hiene bestæl se here into Werham Wesseaxna fierde, and wiþ þone here se cyning friþ
nam, and him þa aþas sworon on þam halgan beage, þe hie ær nanre þeode noldon, þæt
hie hrædlice of his rice foren. This text is quoted by Kabell 1975, 36. In the Latin transla-
tion of this text, made about 100 years later, the ring is called sacrum armillum. See the
Peterborough Manuscript 876 [875], The Anglo-Saxon Chronicle, (ed.) Swanton 1997.
58 “ ‘First you shall give me all these oaths,/ by the side of the ship and the rim of the shield,/
the back of the horse and the edge of the blade.” Trans. Larrington. See also Beck 1986,
537f.
59 See Habbe 2005, 141f.
60 “Þess strengi ek heit”, sagði hann, “at áðr alþingi er úti í sumar, skal ek hafa fullsekðat
Arngrím goða . . .”
assembly.61 Grágás records that oaths were sworn on the cross: “I name wit-
nesses to witness that I swear an oath on the cross, a lawful oath, and declare
before God . . .”62 These customs may be seen as rituals of continuity, which
had their origins in the pre-Christian context, but were accommodated after
Christianization to the new circumstances in Iceland.
There is thus good written evidence from Iceland that ritual rings appeared
in ceremonial buildings during the Viking Age and that different types of ritu-
als were related to them. But is it possible to relate “oath-rings” or other ritual
rings with ceremonial buildings also in eastern Scandinavia and particularly
in Hälsingland? According to Carl Löfving, there is nothing to indicate this.63
Löfving argues that traits of more specific cultic buildings at pagan sanctuaries
are not known in Scandinavia. The pre-Christian cult was performed outdoors
or in the chieftain’s hall. This is a problem for those who argue that the Forsa
ring should be dated to the Viking Age, according to Löfving, since it is impos-
sible to explain how the ring can have been preserved in such good shape dur-
ing this long period.64 In my opinion, there is much support for the theory
that the Late Iron Age halls found at several places in Scandinavia really were
regarded as protected religious spaces.65 It was in such spaces the ceremonial
banquets were celebrated. As noted above, archaeologists have shown that
these halls often have continuity over time, which indicates that they could
contain objects of symbolic value for a long period. At the assembly place of
Hög a large Iron Age building has recently been found, as well as an outdoor
sacrificial place dated to the Viking Age.66
We will now turn to some other Late Iron Age ceremonial buildings and
sanctuaries in the Mälaren area and see if it possible to find ritual rings at these
places. These buildings and places are mainly attested in the archaeological
record. The following survey is far from complete. I will only highlight some
61 . . . at ek vinn eið at bók, lǫgeið, ok segi ek þat guði . . . Ísl. Fornr. 12, 392.
62 . . . nefna ser vatta. Nefni ec i þar vætti. At ec vin eið at krossi log eið ok segi þat Guþi. . . . GrK
ch. 25. See also GrK ch. 41–42. Quoted from Habbe 2005, 144 and 241.
63 Löfving 2005, 113; 2010.
64 Stefan Brink (2010) has pointed out, on well-founded grounds, several flaws in Löfving’s
way of reasoning.
65 See Herschend 1993; 1997; 1998; 2009.
66 Personal communication from the archaeologist Katarina Eriksson in November 2011.
67 K. Ström 1970, 1. It should be noted that the amulets on these rings do not always look like
hammers.
68 Bäck et al. 2008, 45.
69 G. Andersson 2005, 63f.
70 Nielsen 1996, 98; Vikstrand 2001, 263.
71 For a complete investigation of fire-steel-shaped rings in the Mälaren region, see Jonsson
2006.
rings found at Borg had been forged together from two separate parts and had
thereby been shaped like fire-steels . . .”72 Some of them have pendants which
resemble Þórr’s hammers and some have miniature axes. This could also be
related to the decoration of the Forsa ring. Some of the rings have a diameter
of 8 centimetres.73 According to Landnámabók (Úlfljótslǫg), the “oath-ring”,
which appeared on the altar (stalli) in the hof sanctuary weighed at least two
eyrir, i.e. 54 grams. According to Víga-Glúms saga, “oath-rings” had to weigh at
least three eyrir, i.e. 81 grams. If this information is taken into consideration,
we cannot rule out that the small rings which appeared at the “altar” founda-
tion in the cultic house at Borg may have been some kind of oath or cultic rings
of a type similar to those referred to in the Old Norse literature.
buildings were most likely erected. The shape of the Þórr’s hammer ring of
Uppsala resembles in one sense the Forsa ring, with its three-branched figure.
But there are also some differences. The Þórr’s hammer ring of Uppsala was a
grave gift, while the Forsa ring has been preserved in other contexts. It is pos-
sible, however, that the Þórr’s hammer ring of Uppsala in a previous phase
played a ritual role in the context of the sanctuary of the site.
During the summer of 2012 excavations about 300 metres to the east of the
church discovered several hoards of amulet rings. Eighty amulet rings, includ-
ing Þórr’s hammer symbols, were found. These finds emphasize the religious
significance of this site.
10.3.5 Adelsö and Other Places with Ritual Rings in the Mälaren Area
There are several other places where ritual rings, fire-steels and pendants have
been found in the context of possible ceremonial buildings at Late Iron Age
settlements in the Mälaren area. Under the floor of one of the long-houses,
built on a foundation just beside the assumed banqueting hall at Adelsö,
Uppland, a miniature of a fire-steel was found.85 Close to the hearth in the
same house, three linked miniature sickles were found. These terraced build-
ings seem to have been in use from 700 to 900. The place appears to have been
a royal site, but some cultic activities may have also been performed there. One
fire-steel-shaped ring (approx. 7 × 8 cm), with three smaller fire-steel-shaped
rings hanging on it, was also found at Adelsö in 1942.86 Information about
the find context is missing, unfortunately. At the farm of Sanda (RAÄ 147) in
central Uppland, three Viking Age fire-steels and four amulet-rings were like-
wise found in the context of a hall and an outdoor ritual space area, includ-
ing an assumed equivalent to an ON hǫrgr.87 Most of the rings were found at
the assumed hǫrgr. This farm was definitely an important cultic place in the
local context during the Merovingian and Viking Periods. Late Iron Age amulet
rings have also been found in the context of settlements at several places in
Uppland and Södermanland.88 A Þórr’s hammer ring was discovered at a set-
tlement called Viby gård (RAÄ 308) in Sollentuna parish, Uppland.89 It is quite
unusual for such a ring to appear outside grave contexts. At Sanda in Vansö,
Södermanland, two large fire-steel-shaped “neck-rings” (approx. 16 cm in
diameter) were found. In one of them there were holes for hanging pendants.90
Some of the fire-steel-shaped rings found in the Mälaren area were made of sil-
ver, such as the ones discovered at Eketorp, Edsberg, Närke.91 Such silver rings
are more common on Gotland.
in Gästrikland, not far from the border with Hälsingland (fig. 21).95 This ring
was discovered in 1887 in connection with road construction. Since 1920 it has
been kept at the Länsmuseet in Gävle. The ring is made out of bent iron rod,
and it has a diameter of 30.8 cm. The ends are hook-shaped so they may grip
each other. The inside of the ring has hook-bent iron points, alternately turned
towards each other, alternately away from each other. One of them has a hole,
which may have made it possible to put a ring into it. The man who discovered
the ring remembers that he also saw other smaller rings beside the Häckelsäng
ring. He also remembers a hammer made out of iron weighing about 1 kg. This
is lost, however, but there is a drawing. According to the archaeologist Hanna
Rydh, the ring should probably be dated to the Viking Age, and most likely it
had a symbolic-religious significance. There are several Late Iron Age burial
fields at Häckelsäng and in Hamrånge parish. At Hamrånge church, there
is a place called Vi, which most likely is the assembly and cultic site of the
settlement.96 It is not impossible that the ring played an important role at this
site during the Viking Age. The situation at Hamrånge resembles the context of
the Forsa ring, i.e. a ritual ring which appears at a cultic place called Vi.
The survey of the archaeological find sites in the Mälaren area and its envi-
rons clearly indicates that ritual rings and fire-steels often appeared at places
which could be regarded as pre-Christian cultic sites or ceremonial buildings.
Some of the rings with a fire-steel shape could perhaps be related to the cult of
the god Þórr, for instance, the Häckelsäng ring and the Forsa ring. I agree with
Brink that the Forsa ring may have been kept at a site in Hög parish, perhaps at
Kungshögen, which was both a sanctuary and an assembly place. At this sacred
place the ring played an important role in ritual contexts, perhaps at sacri-
fices and when swearing oaths. In Norway too, Þórr’s hammer rings are quite
common. A ring with amulets, for instance, has been found at Hilda, Innvik.97
Viking Age golden rings were also found in the “temple treasure” at Hoen Øvre
Eiker, Buskerud.98 Other types of ritual rings appear in the context of sanctuar-
ies in Trøndelag (see ch. 11 below).
The sources thus indicate that ritual rings appeared at the cultic sites of all three
investigation areas, Iceland, Trøndelag, and Svetjud. I have previously argued
that pre-Christian rulers protected and took care of such sanctuaries. This
may be seen in reliable sources. Like the goðar in the Sagas of Icelanders, the
*gudhar in Svetjud may have acted with judicial functions (see ch. 5 above).
The persons mentioned in the inscription on the Forsa ring, Anundr in Tåsta
and Ofegh in Hjortsta, may very well have been a kind of Viking Age *gudhar
in Hälsingland, with both religious and judicial functions. In this sense these
officials may have had functions comparable to the goðar mentioned in the
Sagas of Icelanders. Maybe they carried the “oath-ring” in their hands during
religious ceremonies, as a sign of dignity and religious power. A picture stone
from Tängelgårda, Gotland, may display such cultic leaders/specialists, in a
procession carrying “oath-rings” in their hands (see fig. 22). It seems as if these
rings contributed to the construction of authority. But do we have any evi-
dence from the Mälaren area indicating that oaths were sworn on such rings in
the sanctuaries as the Icelandic sources report?
The custom of swearing oaths (ON sverja eið, selja eiða) is well attested from
all Germanic areas and cannot be conceived as a local Icelandic tradition.99
The origin of the word oath, ON eiðr (Goth aiþs; OHG eid; OE āð) is debated.100
According to Elof Hellquist, for instance, it is an old loan-word from OI ōeth,101
but if so it must have been borrowed by the Germanic languages at an early
Figure 22
A picture stone from Tängelgårda,
Gotland. SHM.
date.102 The verb ON sverja (OSw sværia; Goth swaran; OE swerian; OSax
swerian), which is connected with the oath (ON sverja eið; OE āðswyrd), also
seems to be old and spread in several Germanic languages.103
The custom of swearing oaths on ritual objects is also attested from all the
Germanic area.104 One source reports that the Saxons swore oaths on swords.105
In Annales Fuldenses an unknown writer says that in the year 873 the Danish
kings Siegfried and Halfdan sent an envoy to Emperor Leo (Ludwig) asking
for peace. They met at the River Eider. There they concluded peace by swear-
ing oaths on weapons.106 The passage is usually conceived as having a high
source value.107 When such oaths were taken, rings or ring swords were cus-
tomarily used as ritual objects. I have already referred to the information in the
Anglo-Saxon Chronicle about the Vikings who swore oaths on a holy ring in
876. Also in the ancient Greek world, oaths and other contractual rituals were
sworn on rings and other sacred objects. The gods were often evoked as wit-
nesses. Such rituals were often circumscribed by a complex of actions, includ-
ing animal sacrifices and libations (σπονδή sg.).108 Often the ring was broken
into two pieces. These pieces were then given to the partners of the contract.
102 Eddic poetry talks about the ‘oath-breaker, perjurer’ (eið-rofa). See e.g. Brot af Sigurðarqviðo
16. Helreið Brynhildar 5. In Hávamál st. 110 a baugeiðr ‘ring-oath’ is mentioned. Atlakviða in
grœnlenzka 30 narrates that an oath was sworn on Ullr’s ring.
103 Cf. H. Beck 1986.
104 H. Beck 1986; Frense 1982, 229ff.; Jacob Friesen 1986; Köbler 1986; Habbe 2005, 134ff.
105 Köbler 1986, 540.
106 Iurabant etiam iuxta ritum gentis suae per arma sua, . . . Annales Fuldenses, MGH,
Scriptorum, Tomus I, 386.
107 Frense 1982, 235.
108 Burkert 1985, 250ff.
It was called σύμβολον (symbolon) ‘sign’ and when used in plural ‘the halves
of a bone or a coin, which two persons broke between them, each keeping
one piece’.109 This word is related to the modern term “symbol”. The ritual oath
testimony was usually performed in the temple.110 In ancient Scandinavian
contexts too, it seems as if the ring was sometimes broken in pieces during
contractual rituals.111
Most likely the swearing of oaths on rings kept at sanctuaries was prac-
tised in Svetjud as well. The archaeological finds made at cultic sites such as
Borg, Helgö and Lilla Ullevi, as well as written sources, support such assump-
tion. The Swedish medieval provincial laws report that oaths were sworn at
the assembly place during the inauguration of kings.112 In connection with
these contractual rituals, certain oath formulae were recited. One possible
pre-Christian oath formula has been preserved in Äldre Västgötalagen.113 The
section Af mandrapi states: sva se mær guþ hol. This is conventionally trans-
lated “may God be gracious to me”.114 Leopold Läffler noted that in a Chris
tian context the word “God”, OSw guþ (nom. masc.), should take an adjective
hulder and not hull. In the form hull the adjective hulder (nom. sg.) appears
in three cases: (1) nom. sing. fem.; (2) nom. pl. neut.; (3) acc. pl. neut. The
third case must be rejected, as guþ is the subject. This word, guþ “God”, is
in general masculine. But when meaning “pagan god”—which often appears in
the plural—it is frequently attested in the neuter form.115 The first case may
therefore also be rejected. Äldre Västgötalagen’s sva se mær guþ hol should thus
109 Liddell & Scott 1888, 759f. See also Braarvig 1997; Gilhus & Mikaelsson 2003, 125.
110 Burkert 1985, 253.
111 The Eddic poem Rígsþula st. 23 mentions that the married couple Karl and Snør split
a ring, bauga deildu. According to Åke V. Ström this was a traditional wedding ritual.
Å. V. Ström 1975, 241. The ring perhaps appears in contractual rituals related to wedding
contexts in other Eddic poems, as well as in Kings’ Sagas. See e.g. Skírnismál st. 21–22;
Hávamál st. 110; Oddr Munk’s Óláfs saga Tryggvason ch. 29, Ísl. Fornr. 25, 226f.; and the
discussion in Steinsland 1991, 148–153.
112 The custom of swearing oaths (OSw eþæ swæriæ) is attested in medieval Swedish provin-
cial laws. In Äldre Västgötalagen (1225) it is stated in Rättlösebalken that the coming king
must swear to be faithful (trolekæn sværiæ) to all the Götar, and obey their law when he
comes to All-Göta Assembly. ÄvgL. R 1, SSGL 1, 36f. Upplandslagen (1296) mentions that
the people who followed the king during the ceremonial royal progress called Eriksgatan
must swear oaths (eþæ swæriæ) that they will be loyal to the coming king. UL Kg 1, SSGL 3,
52ff.
113 Läffler 1905.
114 ÄVgL, Af mandrapi ch. 1, §2. SSGL 1, 10f. Cf. Frense 1982, 256–259. Holmbäck & Wessén 1979
(1933–46) Vol. 5, 16–18.
115 Cf. Fritzner 1972, Vol. IV, 134.
It has been emphasized that the poetic expression “everything in ring is bound”
should be related to a religious-judicial context.119
In this chapter I have argued that ritual rings really appeared in Late Iron
Age sanctuaries and ceremonial buildings in both Iceland and Svetjud. Most
likely they appeared in similar contexts in Trøndelag too, although I have
only occasionally referred to that area in this chapter. These rings were used
by political leaders in ceremonial contexts, for instance when making sacri-
fices and in religious processions, but also during contractual rituals, such as
oath-swearing. These rings were thus essential media or symbols of status and
sovereignty. I have also argued that the Forsa ring from Hälsingland may have
played an important role in religious and ritual contexts as an “oath-ring”. The
small three-branched figure on the ring was interpreted as a Þórr’s hammer.
Together with the form of the runes and the language of the inscription, it
indicates that the ring had a Viking Age background. The fire-steel shape of
this ring likewise points in the same direction. Like the western Scandinavian
goðar, eastern Scandinavian cultic leaders or religious specialists may have
used such rings in ceremonial contexts, at sacrifices and when swearing oaths.
These rings were kept at the sanctuary or the ceremonial building, which was
protected by the cultic leaders. It is possible that the Forsa ring was kept at
such a place before it ended up on the door of Forsa church. Most likely some
religious specialists, cultic leaders and men skilled in law used it during the
pre-Christian assemblies at Hög as a sign of dignity. Perhaps it was Anundr of
Tåsta and Ofegh of Hjortsta who played this important role in the Viking Age
community of Hälsingland.
Sometimes objects of gold appear in the context of Late Iron Age ceremo-
nial buildings in some parts of the investigation areas, i.e. the Mälaren
region and Trøndelag. It has been argued that some of these objects reflect
mythical images or religious symbolism. Since gold also was an expression of
ruling power, we may assume that these objects may be related to a religious-
ideological strategy for gaining authority. Some of these objects only appear
during the Migration Period and the Merovingian Period and thus could not be
expected to be seen in Iceland. I will nevertheless argue that the golden objects
with a presumed religious meaning or symbolism may have been an important
element and medium related to some specific religious strategies which were
exclusive to the very upper stratum in Scandinavia, i.e. the noblest royal fami-
lies in Denmark, Norway and Sweden.
1 The most impressive find of gold foil figures has been made at Sorte Muld in Bornholm. More
than 2,700 gold foil figures have been discovered there since 1985. Watt 1991; 2004; 2007.
2 It has been suggested, however, that even if no gold foil figures were produced after 800 they
may have been used during the Viking Age. The investigations at Slöinge in Halland and Borg
in Lofoten indicate this. See Lundqvist 1996; 1997; 2003; Munch 2003a and 2003b; Carlie 2004,
174ff.; Watt 2007, 133.
Mälaren region and Trøndelag. Some of these foils may have been related to
a specific religious strategy of the highest-ranking royal families in Scandinavia
in these regions. They may signal that the ruling couple were closely related to
the divine sphere, i.e. they descended from the gods and/or mythical beings
(see ch. 3 above). Since they were often placed close to the high-seat, they also
indicated this intimate connection in a concrete and physical way.
In what follows, I will first present the distribution of the gold foil figures, their
find contexts and their appearance in the investigation area. Then I will take
up previous interpretations, and finally present my own opinion about them.
I will argue that some of these object with images represented scenes with
deities/mythical beings, such as Óðinn, Þórr, Freyr, Skaði and Gerðr. Some of
them may also represent scenes with more local mythical beings such as Hǫlgi
and Þorgerðr. In the context of ceremonial buildings these objects underlined
the very close genealogical relation which existed between the mythical world
and the high-ranking ruling families of the Mälaren region and Trøndelag.
6 The term Eriksgata refers to the ceremonial road that the medieval king travelled clock-
wise to each provincial assembly, from Uppland to Västergötland and back again. During
this journey he was acclaimed as king.
7 Nordén 1938; Herschend 1998, 184.
8 Nordén 1938, 228.
9 S. Zachrisson 1963 and 1998, 13f. Personal communication Torun Zachrisson.
10 See A.-S. Gräslund 1992.
11 See J. P. Lamm 2004, 65.
12 Gullmarsberg and Hög Edsten, Bohuslän; Slöinge, Halland; Gårdlösa, Ravlunda, and Vä,
Skåne; Bolmsö and Brahekyrkan (Visingsö), Småland; Eketorp, Öland. J. P. Lamm 2004.
Recently 29 gold foils were also found at Västra Vång, Blekinge.
13 J. P. Lamm 2004, 61f.
14 Lundqvist 1996.
15 See mainly J. P. Lamm 2004, 44ff.
Figure 23a Gold foil figures from Hauge-Tu, Rogaland. Permission from Siv Kristoffersen and
the Museum in Stavanger.
Figure 23b
Gold foil from Krokek, Östergötland. SHM.
buildings. Sometimes they are found at the place where the high-seat is
assumed to have been located.
Gold foil figures have also been found at a couple of places in Trøndelag and
northern Norway. At the remains of a Viking Age ceremonial building under
the church of Mære 19 (or 23) gold foils were found (fig. 8b above).16 Couples
were depicted on these plaques, with the motif of a man and a woman stand-
ing opposite each other. They varied somewhat in size, but most of them were
about 1 square centimetre. Eight were undamaged, while the rest were buckled
and with the impress indistinct. Of these, three can be abstracted as a distinct
group. They were probably impressed with the same stamp. The others vary,
but on all of them the man is shown with a short tunic and hair, whilst the
woman has a long tunic and long loose hair. On four of the images the couple
are holding a staff between them. On the majority the couples are surrounded
by a pearl border. According to Lidén, the gold foil figures from Mære should
be dated to the Viking Age. He based this dating on the clothes of the figures, as
well as on the context of the finds. Most scholars argue, however, that the foils
were made during the Migration and Merovingian Periods, but were perhaps
still in use during the Viking Age.17
Gold foil finds have been used at other places in Trøndelag as well as
elsewhere in Norway. Two gold foil figures (cut-out human figures) were
found at Hemne (Hevne), in Sør-Trøndelag in the early nineteenth century.
Unfortunately, the archaeological context of these finds is unclear.18
Gold foil figures have also been found along the Norwegian coastline, from
the south to the north.19 The most interesting parallel to Mære in the Norwegian
context is probably Borg at Vestervågøy.20 At this place five gold foil figures
were found in the 14-metre-long salr room in the hall-building (fig. 8c above).
All of them were found at a place interpreted as remains of the high-seat in the
northern corner, not far from an 8-metre-long “fireplace area”, which ended
in a round hearth to the north-east. Excavation of the northern post-hole (1)
revealed that the post had been replaced twice. One of the foils came from
the top of this layer. Two were found just outside it. They were probably also
connected with this post-hole. The last two figures were placed under a stone,
deep down in the last phase of the post-hole. They were believed to have been
placed there intentionally. All five gold foil figures depict the couple motif.
Two of them were probably made with the same die, while the others were
quite heterogeneous. One is also unique in that the couple is seen in semi-
profile. Three of them are very tiny, only 7–8 mm high. Gold foils have also been
found in southern Norway. On the farm of Hauge, Klepp, in Rogaland sixteen
foils were discovered (fig. 23a). Some foils were also found at Hov, in Vingrom,
Lillehammer.21 The ones at Hov were related to a possible cultic building.
personified the gods. The disproportionate size and almost grotesque styliza-
tion of the heads supports this interpretation.25
The historian of religions Gro Steinsland has tried to modify Olsen’s and
Lidén’s interpretations of the gold foil figures.26 She toned down the fertility
aspect of these foils and argued that the couple motif should be attached to
religious ruler ideology prevalent in Scandinavia during the Late Iron Age.27
According to Steinsland, the images represent the peculiar marriage between
a god and a giantess, also expressed in Viking Age skaldic poetry such as
Háleygjatal and Ynglingatal (actually elaborated more in Ynglinga saga).28 The
product of this odd marriage was a human son. This son represented the pro-
totypical ruler: the king or the earl. In Háleygjatal the divine couple was con-
stituted by Óðinn and Skaði. Their son was Sæmingr, i.e. the ancestor of the
Earls of Lade. In Ynglinga saga Freyr and Gerðr had a son called Fjǫlnir, who
was the first king in the Ynglinga dynasty. According to Steinsland, the motif
on the gold foil figures reflected this peculiar mythical marriage. The prototypi-
cal marriage myth behind these pictures is the one between Freyr and Gerðr,
found for instance in the Eddic poem Skírnismál. This myth also had ideologi-
cal implications for the rulers, since the royal kin originated from these mythi-
cal parents. Steinsland argues, further, that when the gold foil figures with this
motif were deposited at the high-seat of the ceremonial building in Mære the
intention was to emphasize the status of the building and the family who con-
trolled this house. Their function was to support the ruling power by means
of the hieros gamos myth.29 She cautiously suggests that these objects may
have been used in connection with the inauguration rituals of the ruler, but
the question of the ritual contexts of these finds mostly remains open.30
In an article from 1996, Hans-Emil Lidén maintained his interpretation of
the gold foil figures as sacrificial or votive objects in a fertility cult. He was
also critical of Steinsland’s ruler-ideological interpretation of these finds.
25 See also Back-Danielsson’s (1999, 2002) shamanistic interpretation. She has interpreted
the figures on the foils as masked shamans. Lotte Hedeager (1997a; 1997b; 2011, 126–134)
has also related the foils to seiðr and the category of shamanism. Such interpretations
in Scandinavian contexts are critically considered by Schjødt 2001 and Dillmann 2006.
Håkan Rydving (2011b) is critical of the transcultural and phenomenological concept of
shamanism in a general sense.
26 Steinsland 1990; 1991.
27 Steinsland 1990; 1991.
28 In a recent article Steinsland (2011b) has reconsidered her view on Ynglingatal (see
above).
29 Steinsland 1990; 1991; 2005, 415ff.
30 Steinsland 1991, 160; 2005, 415ff.
According to available sources, Mære was never a chieftain site in the pre-
Christian period. Neither the archaeological materials nor the written sources
indicate that. Most likely the chieftain farm of the district was located at Egge,
just outside Steinkjer, where we have a large burial field dating from the Early,
Middle and Late Iron Age. According to Lidén, the chieftains of Egge, in col-
laboration with other chieftains of Trøndelag, controlled the ceremonial build-
ing at Mære and organized the sacrificial feasts at this site. This cultic site of
Mære was thus an assembly place for many minor chieftains and local groups
without a ruler’s farm.31 In his book from 1999 he also stated that his earlier
interpretation of a high-seat in Mære must be rejected. Therefore Steinsland’s
theory could hardly suit the historical context of this site, according to Lidén.
The Norwegian archaeologist Gerd Stamsø Munch, in opposition to Lidén,
has supported Steinsland’s general interpretation, when discussing the gold foil
figures found at Borg in Lofoten. She added, however, that these foils should be
interpreted as house offerings.32 This ritual action was performed in order to
consecrate the room or this part of the building. She argued that gold foil fig-
ures with the couple motif usually appear at places which were both political
and cultic centres in the Late Iron Age, and followed Steinsland’s theory that
this motif on the foils was connected to ruler ideology. According to her, the
scene on the gold foil figures reflected the hieros gamos between Óðinn and
Skaði as seen in Háleygjatal, i.e. the union between the mythical parents of
the Earls of Lade.33 She also suggested that the chieftains of Borg originally
belonged to the Lade family, since they derived from Hálogaland. Therefore we
may find gold foil figures at the hall of Borg.
Recently it has been argued that the gesture language on these gold foil fig-
ures must also be taken into consideration when interpreting them.34 By relat-
ing the foils to late classical and medieval parallels in iconographic and literary
sources, the Danish archaeologist Margrethe Watt, for instance, suggests that
some figures indicate gestures with a legal or social background.35 The action
of one member of the couple putting his hands around the other person’s wrist
has parallels in the medieval literature and iconography. It refers to the judicial
act of binding or attaching someone to someone else, as in a marriage cer-
emony. It therefore seems reasonable to interpret these figures as iconographi-
cal reflections of marriage scenes. In that sense, Watt’s suggestion supports
Steinsland’s interpretation. Watt argues, however, that the figures should not
be related to any particular mythical being.36
36 “Uden at være helt sikker, er jeg tilbøjlig til at mene, at guldgubber med parfigurer skal
opfattes som ‘neutrale’, d.v.s. ikke nødvendigvis gengivende bestemte navngivne personer
fra den nordiske mytologi.” Watt 2007, 144.
37 See Hákonar saga góða 18, in Heimskringla I, Ísl. Fornr. 26, 172.
38 On kings and earls in Sparbyggvafylki, during the Early Viking Age. See e.g., Heimskringla
I, Ísl. Fornr. 26, 99, 164f.; Heimskringla II, Ísl. Fornr. 27, 241. See also Þórhaddr in 5.2.1 above.
corresponds quite well to the places where gold foil figures have been found
in this region. The Rällinge statuette and the Lunda figures, which most likely
represent Freyr, may also be connected to this area (see ch. 7 and 14). In my
opinion, Steinsland’s interpretation of the foils is thus quite sound, as long as
we relate these objects to this geographic area.39 The mythical wedding on
these foils may thus sometimes be related to Freyr and perhaps Gerðr.
It is possible, however, that the couple on the foils from Trøndelag rep-
resented another divine couple. Perhaps the couple on the foils from Mære
could be related to Óðinn and Skaði, as suggested by Gerd Stamsø Munch and
Gro Steinsland, since these mythical beings, according to Háleygjatal, were
regarded as the divine parents of the Lade Earls. One could also relate the cou-
ple depicted on the gold foil figures to Þorgerðr Hǫlgabrúðr and Hǫlgi, who
may have been considered as the original mythical ancestors of the noble kin
from Hálogaland. This assumption presupposes that the Lade Earls had some
influence on the sanctuary at Mære as early as the Early Viking Age. This idea
must be considered somewhat uncertain. If gold foil figures were still used at
the end of the ninth century this theory is at least possible.40 It is possible too
that in an earlier period there may have been some other noble ruling fami-
lies in Mære and Trøndelag, who regarded Freyr and Gerðr as their mythical
ancestors. Later Óðinn and Skaði were regarded as the mythical parents of the
Háleygja kin.
As noted above, most scholars argue that the gold foil figures could be
regarded as sacrificial objects, since they were found at places which may
be considered as cultic sites.41 Margrethe Watt has a different opinion. She
admits that these objects have been discovered in important settlement com-
plexes, where religious ceremonies also took place.42 However, in her opinion
they played a role as “temple money” in this context. In order to take part in
the religious feasts and ceremonies the participants would have had to pay
with these objects. In my opinion, this interpretation is somewhat doubtful.
At Helgö, for instance, some foils were discovered in post-holes. This indicates
rather that they were sacrificed or ritually deposited in connection with the
foundation of the house or that they had ideological implications. The gold foil
figure no. 1186 at Helgö, for instance, was found in one of the post-holes of the
39 There are some problems of source criticism connected to Steinsland’s theory, that Freyr
and Gerðr were regarded as the parents of the Ynglingar, see ch. 3 above.
40 Some scholars have actually argued that gold foil figures were still used in the Viking
Age. See Lundqvist 2003; Munch 2003a; Carlie 2004, 174ff.; Watt 2007, 133.
41 See e.g., M. Olsen 1909; Lidén 1969; Steinsland 1990; 1991.
42 Watt 2007. Critically considered by Ratke & Simek 2006.
hall.43 Two gold foil figures from Borg in Lofoten were likewise found in post-
holes.44 At the Migration Period hall of Gudme gold foils were found in the
post-holes of the roof-carriers,45 and at the Merovingian Period and Viking Age
hall (houses III and II) at Bergagård in Slöinge such figures were found in post-
holes located in the north-west corner, perhaps the place of the high-seat.46
The relationship between the foils and the presumed high-seat in the hall of
Helgö indicates that they had an ideological function for the ruler. The figures
most likely were not used as visible ruler attributes, since they are so tiny. The
figures there varied between 7 × 7 mm and 11 × 11 mm.47 These objects should
rather be interpreted as ritual and symbolic objects, intended to reinforce the
holiness of a certain place. In my opinion, they were placed at the high-seat in
the hall as a concrete statement of the close relation that existed between the
ruler and the divine world. The iconographic motif may also have emphasized
the divine origin of the ruler, his family (i.e. the genealogical relationship),
and the sacredness of his chair in the hall. The foils may thus have been used
in an inauguration ritual of the new ruler or in connection with the ritual con-
secration of his ceremonial building or high-seat.
In my opinion, gold foil figures found in the ceremonial buildings in the
Mälaren region and Trøndelag (including northern Norway) may thus be seen
as objects and symbols related to the noblest ruling families’ and royalties’
specific genealogical relation to the mythic beings.48 It is possible that this type
of religious strategy never existed among the chieftains of Iceland. Therefore
no gold foil figures have been found there. It is also possible, however, that
the distribution of these archaeological finds reflects temporal variations in
Scandinavian religious ruler strategies. The usage of gold foil figures in aristo-
cratic contexts may have disappeared in the middle of the Viking Age. Since
Iceland was first settled after 870, we therefore never find such objects there.
On the other hand, some scholars argue that gold foil figures were still used at
the end of the ninth century, even if the production of them ceased earlier.49
If these scholars are right, the interpretation that gold foil figures represent a
more specific ideological strategy is much more convincing.
Like the gold foil figures, the gold bracteates appear in both the Mälaren region
and Trøndelag, but not in Iceland.50 The bracteates are usually dated to the
Migration Period (c. 450–550), and one would thus not expect to find them
in Iceland. In my opinion these finds too may be related to the specific ruler-
ideological strategy that existed in Svetjud and Trøndelag (and in southern
Scandinavia). Sometimes they may display mythic motifs, which may reflect
the rulers’ close relation to the divine sphere. Like gold foil figures, these
objects may be found in relation to ceremonial buildings at the rulers’ sites.
The provinces around the Mälaren region are not particularly rich in finds
of gold bracteates. One exception is the hoard found at Söderby, in Danmark
parish, with four identical B-bracteates (IK 176), five identical D-bracteates
(IK 522) and the newly discovered large D-bracteate (IK 583).51 Three fragmen-
tary gold bracteates (IK 270, IK 271 and IK 399) have also been found at the
settlement of Helgö. One of them (IK 270) was very large with an original diam-
eter of roughly 90 mm. According to Morten Axboe, these bracteates can be
attributed to the fifth or the sixth century.52 Two of the Helgö bracteates were
discovered at Building Group 2, Foundation I, associated with the compact cul-
tic hall building IA. IK 270 lay by itself at some distance from the gold foil fig-
ures and the area where it has been suggested that the high-seat was situated.
IK 271, on the other hand, appeared at the place where most of the gold foils
were deposited. Since these bracteates are fragmentary it is almost impossible
to interpret the motifs on them with certainty. Axboe proposed that IK 271 may
be a Nordic C-bracteate, although he admits that this suggestion is conjectural.
Fragments of a third bracteate (IK 399) were found in Building Group 3. The
motif on them, when put together, can be identified as the neck, chest and
foreleg of a C- or F-bracteate animal.
The C-bracteates generally contain an image of a large human head above a
four-legged animal, often interpreted as a horse. Other animals, such as birds
and fish, may also be seen. It has been suggested that this type of image rep-
resents Óðinn in a mythical scene. Karl Hauck suggested that Óðinn here is
curing Baldr’s horse, as mentioned in the Second Merseburg Charm.53 Other
scholars have argued that these images represent Óðinn as a shaman riding on
50 More than 900 bracteates have been found in Sweden, Norway, Denmark, England and
Germany. See Hauck 1985–1986; 1999; Andrén 1991; Axboe 2007.
51 J. P. Lamm et al. 1999; Axboe 2004.
52 Axboe 2004; 2011.
53 Hauck 1985, 99ff., 106; 2002, 81ff.
his horse to the Other World, followed by his ravens, i.e. his helping spirits.54
It has been argued that the medallions had a specific religious-political func-
tion in society and were used by elite groups as expressions of power. The
images on them thus represent not only the god, but also the ruler who saw
himself as descending from the divinity. This mythic-cosmic notion may
have been inspired by ideas occurring in the Roman emperor cult.55 Together
with the archaeologist Anders Kaliff, I have argued that the motif on the
C-bracteates has striking similarities to the iconography associated with
the late classical and Roman god Mithras.56 The bracteates remind us of the
scene where Mithras kills the bull. This does not mean that it is Mithras rep-
resented on them. The bull-killing motif may merely have inspired the artist,
who transformed it into the Germanic style and world of ideas. This may be
why the four-legged animal resembles both an ox and a horse. Perhaps the
intention was to depict the known mythical motif of Óðinn and Sleipnir, or
Óðinn and Baldr’s horse, yet the prototype, the Mithraic scene, still exerted an
influence when it was made.
For the purposes of this work I will only stress that the motifs of the bracte-
ates found at Helgö may have comprised mythical themes. Perhaps the rulers
at Helgö wore these gold bracteates around their necks during the rituals when
sitting in the high-seat. These objects were more visible to the “audience” than
the small gold foil figures. The two bracteates from Helgö may thus have played
an ideological role for the rulers of this site during the Migration Period, per-
haps connected to a mythic-ritual strategy.
As far as I know, two bracteates have also been found in Trøndelag. One
C-bracteate (IK 231) was found in Dalum in Sør-Trøndelag. Another interest-
ing bracteate (IK 282) was found at Hov in Snåsa, Nord-Trøndelag. It is pos-
sible that this place may have been an ancient central place, with a ceremonial
building designated ON hof.57 There is also another type of finds discovered in
Trøndelag, which may be related to the iconography of the golden bracteates
and gold foil figures, namely the golden medallions.
the gold foil figures. It has been argued that this medallion is Germanic but
belongs to a certain group of objects where the intention of the artists was to
imitate Roman medallions from the end of the fourth century.58 There are in
total thirteen examples of these objects in Scandinavia and five were discov-
ered in Norway. A characteristic feature is that these medallions are decorated
on both sides.
The front of the Inderøy medallion displays a full-length portrait of a man
to the right and a woman to the left, turned towards each other in profile. Each
is holding one hand on a tree or a branch with twigs which is placed in between
them. The man is holding a ring in his left hand, while the woman is holding a
staff. The reverse of the medallion displays a male rider, who is also holding
a ring in his hand.
Kristen Møllenhus argued that this medallion probably was produced in
Trøndelag, although it was made as an imitation of Roman portraits of the
Emperor and the Empress.59 The tree between the couple represented “the tree
of life”. The wreath in the hand of the male was interpreted as “the wreath of
victory” belonging to the goddess Victoria, i.e. an attribute which is frequently
used on portraits of the Emperor. Møllenhus never managed to trace the
Roman original of this image, which is a weakness in his argumentation.
The Inderøy medallion is also unique since it depicts the couple as a full-length
portrait. On the Roman medallions only the head/bust of the Emperor is seen.
Gro Steinsland has opposed Møllenhus’s interpretation of the medallion.60
According to her, we must relate the couple motif on it to the motif seen on the
gold foil figures found at Mære. Both of them represent the hieros gamos motif,
known from pre-Christian Scandinavian myth, cult and rulership. Steinsland
admits that the medallion and the gold foil figures display differences in styles
and forms, and perhaps also differences when it comes to function. Anyhow,
the motif may have arisen from the same tradition of ideas.
In my opinion, we may assume that the couple on the Inderøy medallion
represent divinities. The style, the attributes, the poses and the fact that this
full-length portrait is placed on gold indicate this. It is difficult, however, to
relate them to any specific god or goddess/mythical beings. We do not know
whether Óðinn and Skaði, Freyr and Gerðr, or even Hǫlgi and Þorgerðr
Hǫlgabúðr were worshiped in Trøndelag during the fourth and fifth centuries.
Since the medallion was discovered in a grave it most likely had another func-
tion than the gold foil figures. The latter finds have often been unearthed at
Besides golden medallions, bracteates and gold foil figures there were other
golden objects in ancient Scandinavia which may have played a certain ideo-
logical function in the context of the uppermost elite and ceremonial build-
ings, namely the golden door rings of the hof sanctuaries. In the written sources
we hear about a golden ring placed on the door of a sanctuary in Trøndelag.
Snorri Sturluson mentions in his Óláfs saga Tryggvasonar that Earl Hákon had
a large golden ring on the door of the hof sanctuary of Lade. During his mission
in Norway King Óláfr Tryggvason visited this place: “Then he sailed into the
fjord to Lade and had the hof sanctuary there dismantled and all the property
and all the decorations removed from it and from the idol. He took a large gold
ring from the hof gates which Earl Hákon had had constructed, whereupon
he had the temple burned down.”62 It is also mentioned that King Óláfr had
this great gold ring sent to Queen Sigríðr of Svetjud, as an expensive gift. It
was discovered that the ring was not made out of gold and the queen felt that
she had been fooled. It has been argued that this ring story was a late addi-
tion to the tradition, since it does not appear in Oddr Snorrason’s Óláfs saga
Tryggvasonar, which is the major source for Snorri’s text.63
That ritual rings were placed on the doors of ceremonial buildings is, on
the other hand, intimated in other more reliable and direct sources. At the
Middle and Late Iron Age site of Uppåkra, outside Lund, a more specific cul-
tic building was found beside a possible hall building. The cultic house was
13 × 6 metres and probably rather high (fig. 4). Several finds indicating ritual
activities were discovered there.64 At this house an iron door ring measur-
ing 15 cm in diameter was found in a post-hole (see fig. 24a).65 It was deco-
rated with four beads (Sw vulster) and a staple made out of band-iron, which
embraced the ring. Another iron ring was found at the house, which was 23 cm
in diameter and also had four beads (see fig. 24b). Such beads are common on
Romanesque church doors. The church-door ring from Delsbo, Hälsingland,
had this kind of decoration (fig. 18).66 In general the Christian rings have only
three beads, symbolizing the Holy Trinity.67 There is some information in the
Old Norse literature about rings which were nailed to the doors of the pagan
ceremonial buildings.
That ritual rings were placed on the doors of specific buildings may also be
attested from iconographic sources. The Sparlösa stone (Vg 119), from ninth-
century Västergötland, is decorated with images and a runic inscription. On
one side of the stone, at the top, there is an image depicting a building, which
seems to be a hall or perhaps a cultic house (fig. 25). The most salient aspect
of this building is a door with a great ring, which seems to be nailed there.
Several scholars have argued that the door with the ring must have a ritual-
symbolic significance.68 The assumed hall with the ring arouses associations
with a stanza in the Eddic poem Rígsþula. When the god Rígr is out wandering
in stanza 26 he comes to a hall (at sal) where the substantial couple Faðir and
Moðir (i.e. the parents of Jarl) live. He walks towards the door at the southern
part of the building: var hurð hnigin,/ hringr var í gætti. Researchers have inter-
preted these bars in different ways. Ursula Dronke, for instance, suggests this:
“The door was taken down. The handle-ring was in the side-niche.”69 According
to her, Rígr did not have to knock before he entered the building, since the
door was taken down. He passed the ring, which was placed in the side-niche
instead of the door. It is also possible to interpret these bars in another sense.
The door could have been closed, but not locked (hnigin) and in the middle
of the door case (gætti) the ring was placed, i.e. in the middle of the door. The
use of the preposition í makes this interpretation possible.70
There is also information in the written sources that pre-Christian ritual
rings, which had been nailed to the doors of the hof sanctuaries during the
Viking Age, were taken down after Christianization and instead placed on
FigURE 25
The Sparlösa stone (Vg 119), from ninth-century
Västergötland. ATA, Riksantikvarieämbetet.
Photo: Bengt A Lundberg.
the church doors.71 It is attempting to assume that the Forsa ring was also origi-
nally placed on the door of a pagan building and after Christianization was
placed on the church door of Hög or Forsa. The problem is, however, that the
door opening in which the Forsa ring is placed is a secondary construction
(see ch. 10 above). The finds at Uppåkra and the image of the Sparlösa stone
nevertheless indicate that ritual rings appeared at the doors of pre-Christian
sacred buildings in Norway (as narrated by Snorri in Óláfs saga Tryggvasonar)
and elsewhere. When these rings were made out of gold, they may have been
symbols of royal dignity or earldom.
Perhaps Earl Hákon’s golden ring at the sanctuary of Lade should be related
to some other symbols of dignity found elsewhere in Scandinavia, namely
the golden arm- and neck-rings, which may appear in the context of ceremo-
nial buildings and royal persons. Karl Hauck stated about sixty years ago that
such golden rings (Lat torques, monile, diadema, corona, annulus and armilla)
were signs of honour among almost all Germanic tribes.72 Written sources
report that neck-rings were used in royal contexts in Svetjud and elsewhere.
Ynglingatal 12 (9), for example, mentions that King Agni was hung by his
golden neck-ring by his own wife:
Þás gæðing
með gullmeni
Loga dís
at lopti hóf . . .73
Karl Hauck is probably on the right track when he states that the Old Norse
term gullmen ‘neck-ring of gold’ in Ynglingatal refers to a royal attribute.
Another royal attribute related to the Ynglinga dynasty was King Aðils’
golden ring (gullhringr) called Svíagríss. Snorra Edda (Skáldskaparmál 44) tells
how Hrólfr kraki set fire to a hall when he visited his sister Yrsa at Uppsala.
Yrsa, who was married to Aðils, came with a horn full of gold and Aðils’ ring
Svíagríss and gave it all to her brother Hrólfr. The Danes went to their horses
and rode down to Fýrisvellir. Aðils and his men followed Hrólfr and the retain-
ers on their horses. To make them stop, Hrólfr spread out the gold he had stolen
on the road so that Aðils’ men had to pick it up. Hrólfr threw the ring on the
ground; when King Aðils bent to pick it up, Hrólfr said:
Snorri says: “This is the reason why gold is called the seed of Kraki or of
Fýrisvellir”,75 and he refers to a stanza from Eyvindr skáldaspillir (lausavísa st. 8):
73 “. . . when ‘the wedding dis’ [Skjalf]/ raised aloft/ the ruler [King Agni]/ with the golden
neck-ring”. Heimskringla I, Ísl. Fornr. 26, 38. My trans. Cf. SkP I, 22–25.
74 “Now I have made him the most powerful of the Svear to make a Swine-bend.”
Skáldskaparmál 44.
75 Af þessi sǫk er gull kallat sáð Kraka eða Fýrisvalla. Skáldskaparmál 44. Cf. Skáldskaparmál
32 on kennings for gold, and Ynglinga saga 29, in Heimskringla I, Ísl. Fornr. 26, 57.
76 “Battle-leek Ull [sword warrior], we used to wear on hawk-fells [arms] the seed of
Fýrisvellir [= gold] throughout Hákon’s life.” Text (Ed.) Faulkes, Edda (Skáldskaparmál, 1),
59. Translated by Faulkes, slightly modified. See also SkP I, 226f.
Skalds in tenth-century Norway were thus acquainted with “the Aðils tradition”
and the idea that the king carried a specific golden ring.77 Were it otherwise, it
would have been impossible to understand the kenning for gold, Fýrisvalla fræ.
Norwegian kings from the Ynglinga dynasty were also associated with simi-
lar royal rings. When King Haraldr is designated with a kenning hringberendr
‘the ring carrier’ in the skaldic poem Haraldskvæði st. 1, we may conclude that
it was a common practice that kings either wore rings on their arms or around
their necks, or else held them in their hands during ceremonies. A survival
of the noble ring custom may also be seen in Egils saga 55. When Egill had
helped King Aðalsteinn to conquer King Óláfr rauði, Earl Aðils and the Scots,
he visited Aðalsteinn’s hall with his band.
. . . and when the king saw Egill arrive, he gave an order to clear the lower
bench for his men, and told Egill to sit in the high-seat there, facing
him. . . . And after they had been sitting there like that for a while, the king
unsheathed his sword, took a fine, large ring from his arm and slipped it
over the point of his sword, then stood up and walked across the floor and
handed it over the fire to Egill. Egill stood up, drew his sword and walked
out on the floor. He put his sword through the ring and pulled it towards
him, then went back to his place. The king sat down in his high-seat.78
It seems as if this text reflects an ancient ritual action in the hall of King
Aðalsteinn. In addition this prose text may be supported by a stanza which
may have been composed by Egill himself.
77 Hedeager (2011, 224ff.) argues that the story about Aðils is old and even influenced by
Hunnic traditions.
78 . . . ok er konungr sá, at Egill var inn kominn, þá mælti hann, at rýma skyldi pallinn þann inn
óæðra fyrir þeim, ok mælti, at Egill skyldi sitja þar í ǫndvegi gegnt konungi. . . . ok er þeir sátu
svá um hrið, þá dró konungr sverðit ór slíðrum ok tók gullhring af hendi sér, mikinn ok góðan,
ok dró á blóðrefilinn, stóð upp ok gekk á gólfit ok rétti yfir eldinn til Egils. Egill stóð upp ok
brá sverðinu ok gekk á gólfit; hann stakk sverðinu í bug hringinum ok dró at sér, gekk aptr til
rúms síns, konungr settisk í hásæti. Egils saga, Ísl. Fornr. 2, 143f. Trans. B. Scudder.
Both the golden arm-ring and the high-seats mentioned in Egils saga should
be seen as signs of honour, and only intended for the powerful kings and the
most outstanding chieftains. The story of Earl Hákon’s large golden ring at
the sanctuary of Lade points in the same direction.
Late Iron Age golden arm- and neck-rings have been found in Sweden,
Norway and Denmark, but not in Iceland.80 Occasionally golden rings have
been discovered in the context of cultic sites with ceremonial buildings. At
Tissø, in Denmark, for instance, a golden neck-ring, weighing 1.8 kilos and
dated to c. AD 900, was discovered in the workshop area.81 It is interesting to
note that this neck-ring was found at a place which included both a hall and a
cultic house. In my opinion, the reason these rings sometimes had this impor-
tant role in cultic contexts is that they symbolized and represented aspects of
the divinity and the mythical world. Karl Hauck’s investigation indicates that
not only kings used rings as honourable objects. They were also carried by the
gods. Hauck refers, for instance, to a Migration Period wooden image found in
Rude Eskildstrup, Zealand, which has been interpreted as a representation of
a divinity.82 It has a ring around its neck. Also interesting in this context is the
golden figure from Slipshavnskogen, Funen, which is almost 7 cm high. It has
been described as either an image of god or an aristocrat wearing a double ring
around his neck.83 Since it is naked I think that it represents a divinity. The
small figurine from Kymbo, Västergötland, also has a double ring made of gold
(see fig. 26).84 It has been suggested that it reflects a hanged person. I agree
79 “The god of the armour (the king) hangs/ a jangling snare (a ring) upon my clutch/ the
gibbet of hunting-birds (arm)/ the stamping-ground of hawks/ I raise the ring, the clasp
that is worn/ on the shield-splitting arm,/ on to my rod of the battle-storm (sword)/ in
praise of the feeder of ravens (warrior, i.e. the king).” Egils saga, Ísl. Fornr. 2, 144. Trans.
B. Scudder.
80 See e.g., Hårdh 1996, 132–134; Utne Nilsen 2002. See also Kristján Eldjárn 1956 and the new
edition Kristján Eldjárn & Adolf Friðriksson 2000.
81 Jørgenssen 2002.
82 Hauck 1954; Van der Sanden & Capelle 2000, 53.
83 G. Andersson et al. 2004, 79ff.
84 T. Zachrisson 2003; G. Andersson et al. 2004, 82ff.
Figure 26
The small figurine from Kymbo, Västergötland.
with Torun Zachrisson’s suggestion, that it may represent Óðinn, hangatýr, ‘the
hanging god’.85
In the written sources too, we see that the images of gods could be decorated
with rings, mostly arm-rings. Saxo, for instance, mentions a statue represent-
ing Óðinn, where the god had heavy golden rings on his arms.86 The Old Norse
mythic traditions mention the golden arm-ring Draupnir, which belonged to
Óðinn.87 From it another eight equally heavy rings drip every ninth night. In
the traditions about Earl Hákon’s and Guðbrandr’s hof in Guðbrandsdalr the
cultic images were decorated with rings. Njáls saga 88 reports that Þorgerðr
had a huge golden bracelet in her hand or ‘on her arm’.88 The historical value
of this episode must be questioned. But other sources state that Þorgerðr had a
ring as an attribute on statues representing her in Trøndelag.89 In this context,
we must also mention Freyja’s necklace, which according to late Old Norse tra-
ditions was called Brísingamen.90
The rings appearing in the context of sanctuaries could very well be seen
as symbols of dignity. Since the images of gods also carried similar objects, it
is possible that these rings in the earthly sanctuary context also referred to a
divine aspect, and therefore also gave the leaders using them as attributes in
the ceremonies a divine dignity and religious sanction for their social position.
85 T. Zachrisson 2003. See also the small figurine with a double neck-ring from Viksø,
Zealand. It has been interpreted as a representation of a goddess. This statue is much
older, however, perhaps from 500 BC.
86 Cuius etiam brachiorum lineamenta confertissimo armillarum pondere perstringebant.
Gesta Danorum 1.7.1.
87 See Gylfaginning 48; Skáldskaparmál 33.
88 Hann sá Þorgerði hǫldabrúði sitja, ok var hon svá mikil sem maðr roskinn; hon hafði
mikinn gullhring á hendi ok fald á hǫfði. Hann sviptir faldinum hennar, en tekr af henni
gullhringinn. Þá sér hann kerru Þórs ok tekr af honum annan hring. Hann tók inn þriðja af
Irpu . . . Brennu-Njáls saga 88, Ísl. Fornr. 12, 214.
89 See e.g., Færeyinga saga 23, Ísl. Fornr. 25, 50f. and below.
90 Þrymskviða 13; Gylfaginning 34; Skáldskaparmál 20.
For some people these impressive golden rings, which were carried by the rul-
ers or placed on cultic images, were perceived as agencies of power in a more
direct way, without making symbolic interpretations.
There are plenty of great burial mounds from the Late Iron Age situated beside
the ceremonial buildings in the Mälaren area and Trøndelag.1 Recently schol-
ars have emphasized their ideological dimensions.2 Besides being the last
dwelling place for wealthy people, they were also important means for the
elite to distinguish themselves from other groups in these regions.3 The mon-
umental impression of these graves was thus used as a visual marker of the
presence of power in the landscape. Most likely they also had other functions.
Some of them may have served as meeting places for the judicial assembly and
quite often they occurred at central places which were later considered as royal
demesnes, i.e. places which played an organizational role for society. In his
classical study Uppsala högar och Ottarshögen (1936) Sune Lindqvist argued
that these graves often functioned as thing mounds. Different assemblies
and ceremonies took place on these mounds, including the inauguration of
the ruler.4
1 In archaeology a mound which is at least 20 metres in diameter at the base is usually called
a “great burial mound” (Sw storhög). A mound which is at least 30 metres in diameter at
the base is usually called a “royal burial mound” (Sw kungshög). See Hyenstrand 1974. When
using these terms scholars usually refer to mounds belonging to the Iron Age, while monu-
mental Bronze Age mounds are excluded. Silver 1996. For a critical discussion of these con-
cepts, see Bratt 2008, 43ff. Surveys of the existence of great burial mounds in the Mälaren
area and Trøndelag may be found in Bratt 2008 and Stenvik 1996.
2 Bjørn Ringstad (1987; 1991), for instance, has discussed the ideological implication of great
burial mounds from western Norway in the Migration Period, while Peter Bratt (2008)
focused on large mounds and power structures in the Mälaren basin during the Iron Age. On
similar interpretations of great burial mounds, see also e.g. Ramqvist 1987; 1995; Müller-Wille
1992; T. Zachrisson 1994; Gansum 1995; 2002; 2004; 2013; Silver 1996.
3 Snorri wrote in Ynglinga saga (regarding Svetjud): “For notable men burial mounds were to
be thrown up as memorials.” Heimskringla I, Ísl. Fornr. 26, 20. Cf. Heimskringla I, Ísl. Fornr. 26,
4. It is quite clear that Snorri was acquainted with the tradition in Svetjud of erecting large
mounds after departed magnates.
4 See Lindqvist 1936, 3–17. Cf. Ringstad 1991, 143; Brink 2000a; 2000b; Bratt 2008, 141; Gansum
2013.
During the Iron Age great burial mounds appear frequently in Sweden and
Norway. According to available records there are at least 268 great burial
mounds in the Mälaren region.6 Most of them (79%) have a diameter of 20–29
metres. There are also some large royal mounds with a base over 40 metres in
diameter. But they make up only 7.5% of the total (i.e. 20 mounds). The largest
mounds are located in the central part of the Late Iron Age settlements, for
instance, Uppsa kulle and Norsborgshögen in Södermanland, Uppsala högar
and Nordians hög in Uppland, and Anundshögen, Ströbohögen and Östens
hög, Västmanland. The great burial mounds are often situated at lines of com-
munication, especially water routes. They are often located at burial fields of
important settlements and places close to ancient chieftain sites designated
S kalla-Grímr’s mound at Borgarnes, but they are quite small compared to the
great burial mounds found in Trøndelag and Svetjud. Skalla-Grímr’s mound is
only 5 metres in diameter at the base and about 1 metre in height.14 Sometimes
the Sagas of Icelanders and Landnámabók also mention that people were bur-
ied in mounds.15 Whether they should be regarded as great burial mounds is
uncertain. At any rate, no archaeological finds support the existence of such
monuments in Iceland.
As noted above, the great and royal burial mounds in the Mälaren region and
Trøndelag could be seen as a ritual-symbolic strategy used by elite groups.
Whether they appeared at cultic sites is somewhat debated. The archaeolo-
gist Peter Bratt, for instance, has recently argued, in a detailed and impressing
investigation, that the burial mounds in the Mälaren region rarely coincide
with cultic sites.16 The cultic sites are often situated on the periphery of
settlements with mounds, rather than at the sites of the mounds themselves.
Bratt supports his argumentation mostly with theophoric place-names. In
my opinion, his documentation is too narrow in this case. If we include the
halls of these settlements, where the ceremonial feasts took place, we may
find a stronger connection between burial mounds and cultic sites, at least
in the foremost aristocratic milieus. There is, for instance, a physical relation-
ship between banqueting halls and great mounds at several important cen-
tral places, such as Uppsala, Tuna, in Vendel and at Hovgården, Adelsö. Since
these important central places (which probably included cultic and ceremo-
nial functions) have names which do not contain obvious theophoric or cultic
elements, Bratt’s conclusion is somewhat uncertain, at least when it comes
14 Kristján Eldjárn (1956, 206) writes: “Það virðist leyfilegt að álykta, að stórir haugar hafi
verið fátíðir eða óþekktir á Íslandi. Haugar hafa að vísu verið, en mjög lágir, líklega sjaldan
meira en 1 m á hæð og 5 m í þvm., oft minni.” Cf. Kristján Eldjárn & Adolf Friðriksson 2000.
15 See e.g., Laxdæla saga, Ísl. Fornr. 5, 13, 72; Gísla saga, Ísl. Fornr. 6, 58; Brennu-Njáls saga,
Ísl. Fornr. 10, 193; Landnámabók, H60, Ísl. Fornr. 1, 103ff.; See also Sveinbjörn Rafnsson
1974ff.; Gansum 2004, 102f. Sveinbjörn argues that many “mound narratives” (högberät-
telser) are literary constructions in Landnámabók. They were made up because of the
judicial right to land or in an apologetic manner against pagan practice.
16 Bratt 2008, 137–141.
Figure 27a The three royal burial mounds from Old Uppsala.
Photo: Marie-Louise Sundqvist.
17 Bratt (2008, 185) is aware of these conditions when he writes: “The case of (Gamla)
Uppsala does show that cult and Thing could be located in one and the same place, but
this may have been reserved for the foremost ruling dynasties.”
18 L. Larsson & Lenntorp 2004; Christensen 1991; 1997; 2001; on Borre, personal communica-
tion T. Gansum.
halls,23 combined with royal and great burial mounds containing extensive
finds.24 At these burial mounds there are also “thing mounds” with plateaus,
i.e. places where assemblies could possibly have been held in ancient periods.25
The locations of these thing mounds in relation to the burial mounds are sur-
prisingly similar at these sites. At both Uppsala and Hovgården they constitute
a complex with three mounds in a straight row, and a thing mound somewhat
separated from the burial mounds, at the end of the line.26
A similar ritual structure is partly also visible at Högom, Medelpad. There
we can see four great mounds from the Migration Period associated with an
Iron Age hall.27 Three of these mounds were excavated in 1949 to 1960. In one
of them a wooden chamber (5 × 2 m) was found with the remains of a man,
an imported sword, axe, spear, expensive glass, and beakers with gold and
silver decoration. These finds indicate that the man belonged to the upper-
most social stratum in this region. Below burial mound no. 3, well-preserved
parts of a house foundation were discovered. It was interpreted as a hall 40 m
long and 7 m wide in the middle. The archaeologist Per H. Ramqvist identified
one central part of the foundation and some post rows as a well built stage-
like structure. Near it, he argued, the high-seat was situated.28 Most artefacts
were concentrated in this area of the house. Also under the burial mound
no. 4 there were remains of a house.29
Three royal mounds in a row are also visible at Bertnem in Nord-Trøndelag
(fig. 27c). They are about 40–45 metres in diameter and 5–6 metres high. These
23 Brunstedt 1996, 20, 23ff.; Damell 1993 and A. Carlsson 1997.
24 The burial mound Skopintull (RAÄ 48) is dated to the Viking Age. It displays rich finds
belonging to a male. It is possible that he was a king or a royal. See Rydh 1936, 104ff.,
211; Bratt 2008, 231ff. The other burial mounds at Adelsö have not been excavated.
25 Bratt 2008, 145f.
26 There is also a similar pattern at Fornsigtuna. Close to the halls in Fornsigtuna a great
mound was erected during the Iron Age. This mound did not contain a grave, however.
Three or more great burial mounds in a row may be found at ten places in the Mälaren
region. This tradition is attested from the Migration Period or Early Merovingian Period
to the Viking Age. See Bratt 2008, 122ff., 185.
27 Baudou 1989; Ramqvist 1987; 1995.
28 Ramqvist 1995, 37.
29 Evert Baudou (1989) argued that ancestor worship in general was practised at grave
mounds, which were intentionally placed over houses. Finds connected to these mounds
indicate that cultic activities took place there. Good examples of this phenomenon could
be seen at the burial fields of Högom, Medelpad, and Trogsta, Forsa parish, Hälsingland.
Some of the mounds at these sites did not even contain graves. Baudou states that burial
mounds were erected over houses in south-western Norway as well, for instance, at
Ullandhaug in Rogaland. On this phenomenon, see also Hållans Stenholm 2012, 195ff.
prominent mounds have not been excavated. It has been suggested, however,
that they should be dated to the Migration Period. There are also traces of a
chieftain house at a settlement situated close to these mounds which has been
excavated and dated to this period. Most likely this settlement is contempo-
rary with the mounds.30 Perhaps the ritual structure of the burial mounds at
Bertnem was based on a common cultural model, which was spread among
elite groups in Svetjud and Trøndelag in the the Late Iron Age (including
Viking Age).31
12.2.2 Other Monumental Graves and Cultic Sites in the Mälaren Region
Extremely rich and spectacular graves associated with hall buildings may also
be seen in Vendel. The funeral field there is divided into three groups: the
southern, the middle and the north cemetery groups.32 The southern group
is connected to the famous boat graves, the great burial mound of Vendla and
two smaller burial fields. The burial finds comprise high-status objects. Most
of them were found in the boat graves. There are 15 boat graves in all. Close
to the hall at Vendel (Tuna) is the great burial mound of Vendla, which actu-
ally comprises two interlinked mounds. It also contained high-status finds.33
A couple of kilometres from the hall is Ottarshögen at Husby, containing rich
materials.34 Also at Valsgärde there are boat graves close to a large Late Iron
Age banqueting hall. A majority of the graves at this site were, however, more
common cremation graves.35 This burial field displays continuity from the Pre-
Roman Iron Age to the Viking Age. There are five chamber graves, which are
older than the boat graves. The rich boat graves were built from the end of the
sixth century up to the eleventh century. In the boat graves there are only men,
one individual from each generation. There are also extremely rich female
graves at Valsgärde. Grave 57, from the Merovingian Period, and graves 85 and
94, from the Viking Age, were clearly high-status female graves.36 Grave 57 is
located at the highest point of the cemetery. The location of the monument
clearly indicates that this woman was particularly honoured (see ch. 9 above).
30 Farbregd 1980; Ramqvist 1987, 114f.; Müller-Wille 1992, 4; Østmo & Hedeager 2004, 37; Bratt
2008, 39.
31 Müller-Wille 1992, 2ff.
32 See Seiler 1997; 2001. On boat graves see Stolpe & Arne 1912; Arne 1932; Müller-Wille 1970.
33 Ljungkvist 2006, 65ff.
34 Lindqvist 1936, 162ff.
35 Arwidsson 1942; 1954; 1977; 1980; A.-S. Gräslund 1997; Schönbäck & Thunmark-Nylén
2002; Ljungkvist 2006, 72.
36 A.-S. Gräslund 2008.
37 Bratt (2008, 149) states: “Min analys av de sakrala ortnamnens relation till storhögsenhet-
erna visar att närmare 25 % av dessa har ett direkt eller nära samband med heliga platser.”
38 For a complete investigation of the relationship between pagan cultic sites and great
burial mounds in Trøndelag, see Brendalsmo 2006, 137ff.
39 See e.g., Brendalsmo 2006, 137.
40 Gerhard Schøning II:235. Quoted from Bjørk Birgisdottir & Strøm 2002, 19.
They could also be dated to the Viking Age.41 Most likely Melhus was both a
chieftain farm and a public cultic site during this period. Like Lade, Melhus
became the site of a “shire church” ( fylkiskirkja) in the twelfth century, and
thus continued to be an important cultic site for the people of this fylki.
There are some other possible cultic sites in Ut-Trøndelag where great burial
mounds from the Late Iron Age occur. Since some of them are designated with
place-names including the ON term hof/hov, they may have had ceremonial
buildings. The burial mounds at these sites indicate the presence of wealthy
and powerful families. These families were probably not as powerful as the
Earls of Lade and the rulers of Melhus, yet they may have tried at some time
to gain a certain position by means of public cult and great burial mounds. At
the farm of Hove in Åsen, Strindafylki, for instance, finds have been made indi-
cating cult (see ch. 5). At the neighbouring farm of Vang there is a great burial
mound.42 This mound is 35 metres in diameter at the base and 3 metres high.
fylkir ‘king’, but he preferred the latter interpretation. He thus interpreted these
names as ‘the royal burial mound’ (‘kongshaugen’ or ‘haugen hvor en konge
ligger (eller menes å ligge) begravet’). Also interesting in this context is the farm-
name Hauge in Klepp, Rogaland, which may refer to the royal burial mound
called Krosshaug. At this farm 16 gold foils were also found, indicating that this
farm too was a cultic site, perhaps with a ceremonial building.44
Alstadhaug (ON Ǫlvishaugr) is located at Skogn in Skeynafylki (fig. 27b).45
The first element in the place-name contains Ǫlvir- in the genitive case, which
may derive from a proper name Ǫlvir or an appellation and designation of a
cultic leader called *ǫlvir (< *alu-wīhaz ‘Priester eines alu- (alh) Heiligtums’).46
The last element haugr obviously refers to the great burial mound located
at the churchyard. This mound is dated to the Early Iron Age.47 The medi-
eval name Ǫlvishaugr is mentioned in Hákonar saga góða 18. Snorri says that
Blótólfr of Ǫlvishaugr was one of the eight chieftains who had most to do with
the sacrifices in the whole of Trondheim. Thirteenth-century sources report
that Ǫlvishaugr was an ancient political and religious centre. This informa-
tion may to some extent be corroborated by archaeology. There are several
Iron Age burial mounds at this place, including stray finds, such as swords,
remains of a shield boss and a bronze bowl. These finds indicate high status
and the presence of chieftains. Whether cultic activities took place is difficult
to say. Sometime in the twelfth century, a Romanesque church was erected at
Alstadhaug, and it became a fylkiskirkja.
The name Sakshaug (medieval Saurshaugr), on Inderøy, in Eynafylki, most
likely refers to “Saur’s (or *Sorr’s < *Sarwa’s) burial mound”,48 but this mound
does not exist today. Whether a great burial mound ever was located at this
site is somewhat debated.49 There are several small Iron Age burial mounds
documented on this farm, as well as burial cairns. Other finds have been made,
such as swords and bowl-buckles made of bronze. These finds do not give any
clues to the general status and function of the farm. Written sources indicate,
however, that Sakshaug had a political and religious significance during the
pre-Christian period. A peculiar story is related in Hákonar saga góða 12 about
44 Cf. Særheim 1989; 2015, 53; Kristoffersson 2006; Sundqvist 2015.
45 See e.g., M. Olsen 1926, 265ff.; Sandnes & Stemshaug 2007, 69; Røskaft 1997, 232; 2003, 119ff.
46 De Vries 1932; 1956–57, §278; 1977, 687; Kousgård Sørensen 1989; See also Sundqvist 2003a,
113; 2007, 34f.
47 Stenvik 1996, 86.
48 See Sandnes & Stemshaug 2007, 379. Cf. M. Olsen 1926, 265ff.; Røskaft 2003, 131. Other
interpretations have been suggested, see Stenvik 1996, 86.
49 Stenvik 1996, 86.
King Eysteinn from Oppland, who ravaged in Trøndelag and conquered both
Eynafylki and Sparbyggvafylki, He put his dog Saur as ruler there in the high-
seat, which was situated on the mound called Saurshaugr (i.e. Sakshaug). This
story is of course some kind of migratory legend, where the name of the dog
has been associated with the old name of the mound. In any case, it indicates
that people during the thirteenth century associated Saurshaugr/Sakshaug
with ancient ruling power and pre-Christian ceremonies connected to the
mound.50
Finally, there is also a place called Haug (< Haugr) in Verdølafylki, close
to Stiklastaðir. The name was probably related to a great burial mound. This
mound does not exist today and some scholars think it never did.51 There is
an Iron Age burial field with some mounds on this farm.52 Some of them are
quite big. One measures about 25 metres in diameter and it is 3.5 metres high.
Some of the finds from Haugr indicate high social status. In one boat grave
a sword and the point of a spear were discovered. Most likely this farm was
rich in the Late Viking Age and the Early Middle Ages. Haugr is mentioned
a couple of times in Heimskringla. According to Óláfs saga Helga 109, the
farm at Haugr was owned by the Christian king, and the people living there
seem to have been Christians.53 Historical sources report that the farm was
regarded as crown lands all through the Middle Ages. Whether there was a pre-
Christian cultic building or banqueting hall in Haugr is uncertain. There are
perhaps some indications of a Late Viking Age ceremonial building at Haugr
in Magnúss saga in góða 14. Snorri reports that King Magnús was at a ban-
quet (veizla) in Haugr together with Kálfr Árnason and Einarr Þambarskelfir
Einriðason.54 It seems as if the king sat in the high-seat, with Kálfr on one side
and Einarr on the other side. Most likely this banquet took place in some kind
of hall at Haugr. Of course this story is just based on Snorri’s imagination. At
least in the Middle Ages, Haugr was a religious centre with a fylkiskirkja. It was
a stone church, probably built in the twelfth century. In the fourteenth century,
however, this church burned down, and the church of Stiklastaðir took over its
role as a fylkiskirkja.
50 See further M. Olsen 1926, 265–270. Based on the medieval information Magnus Olsen
suggested that Sakshaug was also a cultic and political centre during pre-Christian era.
Most scholars today think that this assumption is not possible, since the status of the
available sources is so poor. See e.g., Røskaft 2003, 132.
51 See further Stenvik 1996, 86f.
52 Røskaft 2003, 128.
53 Heimskringla II, Ísl. Fornr. 27, 179f.
54 Heimskringla III, Ísl. Fornr. 28, 24.
Nobody shall invoke groves and mounds, or heathen gods (or idols), sanc-
tuaries or stave-enclosures.56
The verb OGut haita ‘invoke’ (cf. OSw heta, hæta, haita, ON heita, Lat vocare),
with the postpositive preposition a, indicates that it was the groves and
mounds themselves which were regarded as divine and thus worshipped.57 It
is most likely that the noun OGut haugr refers to ‘burial mound, grave howe’.58
This law thus suggests that cultic activities really did take place at pagan burial
mounds. This may also be supported by the Guta Saga, which states: “Prior to
that time, and for a long time afterwards, people believed in groves and grave
howes, holy places and ancient sites, and in heathen idols.”59 There were thus
medieval traditions in Gotland mentioning that people in earlier times vener-
ated mounds (haugr sg.).60
Most likely Norwegian medieval laws also indicate the existence of cult at
burial mounds.61 Den ældre Gulathings-Lov ch. 29 mentions that: “heathen sac-
rifices are also banned for we are not permitted to worship any heathen god,
or [any] mound, or [in any] heathen sanctuaries.”62 Perhaps the implicit
meaning of the term haugr is ‘burial mound’ and that sacrifices to such struc-
tures were forbidden. A similar formulation is repeated in the Ecclesiastical
Law of King Sverrir (Kong Sverres Christenret), which is directly derived from
55 The oldest manuscript (ms.A) is dated to c. 1350. It goes back to an original dated to c. 1220.
Holmbäck & Wessén 1979 (1933–46) Vol. 4, lxiv ff.
56 Engin ma haita a huathci a hult eþa hauga. eþa haþin guþ. huatki a vi eþa stafgarþa.
Gutalagen, Af blotan. (Ed.) Pipping, 7. Cf. GL I: 4, SSGL 7, p. 14. My trans.
57 Cf. Lindqvist 1962–1963, 57–60.
58 Cf. Peel 1999, 68.
59 Cf. Peel 1999, 4f.
60 Cf. Sveinbjörn Rafnsson 1974, 197.
61 See e.g., Nilsson 1992, 34ff.; Sanmark 2004, 158f.
62 Blot er oss oc kviðiat at vér scolom eigi blota heiðit guð. ne hauga. ne horga. NGL 1, 18. Trans.
Larson.
Den ældre Gulathings-Lov.63 Thus Norwegian laws likewise seem to have regu-
lations against sacrifices performed at burial mounds. They therefore support
the assumption that these burial mounds were regarded as cultic places in
Norway during the transition period.
The sources thus seem to indicate that great burial mounds in Svetjud and
Trøndelag were regarded as ritual elements at aristocratic sanctuaries.64 In
what follows I will argue that the reason they had this status may be related to a
specific ruler-ideological strategy, which was applied by rulers in these regions.
In my opinion, some mounds were regarded (at least among certain people
with cultural and religious knowledge as well as among others) as a material-
ized medium or symbol of noble descent. The top figure of the genealogy was a
forefather or a foremother, who was buried in the mound. This figure may even
have had a divine status and been regarded as a kind of hero.65 Qualities such
as noble or divine descent seem to have been crucial for a legitimate ruler, and
also for land rights.
63 Blott er os kuiðiat at ver skulum æigi blota hæiðnar vetter. oc æigi hæiðin guð ne hauga ne
horgha. NGL 1, 430.
64 The expression hǫlga grǫf in Eddic poetry also supports the idea that the mound was
considered as a sanctuary. See Baetke 1942, 129f, 137.
65 See Widengren (1969, 185f.) on hero-cult. Cf. Ringgren & Ström 1984, 299. See also below.
66 Cf. Ringstad 1991, 144f.; T. Zachrisson 1994.
67 All the finds in this mound are presented by Lindqvist 1936, 168–177, 341f.
68 Duczko 1996a, 413.
may have been attached to a boy’s leather helmet. Birgit Arrhenius therefore
suggests that this mound was erected over a young, high-ranking person.69
According to Arrhenius and Sjøvold, a jawbone probably belonged to a boy.70
If their assumptions are correct, this constitutes strong indications of dynastic
kingship at Uppsala. Osteological investigations of the bone fragments in the
Eastern Mound show, however, that some human bones in this mound prob-
ably belonged to a woman aged about 20–30.71
The Uppsala mounds may in fact have belonged to one kin or dynasty.
Behind the mounds there was a lot of labour and organization.72 To raise a
royal mound, 30 metres in diameter at the base, took 10,000 man-days.73 Very
few could afford such graves. They were thus made in order to raise this family
above other chieftain families as well as ordinary people. They may also have
functioned as “genealogical markers”, i.e. they enumerate the forefathers of a
certain family with claims to kingship at Uppsala. Other places in the area of
the Svear also had mounds, perhaps with similar functions.74 As mentioned
above, Adelsö has three great mounds in a row, like Uppsala, which may have
expressed genealogical concerns.75 Similar conditions may have existed at
Högom in Medelpad and Bertnem in Nord-Trøndelag, where we also find great
or royal burial mounds situated in a row.
A genealogical strategy may perhaps be attested in the act of burying dead
relatives on top of the graves of early ancestors, i.e. overlying graves (Sw överla-
grade graver). This phenomenon is seen at the cemetery of Ormknös on Björkö
in Lake Mälaren. There are three great mounds on the crest of the ridge. Birgit
Arrhenius has investigated one of these. This mound, which is 21 metres in
diameter and 2.7 metres high, seems to have been constructed in two phases.
The first mound was built in the early Roman Iron Age, about 17 metres in
diameter and 1 metre high, covering a single grave, which has been plundered.
In the Viking Age this mound seems to have been enlarged to its current size.
Arrhenius states that it has the impression of being “erected as a sort of marking
with the earlier ancestors”.76 A parallel to the mound at Ormknös is perhaps the
Gullhögen mound in Husby-Långhundra, Uppland. Torun Zachrisson points to
several other cases where mounds overlie older graves, e.g. Valsta in Norrsunda
parish (RAÄ 59), Uppland, Sweden.77 Perhaps these graves are visible signs of
genealogies within the landscape.
it is probable that these stones lined the processional road called Eriksgatan
which passed by this monument. There are many other monuments at this
site, such as five stone ships (skeppssättningar). It is most likely that different
types of ceremonies also took place there, at least in the Viking Age. During the
Middle Ages Anundshögen was the assembly place for Siende härad.82 It is thus
possible that different types of religious and judicial activities were related to
this place. The main monument there represented the great ancestor. Whether
he was called Anundr or something else is unknown.
There are also other great burial mounds in the Mälaren region which seem
to be designated with names referring to ancient ancestors. An interesting
example is the royal burial mound called Ottarshögen (sixth century) at Husby,
Vendel parish, Uppland. It has been argued that the name of this royal burial
mound refers to the dwelling place of the Migration Period Ynglinga king
called Óttarr vendilkráka.83 According to Ynglinga saga, Óttarr was killed at
Vendil in Denmark.84 There is evidence to indicate that Óttarr actually could
be connected with Vendel in Uppland. In two independent records from the
seventeenth century this burial mound is called Utters hög and Otters hög.
The information stems from people who probably could not have been influ-
enced by the name Óttarr vendilkráka attested in Ynglinga saga. Snorri’s text
became known in Sweden through Worm’s 1633 edition of Ynglinga saga.85 In
1677 the name is mentioned in an official inquiry into antiquities, in the form
Otters hög.86 The information was submitted by the vicar of Vendel church.
More important is the testimony in court two years earlier (1675) at proceed-
ings held in Vendel. At this trial, two women were accused of sorcery, and one
of them stated that Utters hög in Vendel was one of their places for rendezvous
with the devil.87 It thus seems as if the name was rooted in a popular tradition
82 The great mound at Aspa löt, in Ludgo parish, was also probably located beside a thing
place. The runic inscription Sö 137 indicates this. S. B. F. Jansson 1987, 123ff. This struc-
ture may be seen at some other places in the Mälaren region, see Bratt 2008, 142ff. and
Sanmark & Semple 2008.
83 Heimskringla I, Ísl. Fornr. 26, 53–55.
84 “Then he steered west in the Öresund (Eyrarsund), and then south to Jutland ( Jótland) and
entered the Limfjord (Limafjǫrðr). Then he harried in the Vendil District (á Vendli). . . . in
Heimskringla I, Ísl. Fornr. 26, 54. Also Historia Norwegie (IX,25) places Vendil in Denmark.
Cui successit in regnum filius suus Ottarus, qui a suo equiuoco Ottaro, Danorum comite,
et fratre eius Fasta in una prouinciarum Danie, scilicet Wendli, interemptus est.
85 Linderholm 1919; Lindqvist 1936, 37–47; Vikstrand 2004b.
86 See Rannsakningar I, 24.
87 Linderholm 1919, 36ff.; Vikstrand 2004b, 379f.
Herbjørn ligger i haugen i Drifsvig. Aatte til eigu och edel Helgabostad och
Holm och Starrin. Barder Nessjakonnig ligger i haugen paa Bardarstad;
hand aate all Roma och alla Strandarvig och Gudøj och Hundeidavig. Eilifr
jarl ligger i Eilifshaug i Gudø. Thorsten Bláfotr ligger indeni Hundeid. Rafn
som ligger i Hval i Heimdal. Grimr ligger i haugen paa Hjortdal paa
Sundmør.94
The great burial mounds of Sunnmøre thus had the function of a concrete
marker of territorial requirements. Magnus Lagabøters landslov (1273) says
that if someone discovers goods in the ground (“jordgravet gods”), the king
should have one third of them, but the one who is closest to haugóðalsmaðr,
i.e. the person who can prove that he is descended from the people buried in
the mound, should have one third of the goods. One third goes to the owner
of the land. Sometimes the mounds are not mentioned in these laws. In those
cases the heir should be able to count his forefathers in five or six generations.
Den ældre Gulathings-Lov states that the litigants disputing óðal, ancestral land
property, had to prove their claims by enumerating their forefathers:
Now when the doom is set, [the claimant] shall enumerate his ancestors,
the five who have owned the land, and the sixth who had it both in own-
ership and in óðal.95
Then witnesses were brought into court. The one who had the best witnesses
won the lawsuit. The óðal right was proved by referring to the ancestors who
had possessed the estate in six successive generations.96 According to the
Frostathings-Lov, however, the land must have been inherited through four
generations of kinsmen when óðal is claimed.97 Since the Ældre Gulathings-
Lov is older than Frostathings-Lov, historians have argued that the earlier rig-
orous demands on óðal were somewhat eased at the beginning of the Middle
Ages.98 These examples show that the laws, long after Christianity was intro-
duced in Norway, were still influenced by the old óðal system, in which the
pre-Christian burial mounds also played an essential role.
Most likely the great burial mounds played a similar role in the óðal sys-
tem of the Mälaren region during the Late Iron Age. The Runic Swedish term
ōðal is attested on several runic stones there. On a rock at Nora, in Danderyd
parish, Uppland (U 130), an inscription tells us about Bjǫrn Finnvidsson who
had the rock carved with runes in memory of his brother Olev: “This farm is
their ōðal and family inheritance, the sons of Finnvid at Älgesta (Er þessi byr
þæira oðal ok ættærfi, Finnviðar suna a Ælgiastaðum).”99 The genealogical and
judicial concept of ættærfi “family inheritance” appears here almost as a syn-
onym to ōðal. RSw ōðal also occurs in the runic inscription at Eneby, Runtuna
parish, Södermanland (Sö 145).100 Two important runic inscriptions from
eleventh-century Hälsingland and Småland indicate a connection between
genealogies and land rights, namely the Malsta stone (Hs 14) and the Sandsjö
stone (Sm 71). By enumerating the five forefathers of He-Gylve in the Malsta
inscription, Romund is declaring his legal right to inherit the estates that his
father had possessed.101 Most likely judicial claims of ōðal ‘inherited land’ or
ættærfi ‘family inheritance’ connected to genealogy are also attested in the
Sandsjö inscription.102 This inscription counts six generations of kinsmen.
The royal burial mounds as expressions of genealogy probably played a
similar role throughout Norway and Svetjud also when royal patrimony was to
be inherited. It proved a ruler’s nobilitas, i.e. that he was a konungsefni, and his
Cöllen 2011. On the age of Hyndlolióð, see e.g., Steinsland 1991, 247ff. It has been debated
whether Hyndlolióð can be regarded as a uniform poem. Some scholars think that the
poem is composed from two independent sources. E.g. Gurevich 1992, 190–199. Steinsland
has convincingly shown that this lay is uniform. Steinsland 1991, 248–251. Cf. Cöllen
2011, 192ff.
97 Engum manni verðr iörð at óðali fyrr en .iij. langfeðr hafa átt oc kemr undir hinn .iiij. samf-
leytt. “Land does not become any man’s óðal until three of his ancestors have owned it and
it comes to the fourth in unbroken descent.” Frostathings-Lov, NGL 1, 237.
98 Lindkvist 1979, 142.
99 S. B. F Jansson 1987, 97. Cf. T. Zachrisson 1994, 222f.; Sundqvist 2002, 173.
100 Peterson 1994, 37. Cf. T. Zachrisson 1994, 223; Sundqvist 2002, 173.
101 Brink 1994; Sundqvist 2002, 151ff.
102 Sundqvist 2002, 151ff.; T. Zachrisson 2002, 39f.
land rights, i.e. his right to the fǫðurleifð and óðal.103 That kingship was consid-
ered as leifð or óðal can be seen in Late Viking Age kennings and poetic expres-
sions. Following skalds state thus:
Also according to medieval sources, the Norwegian rulers counted the king-
dom (konungs dómr) and the land as their óðal.107 Snorri states, for instance,
that Óláfr Tryggvason was entitled by óðal birth to the kingdom (er óðalborinn
var til konungdóms).108 Similar expressions occur in the context of Óláfr
Haraldsson. When he returned to Norway and spoke to his stepfather Sigurðr,
his mother Ásta and foster-father Hrani, Óláfr mentioned that foreigners “dis-
pose of the possession which my father, and his father, and one after the other
of our kinsmen owned, and to which I am entitled (ok em ek óðalborinn til).
Nor are they satisfied with that, but have appropriated what has belonged to
our kinsmen who are descended in direct line from Haraldr Finehair (at lang-
feðgatali erum komnir frá Haraldi inum hárfagra). . . . Now I shall disclose to
you what has been in my mind for a long time, which is that I mean to regain
my paternal inheritance (at heimta fǫðurarf minn); . . .”109 It should be noted
that important terms such as óðalborinn, langfeðgatal and fǫðurarfr appear in
this passage. They indicate a connection between royal genealogies and the
land/kingdom as paternal inheritance.110 That Snorri’s description of these
conditions is based on sound information is indicated by some words in the
Norwegian laws, which say this about the king and his land: “. . . þui at hans
103 Cf. Gurevich 1992, 198; Sundqvist 2002; Cöllen 2011, 192–247.
104 Skj. B1, 232.
105 Skj. B1, 235.
106 Skj. B1, 361.
107 T. Zachrisson 1994, 221.
108 Heimskringla II, Ísl. Fornr. 27, 47.
109 Heimskringla II, Ísl. Fornr. 27, 43f.
110 The Eddic poem Rígsþula tells us that Rígr (perhaps Heimdallr) taught Jarl, i.e. sover-
eignty, runes, gave him his own name—declared he had a son. “That son he bade appro-
priate ancestral plains—ancestral plains, long-dwelt-in countrysides (þann bað hann
eignaz óðalvǫllo, óðalvǫllo, aldnar bygðir).” Rígsþula 36. Trans. Dronke.
æign oc oðall er alt landit.”111 The kings of Norway thus counted the kingdom
and the land as their óðal, and their ancestors’ mounds were the visible proof
of this.112 Just like any óðal farm, the royal estates also had their mounds in the
yards.113 Most likely a similar system was applied in Svetjud. The royal mounds
in Uppsala, for instance, may have been related to the rights of a certain noble
family to acquire the early crown lands called “Uppsala öd”.
It is thus possible that some great burial mounds in Norway and Svetjud
can be interpreted as symbolic expressions of the ruler family’s genealogy and
their noble descent. Perhaps they also symbolized the ruler’s right to acquire
certain land related to his office.
The great burial mounds in Svetjud and Trøndelag may thus have been the
materialized and symbolic expression of a great genealogy, a memory of a
famous ancestor and/or land rights related to the ruler family. Perhaps some of
the small burial mounds in Iceland were related to similar ideas. In my opin-
ion, this custom was more restricted in Iceland, however, since allodial land
had no long continuity there.
In both the Mälaren region and Trøndelag the royal or great burial mounds
seem to have constituted an important ritual element of the central and
important assembly sites and sanctuaries. One reason that monumental bar-
rows and rich boat graves sometimes are situated inside the ritual area of these
sanctuaries may be connected to the cult of dead royal ancestors. There are
signs in the written sources that the dead rulers were sometimes honoured
with a common and public cult. Most likely these royals were even deified or
regarded as some kind of heroes or mythical beings and expected to ensure
good harvests etc.120
120 See e.g., Birkeli 1938, 89; Lid 1942, 97ff.; F. Ström 1954, 34; de Vries 1956–57, §184. Cf.
Nordberg 2013, 280ff. and 382. Geo Widengren (1969, 185f., 419) describes the hero
cult thus: “Die Trennungslinie zwischen Göttern und Menchen wurde in der Antike
nicht sonderlich scharf gezogen. Zwischen der Gottheit und dem Menschen stand der
theíos anér, der göttliche Mensch, der sich aus der menschlichen Sphäre zur göttlichen
erhob. Besonders die Heroen gehörtenzu dieser Klasse von Übergangswesen, mehr als
Menschen, aber gänzlich Götter. Der Heroenkult war örtlich gebunden und an die Gräber
geknüpft, wo ihre sterblichen Überreste verwahrt wurden und wo sie ihre Wirkung aus-
übten. Ihr Kult war sehr volktümlich. Zu den Gräbern der Heroen kan man, um Hilfe in
allerei Angelegenheiten zu suchen, vor allem aber gegen Krankheiten. So sind die Heroen
Heilsgestalten geworden. Ihnen brachte man blutige Opfer, Gewänder und Früchte dar.
Typisch für den Heroenkult waren jedoch Mahlzeiten, gefeiert in den Heiligtümern der
Heroen, die über ihren Gräbern errichtet waren.” For recent treatments on Greek hero
cult, see e.g., Whitley 1995; Ekroth 2002; Hållans Stenholm 2012.
For they [the Svear] had resolved to have a temple in honour of the late
king, and had begun to render votive offerings and sacrifices to him as to
a god.124
121 Colunt et deos ex hominibus factos, quos pro ingentibus factis immortalitate donant, . . .
Adam IV,26.
122 “. . . sicut in Vita sancti Ansgarii legitur Hericum [Eiríkr] regem fecisse. Adam IV,26.
123 Porro, si etiam plures deos habere desideratis, et nos vobis non sufficimus, Ericum quondam
regem vestrum nos unanimes in collegium nostrum asciscimus, ut sit unus de numero deo-
rum. Vita Anskarii 26.
124 Nam et templum in honore supradicti regis dudum defuncti statuerunt et ipsi tanquam deo
vota et sacrificia offerre coeperunt. Vita Anskarii 26.
125 See Hultgård 1997, 21.
126 Epistula Jeremiae st. 50–51, 72.
same cliché or topos when he writes about the Slavic sanctuary in Riedegost. In
this sanctuary there are gods made by human hands (dii manu facti).127
Rimbert’s information, on the other hand, may not be totally out of order,
for there are indications in the Old Norse skaldic poetry that kings after their
deaths were considered as mythical beings or deities. The panegyric poem
Hákonarmál (c. 960), for instance, mentions that King Hákon the Good fell at
the Battle of Fitjar. When the king died with his retinues on the battlefield one
valkyria exclaimed: Vex nú gengi goða.128 According to Edith Marold, this phrase
may be interpreted as “die Schar der Götter wächst”, i.e. the pantheon was
increased when Hákon died.129 Marold states that an apotheosis took place, i.e.
the glorification and elevation of the king to the rank of god. Perhaps the king
was merely transformed to one of the einherjar, i.e. a dead warrior who dwelled
in Valhǫll.130 The term gengi goða should probably be translated as ‘the retinue
of the gods’, which may be synonymous with the einherjar. Both interpreta-
tions, however, harmonize well with other expressions in Hákonarmál, such
as “when all those who have domination and power [the gods] bade Hákon
welcome” (es Hǫ́ kon bǫ́ ðu heilan koma rǫ́ ð ǫll ok regin).131 When Hákon arrives
at Valhǫll it is stated: “You may share the ale of Æsir” (þigg þú at ǫ́ sum ǫl); and
stanza 21 says that Hákon “to the heathen gods fared” ( fór með heiðin goð).
I partly agree with Marold when she concludes: “Hákon kommt also nicht nur
nach Walhall und zu Odin, er wird vielmehr unter die Götter aufgenommen.”132
It is also possible that Hákon after death was just considered as a mythical
being designated einheri.
Since the notion of a deified/mythified king probably appears in
Hákonarmál, we cannot, in my opinion, exclude that Rimbert’s text (and
Adam’s notice) contains authentic information when saying that the dead
King Eiríkr was worshipped with sacrifices. But still it seems quite odd that
this cult took place at a “temple”, as most translations declare.133 In the ancient
Scandinavian context we hear that dead kings were honoured with sacrifices at
their burial mounds (see below). If we focus on the Latin text we may note that
Rimbert used the term templum, which in classical contexts meant ‘a space
marked out; hence, in partic., in augury, an open place for observation, marked
out by the augur with his staff’, ‘an open, clear space, a circuit’, ‘a consecrated or
sacred place, a sanctuary’. In addition to these meanings the dictionaries also
give the following: ‘a place dedicated to some particular deity, a fane, temple,
shrine’.134 The term templum could thus have designated a cultic place under
the open sky, where a burial mound was also erected. One important aspect is
probably that cultic activities took place there, as Rimbert’s statements clearly
indicate. The verb statuere means ‘to establish, charter, determine’ besides the
basic meaning ‘to erect, to build’.135 The question is whether Rimbert’s expres-
sion templum . . . statuerunt in a medieval context could be interpreted as “that
they [the Svear] established a cultic place under open sky”, where (implicitly)
a burial mound was also erected. Such an interpretation would fit the ancient
Scandinavian context perfect in a purely factual sense.
b uilding or erecting churches he normally uses the verbs fabricare, fiere, facere and perfi-
cere. See Vita Anskarii 11, 24, 28, 31 och 32.
134 Lewis & Short. 1879. Ahlberg et al. (1964, 830) has ‘ritualenligt bestämt område e. rum’,
‘invigd plats’, ‘helgedom, tempel, kapell’. Also in Georges’ dictionary (1951, 3049) we find a
similar explanations.
135 Ahlberg et al. 1964, 830. Cf. Georges 1951, 2790.
136 The þáttr has survived in many text versions published in The Great Saga of St Olaf (“Den
store saga om Olaf den helige”) (all dated to the 14th century): Den selvstændige tot
(AM 75e); Flateyjarbók; Bæjarbók; AM 61; Bergsbók. There is also a more comprehen-
sive version of the þáttr in The Legendary Saga of St Olaf (“Legendariska Olavssagan”)
(13th century). Most known is the version in Flateyjarbók, where the text is titled Óláfs
þáttr Geirstaðaálfs. On the preservation of the þáttr in the medieval manuscripts and the
different text variations, see Røthe 1997 and Wellendorf 2003. In present study I mainly
follow the version from Flateyjarbók II, 6–9 and 134–135. Occasionally I refer also to the
other versions of the þáttr.
. . . var þa þat rad tekit at þeir blotudu Olaf konung til árs ser ok kolludu
hann Geirstada álf.
. . . then they took that advice and sacrificed to King Óláfr for a good years’
crops and they called him Geirstaðaálfr.138
The second part of the þáttr takes place later during Óláfr Tryggvason’s and
Óláfr Haraldsson’s reigns (i.e. c. 995–1028) and is mostly related to the strange
events which took place at the burial mound of the Óláfr Geirstaðaálfr. In
this part the farmer Hrani is playing a central role. After a notice in a dream,
he opened the burial mound and took a belt with a knife, a sword and other
objects (e.g. a golden ring) from the Geirstaðaálfr.139 These precious objects
were later given to queen Ásta and her son St Óláfr. There is also another pas-
sage about the close relationship between Óláfr Geirstaðaálfr and St Óláfr.140
137 In, for instance, The Legendary Saga of St Olaf 2 (ed. Johnsen, 1) it is stated thus: oc var
kallaðr digrbæinn eða geirstaða alfr.
138 Flateyjarbók II, 6f. My trans.
139 . . . briota hauginn . . . iij. gripe þa er þu kyss . . . taka gullhring af þeim manni . . . knif ok
bellti . . . taka suerd . . . Flateyjarbók II, 8.
140 This part of the tradition is placed in Óláfs saga hins Helga, Flateyjarbók II, 134–135.
When St Óláfr and his retainers went past Geirstaðaálfr’s mound once, a man
asked him: “ ‘tell me, Lord, were you buried here?’ The King answered: ‘never
did my soul have two bodies, and it never will have, neither now nor on the day
of resurrection, and if I say anything else, then the common faith is not truly
implanted in me.’ ”141 The text continues: “Then the courtier said: ‘people have
said that when you came to this place before you exclaimed; “here we were,
and here we go.” ’ The king answered. ‘I never said that and I never will.’ The
King was deeply disturbed at heart; he pricked his horse and sped from
the place as fast as he could. It was easy to see that King Óláfr wished to uproot
and blot out this heretical superstition.”142
This þáttr is with no doubt permeated with literary medieval ideas and
Christian topoi.143 The motif of Óláfr Guðrøðarson as “a noble heathen”, for
instance, is an example of such topos.144 Before dying (in the first part of the
þáttr) King Óláfr calls his people to avoid sacrifices to his burial mound when
he is dead (i.e. pagan customs). This is a bit strange, since Christianity and
the new cult not yet had reached Norway at that time. He thus acts as a true
Christian, though he had never heard the gospels. This is a well-known theme in
medieval texts treating great pagan ancestors and beloved rulers from the
past.145 The þáttr also reflects the common principle of interpretation called
interpretatio christiana typologica or the “principle of prefiguration”. Events
and characters in Jewish history and in the history of other “pagan religions”
are seen as precursors of the true God’s revelation. The tradition about the
pagan king has thus been mixed up with the medieval veneration of St Óláfr.146
This Christian elaboration about St Óláfr does not preclude that some ele-
ments of the story of King Óláfr Guðrøðarson may be based on authentic
pre-Christian beliefs and customs.147 In this story, the dead king seems to be
regarded as an álfr, i.e. a mythical being, who was worshiped with sacrifices in
order to secure fertility and the crops for the people.148 In the Eddic lays the
álfar are described as a category beside the Æsir and the Vanir, indicating that
they since ancient times had a prominent position in the old religion.149 We
often meet the alliterated expression ásar ok álfar in oblique forms in these
poems.150 Sacrifices to the álfar (álfablót) are known from several sources, such
as Sigvatr Þórðarson’s Austfararvísur (see ch. 9 above). Even if these stanzas
have been regarded as somewhat suspicious, the reference to the álfablót is
usually perceived as reliable.151 However, this passage does not say much about
the character and function of this sacrificial feast and the mythical beings
called álfar. Another reference to the cult of álfar appears in Kormáks saga,
where the injured hero Þorvarðr was recommended to pour the blood of a bull
on to one of the hills (or burial mounds) (hóll einn) inhabited by álfar and to
prepare a meal for them from the meat of the bull in order to heal his injury.152
Some scholars argue that this text reflects a memory of sacrifices performed at
burial mounds153 to dead human beings called álfar,154 while others think that
it has been somewhat distorted by secondary elements.155
Evidence in Austfararvísur indicates anyhow that sacrifices to the álfar (i.e.
the álfablót) may belong to the pre-Christian religion. This evidence harmo-
nizes also with the information about sacrifices to the dead king (the álfr) in
Óláfs þáttr Geirstaðaálfs. There are some few examples of other rulers who
were designated álfr after death, for instance Hálfdan hvítbeinn in Ynglingatal
(see below).156 It is, however, much uncertain whether the worship of álfar,
in a general sense, was developed from a belief of the spirits of dead people
148 Birkeli 1938, 87–89. Cf. de Vries 1956–57, §184; E. O. G. Turville-Petre 1964, 231; Røthe 1997,
34; Steinsland 2005, 248, 342ff.
149 Cf. F. Ström 1985, 199; Lindow 2001, 109f.; Gunnell 2007 and Nordberg 2013, 248.
150 See e.g., Þrymsqviða 7 and Locasenna 13.
151 On álfablót, see e.g., F. Ström 1985, 89; Holtsmark 1992, 58; Lindow 2001, 54; Simek 2006, 8.
For critical assessments, see e.g., Lönnroth 1996 and Vikstrand 2001, 265f.
152 Hon [Steingerðr] segir: “Hóll einn er heðan skammt í brott, er álfar búa í; graðung þann, er
Kormákr drap, skaltu fá ok rjóða blóð graðungsins á hólinn útan, en gera álfum veizlu af
slátrinu, ok mun þér batna.” Kormáks saga 22, Ísl. Fornr. 8, 288.
153 Whether the concept hóll could be interpreted as ‘burial mound’ is uncertain. See
Nordberg 2013, 248.
154 De Vries (1956–57, §184) states: “Die in Hügel hausenden Wesen, die in Lebensgefahr Hilfe
bringen können, sind sicherlich als Totengeister zu verstanden.”
155 Lindow 2001, 54 and Simek 2006, 8.
156 See de Vries (1956–57, §184) for some examples.
157 I certainly agree with Nordberg (2013, 249) when he states: ”Tillnamnet Geirstaðaálfr kan
dock knappast ses som bevis för att alverna som kategori generellt skulle ha utvecklats
ur de dödas andar.” There are two lines of interpretation regarding the background of
these mythical beings: Some scholars relate the worship of álfar to a more general pre-
Christian ancestor worship. E.g. Birkeli 1938 and Steinsland 2005, 247, 344ff. Others reject
the existence of such cult and interpret the álfar as developed from ancient fertility beings.
F. Ström 1954, 34 and 1985, 199. For a qualified overview and discussion on the existence
of ancestor worship in ancient Scandinavia, see Nordberg 2013, 239–302. Nordberg sug-
gests that the term ancestor veneration would suit the research on ancient Scandinavian
religion better than the concept ancestor worship.
158 Nordberg 2013, 249, 280f.
159 Great Saga of St Olaf (eds.) Johnsen & Helgason, 730. Røthe (1997), Wellendorf (2003) and
Hultgård (2014) makes the same interpretation.
160 Anders Hultgård (2014, 34) has also produced a somewhat different suggestion in addition
to the common interpretation that Óláfr after death was elevated to the rank of gods and
became object of a sacrificial cult: “Det kan dock enligt min mening även vara så att en
lokal fruktbarhetsgud sekundärt kom att knytas till en hövding eller kung med namnet
Olav. Hur det än förhåller sig med detta, utgör berättelsen om Olav Geirstadalv ett exem-
pel på en i förkristen tid dyrkad lokalgudom.” Cf. Hultgård (2001c, 578): “Der Beiname
Geirstaðaálfr eines vorchrist. Herrschers in Vestfold deutet auf die Existenz des Kultes
einer Lokalgottheit (álfr) hin, bennant nach Geirstad, dem Haupsitz der Herschenden
Familie”. Nordberg (2013, 281) presents a similar interpretation: “Snarare kom dessa kun-
gar och stormän att övergå till en form av lokala eller regionala platsrådare (genius loci),
och lika nära som de stod till en allmän kategori anfäder, stod de förmodligen till andra
typer av lägre gudomligheter med snarlika funktioner, men utan mänsklig bakgrund.”
The byname Geirstaðaálfr appears in texts which are older than Flateyjarbók,
namely Ynglinga saga (c. 1230) and the þáttr called Af Upplendinga konungum
preserved in Hauksbók (compiled by Haukr Erlandsson before 1334).161 But nei-
ther Ynglinga saga nor Af Upplendinga konungum say anything about sacrifices
at King Óláfr’s burial mound. The byname Geirstaðaálfr also appears in The
Legendary Saga of St Olaf (13th century),162 which is based on The Oldest Saga
of St Olaf (12th century) (which only is preserved in fragments). Ynglingatal
(c. 890) knows about this king, but not his byname.163 It says in stanza 26
only that King Óláfr had his site at Geirstaðir (á Geirstaðum). In this stanza
some interesting information about this king appears, however. It says that
King Óláfr’s family belonged to niðkvísl . . . þróttar Þrós (see ch. 3 above). The
name Þrór has been related to both Óðinn and Freyr, and the expression could
thus be interpreted as ‘the lineage of the strong Óðinn (or Freyr)’, i.e. a fam-
ily who descended from the gods. Thus Ynglingatal intimates that King Óláfr
was according to ancient tradition related to a mythical sphere. King
Óláfr’s great-grandfather, King Hálfdan hvítbeinn, was furthermore called
brynjálfr ‘mail-álfr’ in Ynglingatal after being buried at Skíringssalr in Vestfold,164
but this expression may be a general kenning and not a specific designation of
the deified king.165
It was previously argued that the historical King Óláfr Guðrøðarson was
buried at the great burial mound of Gokstad in Vestfold dating back to the
Viking Age.166 Such interpretations are nowadays usually not accepted, since it
has been discovered that the grave probably is from the tenth century, i.e. too
late to fit the time of Óláfr’s reign.167 Anyhow, we cannot dismiss the possibil-
ity that the traditions about Óláfr Geirstaðaálfr reflects a memory of a cult of
deified rulers, comparable with the Greek cult of heroes.168 These dead rulers/
mythical beings were honoured with sacrifices at their great burial mounds in
order to give prosperity (til árs) (on the formula til árs ok friðar, see ch. 9 above).
It should be admitted that such cult has some affinities with the medieval
161 Heimskringla I, Ísl. Fornr. 26, 4, 79, 82–84; Hauksbók, (ed.) Eiríkur Jónsson & Finnur
Jónsson, 456f.
162 As noted above the name also appears in The Legendary Saga of St Olaf 2, (ed.) Johnsen, 1.
163 On the dating of Ynglingatal, see e.g., Sundqvist 2002, 43–47 and McKinnell 2010 and SkP I, 5f.
164 See de Vries 1956–57, §184.
165 Birkeli 1938, 87; Simek 2006, 8. It is possible that we should relate Óláfr Geirstaðaálfr’s
name also to his mother Álfhildr and his grandfather Álfarinn from Álfheimar. See
Heimskringla I, Ísl. Fornr. 26, 79ff. Cf. Gunnell 2007.
166 See e.g., Brøgger 1916; Birkeli 1938, 88.
167 Myhre 1992a, 43; 1992b, 276.
168 See e.g., Birkeli 1938, 89; de Vries 1956–57, §184; Nordberg 2013, 280ff. and 382.
cults of royal saints, which often were local and related to the burial places
and reliquaries of the holy king.169 The Greek cult of heroes indicates that
such ritual practices were not restricted to Christian contexts, however. Walter
Burkert argued that the pre-Christian cult of heroes was transformed to the
Christian cult of saints in the Mediterranean area.170 It is not impossible that
something similar also took place in Scandinavia. When the Christian cult of
St Óláfr appeared it could easily be recognized by people, who were used with
a traditional hero cult, such as the one performed to Óláfr Geirstaðaálfr.
He [Hálfdan] had been most blessed with good years’ crops among all
kings (hann hafði verit allra konunga ársælstr); and people were so
affected by his death that when they learned of his demise and that his
body was being taken to Hringaríki in order to be interred there, men of
influence from Raumaríki, Vestfold and Heiðmǫrk came and prayed, all
of them, to take the body with them to be buried in their lands; for it was
thought that he who got possession of it could expect good seasons (ok
þótti þat vera árvænt þeim, er næði). They reached an agreement in this
wise, that the body was assigned to four places: the head was laid in a
mound in Steinn in Hringaríki, but each of the others carried away their
share and interred them in burial mounds in their homelands, and all are
called the Mounds of Hálfdan.171
The account of King Hálfdan the Black is also preserved in other texts, but in
different versions.172 In Fagrskinna’s redaction A of this story, the body was
dismembered into three parts.173 The lost manuscript *A is dated to the four-
teenth century.174 In the older redaction B (based on manuscript *B)175 this
theme is lacking; it says only that Hálfdan drowned at Rǫnd í Rykinsvík and
that he was taken to Steinn in Hringaríki and buried in a mound there ( fœrðr
til Steins á Hringaríki ok þar heygðr). Neither does Nóregs konunga-tal, dated
to the last decades of the twelfth century (and probably the oldest version of
Hálfdan’s death), mention any ritual dismemberment. It says only that Hálfdan
drowned under the ice, and that he was buried in the “warlike Hringaríki”
(ok es hann í haug lagiðr á herskǫ Hringaríki).176 Ágrip is also silent regard-
ing the dismemberment. It only mentions that he was taken to Steinn in
Hringaríki and buried in a mound there.177 Historia Norwegie mentions that
Hálfdan drowned at a lake called Rand, but it does not say anything about his
burial or burial mound.178
The manuscripts of Hálfdanar saga svarta in Heimskringla, Fagrskinna A,
and Flateyjarbók (Þáttr Hálfdanar svarta) mention that the people believed
that Hálfdan was ársæll both in life and in death, and for this reason all the
“shires” wanted his body.179 According to Flateyjarbók, sacrifices were made to
the mound: “Each chieftain brought his part (of the body) to his home and
they built a burial mound in each fylki, which they called the Hálfdan mound.
Many people who had that belief performed sacrifice, until it was forbidden
by kinsmen.”180
The worship of a dead ruler’s relics at his grave resembles the cult of royal
saints. Hagiographical traditions may have influenced the medieval histori-
ans who wrote the late stories about Hálfdan. In the oldest texts, this motif
is lacking. On the other hand, the theme of (ritual) dismemberment appears
in mythical traditions, such as the cosmogonic myth of Ymir.181 This mythic
theme is probably old and need not have been affected by Christian traditions.
It has, for instance, parallels in several Indo-European traditions, and some
scholars think that it belongs to a common heritage.182 Ritual dismember-
ment of a royal (mythic) person’s body also has parallels in an Indo-European
context. Plutarch, for instance, mentions how Romulus was dismembered
in the temple of Vulcan; thereafter the pieces were distributed to the sena-
tors, who carried them away hidden in their robes.183 This narration has been
interpreted as both a cosmogonic and a sociogonic myth.184 The theme of dis-
memberment and venerating the dead body of a royal person need not be con-
sidered as a Christian notion. I thus agree with François-Xavier Dillmann when
he cautiously suggests in his thorough analysis of the Hálfdan traditions that
the motif of dismemberment might be old and could be related to the ritual
repetition of the cosmogony.185
know whether Auðr was buried at Krosshólar. In Sturlubók we read that Auðr
was “buried at the high water mark as she had ordered, because having been
baptized, she did not wish to lie in unconsecrated earth”.192 The only thing
we can say with the support of this text is that the people believed that they
would go into the hills when they died. Birkeli has no support for his statement
that “den ‘átrúnaðr’ som tok til på haugen, var knyttet til dyrkningen av Aud
personlig”.193 Nothing in Sturlubók actually supports this claim.
In my opinion it is thus possible that Late Iron Age rulers of the uppermost
elite in Norway and Svetjud were deified and honoured with cult after their
death. This cult may sometimes have taken place at their great or royal burial
mound, close to the cultic building/ceremonial hall. It seems as if the dead rul-
ers could produce welfare for living people in several ways. Whether a similar
cult appeared in Iceland is uncertain. In what follows I will give some exam-
ples from semi-mythical traditions suggesting that divine and royal ancestors
had such cult at their great burial mounds in Svetjud and Trøndelag.
Freyr took sick; and when the sickness gained on him, his followers hit
upon the plan to let few men see him, and they threw up a great burial
mound with a door and three windows. And when Freyr was dead they
carried him secretly into the mound and told the Svear that he was still
alive, and kept him there for three years. But all the tribute they poured
192 . . . ok var grafin í flæðarmáli, sem hon hafði fyrir sagt, því at hon vildi eigi liggja í óvígðri
moldu, er hon var skírð. Landnámabók, S110, Ísl. Fornr. 1, 146f. Trans. Hermann Pálsson &
Paul Edwards.
193 Birkeli 1938, 90.
into the mound—gold by one window, silver by another, and copper coin
by the third. Thus good seasons and peace endured.194
194 Freyr tók sótt, en er at honum leið sóttin, leituðu menn sér ráðs ok létu fá menn til hans
koma, en bjoggu haug mikinn ok létu dyrr á ok þrjá glugga. En er Freyr var dauðr, báru þeir
hann leyniliga í hauginn ok sǫgðu Svíum, at hann lifði, ok varðveittu hann þar þrjá vetr. En
skatt ǫllum helltu þeir í hauginn, í einn glugg gullinu, en í annan silfrinu, í inn þriðja eirpen-
ningum. Þá helzk ár ok friðr. See Ynglinga saga Prologus: “Freyr was put to rest in a burial
mound at Uppsala” (Freyr hafði heygðr verit at Uppsǫlum). Heimskringla I, Ísl. Fornr. 26, 4,
24. Trans. Hollander.
195 Óláfs saga Tryggvasonar, in Flateyjarbók I, 400–405.
196 See e.g., Lindqvist 1936, 112, 248f.
197 See e.g., G. W. Weber 1994, 9–12. Critically considered by Schjødt 2009.
198 Cf. Weber 1994, 3f. and Schjødt 2009, 569.
199 Gylfaginning 51. Cf. Vǫluspá 53; Lokasenna 42.
200 Saxo 5.16.3.
identical with Freyr.201 During his reign the “Peace of Frotho” (cf. Fróða friðr)
was established.202 This peace also included a long series of good years’ crops
which made the farmers prosperous. The peace was maintained as long as
nobody removed a golden ring lying on the hearth at the royal hall.203 When
Frotho was dead, he was carried around in a vehicle for three years, in order to
keep the information about his death secret. As long as they thought that he
was alive they continued to pay tribute to him to maintain the peace of Frotho.
Thus this narrative resembles the traditions about Freyr’s death, with the trib-
ute and the maintained peace.
Whether this motif is built on an ancient Scandinavian tradition is
somewhat uncertain. There are similar stories in the classical literature.
Herodotus reports on a deity of the Getae people (Thracian Dacians) called
Zalmoxis.204 He was a god of death, and hid himself for three years in the
underworld, while the people mourned his death. However, on the fourth
year he appeared again to the Getae and taught them about the immortality
of the soul. It has been argued that the Zalmoxis myth may have influenced
the Scandinavian tradition about Freyr’s death.205 In my opinion, the Greek
and Scandinavian traditions do not resemble each other in detail.206 In
Ynglinga saga Freyr does not return to his people after three years as Zalmoxis
does, for instance. Therefore these traditions need not necessarily be linked.
Even if the accounts of Freyr’s death at Uppsala and his burial in a mound
have the character of euhemeristic fictions, they may reflect a misty memory
of sacrifices to a dead deified ancestor or hero, which were performed at his
mound in Uppsala.207 As noted above, it is quite likely that Freyr was regarded
as the specific sacrificial god of the Svear (blótgoð svía) and it seems as if he had
a specific relation to the royal Svea kin. Some of the rulers of the Ynglinga fam-
ily, according to Ynglingatal, were called Freys afspringr and Freys ǫ́ ttungr, i.e.
Freyr’s offspring. Snorri’s account of Freyr may have developed from a tradition
mentioning that an ancestor of the royal family received cult at his mound.
201 There is support for this conclusion in Ynglinga saga 10, where Fróða friðr is mentioned
in connection with Freyr. See also e.g. de Vries 1956–57, §§460f.; Å. V. Ström 1975, 143, 148;
Näsström 2001b, 88f.
202 See also Vellekla 18; Grt. 1–6; Ynglinga saga 10–11; Skáldskaparmál 40.
203 See Simek 2006, 95.
204 Herodotus 4,94–96.
205 See e.g., Neckel 1920, 119; see also de Vries 1956–57, §460 and the literature referred to
there.
206 Cf. Schier 1968; Eliade 1972; Polomé 1995; Schjødt 2009.
207 The tradition of the Danish king Frotho mentions that he was also honoured with cer-
emonies, such as the reciting of skaldic poems, at his burial mound. See Saxo 6.1.1.
This human ancestor may have been deified after death and related to, or con-
fused with, the mythical forefather Freyr, who also secured good seasons and
peace (ár ok friðr) for the people.
208 Brennu-Njáls saga 88 for instance, mentions Earl Hákon’s and Guðbrandr’s hof in
Guðbrandsdalr, where the deities Þórr, Þorgerðr and Irpa were worshipped. In Harðar
saga ok Hólmverja 19, a sanctuary of Bláskógar in Iceland is mentioned, called hof
Þorgerðar Hǫlgabruðar.
209 E.g. Flateyjarbók I, 144, 213f., 407–409.
210 . . . þa brend bæde likneskin Þorgerðar ok Freys . . . Óláfs saga Tryggvasonar, Flateyjarbók I,
409.
211 De Vries 1956–57, §562; cf. Steinsland 2005, 293.
212 See Færeyinga saga, Ísl. Fornr. 25, 50f. The text is quoted in ch. 7 above. See also a similar
version in Færeyinga þáttr, Flateyjarbók I, 144f.
213 Jómsvíkingasaga ch. 34. (Ed.) Petersens, 115. Cf. Jómsvikinga Þáttr, Flateyjarbók I, 191.
Jómsvíkinga drápa, tells that the earl sacrificed his son to Þorgerðr Hǫlgabúðr. Skj. B2, 7.
old, and that she accepted. The fight went on and suddenly a great and unusual
hailstorm arose. It worked to the advantage of Earl Hákon, and his enemies,
Earl Sigvaldr and the Jómsvíkingar were put to flight.214
In the nineteenth century it was often argued that the intimate cultic rela-
tionship between Earl Hákon and Þorgerðr Hǫlgabúðr reflected historical
conditions. It was also suggested that Þorgerðr’s father or lover Hǫlgi (Helgi
or Hølgi) was regarded as the eponymous ancestor of the Háleygja kin, i.e. the
mythical father of the Earls of Lade. Gustav Storm, for instance, believed that
Þorgerðr was the original divine “tribe mother” of this family, but this notion
was replaced by a more “distinguished” ( finere) genealogical tree for the earls
when Eyvindr skáldaspillir in Háleygjatal made the earls into descendants
from Óðinn and Skaði.215
A more sceptical attitude towards the traditions about Hǫlgi and Þorgerðr was
seen in the twentieth century. Walter Baetke, for instance, stated that Þorgerðr
was a late fabulous being with no basis in the pre-Christian Scandinavian
religion.216 Other scholars likewise felt that Þorgerðr was a late literary figure
who did not have a place in older traditions. Since the contemporary skalds of
the Earls of Lade, e.g. Kormákr, Eyvindr skáldaspillir, Einarr skálaglamm and
Tindr Hallkelsson, never mentioned Þorgerðr in their poetry, she cannot have
played any significant role for this family during the tenth century.217
In her doctoral thesis, Gro Steinsland opposed these sceptical opinions
about Þorgerðr. According to her, there may very well be allusions to Þorgerðr
in the contemporary skaldic poetry. Steinsland referred, for instance, to Tindr
Hallkelsson’s poem (drápa) on Earl Hákon where the name Gerðr appears,
Cf. Heimskringla I, Ísl. Fornr. 26, 286 and Fagrskinna. Ísl. Fornr. 29, 129, 131, 138. In the latter
texts nothing is said about a sacrifice.
214 Scholars have in general been sceptical of these traditions about Þorgerðr’s role at the
battle of Hjǫringavágr and many have classified them as Christian inventions. See e.g.,
Baetke 1951, 45; Weibull 1948, 349–358; Näsström 1996. Cf. de Vries 1956–57, §562, note 2;
F. Ström 1954, 52; Hultgård 2002. The relation between Earl Hákon and Þorgerðr Hǫlgabúðr
is also evident in the story about the skald Þorleifr jarlaskáld rendered in Þáttr Þorleifs
jarlaskálds in Flateyjarbók I, 213ff. In addition to the saga literature we may perhaps also
find references to Þorgerðr in Saxo’s Gesta Danorum, i.e. the being called Ladgerd (Latin
Lathgertha). See Chadwick 1950, 408 and Näsström 1995, 159.
215 Storm 1885. Cf. Steinsland 2011b, 33ff.
216 Baetke 1950, 32 and 45.
217 Olrik & Ellekilde 1925–1957, 458f. Olrik and Ellekilde argued that Þorgerðr belonged to
the lower mythology (“den lavere mytologi”). According to them, the worship connected
to her was described as a primitive fisher and seaman cult. Motz (1997, 473ff.) identified
Þorgerðr as a giantess who controlled the elements of nature.
Christian attitudes to the old pagan goddess.225 In the oldest texts, the com-
pound always has the last element ‑brúðr, meaning plainly ‘bride, wife,
woman’.226 But it should be admitted that Snorri lists the name Hǫlgabrúðr
among designations of giantesses.227
They say that the king known as Hǫlgi, after whom Hálogaland is named,
was Þorgerðr Hǫlgabrúðr’s father. Sacrifices were offered to them both,
and Hǫlgi’s mound was raised with alternately a layer of gold or silver—
this was the money offered in sacrifice—and a layer of earth and stone.228
Þá er ræfrvita Reifnis
rauð ek fyrir Svǫlð til auðar,
herfylgins bar ek Hǫlga
haugþǫk sama<n> baugum.229
The kenning for gold, Hǫlga haugþak ‘Hǫlgi’s mound-roof’, indicates that there
were old traditions about Hǫlgi’s (rich) burial mound.
230 See Codex Upsaliensis of Snorra Edda, edited by Heimir Pálsson 2012, 244.
231 Røthe 2007.
232 See NGL 1, 430.
233 Røthe 2007, 50. Cf. O. Olsen 1966, 105f.
234 Þorgerðr is called Hǫrgabrúðr, for instance, in one of the manuscripts of Njáls saga,
i.e. Kálfalækjarbók. See Ísl. Fornr. 12, 214, note 2. See also Harðar saga Hólmverja ch. 19;
Flateyjarbók I, 214.
approaching him.235 Like the story about Freyr in Uppsala, this tradition indi-
cates that some kind of rituals were performed to the dead people “living” in
the mound.
There are some weaknesses in Røthe’s argumentation. It is highly uncertain
whether a burial mound could be designated hǫrgr in Old Norse. It is possible
that the term hǫrgr in manuscript AM 78 4° may only refer to ‘house’ (hus)
and not to ‘mound’ (haugr). Her assumption that Þorgerðr was worshipped
at a mound called hǫrgr, further, is based on Þorgerðr’s byname Hǫrgabrúðr.
This name might be secondary to the name Hǫlgabrúðr, which seems to be the
most common name-form in the oldest texts. Røthe’s suggestion that Þorgerðr
originally was a human woman is based on more solid ground. In Njáls saga
113 we have the genealogical information that Þorgerðr was daughter of a king
called Háleygr of Hálogaland (i.e. Hǫlgi).236 Snorri too believed that she was
the daughter of King Hǫlgi. In one manuscript Þorgerðr was buried in a mound
like her father, thus indicating that she originally was regarded as a human.
This may of course be an expression of learned euhemerism on Snorri’s part.
A connection between Þorgerðr and a burial mound also appears, however,
in Harðar saga Hólmverja 15. In this text Þorgerðr is described as the sister of
the haugbúi Sóti. If Sóti was regarded as a mound-dweller then his sister could
also be regarded as such. This tradition indicates that Þorgerðr was believed to
be an ordinary woman, most likely a queen or a princess, who was linked to a
burial mound.
All these traditions about Hǫlgi and Þorgerðr may perhaps be related to
the idea mentioned in several sources that rulers in Norway sometimes were
deified after their death. Such rulers were worshipped with sacrifices at their
burial mounds. Even if several imaginative aspects were gradually added to
the traditions about Hǫlgi and Þorgerðr, it seems as if they indicate a pat-
tern visible in other sources. Evidence thus indicates that there really was a
pre-Christian cult related to these local mythical beings (heroes/heroines)
at a great burial mound.237 Whether Hǫlgi and Þorgerðr were considered as
father and daughter or husband and wife is hard to know. The term bruðr has
the primary meaning ‘bride, woman’ but not ‘daughter’. In my opinion, this
means that their relationship should be described as the one existing between
235 Haralds saga ins Hárfagra, in Heimskringla I, Ísl. Fornr. 26, 99f.
236 Þorgerðr, dóttir Háleygs konungs af Hálogalandi. Brennu-Njáls saga 113, Ísl. Fornr. 12, 285.
Cf. Storm 1885, 128.
237 Cf. Storm 1885; de Vries 1956–57, §562; Ström 1983; Steinsland 1991, 220–226; Hultgård
2001c, 578.
husband and wife.238 Most likely they were also regarded as the protectors
(tutelary spirits) of the noble kin of Hálogaland.239 It also seems plausible that
Hǫlgi at some stage in history was regarded as the eponymous hero of this kin,
although a competing genealogy related to the Earls of Lade made Óðinn the
divine ancestor of this family. Based on these assumptions, it is not unreason-
able to presume that Earl Hákon had a certain relation to Þorgerðr. But I must
admit that the descriptions of this relation in the medieval sources most likely
are results of literary constructions.
In Ynglinga saga 36 Snorri tells the story of King Ingjaldr and the funeral feast
he gave for his father King Ǫnundr at Uppsala:
King Ingjaldr had a great banquet prepared at Uppsala for the purpose of
honouring King Ǫnundr, his father, with a funeral feast (erfi). He had
1 There are few studies on this topic, although one exception is an essay by Ottar Grønvik (1982)
dealing mostly with etymology and linguistic aspects. Cf. de Vries 1956–57, §292f.; Cahen 1921,
51f., 86–91, 150–161; Düwel 1985, 75–82. See however Sundqvist 2002, 259ff. General drink-
ing ceremonies have been treated more frequently in the research. The classical work is by
Cahen 1921. Cf. de Vries 1956–57, §292f. L. Carlsson 1965, 125–131; Å. V. Ström 1975, 234–239;
Düwel 1985, 48–118; Grønvik 1990; Drobin 1991; Hultgård 1996, 42–44; Rydving 1996; Dillmann
1997, 59–61; Sundqvist 2002, 191ff.
2 This is a revised version of the chapter “The ruler’s funeral feast (erfi)” in my dissertation,
Sundqvist 2002.
made ready a hall in no wise smaller or less stately than [the hall] Uppsalr,
which he called the Hall of Seven Kings (sjau konunga sal). In it were
erected seven high-seats. King Ingjaldr sent messengers through all of
Svetjud, inviting kings, earls, and other prominent men. To this funeral
feast (erfi) came King Algautr, Ingjaldr’s father-in-law, and king Yngvarr of
Fjaðryndaland, and his two sons, Agnarr and Álfr. Also King Sporsnjallr
of Næríki and King Sigverkr of Áttundaland. Only King Granmarr of
Suðrmannaland did not come. There, the six kings were assigned seats in
the new hall. One high-seat that King Ingjaldr had had erected remained
empty. All the host that had come there were given seats in the new hall.
But for his bodyguard and all his people King Ingjaldr had made room in
the [old hall called] Uppsalr. It was custom at that time, when a funeral
feast was prepared to honour a [departed] king or earl, that the one who
prepared the feast and was to be inducted into the inheritance, was to sit
on the step before the high-seat (á skǫrinni fyrir hásætinu) until the bea-
ker ( full) called the bragafull was brought in; and then he was to stand up
to receive it and make a vow (strengja heit), then quaff the beaker, where-
upon he was to be inducted in the high-seat (í hásæti) which his father
had occupied. Then he had come into the [rightful] inheritance to suc-
ceed him. So was done here; and when the beaker was brought in, King
Ingjaldr stood up, seized a large drinking horn (dýrshorni), and made the
vow that he would increase his dominion to double its size in every direc-
tion, or else die. Then he emptied the beaker. Now when everyone was
drunk, King Ingjaldr told Fólkviðr and Hulviðr, the sons of Svipdag, to
arm themselves and their men when evening approached, as was
planned. They went out to the new hall and put it to the torch; the hall
blazed up, and the six kings and all their followers were burned [inside].
Those that tried to come out were quickly cut down. Thereupon King
Ingjaldr took possession of all realms these kings had ruled, and levied
tribute on them.3
The text tells further about the banquet that King Granmarr of Södermanland
held for King Hjǫrvarðr of Denmark. It resulted in a companionship between
them. King Ingjaldr thus felt threatened and therefore collected a force intend-
ing to proceed against these kings. Then they fought a hard battle, but Ingjaldr
had to return to Uppsala. When Granmarr and Hjǫrvarðr made a royal progress
at their estates in the island called Selaön (Sili) in Lake Mälaren, Ingjaldr fell
upon them one night with his army. They surrounded the house and burned
them inside with all their followers. He also subjected all their dominions and
set chieftains over them. Later, the same fate struck Ingjaldr himself. When
visiting a hall at Ræningr he was surrounded by King Ívarr and his force. So he
and his wife Queen Ása made the decision which has since been famous. They
got all their men dead drunk and then set fire to the hall. The hall burned down
with all people inside.
The story of Ingjaldr is the most elaborate about any single ruler in Ynglinga
saga. In Ynglingatal there is one stanza about Ǫnundr’s death and burial,4
and one about Ingjaldr himself.5 None of them mention Ǫnundr’s erfi feast.
The latter stanza tells only that Ingjaldr committed suicide and let himself be
burnt inside a house at Ræningr. Historia Norwegiæ does not add much more
information about these kings.6 Ingjaldr inn illráði is also enumerated among
Ari’s forefathers in his langfeðgatal, but we do not get any detailed information
about him there either.7
Although Snorri’s story about Ingjaldr’s erfi feast cannot be verified by other
sources, some scholars have been inclined to regard it as partly built on old
oral traditions, perhaps with roots in eastern Scandinavia. Elias Wessén, for
instance, thought that the naming custom in Snorri’s text, e.g. Fólkviðr and
Hulviðr, is authentically Swedish, and thus indicates that the tradition had
an eastern origin.8 He argued that names with -viðr are unknown in Iceland
and rare in Norway. They are on the other hand very common in the region of
4 “Ǫnundr was stopped by the mass of falling stones at Himinfjǫll. The bastard’s hatred was
very heavy for the enemy of the Estonians, and the one who killed Hǫgni was covered under
the bones of the earth.” Varð Ǫnundr/ Jónakrs bura/ harmi heptr/ und Himinfjǫllum,/ ok
ofvæg/ Eistra dolgi/ heipt hrísungs/ at hendi kom,/ ok sá frǫmuðr/ foldar beinum/ Hǫgna hrørs/
of horfinn vas. My trans. based on Wessén 1964. Cf. SkP I, 42.
5 “And at Ræningr the one who sends smoke gushing trod down the still living Ingjaldr, when
the housebreaker on stockinged feet of flame stepped right through the man of divine
descent. And this fate most fitting seemed to all Svear for scion of kings: to die first in fiery
death and end first his own brave life.” Ok Ingjald/ ífjǫrvan trað/ reyks rǫsuðr/ á Ræningi,/ þás
húsþjófr/ hyrjar leistum/ goðkynning/ í gǫgnum steig.// Ok sá yrðr/ allri þjóðu/ sanngǫrvastr/
með Svíum þótti,/ es hann sjalfr/ sínu fjǫrvi/ frœknu fyrstr/ of fara vildi. Trans. Foote, in Jónas
Kristjánsson 1988, 95 and Hollander. Cf. SkP I, 44.
6 It has been suggested that Snorri took his information for this account from the original
Skjǫldunga saga or possibly Þáttr af Upplendingakonungum. The lacuna in Arngrim’s manu-
script may possibly have covered this story. Lindqvist 1936, 270–273; Beyschlag 1950, 80–83;
Bjarni Guðnason 1963, 90–95. Ingjaldr is mentioned in the Sǫgubrot af Skjǫldungasaga
(Ísl. Fornr. 35, 56) and he appears among Haraldr Finehair’s forefathers in Ágrip af Sǫgu
Danakonunga (Ísl. Fornr. 35, 325).
7 Íslendingabók, Ísl. Fornr. 1, 27.
8 Wessén 1964, xi, 72.
the Svear. There are also authentic place-names in the Ingjaldr account that
can be identified with toponymy in eastern Sweden, such as Myrkvafjǫrðr, i.e.
Mörköfjärden in Södermanland, and Sili (OSw Sila 1327), i.e. Selaön in Lake
Mälaren.9 At Selaön both place-names and evidence from archaeology reflect
an organized Late Iron Age society, including royal farms. Close to each other
are Tuna, Husby and Karleby, which indicate power and centrality. Other names
indicate pre-Christian beliefs and cultic practices, e.g. Fröberga, Fröslunda,
Kilfröslunda, Lytislunda, Ullunda and Odensicke.10 Snorri’s information that
King Granmarr of Södermanland and King Hjǫrvarðr took banquets (taka
veizlu) on their farms (at búum sínum) on the island Selaön (Sili) in Mälaren,
is not historical, of course. On the other hand, it is reasonable to assume that
such royal banquets may have taken place there during the Late Viking Age
and the Early Middle Ages. Snorri’s comment that Uppsala was an important
royal site, with at least one great hall building, harmonizes with archaeologi-
cal evidence (see ch. 4 above). This information is probably also built on more
reliable oral traditions from the east.
The strongest argument for the suggestion that the Snorri’s account was
based on eastern traditions is a quite obscure place-name mentioned in
Ynglingatal’s account of Ingjaldr, which could not have been common knowl-
edge for the ninth-century Norwegian skalds. The place-name included in
an expression of Ynglingatal’s Jǫfraskinna version, á rauningi, corresponds
directly to the sequence a rauniki (a Rauningi) quoted from a Viking Age runic
stone from Aspa löt in Södermanland. This name is related to Rönö hundred
(OSw Røna hundare) and the parish name Runtuna (OSw Røntuna), which
both appear in the same area of Södermanland.11
Nevertheless there are several fictious and literary elements at stake in
the Ǫnundr-Ingjaldr-account, which do not have support in direct sources. It
seems as if Snorri wanted to describe how the decentralized Svetjud, with sev-
eral district kings, was unified under one monarch, in a similar way as King
Haraldr unified Norway to one kingdom.12 Some scholars have argued that
Ynglinga saga in this passage was subjected “to typification on the basis of the
ideology of early medieval kingship” and that King Ǫnundr was depicted as an
archetype of the good medieval ruler.13 He cleared new land and built many
roads. He established great provinces and had a mansion (bú) in each province
(stórherað). He went around the land on circuit (at veizlum), arranged assem-
blies (þing) and held great sacrifices (blót) at Uppsala.14 Snorri’s statement that
Ǫnundr established great provinces and mansions in each province is not reli-
able. Such a political process must have taken several generations to build up.
Bruce Lincoln has recently empasized the literary function of the Ingjaldr
account not only in Ynglinga saga, but for Heimskringla as a whole text. It
heralds the story about King Haraldr in Haralds saga ins hárfagra. The text
thus constructs a clear parallel between Ingjaldr and Haraldr, “two men who
began their careers as district kings, but who, by virtue of vows and subse-
quent ruthless action, transformed themselves into full-fledged monarchs,
conquering and uniting their nations . . . Heimskringla’s portrait of Ingjald’s
viciousness and treachery, and also its description of the oath he swore at
the start of his reign, seems to have been carefully crafted to resonate with
later events: not only Harald Fairhair’s state-founding conquest—that much is
obvious—but also Norway’s designs on Iceland in the mid-thirteen century.
Thus, through the story of Ingjald, the text argued that national unification
under a single monarch was a project that could be accomplished only by vio-
lence and treachery, and that a king capable of this project was a wolf-hearted
miscreant, situated halfway between man and beast.”15 It is thus true that
Snorri’s account cannot be trusted as factual history, since it to a great extent
has an ideological and literary function in Heimskringla.
Even if Snorri’s account of Ingjaldr has been revised through medieval per-
spectives, methods, and “patterns”,16 it seems to be at least partly based on older
traditions, perhaps deriving from the Mälaren area. This tradition included
authentic Swedish place-names and personal names, for instance. Also the
existence of a great banqueting hall in Uppsala, and dated to the Merovingian
Period, may indicate that the Ingjaldr story in some details is old and based
on eastern oral traditions. In my opinion also the descriptions of some rituals
may reflect Late Iron Age and Early Medieval conditions, and thus be based on
older traditions.
Snorri describes in general terms the memorial drinking rites (ON minni,
full) connected with the erfi feasts. These rituals have been hotly debated in
research. The most common view among scholars is that the practice had
its origin in the pre-Christian period but continued into the Middle Ages.17
Maurice Cahen argued that minni (n.)18 replaced the older word full (n.),19
sometime between the tenth and the thirteenth centuries: “Minni s’est substi-
tué à full dans tous les sens et dans toutes les expressions de la phraséologie.”20
When the pagans toasted their gods they called it Óðins full, Freys full, while
Christians drank Krists minni and Mikjáls minni. There was thus a cultic con-
tinuity in the libation rituals, but not in the terminology: “Le rite n’a pas varié,
seul le nom a changé.”21 Klaus Düwel opposed this line of reasoning; when the
saga writers were describing pre-Christian drinking ceremonies, they actually
had in mind customs common to the medieval guilds.22 According to Düwel,
full is an old word but the meaning is unclear in the earliest sources, while
the meaning of minni—‘drink, toast in memory of a person’—first appears
clearly in the Middle Ages (1200) in Old Norse prose literature.23 The meaning
of minni ‘drink, toast in memory of a person’ appears clearly first during the
Middle Ages (AD 1200) in Old Norse prose literature.24
17 E.g. Cahen 1921; Bø 1960; B. Stoklund 1960; L. Carlsson 1965, 125; Grønvik 1990; Dillmann
1997. On mythical aspects of ritual drinking, see particularly Drobin 1991. Cf. Rydving 1996.
18 O N minni (n.) ‘1) hukommelse, evne til at mindes; 2) minde, hvad man mindes; 3) fyldt
bæger som naar man drak nogen til etc.’ Fritzner 1954, Vol. II, 701–703.
19 ON full (n.) ‘fyldt bæger’. Fritzner 1954, Vol. I, 500.
20 Cahen 1921, 179.
21 Cahen 1921, 179.
22 Düwel states: “Für die detallierte Beschreibung von Trünken auf die Götter als einer fes-
ten Einrichtung im Opferablauf rechne ich mit einer Übertragung, einer Rückprojektion
der christlichen Sitte des Minnetrinkens ins Heidentum.” Düwel 1985, 52–60, 119. Cf.
O. Olsen 1966, 60; Holtsmark 1992, 32.
23 Düwel 1985, 70–118. Simek (2006, 217) follows a similar line of reasoning: “The custom of
‘drinking in memory of someone’ only reached Scandinavia in the High Middle Ages from
Germany, and therefore it should not be seen as a heathen custom.” Other scholars have
partly moderated Düwel’s criticism. See e.g., Hultgård 1993; Dillmann 1997; Sundqvist
2005f. and 2013b.
24 Snorri distinguishes between minni and full in Hákonar saga góða. He uses full for the
gods (Óðins full and Njarðar full) and minni for kinsmen (Menn drukku ok full frænda
sinna, þeira er heygðir hǫfðu verit, ok váru þat minni kǫlluð). Heimskringla I, Ísl. Fornr. 26,
168. But Snorri is not consistent. In Óláfs saga helga he uses minni for the gods: at þar
væri minni ǫll signuð Ásum at fornum sið. Heimskringla II, Ísl. Fornr. 27, 177. Cf. Cahen
Düwel is probably right when arguing that the term minni with the mean-
ing ‘drink, toast in memory of a person’ never appeared in pre-Christian
Scandinavia. This word probably arrived in Scandinavia in a Christian context.
But even if the term minni is young and has a Christian origin, it is very likely
that some type of drinking ceremonies were held when celebrating dead kins-
men and the gods in ceremonial buildings during the Viking Age.25 The term
full or bragafull could very well have been applied in such contexts.
to early medieval conditions thus show that libations were made during the
inheritance feast. Such a ceremony ought to take place soon after death.31 It
could be held on the same day as the funeral, or on the seventh or the thirtieth
day after death or even later. Common people in Norway did not need to hold
funeral feasts; it was required only for rich farmers and rulers, as it carried legal
implications for the inheritance.32
According to several runologists, an early ceremonial funeral feast (erfi),
including beer ceremonies, perhaps occurs also in an aristocratic context in
the Tune inscription (fifth century), Norway. James E. Knirk recently read and
interpreted it thus:
A: ekwiwazafter · woduri /
– – dewita(da(halaiban : worahto : ? . . .
B: – ????<z>woduride : staina · /
– – – þºrijozdohtrirz}dalidun /
B: arbijaºsijostezarbijano
36 See Krause 1966, 166. Cf. Grønvik 1981; Marstrander 1930.
37 Krause 1966, 164f.
38 Krause 1966, 165. Cf. Peterson 2004, 19, who also presents alternative interpretations. See
also Vikstrand 2009, 10f.
39 Grønvik 1981, 162–191. Krause argued that Woduridaz was a goði who kept a hof at Tune
where Óðinn was worshipped. Before he died he was an ársæll chieftain, called *vitand-
hleif i.e. ‘he who takes care of bread’, ‘bread-securer’. Krause 1966, 166f.
40 Grønvik 1987, 54f.; cf. 1981.
41 Ísl. Fornr. 5, 13. On the great erfi feast after Hǫskuldr, see Laxdæla saga, Ísl. Fornr. 5, 73ff.
42 . . . at þá skyldi standa veizlan enn þrjár nætr; hon kvað þat vera skyldu erfi sitt. Landnámabók,
S110, Ísl. Fornr. 1, 146.
43 Flóamanna saga, Ísl. Fornr. 13, 257. Fritzner 1954, Vol. I, 349.
off Nes. Soon after this event his family arranged a funeral feast, at which they
used the jól ale for the drinking ceremonies (var þá tekit jólaǫl þeira ok snúit til
erfisins). On the first evening of the feast, when all guests were seated, Þóroddr
and his companions came into the room drenched to the skin. “Everyone wel-
comed Þóroddr and his men, and thought this a happy omen because in those
days it was believed that drowned people had been well-received by the sea-
goddess, Rán, if they came to their own funeral feast.”44 It has been argued
that this text indicates that the erfi feast, in previous periods, clearly was set in
a pre-Christian religious context.45 The detail about Rán has no counterpart
in any other tradition about pre-Christian religion, however. I must therefore
agree with Thomas A. DuBois when he states that this account “seems created
to underscore the conflict of old and new understandings of death during the
era of conversion”.46
Anyhow, we may conclude that drinking ceremonies were performed dur-
ing the pre-Christian funeral feasts in all three investigation areas, Svetjud,
Trøndelag and Iceland. Probably such customs were more common in elite
milieus, for instance, among kings, chieftains and rich farmers. The Eddic
poem Guðrúnarhvǫt 8, which probably is quite ancient, indicates such customs
when it says: “that you are drinking a funeral feast for us all” (at þú erfi at ǫll oss
dryccir).
The Kings’ Sagas and the Sagas of Icelanders report that the pre-Christian
drinking ceremonies took place in specific buildings for the purpose. Accord
ing to these medieval texts these buildings could be referred to as drinking
halls. Snorri states in Ynglinga saga 36 that King Ingjaldr “made ready a hall
in no wise smaller or less stately than [the hall] Uppsalr, which he called the
Hall of Seven Kings (sjau konunga sal)”. In this hall seven high-seats were
erected for minor kings representing the provinces of Svetjud, Fjaðryndaland,
Áttundaland, Næríki and Suðrmannaland. This story is impaired by fictional
elements (see above), but it is nevertheless interesting that Snorri thought that
44 Menn fǫgnuðu vel Þóroddi, því at þetta þótti góðr fyrirburðr, því at þá hǫfðu menn þat
fyrir satt, at þá væri mǫnnum vel fagnat at Ránar, ef sædauðir menn vitjuðu erfis síns; . . .”
Eyrbyggja saga 54, Ísl. Fornr. 4, 148.
45 See Ólafur Briem 1956.
46 DuBois 1999, 90.
the royal erfi feast should take place in a specific building which was equipped
with high-seats for the ruling elite.
In Ynglinga saga, Snorri associated the term ON salr in the place-name ON
Uppsalr with a hall building. This is one of few times the name is written in
the singular. It is usually attested in the plural, Upsalir, Upsala or Upsǫlum.47 The
meaning of the place-name element Old Swedish sal in Uppsala is in fact hotly
debated.
Snorri’s idea that ON -salr at Uppsala refers to a hall building(s), has gained
support from philologists (Stefan Brink) and archaeologists (e.g. Bo Gräslund
and Frands Herschend). Since features of a Merovingian Period hall has been
discovered at Uppsala, it has been suggested that the meaning of the Old
Swedish place-name refers to it, and that sal is used in the sense of ‘banqueting
hall’.48 The most common and traditional interpretations of sal in the name
Uppsala among specialists (e.g. Jöran Sahlgren and Mats Wahlberg) is ‘occa-
sional dwelling-place’, ‘simple building’, perhaps some kind of meadow-barn.49
There is, however, strong evidence that ON salr in skaldic poetry often denotes
‘banqueting hall’.50 In Eddic poetry its meaning is more ambivalent,51 but in
other Germanic languages equivalents of the word denote ‘hall of banquets
47 The singular form appears in Ynglinga saga, in Heimskringla I, Ísl. Fornr. 26, 66 and per-
haps in Vg 119.
48 B. Gräslund 1993, 179–184; Brink 1999a, 38f., 48f., particularly note 15; Herschend 2001, 39ff.
49 E.g. Sahlgren 1953, 26–32; Wahlberg 1994, 15. See also SOL, 267, 338, which refers to both
meanings.
50 In e.g. Sigvatr’s Austrfararvísur st. 16, the expression hilmis salr is used as a synonym for
the word hǫll in a royal context. Also in Eiríksmál the word salr appears as a synonym
for hǫll, but in a mythical-royal context.
51 O N salr does not always refer to a hall in the Eddic poetry, but to a house consisting of
one room, thus also used for a house of a poor man, eigi . . . taugreptan sal “own . . . a tow-
raftered hut” (e.g. Hávamál st. 36). It might also represent an abode of gods, giants and
other beings, e.g. í Suttungs sǫlom (Hávamál st. 104) or just a ‘ground’ (cf. Lat solum) as
in Vǫluspá st. 4 á salar steina “on the stones of the ground”. Eddic poetry thus allows for
wider connotations of the word. However, it often has the meaning of “banqueting hall”.
In Vǫluspá st. 37 the biórsalr ‘beer hall, drinking hall’ must be a banqueting hall, like the
salr in Skírnismál 16. References in Vǫlundarkviða st. 7, 16 and 30 and in Þrymskviða st. 27,
to endlangr salr “the length of the hall” seem to connote a long building. Hávamál st. 152
has a salr which must be interpreted as ‘hall’. On salr in Gylfaginning 14, see ch. 4 above).
Gylfaginning 35 reports that Frigg lived in a place called Fensalir. Cf. Brink 1996b, 255–258;
Brink 1999a, 27–30. For the expression eigi . . . taugreptan sal in particular, see Herschend
2001, 39–60.
52 O E sele refers to “hall, house, dwelling, prison”. Clarke Hall 1916, 260. OHG, MHG sal ‘die
alte Bezeichnung des germ. Einraumhauses, auch des bäuerlichen’. Kluge 1957 (1883), 617.
53 Vikstrand 2013.
54 Þeir Þóroddr gengu eptir endilǫngum setaskálanum, en hann var tvídyrðr; þeir gengu til
eldaskála ok tóku einskis manns kveðju, settusk þeir við eldinn, en heimamenn stukku ór
eldaskálanum, en þeir Þóroddr sátu þar eptir, þat til er eldrinn var fǫlskaðr; þá hurfu þeir á
brott. Ísl. Fornr. 4, 148.
55 Heggstad et al. 1993, 362.
56 Heggstad et al. 1993, 91.
57 See also Meulengracht Sørensen 2003.
during the evening when the combined wedding and funeral feast (brullaup
Óláfs ok erfi Unnar) was celebrated.58 Most likely this house (skáli) resembled
the setaskáli or eldskáli mentioned in Eyrbyggja saga 54. When funeral feasts
were held in memory of great magnates it was probably necessary to hold such
feasts in large buildings in all parts of Scandinavia. It should be noticed, how-
ever, that in Old Norse sources the Icelandic farmer’s banqueting house called
skáli was set in direct opposition of the king’s feasting building called hǫll in
Norway and Svetjud.59
As noted in chapter 6 above, several Old Norse sources indicate that the high-
seat was the ritual-symbolic centre during the ceremonies performed in the
hall building. The rituals which took place at the high-seat had both judicial
and religious dimensions.60
If the funeral feast was held according to old custom, it had to take place
in the same year as the death of the man from whom the inheritance
came. And he who had arranged the funeral feast was not to sit in that
man’s seat from whom he inherited until the men had drunk the funeral
beer. And the first night, when the men came to the funeral feast, many
drinking cups should be filled, in that way that memorial cups (minni)
are drunk nowadays; and they devoted that cup ( full) to their most pow-
erful kinsmen or to Þórr or some other of their gods, as long as they were
pagan; and in the last place there should be filled the bragafull and then
he who held the funeral feast had to make a promise at the bragafull and
so should all who participated in the funeral feast. And then the heir
would enter into the seat of the man who had left the inheritance. And
then the inheritance was fulfilled and the praise was accomplished after
the dead man, but not before that.61
Not until the heir had drunk the bragafull and made the oath was he allowed
to enter the high-seat. Snorri was also acquainted with this tradition about
King Sveinn and reproduced a similar ritual of the erfi feast in his Óláfs Saga
Tryggvasonar 35:
King Sveinn arranged a great feast, requesting the presence of all the
chieftains in his realm. He intended to honour his father Haraldr with a
funeral feast, and enter into his inheritance. . . . On the first day of the
banquet, before King Sveinn ascended the high-seat of his father, he
drank to his memory and made the vow that before three years had
passed he would invade England . . .62
61 En þá er erfin váru gǫr eptir fornum sið, þá skyldi þat skylt at gøra þau á því ári, er sá hafði
andazk, er erfit var eptir gǫrt. En sá er gøra léti erfit, skyldi eigi fyrr setjask í þess manns
sæti, er hann erfði, en menn drykki erfit. Et fyrsta kveld, er menn kómu til erfis, skyldi skenkja
upp mǫrg full með þeima hætti sem nú eru minni, ok eignuðu þau full enum ríkustu fræn-
dum sínum eða Þór eða ǫðrum guðum sínum, þá er heiðni var, en síðast skyldi upp skenkja
bragafull ok þá skyldi sá, er erfit gørði, strengja heit at bragafulli ok svá allir þeir at erfinu
væri, ok stíga þá í sæti þess, er erfðr var, ok skyldi þá fullkominn vera til arfs ok virðingar eptir
enn dauða, en eigi fyrr. Fagrskinna, Ísl. Fornr. 29, 124f. My trans. There are two versions of
this story in the manuscripts (*A and *B). I follow the older version *B, which is given in
Ísl. Fornr. 29. On the witnesses, see ch. 6 above.
62 Sveinn konungr gerði mannboð ríkt ok stefndi til sín ǫllum hǫfðingjum þeim, er váru í ríki
hans. Hann skyldi erfa Harald, fǫður sinn. . . . Fyrsta dag at veizlunni, áðr Sveinn konungr
stigi í hásæti fǫður sins, þá drakk hann minni hans ok strengði heit, áðr þrír vetr væri
liðnir, at hann skyldi kominn með her sinn til Englands . . . Óláfs Saga Tryggvasonar 35,
in Heimskringla I, Ísl. Fornr. 26, 273f. See also Jómsvíkingasaga (exists in five versions).
Flateyjarbók I, 179–181. These sagas, as well as the general part of the erfi in the Ingjaldr
account, probably depend on the Fagrskinna redaction A. Cf. Düwel 1985, 76–79.
Snorri’s and Fagrskinna’s texts thus intimate that the funeral feast concerned
not only the dead ruler and the inheritance but also the inauguration of the
successor. After the libations and the oaths, the heir was to take his place in
the high-seat. The ceremony was then complete in both the legal and the reli-
gious sense.
There are several attestations of the heir being placed in the high-seat at
inheritance ceremonies in the medieval Scandinavian literature. However,
most of them describe medieval conditions, especially in Norway. Den ældre
Gulathings-Lov says: “Now a man is dead and the heir must be seated in the
high-seat.”63 Den ældre Frostathings-Lov states: “A person should dispose his
estate as long as he is able to sit in the high-seat, both woman and man.”64
Similar ideas are expressed in other medieval sources referring to Norway.65
Flateyjarbók tells about King Hákon Hákonarson: “then he [Hákon] could go
there and require from them inheritance and the high-seat and instead be
seated there himself.”66
In Iceland too, it seems as if the high-seat played a ritual role during the
inheritance. The high-seat of the murdered chieftain Ingimundr of Vatnsdalr,
in north-western Iceland, plays an important role in Vatnsdœla saga. Þorsteinn
told his brother that they should not be seated in their father’s high-seat,
neither at home nor in the house of men, until their father was avenged.67 The
sons of Ingimundr were at home that winter and sat on the benches of lower
rank (sátu á inn óœðra bekk).68 When the father had been avenged, Þorsteinn
said that they could now enter their father’s high-seat.69
Emil Birkeli was thus partly right in his statement that it was in a concrete
sense the hásæti was inherited. The heir inherited not only his predecessor’s
63 Nu er maðr dauðr. arve scal i ondvege setiazt. Den ældre Gulathings-Lov. NGL 1, 51.
64 Siálfr scal hverr ráða fé síno meðan hann má sitia í öndvegi síno. svá cona sem carlmaðr. Den
ældre Frostathings-Lov. NGL 1, 213.
65 Cf. Ældre Eidsivathings-Lov: Sitiæ skal han i andueghe ok kona hans i hia honum. NGL 1,
404. Nyere Lands-Lov states: Nu er maðr dauðr. Þa skal arfe i anduegi setiazt. NGL 2, 86. Cf.
Birkeli 1932, 44–46; Frense 1982, 195–197; Steinsland 1991, 78–81.
66 Þa mætti hann ganga til ok krefia (þa) arfs ok or haasæti enn setiaz sialfr i stadinn.
Flateyjarbók, III, 12.
67 “Þat sýnisk mér ráð, at vér setimsk eigi í sæti fǫður várs, hvárki heima né í mannboðum,
meðan hans er óhefnt.” Vatnsdæla saga 23, Ísl. Fornr. 8, 63.
68 Vatnsdæla saga 24, Ísl. Fornr. 8, 65.
69 “. . . nú þykki mér ráð, at vér fœrim sess várn í ǫndvegi fǫður várs.” Vatnsdæla saga 27,
Ísl. Fornr. 8, 71.
dignity and property, but also the high-seat. It is therefore interesting to note
that the high-seat in Old English was called yrfestōl ‘hereditary seat, home’.70
Medieval sources often report that Christian kings and earls in Norway and
elsewhere were placed in the high-seat at their investiture.72 The most
detailed version of this ceremony can be seen in Magnús Lagabøter’s Hirðskrá
(c. AD 1270),73 which describes a ceremony called konungstekja74 that could
only be performed at the val-thing. Only a rightful claimant (kongsefni) was
taken as king;75 after a common service, all men processed to the thing place,
where there were several high-seats for the nobles. One of the seats in the
middle was the most splendid one, and the claimant was supposed to sit on
the step below it. When he was chosen king, he was raised by the nobles and
placed in the high-seat.76 The last part of the procedure corresponds to Snorri’s
and Fagrskinna’s accounts of pre-Christian ceremonies.
Genealogical knowledge is sometimes related to the ritual induction in the
high-seat in early Christian contexts. When King Magnús appointed Sveinn
Úlfsson (also called Sveinn Ástriðarson) to earl (c. 1047), Sveinn had to sit on the
foot-board (á fótskǫrinni) in front of the king who sat in his high-seat (í hásæti),
according to Snorri.77 Then the king declaimed in public, that Sveinn “has
the birth to be a chieftain” (Hefir hann ætt til þess at vera hǫfðingi). Thereupon the
king stood up and gave a sword and a shield to Sveinn, and last he set a helmet
on Sveinn’s head and bestowed the title of earl on him. Therupon the king led
the earl to share his high-seat with him (Síðan leiddi konungr jarl til hásætis
með sér). In this context Snorri also describes Sveinn’s paternal and mater-
nal ancestors: “Sveinn was the name of the son of Earl Úlfr, the son of Þorgils.
Sveinn’s mother was Ástríðr, the daughter of Sveinn Forkbeared. She was the
sister of King Knútr the Powerful by the same father and of Óláfr Eiríksson of
Sweden by the same mother. Their mother was Queen Sigríðr the Haughty, the
72 See e.g., . . . þá leiddi Sigurðr konungr Rogeir jarl til hásætis með sér ok gaf hónum konungs
nafn. Fagrskinna, Ísl. Fornr. 29, 318. . . . þá tok Sigurðr konungr í hǫnd hertoganum ok leiddi
hann upp í hásæti. Magnússona saga 8, Ísl. Fornr. 28, 247. Síðan leiddi konungr [Magnúss]
jarl [Sveinn] til hásætis með sér. Magnúss saga ins góða 23, Heimskringla III, Ísl. Fornr. 28,
37 f. . . . sattis hedherlica a Konungx stool vidh Vpsala “The Legend of St Erik”, SRS II, p. 318.
73 Hirðskrá 5–11, NGL 2, 395–399.
74 Bøe 1964.
75 See Flateyjarbók I, 239–242; Sverris saga 16 (Cod. AM 327 40), (ed.) Indrebø, 17–19;
NGL 2, 309.
76 Hirðskrá 5 says: skulu up standa byskupar oc lender men hirðstiorar oc loghmen. oc hæfia
konong up i hasæti. NGL 2, 396.
77 Magnúss saga ins góða 23, Ísl. Fornr. 28, 37f.
78 Magnúss saga ins góða 22, Ísl. Fornr. 28, 36. Trans. Hollander.
79 Fagrskinna, Ísl. Fornr. 29, 218f.
80 Dumville 1979, 73.
81 Dumville (1979, 73) bases his argument on Duncan 1975, 555. Duncan, in his turn, refers to
Chron. Fordun (Chronica Gentis Scotorum), I, 293–295.
82 Olrik 1909; de Vries 1956–57, §246; Ringstad 1991, 142.
83 Olrik 1909, 3.
A high-seat was prepared for him, and he sat on a hill, as kings do, and
lived on the Inner Island, residing in the place called Saurshaugr (“Saur’s
mound”).84
Even if this story is fictional it is likely that Snorri built on older traditions
when he stated that kings sat in their high-seats on the royal mound. Heroic
and mythical traditions also indicate that the rulers exercised their power from
a (grave) mound (ON haugr). Saxo mentions that King Høther, after being
defeated by Baldr (Balderus), travelled around in Svetjud. He was now filled
with sorrow and he was blamed for not doing his duties. Previously he used
to give his people advice from the top of a mound (consueuerat autem in editi
montis uertice consulenti populo plebiscita depromere), but now he neglected
this important societal function.85 The Fornaldar sögur likewise mention that
the kings used to sit on their mounds as an expression of their power and
authority. In chapter 2 of Vǫlsunga saga King Rerir sat on the mound (sat á
haugi) when Hrímnir’s daughter arrived at his home. Gautreks saga 8 tells
how King Gautrekr mourned his dead wife and spent all his days on her grave
mound (Hann sat á haugi hvern dag). According to Þrymskviða 6, Þrymr was
sitting on a (burial) mound when Loki arrived in the land of giants:
Þrymr must here be considered as a ruler of the giants, since he is called “the
lord of the ogres” (þursa dróttinn).
Some traditions also intimate that the inauguration rituals, and the acta
contraria ceremonies, were related to the high-seat on the burial mound or the
thing mound. A very interesting custom is mentioned, for instance, by Snorri
in Haralds saga ins hárfagra, where he describes two petty kings of Naumudalr
(just north of Trøndelag) in Norway. When they heard that King Haraldr was
falling upon them, Herlaugr went into a mound with eleven companions and
a lot of food. He ordered the mound to be closed up. The other brother acted
as follows:
King Hrollaugr went up on the mound on which the kings were wont to
sit. There he had a king’s high-seat prepared for himself, and seated him-
self on it. Then he had down pillows laid on the footstool where it was the
custom of earls to sit. Thereupon King Hrollaugr rolled himself down
from the king’s high-seat and onto the earl’s seat and gave himself the
title of “earl”. Then he went to meet King Haraldr and gave to him all his
realm, offering to become his follower and informing him about the pro-
cedure he had taken.87
In connection with this passage the author of Ágrip refers to the skaldic poem
Háleygjatal composed by Eyvindr Finnsson (skáldaspillir) in the tenth cen-
tury. The stanzas mentioning this story are unfortunately lost today.89 The
87 Hrollaugr konungr fór upp á haug þann, er konungar váru vanir at sitja á, ok lét þar búa
konungs hásæti ok settisk þar í. Þá lét hann leggja dýnur á fótpallinn, þar er jarlar váru
vanir at sitja. Þá veltisk Hrollaugr konungr ór konungshásætinu ok í jarlssæti ok gaf sér sjálfr
jarlsnafn. Heimskringla I, Ísl. Fornr. 26, 99f. Trans. Hollander. According to several schol-
ars, these degradation-rituals were based on an ancient tradition, as well as the opposite
act to place the new king in his high-seat or on the rock. In connection to Snorri’s text
Dillmann (2000c, 442) states thus: “. . . cette coutume a été expliquée, avec une grande
vraisemblance, comme une manifestation du pouvoir à la fois juridique et législatif de
ces rois, sur l’arrière plan du caractère sacré de royauté chez les anciens Scandinaves”.
Vestergaard (1990, 122) argued in the same vein as Dillmann: “The rulers of the Viking Age
were persons of royal descent who were ‘taken’ kings by popular courts and placed upon
a stone or highseat . . . However, they did not have power over their seats. Kings could be
removed, . . .”.
88 Ok hann fór þá á haug nekkvern ok veltisk fyrir ofan ok kvaðsk þá hafa velzk ór konungs nafni
ok hengði sik síðan í jarls nafni, . . . . Ágrip, Ísl. Fornr. 29, 18.
89 In all probability these customs are ancient. However, similar judicial-ceremonial acts,
known as acta contraria, are known from Roman Law, and also in Christian contexts.
Holmgren 1937b, 24f.; Olivecrona 1942.
i nformation that the king was called Hersir has been observed in research and
taken as an indication of a late addition to the tradition.90
These stories about acta contraria ceremonies do not reflect historical
events. But the custom whereby the king threw himself down from the high-
seat may correspond with the oldest preserved medieval law in Sweden, Äldre
Västgötalagen (c. 1225), where the Old Swedish term vræka ‘heave, evict, throw,
reject’ occurs:
Svear have the right to take and also reject the king.
The opposite act of vræka was thus “taking a king” (taka konung). Evidence
from the fourteenth century shows that when the Svear, in a concrete sense,
took a man as king they elevated him on the Mora stone south of Uppsala.92
This ceremony is described in Erikskrönikan (Rimkrönikan), dated to 1320–1350
but preserved only in fifteenth-century manuscripts,93 Kristoffers Landslag
(1442)94 and Olaus Magnus (1555).95 A note from 1434 intimates that each royal
election was documented by an inscription on a little stone that was placed
upon the Mora stone.96 According to a picture in Olaus Magnus’ book the king
was lifted up on to the stone. This is also attested in written documents.97 There
is a house, built in 1770, which contains Mora stenar. Nine stones are preserved,
two of which may be genuine documentation stones.98
90 Driscoll (2008, 92) writes thus: “It is highly unlikely that Hersir was ever the name of any
particular king and there is no other record of any king bearing this name.”
91 ÄVgL R 1. SSGL 1, 36.
92 For a more detailed and critical description of this ceremony, see Sundqvist 2002, 306–333.
93 See Holmgren 1937a, p. 4.
94 KrL, Kg III, SSGL 12,14.
95 . . . lapis ingens, & rotundus . . . Morasten dictus. . . . [. . .] . . . super latum, & ingentem lapidem
Rex electus stans, . . . . Historia De Gentibus Septentrionalibus I:31, 52; XI:45, 401. According
to Olaus Magnus Morasten was surrounded by twelve smaller rocks: . . . est lapis camp-
estris amplus, ab incolis perpetuo tempore Morasten appellatus, in circuitu XII. continens
lapides paulo minori forma humi firmatos. . . . [. . .]. Historia De Gentibus Septentrionalibus
VIII: 1, 243.
96 The notary public made a notation regarding the “investiture stone” of Erik of Pomerania,
which was placed on the Mora rock. See Holmgren 1954, 5.
97 In the annotations to Vadstenadiariet the elevation of King Kristoffer on the Mora rock in
1441 is mentioned. Holmgren 1937b, 21. See also Libellus Magnipolensis, SRS III: 1, 196. Cf.
Rosén 1939, 384f. See also Olaus Magnus above.
98 Cf. Holmgren 1954; Liedgren 1966; Lönnerholm 1993.
The stone of Mora was probably moved from its ancient site in the sec-
ond half of the fifteenth century. King Gustav Vasa had a search made for it
but could not find it.99 It is not improbable, however, that the Uppsala burial
mounds, and/or the Mora stone, in more ancient times, were regarded as a
kind of high-seat for all Svetjud.
According to several traditions the Danish kings were taken from the royal
kin at the thing at Viborg. As in Sweden, the king was lifted up on to a stone at
Danaerigh (Danaerygh, Danaerugh), where he was also given his royal name.100
The royal authority of Danish kings is sometimes called Hleiðrarstóll ‘the seat
of Lejre’.101 A similar custom of placing the new king on a stone or a seat is
found in sources referring to medieval Europe.102
To sum up, Fagrskinna’s and Snorri’s descriptions of the royal funeral feasts
may include a type of royal inauguration. It seems as if this ritual action could
be performed in the ceremonial hall buildings, but also outdoors at rocks,
thing mounds or the burial mound of some ancestor or relative. In cases where
the high-seat was placed on the burial mound or thing mound, the purpose
was probably to make the ceremony more public for a wider crowd. Perhaps
there was a religious argument behind this action too, especially when the
high-seat was placed on a burial mound. The high-seat in itself was the link to
the ancestors, who had previously sat in this seat. By placing the high-seat on
the burial mound, this belief was emphasized even more, since the present
ruler was in a concrete sense sitting on the (divine) forefather. When the high-
seat was placed inside the hall the link to the (divine) ancestors was perhaps
established by means of the gold foils depicting these mythical forebears
which were placed in the holes of the high-seat posts. During the performa-
tive inauguration ritual, legitimacy was thus sought from both mythical and
worldly regions.
Some scholars are sceptical about these medieval descriptions and state that
“there is no evidence of pre-Christian inauguration rituals in Scandinavia”.103
They feel that the descriptions mentioned in the Kings’ Sagas only refer to
medieval customs. Other scholars, however, have argued that the medieval
ceremony of placing the king in the high-seat was a memory of pre-Christian
customs. Elisabeth Vestergaard, for instance, has suggested that “the placing
of the future ruler, with the consent of the people, upon the stone or in the
high-seat endowed him with regal qualities and royal power. No other ritual
was necessary for granting the qualities that belonged to the seated person.”104
In my opinion, Vestergaard’s stance is quite sound. The custom of placing the
future ruler on a stone or in a high-seat is most likely older than the medieval
sources. However, I cannot agree with her statement that no other rituals were
performed to grant the qualities that belonged to a king. Most likely there was
a complex of rites and ideas associated with the inaugurations, which are not
accessible for us today because of the state of the sources. Some scholars have
searched for such rituals and notions in the mythical traditions. Jere Fleck,
for instance, suggested that some Eddic poems, especially Rígsþula, reflect
an ancient ritual education, containing genealogies, magic spells and runic
knowledge, which a prince had to receive before entering upon the office of
sacred kingship.105 Jens Peter Schjødt partly followed Fleck, as he stated that
the king received his divine character and numinous knowledge through an
inauguration, described as a rite de passage or rather an “initiation”.106 He con-
cludes that the pre-Christians Scandinavians certainly had specific ideas about
“what a future king or chieftain had to master in the field of numinous knowl-
edge and numinous powers and something of how these could be acquired”,
even if the material about consecration of kings is sparse.
In my opinion, it is quite natural for royal funeral and inheritance ceremo-
nies to be attached to inaugurations as intimated in the Ingjaldr account. Such
performative rituals often have the character of rites de passage, common in
cultures all around the world. Arnold van Gennep identified three succes-
sive and distinct factors in such rituals: separation, margin, and aggregation.107
Victor Turner proceeded from van Gennep’s framework, but went further and
focused on an aspect of these rituals that had previously been neglected, the
marginal or liminal period. In initiation rituals the neophytes are—during a
well-marked liminal period—removed, hidden, without rank or insignia etc.
“The subject of passage ritual is, in the liminal period, structurally, if not physi-
cally, ‘invisible’ ”.108 He is “betwixt and between”, neither here nor there. The
neophyte is thus regarded as symbolically dead. He is leaving his former being,
but he has yet not achieved what he will become. Only after the “exhibitions”,
“actions” and “instructions” (of the “knowledge”), i.e. the ritual actions that
comprised the initiation, can he become what he is going to be.109
If we go back to Snorri’s account we can notice that when the future king
“sit[s] on the step” in the hall, (the skǫr), the most sacred object of his father’s
office and power is displayed for him, namely the high-seat. In this marginal
position he can be perceived of as passing through the liminal phase of a rite
de passage. He is no longer a “prince” and he has not yet become the new king
or ruler. He cannot sit in his former seat on the long bench (which followed
a hierarchic order),110 yet the dead ruler’s high-seat is still empty. Not until
he drinks the bragafull and swears the oath can he be a righteous ruler and
allowed to enter the seat. These rituals transform his identity so he can become
the heir of his father’s possessions and power. Perhaps he also acquired some
numinous knowledge during this ritual process, although evidence in these
particular sources is lacking.
and Fagrskinna, may thus be part of the combined erfi and inauguration cer-
emony. Medieval laws describe how oaths were sworn during the investitures
of kings and earls in Sweden and Norway. According to Äldre Västgötalagen the
Swedish king had to swear fidelity to All-Götar and promise to uphold
the law.112 Hirðskrá states that Norwegian kings and earls swore oaths dur-
ing the ceremony called konungstekja.113 These ceremonies could perhaps be
regarded as survivals from the pre-Christian period.
Libations (bragafull/bragarfull) combined with oaths (strengja heit, heit-
strengingar) occur in Old Norse prose traditions. These rituals seem to be per-
formed in banqueting halls, especially at Yule Eve (iólaaptann). The term bragr
‘the most outstanding’ indicates that this libation (bragarfull) was made in
honour of the “high-seat man”, i.e. the one who was leading the ceremonial
drinking in the hall building. Hervarar saga ok Heiðreks narrates that Arngrímr
and his twelve sons were celebrating Yule ( jól) at the farm of Bólmr ey (Bolmsö
in present-day Småland). As custom required, oaths were made at the liba-
tions (heit strengja at bragarfulli). The son of Arngrímr, Hjǫrvarðr, promised
that he either would marry the daughter of King Ingjaldr of Svetjud, or remain
unmarried. During the spring the brothers travelled to Uppsala and Hjǫrvarðr
stepped forward to the king’s table, where he related his promise. He was then
informed that another man, namely Hjálmarr, also wished to marry the daugh-
ter of the king. This led to the fall of both Hjǫrvarðr and Hjálmarr.114 Whether
the customs described in this saga reflect pre-Christian rites is very uncertain.
There are some similar descriptions of libations (bragarfull) combined with
oaths on Yule Eve (iólaaptann) in the Eddic poem Helgakviða Hiǫrvarðzsonar.115
Similar customs also occur in Hervarar saga ok Heiðreks (redaction R).116 All the
c omposé, mais plus loin le ms. J1 donne la leçon bragar-full (en revanche, au chap. XIV
de l’Histoire de Hakon le Bon, tous les mss donnent Bragafull). Dans la leçon bragar-full, le
premier élément est le génitif sing. du mot bragr (‘le premier, le meilleur’, cf. anglo-saxon
brego ‘souverain, roi’), en sorte que bragar-full doit posséder le sens de ‘coupe du person-
nage le plus éminent’.” Dillmann 2000c, 412. Cf. Düwel 1985, 84–89.
112 ÄVgL R 1. SSGL 1, 36.
113 NGL 2, 395f. Cf. Heimskringla III, Ísl. Fornr. 28, 37f.
114 See Hervarar saga ok Heiðreks, ch. 2, (ed.) E. O. G. Turville-Petre, 2ff.
115 See also the prose passage before st. 31. On this ritual complex, see in particular H. Beck
1965, 177ff.
116 The R-version of the saga mentioned that King Heiðrekr’s large boar was used when mak-
ing promises and oaths. According to some other manuscripts of this saga (redaction H
and U), Heiðrekr worshipped Freyr (Freyja in U) and sacrificed a boar to the god during
the annual feast in February, in order to procure good growth (gefa Frey til árbotar). The
king claimed that this boar was so holy that it could give advice in all cases. On Yule Eve,
texts are late and hard to evaluate as sources for pre-Christian religion. They
may nevertheless show that the information about the rituals at Ǫnundr’s erfi
is not totally isolated in the Old Norse text traditions.
In Ynglinga saga King Ingjaldr made the vow (strengði . . . heit) that he would
increase his dominion (auka ríki) to twice its size in every direction, or else die.117
Then he emptied the cup. This ritual is directly connected with his new office
and the fortune of his forthcoming reign. Fagrskinna says that Sveinn vowed
that before three winters were gone he would travel to England and either kill
King Aðalráðr or drive him out of his land, while Earl Sigvaldr of Jómsborg
vowed that he would kill Earl Hákon in Norway.118 This story, too, features vows
associated with the deeds to be accomplished in the forthcoming reign.119 It is
probable that the connection of vows and memorial drinking to dead ances-
tors and gods referred to succession and the continuity of power and luck. The
anxious transition from one reign to another was probably a crucial phase of
the political process and it had to be resolved in the most appropriate way. The
toasts made to the family’s mythical ancestor(s) might be regarded as expres-
sions of the specific links which existed between the ruler, the god(s) and the
dead forefathers. Some scholars would probably consider them to be an “indi-
vidual consecration” (Individualweihe) to a certain god or gods.120 It has previ-
ously been argued by scholars that ancient Scandinavian inauguration rituals
consisted of a ritual education, which besides royal genealogy also consisted
of runic knowledge and magic spells (see above). It is thus possible that some
types of “esoteric knowledge”, such as information about genealogy, including
divine ancestors, played some role at this moment of the inauguration.121 As
noted above, genealogies, burial mounds and high-seats were important at
inheritance processes in general. The genealogy could have been part of the
“exhibitions”, “actions” and “instructions” which often occurred in the liminal
phase in general of initiation rituals.122 Remains of such genealogies may be
the boar had to be led to the king, so that men could put their hands on his bristles and
make promises (logdu menn þá hendur yfir bust honum ok streingdu heit). See the text
in Hervarar saga ok Heiðreks, (ed.) Helgason, 129 and the comments in Hervarar saga ok
Heiðreks, (ed.) E. O. G. Turville-Petre, 78. See also H. Beck 1965, 177.
117 Heimskringla I, Ísl. Fornr. 26, 67.
118 Fagrskinna, Ísl. Fornr. 29, 125.
119 Cf. Lincoln (2014) who also brings in the medieval perspective of these rituals as they
were presented in Heimskringla. See above.
120 Cf. Höfler 1952. Grønvik 1983.
121 Cf. Schjødt 2008.
122 Gro Steinsland states in a recent article: “Kinship was probably more important for the
elite than for other classes of population—note that for example Germanic titles of rulers
are derived from kin-groups. The importance of kin for the elite can also be seen in the
many genealogies: on the level of kings and earls ancestors were counted back to the gods
and other non-human powers.” Steinsland 2011a, 6.
church parishes. There was thus a public sanctuary centrally located in every
“pagan parish”, where people regularly gathered to celebrate pagan feasts and
perform sacrifices. There was also a type of chief sanctuaries above this level,
which concerned many people living in a wider area, such as the fylki ‘shire’.
At these places not only cultic activities took place; administrative-political
matters were also dealt with there. In Trøndelag these cultic-political central
places were identical to the sites where the shire churches (fylkiskirkjur) were
later erected (see ch. 12).
Olsen’s theory about the pre-Christian religious organization in Norway and
Trøndelag is quite obviously influenced by the medieval and clerical cultic
organization. The terminology, for instance “hedenske kultsogn”, bears a clear
Christian impression.2 In a more general sense, however, Olsen’s theory is quite
sound. Most likely there was both a private and public cult in pre-Christian
Norway and Trøndelag.3 The public cult was probably also hierarchical in some
sense; there were sanctuaries for more local settlement districts and there were
chief shrines which were important for an entire shire (fylki). The “fylki sanc-
tuaries” were situated at the central farms (No sentralgårdar), while some of
the “local sanctuaries” appeared at the hof sites, particularly in Ut-Trøndelag.4
Most likely there were also regional sanctuaries, which were important for all
people living in Trøndelag, e.g. Mære. Some public central sanctuaries may
have even been of great significance for all of Norway, for instance Lade.
Specialists in place-names, such as Lars Hellberg, have argued that Svetjud
also had a similar hierarchical cultic organization to that in Norway.5 On the
lowest level there were local sanctuaries, which were important either for a
district equal to the medieval “hundred” (Sw hund, hundare, skeppslag) or for
the “parish” (Sw socken).6 There were also regional or inter-regional sanctu-
aries, which were important for all the Svear, such as Old Uppsala. As noted
above, few scholars accept today that the Mälaren area had prehistoric dis-
tricts equal to medieval parishes.7 Hellberg’s suggested that pre-Christian
cultic organization nevertheless harmonizes quite well with Stefan Brink’s
more recent proposal of spatial or horizontal division of society in prehistoric
Svetjud. People lived in settlement districts (OSw byghþ sg.) of various sizes.
These local districts were loosely linked to some larger regional formation
called “land” (OSw land) (e.g. Uppland, Västmanland and Södermanland)8 and
perhaps an inter-regional area called “realm” (OSw rike), including all Svetjud
(Svíþjóð, Svíaríki or Svíaveldi).9 Brink writes: “For each and every one of these
settlement districts, as well as all the social formations on the regional and
inter-regional levels, there were one or more focal sites that may be character-
ized as central places (of different kinds).”10 At some of these central places
public cultic activities also took place.
Per Vikstrand has made important contributions to the discussion of pre-
Christian cultic organization. He argued that the sacral place-names in the
Mälaren region represent three different types of locations: (1) “a normal Iron
Age settlement”; (2) “the centre of the community type” and (3) “the aristo-
cratic centre”.11 For the present study it is important to make a distinction
between “the cultic centre of communal type” and “the aristocratic cultic cen-
tre”. In my opinion, both types were controlled by rulers.
We may conclude that cultic organization in both Norway and Svetjud was
hierarchical in some sense. Most likely there were both local and regional (or
inter-regional) sanctuaries in these societies. Some of them could be described
as “aristocratic” and others as “communal” sacred sites. At both types, rulers on
different social levels were in charge and controlled the ritual actions which
took place there.
8 Also smaller lands such as Arland, Soland and *Valand in Uppland may have a similar
background.
9 Cf. Brink 1997, 404–406; T. Andersson 2004.
10 Cf. Brink 1997, 405.
11 Vikstrand 2001, 410–417.
12 See e.g. V. Finsen 1888; Taranger 1890; O. Olsen 1966; Jón Viðar Sigurðsson 1999.
districts in the northern quarter, but three in each of the others.13 So perhaps
there were twelve public hof sanctuaries in the northern quarter. It has been
argued that there must therefore have been thirty-nine public sanctuaries with
goðar in Iceland after 965, besides the smaller private ones, or sanctuaries of
ease.14 But when the General Assembly was established in 930, there may have
been just thirty-six such sanctuaries, before the land was divided into quarters,
as Úlfljótslǫg intimates.15
Olaf Olsen was sceptical of the description of this cultic organization in
Úlfljótslǫg. Most of all he was critical of the concept of hǫfuðhof.16 This term
is not mentioned in Ari’s text, and appears first in the late saga literature.17
He also argued that we have many places with names including the term hof
in Iceland. Totally there are forty-four (?) known hof-names in Iceland, and in
Norway the number is even higher. There is, however, no single name including
the compound hǫfuðhof, neither in Iceland nor in Norway. Olsen says that the
term hǫfuðhof may be a literary construction by the medieval scribes, inspired
by the Christian term hǫfuðkirkja. He argued that the thirteenth-century
scribes were acquainted with the fact that there were many more hof places in
Iceland than goðar. Therefore they had to explain this situation somehow. The
idea that the goðar only were related to hǫfuðhof was one way for them to solve
this problem smoothly.
In my opinion, the term hǫfuðhof may very well be a late construction
inspired by Christian terminology. However, the phenomenon itself, where the
chieftains (goðar) were related to certain public hof sanctuaries, may be old.
Whether there were thirty-six or thirty-nine of these local “public sanctuar-
ies”, or any other number in pre-Christian time, is hard to ascertain. But we
may conclude that they were quite numerous, because of the evidence from
toponymy of hof names. Most likely they were situated at the chieftains’ farms
and spread out in the settlement districts over all Iceland, in all quarters, as
we may see on the distribution map of the hof settlements. How systematic
the organization was in the pre-Christian period and if there was a hierarchy
between different sanctuaries is impossible to say.
The sources clearly indicate, however, that public cultic activities regularly
took place at some of these hof sanctuaries. The archaeological evidence from
the Hofstaðir settlement at Mývatnssveit, for instance, indicates that ritual
slaughter on a seasonal basis took place at this hof farm.18 Several finds and the
size of the hall building indicate also that not only people living on the farm
took part in these feasts, but also people from settlement districts around this
place. We may assume that such public activities must have been organized
and performed by some local authority, perhaps a chieftain, who also took care
of the ceremonial building at Hofstaðir. Whether this chieftain carried the title
goði is hard to say, although we cannot rule out this possibility. In any case,
the cult at Hofstaðir was local and restricted mostly to the people living in the
settlement districts of the Mývatn area.
Public cult thus most often took place at certain local hof sites situated at
chieftain farms in central parts of the settlement districts around Iceland. This
cult was organized by the local chieftains, perhaps called goðar. Whether all
Icelanders had a common regional or inter-regional cult at a chief sanctuary,
comparable with Uppsala for the Svear or Mære and Lade for the Þrœndir, is
more uncertain.19 The decentralized political structure of Icelandic society
probably never gave any favourable conditions for such chief sanctuaries.
Admittedly, there was the title allsherjargoði, ‘goði for the whole assembly’ or
‘goði for the althing’ mentioned in Landnámabók,20 which may indicate a cen-
tral religious institution for all Iceland. Based on information in the Þórðarbók
redaction of Landnámabók,21 some scholars have argued that the allsherjar-
goði was supposed to consecrate the Alþingi at Þingvellir.22 Whether other
cultic activities were performed during this communal assembly is uncertain.
There is actually nothing in the source materials to indicate that Þingvellir was
an important pre-Christian cultic place, nor is there any indication of other
regional or inter-regional public cultic sites in Iceland.
The whole island held the highest sacrifice on its own account, with
human victims, otherwise each third held its own. But smaller assem-
blies held a lesser sacrifice with cattle, food, and drink. Those involved
were called ‘boiling-companions’, because they all cooked their sacrificial
meals together.26
The text does not tell where these “highest sacrifices” took place, however.
Perhaps they were held at the communal assembly place of Roma, where the
medieval general þing was held for all Gutar. The name Roma simply means
‘open place, field’.27 But the pre-Christian cultic site there may have been des-
ignated Gutnal (Gutna + al), which has been interpreted as ‘the sanctuary of
the Gutar’.28 This term appear in Guta saga, as first element in gutnalþing ‘the
Gotlanders’ general assembly’.29 In one Low German translation of Gutalagen
(manuscript dated to 1401) it is stated that this assembly took place at Roma
(czu gutnaldhing das is czu Rume).30 It is not impossible, however, that the
highest sacrifices of the Gotlanders took place in Visby, i.e. the medieval
religious centre. The first element of this compound name, Vis-, may be inter-
preted as the genitive of an appellative vi ‘pre-Christian sanctuary’.31 The Guta
saga states: “Then, later there was a sacrifice at Vi.”32 This place was situated
“below the cliffs” (firir niþan klintu) and is usually identified with present-day
Visby or a site nearby.33
It seems as if the organization of the public sanctuaries and the question of
cultic leadership in Svetjud and Norway varied. According to Old Norse narra-
tive sources, the political leader, the king (konungr), the earl (jarl) or even the
chieftain (hǫfðingi, hersir, goði) could appear in religious roles there and take
care of the sanctuaries (see ch. 5 above).34 The uppermost elite, however, were
probably related to some specific sanctuaries which had a regional or an inter-
regional significance for the Svear and the Þrœndir, for instance Uppsala and
Lade. Yet there was also one “regional sanctuary” in Trøndelag, which seems
to have been a “cultic centre of the community type”, namely Mære. The cult
there seems to have been organized by a cooperative of Þrœndir chieftains at
least during the second half of the tenth century. In what follows I will discuss
the organization of cult and cultic leadership at the inter-regional and regional
sanctuaries of Uppsala, Lade and Mære.
35 During Þórhaddr’s days (ninth century) Mære was probably a local sanctuary controlled
and managed by the local hofgoði.
36 Stenvik 1996, 88.
37 See e.g. Stenvik 1996; Lidén 1999. These scholars do not think that Mære was a chieftain
farm (sentralgård) before the Middle Ages. Probably there was some kind of settlement
there also during the Viking Age. See Røskaft 2003, 143f.
38 Stenvik 1996; Røskaft 1997, 237; 2003, 138f.; Lidén 1999, 45.
39 Cf. Phillpotts 1912–1913, 274.
40 Heimskringla I, Ísl. Fornr. 26, 315–318.
his stay the king said that he wished to journey to the place where the Þrœndir
celebrated their greatest sacrifice, and there see their customary faith (it mesta
blót yðvart, ok sjá þar siðu yðra). He was informed that he then should attend the
midsummer sacrifice (miðsumarsblót) at Mære. When King Óláfr Tryggvason
came to the inner reaches of the Trondheimsfjord and to Mære at midsummer,
all the chieftains who had opposed Christianity had arrived with the rich farm-
ers. At the assembly Járn-Skeggi (Skeggi) Ásbjarnarson of Upphaugr in Yrjar
(today called Ørland located in Ut-Trøndelag) was the leader of the Þrœndir.
He said: “It is our wish, sir king, that you make the sacrifice as other kings have
done here before you.” The king replied by killing Járn-Skeggi and desecrating
the hof building and cultic images there.
It is peculiar that Járn-Skeggi of Yrjar in Ut-Trøndelag represented the
Þrœndir at the hof building of Mære, and not a local chieftain. It seems as
if Snorri believed that the chieftains together managed the communal chief
sanctuary of Mære and together organized cult in Trøndelag. This story could
of course be seen as a projection of the thirteenth-century conditions back to
events taking place around the year 1000, and I agree that it must be partly a
literary construction. It is quite obvious, for instance, that Snorri’s descriptions
of King Óláfr’s missions follows a recurrent pattern; first the king meets the
farmers at an assembly, where he tells them about Christ, then he asks them
to convert or else they must fight the king. In these stories the king is always
successful.
It is a matter of fact, however, that Snorri had good sources when he described
the missions of King Óláfr Tryggvason.41 His most important source for this
text was probably Saga Óláfs Tryggvasonar, written in Latin by the twelfth-
century Icelandic monk Oddr Snorrason at the monastery in Þingey. The Latin
version is lost, but an Old Norse translation of it is preserved. This text is full of
miracles and wonder tales, but there are also some passages which have a more
historical character. One such passage is Oddr’s statement that the major sacri-
fice of the Þrœndir was held at Mære (ok á Mærini í Þrándheimi var hǫfuðblót).42
Another historical aspect in Oddr’s narrative may be his description of the
chieftains, who together organized and protected the pagan cult in Trøndelag
and at Mære.43 In order to eradicate the sacrificial cult there, the Christian
king had to kill a representative of these pagan chieftains and desecrate the hof
sanctuary. As in Snorri’s version, Oddr mentions that it was Járn-Skeggi of Yrjar
who was killed beside the sanctuary.44 Ágrip and Fagrskinna also vaguely indi-
cate that the cult at Mære during the tenth century was common to all Þrœndir,
and perhaps organized by the chieftains together. In the story of King Hákon
the Good’s conflict with the Þrœndir at Mære, as rendered by Fagrskinna, sac-
rificing men (blótmenn) participated on the farmers’ side.45 It seems as if these
men were identical with the Þrœndir chieftains.
There is actually one source which explicitly states that the chieftains of
Trøndelag together organized the cultic feasts at Mære. Óláfs saga Helga 109
reports that a man called Þóraldi was King Óláfr’s steward (ármaðr) on his
estate (konungs bú) at Haugr in Verdælir sometime at the beginning of the
eleventh century. King Óláfr asked him if it was true that the Þrœndir in Inn-
Trøndelag still performed heathen sacrifices. Þóraldi answered him that nearly
all the people in the interior of the Trondheim district were pure heathens in
their belief, and that only a few were baptized:
The chieftain Ǫlvir came from Egge, in Sparabú, close to Mære. That he took
care of the sacrificial feast is not sensational. In any case, we may conclude
from this quotation that in the twelfth- and thirteenth-century saga traditions
there was some information among Icelandic authors about chieftains (“twelve
men”) living in Trøndelag around 1000, who together arranged the common
cultic feasts that were held at the sanctuary of Mære. According to them this
sanctuary of Mære had a regional significance. Since the gap between some
of these texts and the actual events is not more than 150–200 years, these
44
. . . var Járn-Skeggi drepinn hjá hofinu . . . Óláfs saga Tryggvasonar eptir Odd munk
Snorrason, Ísl. Fornr. 25, 280.
45
Fagrskinna, Ísl. Fornr. 29, 80.
46
“. . . En þat er siðr þeira at hafa blót á haust ok fagna þá vetri, annat at miðjum vetri, en
it þriðja at sumri, þá fagna þeir sumri. Eru at þessu ráði Eynir ok Sparbyggvar, Verdælir,
Skeynir. Tólf eru þeir, er fyrir beitask um blótveizlurnar, ok á nú Ǫlvir í vár at halda upp
veizlunni. Er hann nú í starfi miklu á Mærini, ok þangat eru til flutt ǫll fǫng, þau er til þarf at
hafa veizlunnar.” Heimskringla II, Ísl. Fornr. 27, 180.
47 Hann [Guðbrandr í Dala] var inn mesti vin Hákonar jarls; þeir áttu hof báðir saman, ok var
því aldri upp lokit, nema þá er jarl kom þangat; þat var annat mest hof í Nóregi, en annat á
Hlǫðum. Brennu-Njáls saga 87, Ísl. Fornr. 12, 210.
48 Ok þat hit mikla hof er þar stóð hafði hann eignat ótalligum guðum. Ísl. Fornr. 25, 193.
49 Manuscript A (AM 310 4to) states thus: Þar var ok hǫfuðhof þat er mest var í Nóregi. Also in
manuscript S (Stockh. perg 4to nr. 18) it is stated that the sanctuary in Lade was great: þar
stóð mikit hof. Ísl. Fornr. 25, 193.
50 See e.g. Óláfs saga Tryggvasonar 59–60, in Heimskringla I, Ísl. Fornr. 26, 308ff.
interesting to note that in Snorri’s text about the cult in Lade no “cultic leaders/
specialists” are mentioned other than the earl and the king.
Snorri supports his statements about Earl Sigurðr’s generous sacrifi-
cial feasts at Lade with an interesting stanza from Kormákr Ǫgmundarson’s
Sigurðardrápa (AD 960). Judging by this stanza, religious feasts were held at a
sanctuary (vé), which was superintended by Earl Sigurðr. The earl is here called
vés valdr, i.e. “the keeper, owner, ruler, or custodian of the vé sanctuary.” The
Old Norse word valdr (valdi) refers to ‘ruler, owner’ (‘som har magten (over
noget), styrer, besidder),51 and vé to ‘sanctuary’. Most likely this sacred site was
located at Lade, as Snorri assumed. We may thus conclude that the Earls from
Hálogaland owned a great and famous hof sanctuary at Lade. This sanctuary,
at least during Earl Hákon’s reign, probably had an inter-regional significance.
Together with Skíringssalr in Vestfold it was one of Norway’s most important
cultic sites. At Lade prominent guests from all Norway as well as chieftains and
farmers from Trøndelag celebrated religious feasts together, under the surveil-
lance but also the hospitality of the earls.
demons at Uppsala turned blind without gaining help from any deities.”53
Several researchers have been sceptical about Adam’s information on that
point.54 The term sacerdotes evokes the differentiated priesthood of the
Roman Catholic church. Adam may be indulging in an interpretatio romana
or Christiana, in that he interpreted Scandinavian culture and religion through
his Roman-Christian glasses. A centralized differentiated priesthood with reli-
gious specialists who were exclusively devoted to their cult probably never
existed in pre-Christian Svetjud.55 On the other hand, there may have been
some kind of “religious specialists” at the calendrical religious feasts, who were
responsible for specific cultic functions at some sanctuaries. The rituals per-
formed at the public cultic feasts in Uppsala, for instance, were probably very
comprehensive and complicated (see ch. 5 above). Perhaps some “specialists”
were needed for different aspects and separate rites during the public sacrifices
there (see ch. 5 and 9 above). As noted above, there are terms and designations
preserved in Swedish place-names, carrying meanings supporting this notion,
for instance OSw *gudhi, *lytir, and *vivil. Whether these “specialists” were per-
manently employed “priests” at the Uppsala sanctuary is uncertain. Most likely
they had many functions similar to the bailiffs’ and the king’s representatives’
assignments at other royal farms and central places in Svetjud. They took care
of the royal property on behalf of the king. Since they were representatives
of the leading stratum of society, they also temporarily played central roles at
the cultic feasts. As compensation they were protected by the king. The rela-
tionship between the kings and “the religious specialists”/ “cultic leaders” at
Uppsala could perhaps be described as a kind of patron-and-client-system, i.e.
the reciprocal order admitting both clients and patrons to gain benefit from
the other’s support. In my opinion, it was the king who had the overarching
responsibility for the public sacrificial feasts there. It was also he who per-
formed certain important rituals at these gatherings.
It actually seems as if Adam of Bremen was acquainted with the idea that it
was the king who officiated at the sacrifices in the common cult in Uppsala.
Scholion 140 says that King Anunder was driven from the public feast and
assembly (concilium) at Uppsala, since he would not perform as a cult leader
at the sacrifices (see ch. 9). This image also agrees with the information given
in Old Norse literature. According to the U-version of Hervarar saga (c. 1300),
for instance, the Svear removed King Ingi Steinkelsson (1080–1110) from office
53 Quidam e sacerdotibus, qui ad Ubsolam demonibus astare solebat nequiequam iuvantibus
diis factus est cecus. Adam IV,28.
54 See e.g. Hultgård 1997, 19−20 and Sundqvist 2003a; 2003b; 2007.
55 See Sundqvist 2007.
since he would not make sacrifices on their behalf at the assembly which prob-
ably was held at Uppsala.56 Like Anunder, Ingi had converted to the Christian
faith. These sources indicate that the king, according to the tradition, had con-
nections to the cult in the Uppsala sanctuary. However, since some of these
kings had converted to Christianity they did not take part in the sacrifices
there. During the eleventh century the Christian Svea kings appeared at other
places, such as Sigtuna. Perhaps some of the pagan bailiffs and old royal repre-
sentatives took care of the Uppsala sanctuary as deputies for the absent king
during the final phase of the pre-Christian religion. It is possible that it was
these officials that Adam confused with priests.
In my opinion there are arguments supporting the notion that the pagan
Svea rulers were closely connected to the Uppsala sanctuary in the Late Iron
Age and that it should be described as an “aristocratic cultic site”. The sanc-
tuary was thus located on the royal estate. The most important argument is
the strong evidence of archaeology indicating the continuity of outstanding
elite groups in Uppsala from the Merovingian Period to the Early Viking Age
(c. 550–800/900). Archaeological evidence related to the top elite is, however,
lacking from the last part of the tenth century onwards. It is possible that this
reflects some new economic and political situation, where the new Christian
kings now appear at other places. In my opinion, Uppsala was an important
pre-Christian sanctuary up until c. 1060, even if royals did not take part in the
religious activities there. Adam’s text and other sorces clearly indicate that (see
ch. 4 above).
Pre-Christian kings and earls in Svetjud and Trøndelag seem to have owned
and controlled some central sanctuaries, such as Uppsala and Lade. These
sanctuaries most likely had a regional and perhaps even an inter-regional sig-
nificance. They were erected at aristocratic farms. At Mære in Inn-Trøndelag
there was a regional sanctuary which was organized in another fashion, at least
during the second half of the tenth century. It could be described as a “cultic
centre of the community type”. This sanctuary may have been administered by
a cooperative of Þrœndir chieftains.
In the written sources we have descriptions of cultic sites where several deities
were worshiped. The most famous of these is probably Adam’s description of
56
Hervarar saga, (ed.) Jón Helgason 1924: 160. See also Orkneyinga saga 35.
the Uppsala sanctuary (see ch. 4 above). He states that the statues of Óðinn,
Þórr and Freyr were worshipped in the temple there. In my opinion, it is no
coincidence that Adam mentions these specific gods in his description: place-
names and archaeological finds confirm that they were worshipped in this part
of Sweden.57 Adam’s information about the gods may thus be reliable.
The distribution of theophoric place-names in the Mälaren region is very
interesting (see fig. 28). The Freyr names seem to dominate in the south-western
part of the area, more exactly in south Fjädrundaland, south Västmanland, and
Södermanland.58 Place-names with Þórr and Óðinn, on the other hand, appear
throughout the entire area, with a minor dominance in the eastern parts. In
the light of the place-names and Adam’s text, it thus seems as if the common
cult at the Uppsala sanctuary was some kind of synthesis, where different cul-
tic groups, living in different areas of the Mälaren region, were supposed to
Figure 28 The distribution of theophoric place-names in the Mälaren region. Made by Fideli
Sundqvist and based on Per Vikstrand 2001.
57 See Vikstrand 2001, 55ff., 115ff., 141ff., 167ff. and ch. 4 above.
58 Vikstrand 2001, 57ff.
assemble.59 It is possible that this may have been connected to a strategy fol-
lowed by the regional ruler who resided at Uppsala. In order to gain control
over the people living in the Mälaren territory, he promoted the Uppsala cult.
In the ceremonial space of the sanctuary the ruler was surrounded by images
of gods. These images had a symbolic value for several of the groups and indi-
viduals that gathered in Uppsala at the common cultic feast. Here the king
could show that he had a good relation to all these deities. He also took charge
of the sacrifices made to them at Uppsala.
One problem with this hypothesis is that the place-names may reflect differ-
ent periods from those we may isolate by means of Adam’s information and the
written sources. It is most probable, further, that on a local level the cult was
not restricted to just one deity. In polytheistic systems the principle is that the
individual deities guarded different spheres of life and society. Some deities
may have been connected to nourishing and reproductive functions, others
to physical and military protection, while some watched over intellectual
aspects.60 One individual thus needed several gods with different functions to
manage everyday life. However, it is quite possible that a single god may have
been a unifying symbol for an individual, a family, an occupational group or
a cultic community, etc. For instance, Freyr is called blótgoð Svía “the sacrifi-
cial god of the Svear” in Óláfs saga Tryggvasonar,61 and Óðinn is designated
Gauta-týr ‘the god of the Götar’ in the skaldic poem Hákonarmál (960).62 In
medieval sources we hear about pre-Christian Icelandic goðar who had a spe-
cific relation to a certain god. The goði family called Þórsnesingar were related
to Þórr in Landnámabók and Eyrbyggja saga,63 while the goði Hrafnkell was
related to Freyr in Hrafnkels saga Freysgoða.64 This personal devotion to one
god is sometimes expressed in the term fulltrúi in the medieval texts (see ch. 3
above). Most likely such personal devotion also existed in the Mälaren area in
the Viking Age. A person who wore a Þórr’s hammer around his neck in this
region was probably devoted to the thunder god. Therefore, it is not unreason-
able to think that the three deities in the “temple of Uppsala” actually were
part of a ruler strategy, the purpose of which was to gather all local cultic com-
munities and individuals at this sanctuary, in order to display the sovereignty
of the leader by means of ritual expressions. On such occasions the ruler could
also take tribute from his subjects, as suggested by Adam.
65 See e.g. Thrane 1998; Hedeager 2001; 2011; Jørgensen 2011.
66 Kousgård Sørensen 1985, 133ff.; 1992, 233. Cf. SOL, 97. For some critical considerations, see
Brink 2011. He states that -hem could mean ‘an area’, or ‘a district’ as well.
67 For a more thorough argumentation of this, see Kousgård Sørensen 1985, 135.
Most likely several gods were also worshipped at the central sanctuary of
Lade. According to Snorri’s Óláfs saga Tryggvasonar 59, for instance, King Óláfr
dismantled the hof there, and removed all the property and all the decorations
from it and from the images of gods. Several other Norse prose texts give a
similar image (see ch. 7 above). The prose texts frequently mention that indi-
viduals or families on different social levels were devoted to a certain god in
Norway. Earl Hákon had Þorgerðr as his tutelary god at the end of the tenth
century (see ch. 12 above), while Guðbrandr of Hundþorp, in Guðbrandsdalr,
had Þórr as his favourite deity, still at the beginning of the eleventh century.68
Theophoric place-names around the Trondheimsfjord likewise indicate a cult
of a specific deity at local sanctuaries, such as †Óðins(s)alr, used for Beitstad
or Sør-Beitstad, Nord-Trøndelag.69 It is thus possible that the common cult of
many gods at Lade could be seen as a ruler strategy to gather different cultic
groups and individuals of Trøndelag, and elsewhere in Norway, at the central
sanctuary of Lade.
68 Óláfs saga helga 112, in Heimskringla II, Ísl. Fornr. 27, 184.
69 Sandnes & Stemshaug 2007, 341. Magnus Olsen mentions several theophoric place-names
in Trøndelag, e.g. Onsøien (*Óðinsin/ *Óðinsey) at Byneset in Orkdælafylki. M. Olsen 1915,
268. Since I cannot find some of them in Norsk stadsnamnleksikon I guess that his inter-
pretations today sometimes are considered as unreliable.
As noted in chapter 6 above, the Old Norse sources do not only describe cultic
buildings and feasting halls situated in the real world. There are also traditions
describing ceremonial buildings situated in the mythical world, among the
gods. These mythical descriptions have many affinities with the ritual struc-
ture of the real ceremonial buildings, indicating that they had their Sitz im
Leben in such milieus. Most likely they represent some type of ideal ceremonial
buildings, and thus they may contribute information relevant for this study.
In what follows I will discuss the martial dimension of the ceremonial build-
ings. This dimension appears in both mythical and real contexts. In Valhǫll, for
instance, the presence of einherjar ‘single combatants, champions (especially
the dead warriors who dwell in Valhǫll)’ is striking.1 In my opinion, the concept
of Valhǫll should be related to certain hall milieus in Norway and Sweden (and
Denmark), where warrior lords with warrior bands probably appeared quite
frequently. Such comitatus institutions and milieus were most likely less com-
mon in Iceland. For rulers in Svetjud and Trøndelag, the martial and eschato-
logical Valhǫll mythology was very important when constructing power.2 This
martial mythology had its setting in an ideal hall milieu for retinues. When
it was recited in Viking Age Iceland, it probably appeared in other social and
ritual contexts.
The reports from Grímnismál and Gylfaginning3 suggest that Valhǫll refers
to a large aristocratic banqueting hall which is located in the central parts of
Ásgarðr at a place called Glaðsheimr. It is gold-bright (gullbiarta) and it “rises
peacefully, seen from afar” (see ch. 6 above). Grímnismál 9 describes the hall
further thus:
1 On the mythical conception of Valhǫll, see e.g. Neckel 1913; Ellis Davidson 1943; Nordberg
2003; Dillmann 2007; Hultgård 2011.
2 The concept of eschatology refers to both the afterlife (individual eschatology) and the end
and renewal of the world (cosmic or universal eschatology). Hultgård 2011, 298. Cf. Widengren
1969.
3 Grímnismál 8–10, 18–26 and Gylfaginning 37–40.
It is said that Valhǫll has “spear-shafts for rafters” and is thatched “with shields”.
Essential structures of the house are thus made of weapons. When the skald
states that “mail-coats are strewn on the benches” the audience understands
quite clearly that this house was intended for warriors. It is also stated in this
poem that Hroptr (i.e. Óðinn) “chooses every day those dead in combat” (st. 8),
i.e. the einherjar (einheri sg.). It seems as if these einherjar were brought to
Valhǫll ‘the hall of the slain’5 by the valkyrjur ‘those who choose the slain’.6
In Valhǫll these warriors had a paradise-like life, where they spent their days in
the battlefield, but in the evenings they were alive again. In Vafþrúðnismál
(41) the giant Vafþrúðnir replies thus to one of Óðinn’s question:
In Valhǫll the valkyrjur offer the einherjar mead from the goat Heiðrun’s horn,8
and they eat the meat from the constantly renewed boar Sæhrímnir each day.
This boar meat is prepared for the warriors in the cauldron called Eldhrímnir
by the cook Andhrímnir (see ch. 6 above).9 At Ragnarǫk 800 einherjar will
march out of each of the 540 gates of Valhǫll on the gods’ side, in order to
fight against the demonic and destructive power of the wolf (i.e. Fenrisúlfr).
Grímnismál 23 states thus:
4 “It is very easy to recognize for those who come to / Óðinn, to see how his hall’s arranged./
the hall has spear-shafts for rafters, with shields/ it is thatched. Mail-coats are strewn on the/
benches.” Trans. Larrington.
5 The term Valhǫll probably derives from the ON words valr ‘the corpses lying on the battle-
field’ and hǫll ‘hall’. Simek 2006, 346ff.; Hultgård 2011, 301.
6 The term valkyrja (sg.) derives from valr and the verb kjósa ‘to choose’, thus meaning ‘those
who choose the slain’. See Simek 2006, 349.
7 “All the einheriar fight in Óðinn’s courts/ every day/ they choose the slain and ride from
battle;/ then they sit the more at peace together.” Trans. Larrington.
8 Grímnismál 25, 36; Gylfaginning 39.
9 Grímnismál 18; Gylfaginning 38.
Magnus Olsen suggested that the description of Valhǫll in this stanza actually
reflected experiences gained by Scandinavians who had visited the Colosseum
in Rome, with its many combats of gladiators.11 This theory has been doubted
by more recent researchers. Rudolf Simek suggests that the “number of 800
times 540 = 432,000 einherjar mentioned in Grímnismál can possibly be traced
back to Hellenistic influence”.12 He is also open for other interpretations.
Andreas Nordberg states that the skald probably never referred to a cosmic
number. His intention was just to show that the building was huge and that
the number of einherjar was great.13
In the perspective of the present study it is important to emphasize that
Valhǫll in the mythic descriptions is some kind of a large banqueting hall dec-
orated with weapons, and intended for ceremonial feasts. Óðinn invites the
slain warriors to this place. At Ragnarǫk these warriors will march out and fight
on the gods’ side against the giants.
15.2 The Warrior Halls in Late Iron Age Scandinavia and Valhǫll
The situation reflected in the mythical world may be associated with condi-
tions in some human halls discovered by archaeologists in eastern Scandinavia.
One human three-aisled warrior hall has been found at the Garrison of Birka
in Lake Mälaren, on the slope north-west of the stronghold called Borg.14 This
Viking Age building was divided into two large rooms. There were fireplaces
in both rooms, and the building had two entrances. Many finds were made in
the northern room, such as spearheads, shield bosses, webs of rings from mail
coats, as well as mountings and keys for chests. These finds indicate that this
10 “Five hundred doors and forty/ I think there are in Valhǫll;/ eight hundreds of warriors
will go together from one door/when they go to fight the wolf.” It is possible that hundruð
refers to a “long hundred” = 120. Trans. Larrington.
11 M. Olsen 1938 (1931).
12 Simek 2006, 347.
13 Nordberg 2003, 229.
14 Holmquist Olausson 2001; Holmquist Olausson & Kitzler Åhfeldt 2002; Carlie 2004,
172ff., 291.
room was a dwelling place for the warriors of Birka. In the southern room
many animal bones and knives were found, including fragments of glass and
a dragon head made out of bronze. This room was interpreted as an assembly
hall for the warriors, used for feasts and religious ceremonies. In the western
part of this room the most exclusive objects were found. They were concen-
trated in the floor layer in a restricted area. For instance, the dragon head was
found there, a sword handle, and most of the glass fragments. According to
Lena Holmquist, this was the place of the high-seat, i.e. the place “where the
king sat when visiting the Garrison”.15 The sacrificial depositions in the post-
hole of the eastern roof carrier also support the interpretation that warriors
occupied the hall at Birka. In this hole two spearheads made of iron were
found, as well as one decorated ferrule made out of bronze for a sword sheath,
one decorated Þórr’s hammer made of horn, two Arabic coins and about 40
sawn-off comb cases.16 A lot of weapons were also found at other hall buildings
in the Mälaren region, for instance at the halls of Helgö and Fornsigtuna.17 An
enormous amount of weapons were also discovered outside the cultic house
at Uppåkra.18
It could not be ruled out that the weapons in these halls may have been used
for decorating the inside of these buildings, which could be compared with the
appearance of Valhǫll, as suggested by Grímnismál.19 That real warrior halls in
Norway sometimes were decorated with weapons is also attested by Sigvatr
Þórðarson’s Austrfararvísur st. 16. In this stanza the skald praises one of the
retinue-halls of King Óláfr Haraldsson in the early eleventh century:
15 “. . . där kungen satt när han besökte Garnisonen.” Holmquist Olausson 2001, 15.
16 Holmquist Olausson 2001. So too Carlie 2004, 174, 291.
17 Herschend 1995; 1998; Damell 1991; 1993.
18 L. Larsson & Lenntorp 2004.
19 Cf. Nordberg 2003, 203f.
20 Finnur Jónsson translated the stanza thus: “Hirdmændene, som mætter ravnen, udstyrer
kongens sal med hjælme og brynjer; her ser jeg der findes rigeligt af bægge dele på
In Óláfs saga helga 57 Snorri states that Óláfr Haraldsson had a “retinue-hall”
(hirðstófa) at his royal residence (konungsgarðr) in Niðaróss, Trøndelag. This
hall had two doors at both ends, and the king’s high-seat (hásæti) was situated
in the middle of the hall. Opposite the king’s high-seat was the “lesser high-
seat” (ǫðru ǫndugi). This was the seat of the king’s marshal, and next to him the
noble guests were seated. Ale was carried around the fire during the ceremo-
nial drinking feasts and when noble guests visited the king.21
In my opinion, Andreas Nordberg is on the right track when he states that
the social context of the motif of Valhǫll is found in the band of the warrior
aristocracy.22 The warrior band consisted of free men, usually from noble
families, who voluntarily subordinated themselves to a warrior lord in order
to do military service. As long as they belonged to the leader’s band they also
constituted part of his family and household. The chieftain or ruler expected
unlimited support and loyalty from his men, and in return he supplied them
with food, drinking, lodging and weapons. He rewarded them with ceremonial
feasts, at which he also distributed gifts. These feasts took place in the ruler’s
hall, which seems to have been a crucial element in the organization of warrior
bands. Among these groups there was an eschatological belief that those war-
riors who were killed in battle were consecrated or sacrificed to Óðinn. After
death they were taken to his residence called Valhǫll, which was situated in the
world of the gods. There they resided in lustre, light, and warmth, with the gods
and other warriors slain in battle, i.e. the einherjar.
There is thus some kind of relationship between the mythical Valhǫll and
aristocratic warrior halls. In the halls at Birka, Fornsigtuna and Helgö, for
instance, where lots of weapons have been found, the mythical complex related
to the Valhǫll may have played an important role. However, all cultic halls in
the investigation areas may not have had such connections and associations.
In my opinion, this is due to the regional and contextual differences and varia-
tions. The hall at Hofstaðir, Mývatnssveit, for instance, has almost no weap-
ons or other signs of the presence of a warrior aristocracy. At such halls the
myths connected with Valhǫll may have appeared in other social contexts and
played a more literary function, than being embedded in a warrior ideology.
v æggene,—ti ingen ung konge har prægtigere husprydelse at rose sig af; det er udenfor
tvivl; hallen er i enhver henseende herlig.” Skj. B1, 224. Page (1995, 53) translated it thus:
“Men of the king’s guard/ Who feat the wound-swan/ Array the prince’s hall with hel-
mets,/ Mail-coats. Here they hang./ No other ruler/ Can boast richer wall-hangings./ No
fear of that./ Glorious is this hall.”
21 Óláfs saga helga 57, in Heimskringla II, Ísl. Fornr. 27, 72.
22 Nordberg 2003, 85ff., 300.
The c ommunity around Hofstaðir was occupied with livestocking and farming,
not with war (see below).
It is quite striking that the oldest skaldic poems referring to the Valhǫll
motif appear in royal contexts in Norway, where the institution of warrior
bands was essential (see further below). In the tenth-century poems Eiríksmál
and Hákonarmál, history and myth are combined. There we get the image that
Valhǫll was a banqueting hall situated in the divine world, but also intended
for human fallen kings and warriors from Norway.23 Eiríksmál (st. 3–4) calls it
Óðins salr ‘Óðinn’s hall’.24 In stanza 1 we hear that the einherjar must clear up
the hall, since an outstanding person (vísi) is arriving (i.e. the fallen Norwegian
king Eiríkr blóðøx). The valkyrjur must get up to strew the benches and to clean
the drinking vessels and to bring wine (vín bera). In stanza 5 Sigmundr asks
Óðinn why he is expecting Eiríkr rather than other kings. Óðinn answers that
Eiríkr has reddened his sword in many countries (i.e. he was a great warrior
lord). Sigmundr replies: “Why did you then take the victory from him, if you
thought him able (snjallr)?” Óðinn answers by referring to the uncertainty of
knowing when the grey wolf will attack the homes of the gods (i.e. Ragnarǫk).
The other poem, Hákonarmál, describes Valhǫll as grœnna heima goða “the
green home of the gods” (st. 13).25 King Hákon the Good of Norway is wel-
comed to this place after his last battle. His arrival is announced to Óðinn by
the two valkyrjur, Gǫndul and Skǫgul. There he enjoys the peace and truce
of the einherjar (einherja grið) and he drinks beer together with the Æsir (þigg
þú at ǫ́ sum ǫl). This poem tells us that the reason King Hákon was taken to
Valhǫll is that he must support the gods when the wolf breaks free from his
bonds at Ragnarǫk (see st. 20).
Both Eiríksmál and Hákonarmál refer to the theme of a cosmic eschatology,
i.e. that the Norwegian warrior kings will be in “the ranks of the good forces at
the side of the gods during the great battle at the end of time”.26 This was prob-
ably a glorious task awaiting brave kings and warriors in the upper classes in
Viking Age Norway.
Perhaps the Eddic poem Grímnismál (see above) should also be related
to royal contexts in Norway. According to Helmut de Boor, this poem was
23 On these important poems, see e.g. Nordberg 2003, 48ff.; Hultgård 2011, 302ff.; SkP I, 171ff.
Even if these kings have been regarded as Christians in the medieval traditions, the poems
were probably made by contemporary pagan skalds. These skalds intentionally placed
these kings in pagan contexts, perhaps for political reasons. King Hákon may furthermore
have turned to paganism in the last part of his life. On this problem, see Hultgård 2011.
24 Skj. B1, 164ff.
25 Skj. B1, 57ff. Cf. SkP I, 187.
26 Hultgård 2011, 322.
c omposed by one of the skalds appearing in the circle around Earl Hákon at
Lade.27 I will not rule out the possibility that the Valhǫll motif therefore was
more relevant in the context of aristocratic halls in Norway and Sweden where
warrior lords with warrior bands probably appeared quite frequently, than in
Iceland, where such military institutions were less common or totally lacking.
At least it had a different social context in Iceland compared to the setting it
had in Viking Age Norway and Svetjud. In what follows I will look a little more
closely at this warrior institution and outline its distribution in Scandinavia.
27 See de Boor 1930; cf. Meulengracht Sørensen 1991a; 1993, 74–108; Nordberg 2003, 213.
28 E.g. Enright 1996; Norr 1998, 136f. Sawyer & Sawyer 1993, 92–95; Jón Viðar Sigurðsson
2008a, 117–125.
29 Clark Hall 1916, 158.
30 See Rahmqvist 1986; DMS 1.2: 180; DMS 1.7: 218–236; Damell 1991; Ekman 2000.
husabyar.31 When travelling around his land the ruler visited these husabyar
and other royal estates. He claimed hospitality from the brytar for himself and
his retinues.32 This system was called OSw gæstning (cf. ON veizla).33 Most
likely it was in similar contexts that the halls of the warrior bands also played a
certain role in Viking Age Svetjud.
The system of gæstning is actually indicated in sources deriving from
the Late Viking Age. It seems, for instance, that the Svea king had a bryti at the
Viking Age settlement of Adelsö, very close to Birka. The runic inscription of
Hovgården (U 11) from the eleventh century reads as follows:
raþu:runaR:ret:lit:rista:toliR:bry[t]i:i roþ:kunuki:toliR:auk:gyla:
litu:ris···-:þaun:hion:eftiR···k:merki *srni···*hakun*baþ:rista
Rað þu runaR. Rett let rista ToliR bryti i Roð kunungi. ToliR ok Gylla letu
ris[ta. . .], þaun hion æftiR [si]k(?) mærki . . . Hakon bað rista.
You read the runes! Tolir had them cut right, bryti to the king in Roden.
Tolir and Gylla had (these runes) carved, this pair after themselves as a
memorial . . . Hakon bade carve.
31 Schlyter 1887, 97 and 414f.; Hjärne 1951; Lindkvist & Ågren 1985, 20, 38; Brink 2003.
32 The brytar were originally unfree people, but gradually they gained power. OSw bryti
means “som utdelar maten till de öfvriga tjenarne”. Medieval Swedish laws also describe
the royal office of OSw lænsmaþer, i.e. a person who administered great farms (Sw
huvudgårdar) in a skeppslag, hundare, härad etc. Lindkvist & Ågren 1985, 20, 38.
33 Lindkvist 1995; 2009.
34 See Brink (2003, 106) for reading and interpretation.
35 Hjärne 1951, 81. Cf. Brink 2003, 106.
36 See T. Zachrisson 2011a. Cf. T. Andersson 2004, 10, note 3.
located on a plateau there. Perhaps this hall was also used as a dwelling for the
king’s retinues.37
Another possible example of a Viking Age royal bailiff, with a hall for retain-
ers, may be related to the town of Birka, on Björkö in Lake Mälaren, close to
Adelsö. Rimbert mentions in Vita Anskarii (AD 870) that several Svea kings vis-
ited Birka, for instance King Bjǫrn and King Óláfr.38 At Birka the king had a
royal counsellor (consiliarius regis), who also was the prefect of the site (prae-
fectus vici, loci). Perhaps this office had some connections to the fortress called
Borg at Birka and the hall discovered at the Garrison there. As noted above,
the hall was interpreted as a place of assembly for the warriors. In this hall the
place of the high-seat was identified. It has been assumed that the king sat
there when visiting the Garrison.39 Most likely the king and his retainers in
Birka celebrated ceremonial feasts there in the same way that the einherjar
held great feasts together with Óðinn in Valhǫll.
In the Viking Age the rulers of the Svear plundered and took tribute from
the peoples living around the Baltic Sea.40 In Óláfs saga helga, Snorri describes
how Eiríkr Emundarson, king at Uppsala, held a levy (ON leiðangr) every sum-
mer and sailed to various lands.41 The Swedish medieval laws report a military
organization called OSw leþunger, which was led by the king.42 In the medieval
37 It is possible that the aristocratic settlement at Adelsö was developed into crown lands
(bona regalia) belonging to “Uppsala Öd” (Uppsalaauðr) as early as the Viking Age. The
runic stone indicates that. We know for sure that it was a royal estate in the Middle Ages.
A medieval letter from 1200 mentions that Adelsö (Alsnö) was regarded as that: (ad) man-
sionem regiam Alsnu. There are still remains of the medieval royal palace at Adelsö. See
e.g. Thordeman 1920; Ambrosiani & Eriksson 1993, 39ff.; Brunstedt 1996, 10, 14; DMS 1.7:
218–236; SOL, 17, 21;
38 See e.g. Vita Anskarii 11, 26.
39 Perhaps Hergeir also had the function of the cultic leader of the local sanctuary. During
Ansgar’s first visit to Birka the recently converted and pious Hergeir was appointed as
counsellor. At one assembly, when all delegates were gathered for a meeting, Hergeir was
attacked, since he did not praise the pagan gods but held on to his new faith. It seems as if
the participants were upset, since Hergeir neglected and mismanaged his office as bailiff
and representative of the king, i.e. his cultic functions to maintain good relations with the
pagan gods. Hergeir’s function at Birka could be compared to Earl Sigurðr’s role as cultic
leader on behalf of the king at the sanctuary of Lade.
40 E.g. Lindkvist & Ågren 1985, 12f., 22–25; Sawyer & Sawyer 1993, 91f.; Lindkvist 1995; M. G.
Larsson 1990.
41 Heimskringla II, Ísl. Fornr. 27, 115.
42 Nu biuþær kunungær liþ ok leþung utt. biuþær utt roþ och reþ. Þa skal næmpnæ hampn
ok stampn. ok styriman. ok hasætæ allæ. UL Kg 10. SSGL 3, 94. See also UL’s Mh 11:3. Cf.
Hafström 1970 (1959), 49–57. Holmbäck & Wessén 1979 (1933–46), 55–62.
period it functioned as a taxation system. The leþunger may, however, have deep
roots in the native culture. The term hund/hundare, which was associated with
this system, is ancient.43 It seems as if Tacitus was acquainted with it among
the Germanic peoples: “The number of these men [the retainers] is fixed—
one hundred from each canton: and among their own folk this, ‘the Hundred,’
is the precise name they use; what was once a number only has become a title
and a distinction”.44 In addition Tacitus states: “Beyond these tribes the states
of the Suiones [the Svear], not on, but in, the ocean, possess not merely arms
and men but powerful fleets.”45 This information does not necessarily refer to
the leþunger. But Rimbert and many Viking Age runic inscriptions mention
organized maritime expeditions.46 Beowulf mentions “sea-kings” in Swīorīce.47
Other sources also support the idea that great men of Svetjud organized naval
attacks on foreign peoples and took tribute.
Like the Svea rulers, the Viking Age Norwegian kings (konungar), earls (jar-
lar) and chieftains (hersar) had a hirð, i.e. a group of professional warriors.48
These retainers (hirðsveinar, sveinar) had to be supplied with food, equipment,
and weapons. In addition they also had a part of the booty. Therefore an eco-
nomic surplus was important for the Norwegian rulers, in order to maintain
their military power and keep the hirð alive. By plundering and taking trib-
ute, this surplus was achieved. In order to supply his hirð the rulers also had
estates and farms (No hovedgård, sentralgård), which gave economic income.
Such royal farms also appear in medieval Trøndelag.49 Merete Røskaft has
recently presented most of them in her dissertation.50 In Inn-Trøndelag they
43 On the hund/hundare system, see primarily T. Andersson 2004; 2005. Mats G. Larsson
(1993, 122–126) believes that the places where husabyar appeared were ancient gathering
places and connected to the leþunger system.
44 definitur et numerus: centeni ex singulis pagis sunt, idque ipsum inter suos vocantur, et quod
primo numerus fuit, iam nomen et honor est. Germania 6. Trans. Hutton.
45 Suionum hinc civitates, ipso in Oceano, praeter viros armaque classibus valent. Germania 44.
46 See e.g. VA 30. Some ledung expeditions in the eleventh century were organized by
some sort of central leadership. The best-known ledung is “Ingvar’s journey”, attested
on thirty-two runic inscriptions from Sweden and in Icelandic prose. Since this expe-
dition was very complex and involved many people from different districts of eastern
Sweden, one assumes that it needed an organization. Cf. M. G. Larsson 1990; 1997, 9–163
and his literature. Lindkvist 1995, 38–64.
47 Þone sēlestan sæ-cyninga,/ þāra ðe in Swīorīce sinc brytnade,/ mærne þēoden. Beowulf
2382f.
48 Jón Viðar Sigurðsson 2008a, 24ff.
49 Hallan 1981 (1954).
50 Røskaft 2003. Cf. Brendalsmo 2006.
included, for instance, the farms at Alstadhaug, Sakshaug, and Haug, and in
Ut-Trøndelag the farms at Gryting, Meldal, Stein(e), Melhus, Lade, Logtun, and
Værnes. Sometime during the Viking Age or the Early Middle Ages these farms
were confiscated by the king and turned into royal estates.51 Most likely ban-
queting halls were erected there, where the ruler could gather his retainers,
celebrate ceremonial feasts and distribute gifts to them. Perhaps the Middle
Iron Age hall discovered at Hovde, Ørland, had similar functions. It had a stra-
tegic location in a military sense, between Ørland and Agdenes, controlling the
entrance to the Trondheimsfjord.
As noted in chapter 2, there was a tendency towards power concentration
in Norway during the Early Viking Age. At the same time, the hirð system
gradually became something more exclusive for kings, earls and great warrior
lords. Only these social groups could afford to have such band of warriors. In
Norway, as in Svetjud, a royal levy system (ON leiðangr), a seaborne military
organization led by the king, was developed during the Late Viking Age. It has
been suggested that King Hákon Aðalsteinsfóstri first organized this system in
Vestlandet in the tenth century, with a model borrowed from England.52 It was
then introduced to Trøndelag and finally also in Østlandet, when the Danish
influence was defeated. This system was not completely established until the
late twelfth century.
The relationship between the Norwegian king and the members of his hirð
was characterized by loyalty and friendship. According to medieval Norwegian
laws, the king must have at least four retainers (No følgesvenner) accompany-
ing him everywhere, from morning to evening.53 They protected the king and
they were therefore not allowed to drink beer or wine. The royal hirð was one
of the most important elements for constructing power in Late Viking Age and
Early Middle Age Norway and Svetjud. Jón Viðar Sigurðsson states that it even
was the central point of the society.54 It was in this martial milieu the greatest
feasts were celebrated in impressive halls controlled by warrior lords. In these
halls all kinds of cultural activities took place, such as the reciting of poems
and other literary traditions.
15.3.2 Iceland
As noted in chapter 2, most chieftains who came from Norway to Iceland were
recruited from the lower elite. There was some competition between these
chieftains (goðar), but it seems as if they usually did not use weapons and vio-
lence in order to attain a favourable position in society. It has been suggested
that there were mainly three ways to mark high social status in the Icelandic
society: by means of kinship, wealth and generosity.55 These strategies are vis-
ible in the story of Hjalti Þórðarson’s erfi, related in Landnámabók:
In a similar fashion, gifts were distributed to the guests at the erfi of Unnr
(Auðr) djúpúðga Ketilsdóttir, according to Laxdæla saga and Landnámabók
(see ch. 9 above). As a result of the great feasts and the distribution of gifts, the
chieftains established friendship and alliances between each other, and with
the farmers who had pledged allegiance to them.
Even if warfare occasionally also took place in Iceland, the society is often
characterized by its peacefulness and productive life.57 The general assembly
(alþingi) at Þingvellir was probably an important factor for establishing the
peace. Jón Viðar Sigurðsson writes as follows about the Viking Age Icelandic
society and the chieftains there:
55 Jón Viðar Sigurðsson 1999, passim; 2008a, 58ff. Cf. Vésteinn Ólason (1998, 28) who states,
however: “The goði who socially and politically was primus inter pares, had to secure his
own position by creating alliances and by building up an economic position sufficiently
strong to be able to support a small group of fighting men in his home at all times and a
larger number in times of conflict.”
56 Hjalti son Þórðar skálps kom til Íslands ok nam Hjaltadal at ráði Kolbeins ok bjó at Hofi;
hans synir váru þeir Þorvaldr ok Þórðr, ágætir menn. Þat hefir erfi verit ágætast á Íslandi, er
þeir erfðu fǫður sinn, ok váru þar tólf hundruð boðsmanna, ok váru allir virðingamenn með
gjǫfum brutt leiddir. Landnámabók, S 207, H 174, Ísl. Fornr. 1, 238. Trans. Hermann Pálsson
& P. Edwards.
57 Vésteinn Ólason 1998, 30.
There were thus no rulers or warrior lords riding around with a hirð in Viking
Age Iceland, nor was there a leiðangr.59 The Icelandic þingmenn related to a
goði were probably something different from the professional warrior bands
visible in the Norwegian and Swedish contexts. It was not until the twelfth
century, when power was concentrated in a few families, that the chieftains
started to surround themselves with bodyguards and retainers.60
We may thus not expect to encounter great aristocratic warrior halls, decorated
with weapons, in Iceland. There were no Icelandic warrior lords with warrior
bands there who could use such buildings as their dwelling. The halls found by
archaeologists in Iceland, such as Hofstaðir, do not display a lot of weapons,
compared to, for instance, the Garrison at Birka. The social and historical back-
ground of the Valhǫll mythology should therefore not primarily be searched in
these Icelandic hall buildings.
In Svetjud and Trøndelag, on the other hand, the great kings and earls used
military force to express their power. In these regions we can see warrior lords
with hirð retinues ambulating around the country to certain royal estates. At
these estates, martial banqueting halls, decorated with weapons, were erected.
In these halls the lord had his “household”, including the retainers, around him.
He supplied them there with food, drink, entertainment, lodging and weap-
ons. He rewarded them with ceremonial feasts, at which he also distributed
gifts. Mythical traditions were probably recited in these buildings, including
the eschatological Valhǫll mythology.
It is quite possible that this mythology also included an element of ideologi-
cal strategy. As Anders Hultgård has formulated it, the myth of Óðinn welcom-
ing the dead warriors in Valhǫll, to be part of the einherjar troops, “can be part
lord and Óðinn was emphasized when the ruler entered the high-seat. Perhaps
he was considered godlike, since he also was sometimes believed to have
descended from this deity (see ch. 3 above).
Sources presented in this chapter, as well as in all of Part 4, indicate that the
hypothesis of this study could be supported. In the case of religious organi-
zation connected to the ceremonial buildings, there were clear differences
between Iceland, on the one hand, and Sweden and Norway, on the other hand.
In Iceland, ceremonial buildings were controlled by local chieftains. No trans-
regional cultic sites for all Icelanders appeared there. In the Mälaren region
and Trøndelag there were cultic places intended for different levels of society.
In addition to more local shrines, there were also some regional and even inter-
regional cultic sites, such as Uppsala, Mære and Lade. These cultic places were
ruled and exploited by the society’s uppermost elite. Some warlords in Svetjud
and Norway had warrior bands which celebrated feasts in martial halls. Most
likely these halls had some connection to the Valhǫll mythology. Such warrior
halls did not appear in Iceland. When mythic traditions about Valhǫll were
recited in Iceland the social, ritual and ideological setting was probably differ-
ent compared to the one in the Scandinavian homelands.
There were also other regional differences in the ritual and symbolic strat-
egies applied by rulers at cultic sites. In Norway and Sweden, large burial
mounds were often placed adjacent to halls. Such monuments were lacking
in the cultic places in Iceland. There is evidence that Swedish and Norwegian
kings and earls were worshipped with offerings at the burial monuments, as
after their deaths they were seen as mythical beings. A similar cult involving
dead chieftains was absent from Iceland.
Another regional difference is the distribution of gold foils. They are often
found in connection with halls in Sweden and Norway, but so far none have
been found in Iceland. The reason for this is probably linked to chronology.
When Iceland was settled c. AD 870, the tradition of depositing gold foils at the
high-seat had subsided. There may also be another explanation. Gold foils with
a couple depicted on them may have represented the mythical ancestors of the
rulers. The belief that the royal families were closely related to the gods may be
the reason why they were deposited at the high-seat. In this belief we may see a
type of religious strategy of monumentalization of power and rulership, which
had the function of supporting the royal families’ claim to power and prestige.
This belief and strategy existed in Norway and Sweden, but does not seem to
have occurred in Iceland.
Conclusions
This book discusses the significance of Late Iron Age ceremonial buildings and
sanctuaries in Scandinavia as parts of religious strategies to create authority,
legitimation, and empowerment for rulers. The major purpose has been to
analyse how rulers established and expressed their power by means of these
buildings, ritual structures, objects and different types of rituals which took
place in them. In contrast to previous research, this study has emphasized the
context and the process, examining regional variations and developments over
time. A particular focus has been placed on three regions: the Lake Mälaren
Valley, Trøndelag and Iceland. These areas have been chosen because they
index different political systems: kingship, earldom and chieftainship. They
also have plenty of written sources about pre-Christian ceremonial buildings
and sanctuaries. Since the spatial and temporal dimension of the sources is
crucial for my analysis, archaeological finds and place-names also play a sig-
nificant role, not least because they can be related directly to these regions and
can also be connected to specific periods.
Based on the discussion in Part 2, it can be concluded that tendencies
towards concentration of power took place in the Mälaren area from the
Migration Period onwards. Gradually there developed a type of royal power
with ambitions to exercise domination (ríki) over several groups of people
and minor chieftains, with a religious-political centre in Uppsala. Similar ten-
dencies towards concentration of power occurred in Norway, and Trøndelag,
from the Early Viking Age, where powerful kings and earls emerged. As a
consequence of this political development, chieftains and great farmers emi-
grated from Norway to Iceland and other places in the Atlantic. In Iceland the
Commonwealth (the Free State) was established around 930, a decentralized
political system. There were no tendencies of any development there towards
a “centralized state” during the Viking Age, comparable to the kingdoms and
earldoms in contemporary Svetjud and Trøndelag. A concentration of power
did not take place until the end of the Viking Age in Iceland, and no central
royal power appeared on the scene as long as the Free State existed.
By means of a case study, presented in Part 2, it can also be suggested that
the religious strategies for gaining legitimacy and power, in general, diverged
in the three areas, especially in the ways the rulers related to the mythical
world. Old Norse sources indicate that the royal family of the Svear, called the
Ynglingar, and the Earls of Lade claimed divine descent for their families, while
orthern Norway. These ritual-symbolic designs are clearly part of the strategy
n
of monumentalization of power and rulership visible especially in Svetjud and
Trøndelag.
It can also be concluded that the interiors of the ceremonial buildings were
just as impressive as the exterior of these houses, with large rooms and high
roofs. In these rooms there were ritual structures and objects, such as the
high-seats, high-seat posts, cultic images, tapestries, and hearths. They could be
perceived as a ritual-symbolic strategy for legitimacy. The high-seat, for instance,
was regarded as a place of honour in all three regions, and intended for the owner
of the hall: the king, the earl or the chieftain. It was elevated in some way, so
that the ruler actually looked down on the other participants in the room,
thereby creating a spatial hierarchy of power. The other participants seem to
have been seated according to rank and gender at specific places in the cer-
emonial building, but on a lower level than the hall owner. The sources indi-
cate that the high-seat was a social and political furnishing with ideological
dimensions, i.e. a thing materializing hierarchical relationships. In addition, it
played a central role in religious and judicial contexts. It was thus a medium of
tradition, power and authority (see further below).
By means of the investigation in Part 3 it may also be concluded that the
ritual structure of ceremonial buildings and sacred sites in Svetjud, Trøndelag
and Iceland displayed some common features. In all three regions, there were
ritual-symbolic elements or objects which quite regularly appeared at the
ceremonial buildings. This book presents some evidence that these ritual
structures were (at least for certain groups who had access to religious and
culture knowledge) regarded as symbols of mythical or cosmic significance.
Most of them also occur in connection with mythical buildings and mythical
landscapes mentioned in the Old Norse traditions. Evidence indicates a micro-
macro-cosmic relationship between the real ceremonial buildings and the
mythical buildings. It seems as if the cultic places, ceremonial buildings and the
high-seat sometimes were organized according to a prototype or model. Such
sanctuaries were also regarded as thresholds to the Other World, i.e. places
where humans could encounter divine beings. The owner and user of such
a place had an enormous power since he or she appeared on a sacred place
related to the gods. On such occasions, the scene needed to be set with spe-
cific properties. Religious symbols such as representations of the cosmic hall,
the world-tree and the mythical well created the appearance that the ruler’s
authority derrived from a realm beyond politics, society and the natural world.
They were part of the strategies applied by rulers to construct legitimacy by
drawing on mythological concepts and narratives. Some people never noticed
the symbolic dimension of these impressive halls and sanctuaries. The ritual
structures may have been perceived by them in a more direct way as powerful.
Likewise, they provided agency to the elite just by their material appearance.
The sources indicate that cultic images were placed in the ceremonial build-
ings and sanctuaries in the three regions. They also suggest that the rulers
wanted to persuade the cultic communities in a concrete sense that they had a
specific relation to the gods and the mythical world. When the ruler performed
the sacrifices in public, he interacted and communicated with these images by
means of recitations, divinations and other rituals. Some of the cultic images
represented the gods as they were seated in the high-seat. When the gods were
seated in the high-seat they appeared in prototypical ritual roles, which could
be imitated by the earthly rulers.
Evidence presented in Part 3 indicates that rulers in the three areas con-
secrated lands, holy places and ceremonial buildings with a complex of ritu-
als which were performed in public. During these rituals, these magnates
proclaimed certain ritual restrictions which prevailed at these ritual spaces.
By performing these rituals, they gained prestige and demonstrated in public
their close relationship to sacred objects, holy places and the otherworld in
general, which could increase their political ascendancy.
Rituals performed inside the ceremonial buildings also contributed to the
construction of political power and authority. The examination in Part 3 clearly
indicates that rulers in the three areas played a central ritual role during the
public and calendrical sacrifices and ceremonial banquets in the halls and
cultic houses. During the ritual drinking they were seated in the high-seat
and perhaps they consecrated the beaker and recited ritual formulae. It seems
that they also performed divination rituals as part of the sacrifices, in order to
acquire divine legitimacy for the actions to be taken. Some of the rituals were
also related to gift-giving systems, where the religious-political authorities had
the opportunity to manifest their wealth and generosity. The political leaders
built alliances and gained friends on such occasions. These rituals took place in
the ceremonial buildings, where the mythic-cosmic references in them empha-
sized the dignity and religious nature of these actors and actions. Regional sim-
ilarities may also be seen in the oath rituals which probably took place in such
buildings or sanctuaries. Oath-rings were important agencies of power during
the religious ceremonies. They were carried in the hands or worn on the arms of
the rulers while performing the religious and judicial rituals in the sanctuaries
or when participating in public processions with other cultic leaders.
In Part 4 (“Regional diversity”) materials and themes supporting the general
hypothesis were discussed. The investigation indicates that some mythical,
ritual and symbolic strategies for legitimacy related to ceremonial buildings
and sanctuaries in Svetjud and Trøndelag never turned up in Iceland. These
and after the swearing of oaths and holy vows, the heir entered the high-seat.
The commemorative toasts and vows could be seen as constructions of power
continuity and legitimate succession. The high-seat played a central role in this
ritual system; it was the link between the rulers and their realms. It was also
regarded as the ritual entrance both to rulership and to the Other World. When
sitting in the high-seat, the ruler executed his functions as a leader in a judi-
cially legitimate sense. The entering of the high-seat also comprised a complex
of religious ideas. The ruler, for instance, had the ability to make contact with
the gods and divine ancestors from there, which was one of the reasons why
he had to be in his high-seat during the sacrifices at the ceremonial feasts. The
rituals associated with the high-seat at these feasts could also be seen as ritual
strategies for gaining legitimacy and constructing authority and royal power,
which thereby shared in divine agency.
The inquiry in Part 4 indicated that the religious organization related to the
ceremonial buildings and sanctuaries differed in the three regions. Only in
Svetjud and Trøndelag do we find chief sanctuaries with an (inter-)regional
significance such as Uppsala, Lade and Mære. They were related to aristocratic
centres, but some may have had a more communal base, such as Mære in the
Middle Viking Age. Most likely the kings and earls used the chief sanctuaries
and the cultic organization in order to gain legitimacy or claim/invoke power,
sometimes with support from subordinated cultic leaders/specialists. At these
monumental sites they could gather people from the whole region and medi-
ate their sovereignty by means of ceremonies, ritual objects and a cult of many
gods. In Iceland, the situation was different. No chief sanctuary with signifi-
cance for all Icelanders is known. The shrines there were located on chieftains’
farms in the settlement districts, and in general they only played an important
role for local groups. There is nothing to indicate that common religious feasts
were celebrated at the communal assembly of the Alþingi.
The warrior lords in Svetjud and Trøndelag gathered their retainers in the
aristocratic halls in order to strengthen their loyalty to the leaders by means
of religious ceremonies. Such martial hall milieus were probably not common
in Iceland. In connection with this, it can be suggested that the conception
of Valhǫll, and the myths related to it, were formulated in the Scandinavian
homelands among bellicose warrior lords rather than in Iceland, where such
phenomena did not appear. The eschatological Valhǫll myth included ele-
ments of a “warrior ideology” associated with the uppermost strata of Norway,
Svetjud and Denmark. Engaging the Valhǫll myth may have provided support
for the royal families’ claim to power. The narrative that the warrior lord and his
retinues in the afterlife were welcomed by Óðinn to his dwelling place empha-
sized the close connection between the elite groups and the mythical world.
By means of this myth the warrior lord could also attract good and noble war-
riors to his hirð, and ensure them a glorious afterlife which included the hos-
pitality of Óðinn in his dwelling of Valhǫll. The Valhǫll myth also had another
ideological function: when the earthly retinues were seated in the warrior hall
celebrating religious feasts, the warrior lord, who sat in his high-seat, could
have been associated with aspects of Óðinn seated in Hliðskálf together with
the einherjar in Valhǫll. The Valhǫll mythology was probably recited in other
social and ritual contexts in Iceland, since war-lords and warrior bands never
existed there.
The strategies including a monumentalization of power visible at ceremo-
nial buildings and sanctuaries developed gradually in the Mälaren area and
Trøndelag during the Late Iron Age. At the same time the first tendencies to
power concentration may be discerned in these regions, probably after tough
competition between minor kings and chieftains. During this period ceremo-
nial buildings and sanctuaries were erected at many local rulers’ farms. They
became important arenas for competition, where power could be expressed,
claimed or invoked by means of a set of religious and ritual strategies. In the
Migration and Merovingian Periods several powerful petty kings and chieftains
in the Mälaren area and Trøndelag may have used almost the same mythical
narratives, ritual performances, symbolic and organizational devices at these
cultic sites. Gradually the strategies including a monumentalization of power
were monopolized by the uppermost families in these regions. Yet some ele-
ments were still identical at the sanctuaries of chieftains, earls and kings. Since
the uppermost elite never emigrated to Iceland, some of strategies or actions
for claiming and invoking power at sanctuaries never arrived there. They could
only be seen in the old homelands of the Scandinavians.
———. Gisli Sursson’s Saga. Trans. M. S. Regal. In The Sagas of Icelanders. New York.
Glælognskviða. See Skj.
Grágás: Islændernes lovbog i fristatens tid. (Ed.) Vilhjálmur Finsen. Copenhagen:
Berling, 1852. (Newprint Odense 1974).
Gregorii I. Papae. Registrum Epistolarum. Monumenta Germaniae Historica.
Epistolarum Tomi II Pars III. Edidit Ludovicus M. Hartmann. Berlin 1899.
Gregory of Tours. Gregorii Episcopi Turonensis. Historia Francorum. Monumenta
Germaniae Historica. Scriptores rerum Merovingicarum. I. Edidit W. Arndt &
Br. Krusch. Hanover. MCMLI.
———. The History of the Franks. Trans. L. Thorpe. London. 1974.
Grettis saga Ásmundarsonar. In Grettis saga Ásmundarsonar, Bandamanna saga.
Íslenzk fornrit 7. (Ed.) Guðni Jónsson. Reykjavík 1950.
Grímnismál. See Edda.
Gutalag och Guta saga jämte ordbok. (Ed.) H. Pipping. København 1905–07.
Guta saga. The History of the Gotlanders. Viking Society for Northern Research.
University College London. Vol. XII. (Ed.) C. Peel. London 1999.
Gylfaginning. See Snorri Sturluson, Edda.
Hákonardrápa. See Skj.
Hákonarmál. See Skj and SkP.
Háleygjatal. See Skj and SkP.
Haraldskvæði. See Skj and SkP.
Harðar saga. In Harðar saga: Harðar saga Grímkelssonar (Hólmverja saga), Bárðar
saga Snæfellsáss, Þorskfirðinga saga (Gull-Þóris saga), Flóamanna saga, Þórarins
þáttr Nefjólfssonar, Þorsteins þáttr uxafóts, Egils þáttr Síðu-Hallssonar, Orms þáttr
Stórólfssonar, Þorsteins þáttr tjaldstoeðings, Þorsteins þáttr forvitna, Bergbúa
þáttr, Kumlbúa þáttr, and Stjörnu-Odda draumr. Íslenzk fornrit 13. (Eds.) Þórhallur
Vilmundarson & Bjarni Vilhjámsson. Reykjavík 1991.
Hauksbók. (Eds.) Eiríkur Jónsson & Finnur Jónsson. Copenhagen 1892–96.
Haustlǫng. See Skj.
Hávamál. See Edda.
Helgaqviða Hiǫrvarðzsonar. See Edda.
Helreið Brynhildar. See Edda.
Herodotus. Vol. I. (Ed.) J. Loeb. Trans. A. D. Godley. London – Cambridge, Massachusetts.
1960.
Hesiod. Theogony. Works & Day. Testimonia. (Ed.) Loeb. Trans. & (ed.) G. W. Most.
London – Cambridge, Massachusetts. 2007.
Hervarar saga. In Heiðreks saga: Hervarar saga ok Heiðreks konungs. (Ed.) Jón Helgason.
København 1924.
———. In Hervarar saga ok Heiðreks saga. (Ed.) E. O. G. Turville-Petre. Viking Society
for Northern Research. London 2006 (1956).
Kristni saga. In Kristnisaga. þáttr þorvalds ens víðfǫrla. Þáttr Ísleifs biskups Gizurarsonar.
Hungrvaka. Altnordische Saga-Bibliothek 11. (Ed.) B. Kahle. Halle 1905.
———. In Íslendingabók and Kristni saga. Viking Society for Norther Research. Text
Series. Trans. S. Grønlie. London 2006.
Landnámabók. In Íslendingabók. Landnámabók. Íslenzk fornrit 1. (Ed.) Jakob Bene
diktsson. Reykjavík 1986.
———.The Book of the Settlements. Trans. Hermann Pálsson & P. Edwards. Winnipeg.
Laxdœla saga. In Laxdœla saga. Halldórs þættir Snorrasonar. Stúfs þáttr. Íslenzk fornrit
5. (Ed.) Einar Ól. Sveinsson. Reykjavík1934.
———. The Saga of the People of Laxardal. Trans. K. Kunz. In The Sagas of Icelanders.
New York.
Legenda sancti Eskilli. See SRS II, 389–399.
Leges Langobardorum 643–866. Bearbeitet von Franz Beyerle. Göttingen 1962 (1947).
Lex Frisionum. See Clemen, Carolus.
Lex Ribuari. See Clemen, Carolus.
———. Die Gesetze des Karolingerreiches 714–911. Hrsg. K. A. Eckhardt. Weimar 1934.
Lokasenna. See Edda.
Lucretius. On the Nature of Things. (Ed.) Loeb. Trans. W. H. D. Rouse. London –
Cambridge, Massachusetts 1924.
Magnus Lagabøters landslov. Trans. A. Taranger. Kristiania 1915.
NGL = Norges Gamle Love indtil 1387. Vols. I–III. (Eds.) R. Keyser & P. A. Munch.
Christiania 1846–49.
Njáls saga. See Brennu-Njáls saga above.
Ög = Östergötlands runinskrifter. Granskade och tolkade av E. Brate. Sveriges
Runinskrifter. 2. KVHAAH. Stockholm. 1911–1918.
Ǫgmundar þáttr dytts. In Eyfirðinga sǫgur: Víga Glúms saga. Ǫgmundar þáttr dytts.
Þorvalds þáttr tasalda. Svarfdæla saga. Þorleifs þáttr jarlsskálds. Valla-Ljóts saga.
Sneglu-Halla þáttr. Þorgríms þáttr Hallasonar. Íslenzk fornrit 9. (Ed.) Jónas
Kristjánsson. Reykjavík 1956.
Öl = Ölands runinskrifter. Granskade och tolkade av S. Söderberg & E. Brate. Sveriges
Runinskrifter. 1. KVHAAH. Stockholm. 1901–1906.
Óláfsdrápa sœnska. See Skj.
Óláfs saga Tryggvasonar en mesta I, Fornmanna sögur 1–3.
Óláfs saga Tryggvasonar eptir Odd munk Snorrason. In Fœreyinga saga. Óláfs saga
Tryggvasonar eptir Odd munk Snorrason. Íslenzk fornrit 25. (Ed.) Ólafur Halldórsson.
Reykjavík 2006.
Olaus Magnus. Historia de gentibus Septentrionalibus. Farnborough 1971.
Orkneyinga saga. In Orkneyinga saga: Legenda de sancto Magno. Magnúss saga
skemmri. Magnúss saga lengri. Helga þáttr ok Úlfs. Íslenzk fornrit 34. (Ed.) Finnbogi
Guðmundsson. Reykjavík 1965.
———. The History of the Earls of Orkney. Trans. Hermann Pálsson and Paul Edwards.
London 1981 (1978).
Ovid. Ovid’s Fasti. (Ed.) J. Loeb. Trans. J. G. Frazer. London – Cambridge, Massachusetts.
1959 (1931).
Plutarch. Plutarch’s Lives. Vol. I. (Ed.) J. Loeb. Trans. B. Perrin. London – Cambridge,
Massachusetts. 1959 (1914).
Porphyrios. De abstinentia. See Cumont 1896–1899 II, 39ff.
———. De antro nympharum. See Cumont 1896–1899 II, 42f.
———. The Cave of the Nymphs in the Odyssey. Trans. and (ed.) Seminar Classics 609.
Arethusa Monographs, no. 1. Buffalo. 1969.
Ragnarsdrápa. See Skj.
Rannsakningar efter antikviteter. Band I. Uppland, Västmanland, Dalarna, Norrland,
Finland. Häfte I. Text. KVHAA. Utg. I. Olsson et al. Stockholm 1960.
RApp = Documents on Paper in the Swedish National Archives. Stockholm.
Rerum gestarum Saxonicarum. See Widukind.
Rígsþula. See Edda.
Rimbert. Anskar. The Apostle of the North. 801–865. Trans. C. H. Robinson. London. 1921.
Rimbert. Vita Anskarii auctore Rimberto. Accedit Vita Rimberti. (Ed.) G. Waitz. Hanover
1884.
Saxo Grammaticus. Saxonis Gesta Danorum. (Eds.) J. Olrik & H. Ræder. Hauniæ 1931.
———. The History of the Danes, Book I–IX. Edition and Commentary Hilda Ellis
Davidson. Translation Peter Fischer. Cambridge 2008 (1996).
———. Gesta Danorum. I–II. Ed. K. Friis-Jensen. Trans. P. Fischer. Oxford 2015.
Schøning, Gerhard. Reise som giennem en Deel af Norge i de Aar 1773, 1774, 1775 paa
Hans Majestæt Kongens Bekostning er giort og beskreven. Udgivet af Det Kgl.
Videnskabers Selskab i Trondhjem i Anledning af dets 150-Aars Jubilæum. Bind
I–II. Ved Karl Rygh. Trondhjem. 1910.
Sigrdrífomál. See Edda.
SkP I = Skaldic Poetry of the Scandinavian Middle Ages I. Poetry from the Kings’
sagas 1 : From mythical times to c. 1035. (Ed.) D. Whaley. Turnhout 2012.
SkP II = Skaldic Poetry of the Scandinavian Middle Ages II. Poetry from the Kings’
sagas 2 : From c. 1035 to c. 1300. (Ed.) K.E. Gade. Turnhout 2009.
Skáldskaparmál. See Snorri Sturluson, Edda.
Skírnismál (For Scírnis). See Edda.
Skj = Den Norsk-Islandske Skjaldedigtning 800–1400. AI–II, BI–II. (Ed.) Finnur Jónsson.
Köbenhavn 1967–1973 (1912–15).
Skjǫldunga Saga. In Danakonunga Sǫgur: Skjǫldunga Saga. Knýtlinga Saga. Ágrip af
Sǫgu Danakonunga. Íslenzk fornrit 35. (Ed.) Bjarni Guðnason. Reykjavík. 1982.
Sm = Smålands runinskrifter. Granskade och tolkade av Ragnar Kinander. In Sveriges
runinskrifter 4. KVHAAH. Stockholm – Uppsala 1935–61.
Snorri Sturluson. Edda. (Ed.) Finnur Jónsson. København 1900.
Tacitus, Cornelius. Germania. Originalets text med svensk tolkning jämte inledning
och kommentar av A. Önnerfors. Stockholm 1960.
———. Germania. In Agricola, Germania, Dialogus. (Ed.) Loeb. Trans. M. Hutton.
London – Cambridge, Massachusetts. 1992 (1914).
The Anglo-Saxon Chronicle. Trans. and (ed.) M. Swanton. London 1997.
The Earliest Norwegian Laws: Being the Gulathing Law and the Frostathing Law. Trans.
L. M. Larson. New York 1935.
The Great Saga of St Olaf. Den store saga om Olaf den helige. (Eds.) O. A. Johnsen &
Jón Helgason. Utg. För kjeldeskriftfondet. Band 1–2. Oslo. 1941.
The Legendary Saga of St Olaf. Olafs saga hins helga. Efter pergamenthandskrift I
Uppsala universitetsbibliotek, Delagardieske samling nr. 8II utg. Av den Norske
Historiske Kildeskriftkommision. (Ed.) O. A. Johnsen. (Det norske historiske kilde-
skriftfonds skrifter 47.). Kristiania. 1922.
The New English Bible. (Ed.) Collins. 1970 (1961).
The Old English Orosius. Early English Text Society S.S. 6. (Ed.) J. Bately. London – New
York – Toronto. 1980.
The Poetic Edda. L. M. Hollander. Austin 1990 (1962).
———. A new translation by Carolyne Larrington. Oxford 1996.
———. Vol. 2. Mythological Poems. Trans. and (ed.) U. Dronke. Oxford 1997.
Thietmar av Merseburg, Chronik. Thietmari Merseburgensis Episcopi Chronicon. Trans.
and (ed.) W. Trillmich. Darmstadt 1957.
Þjóðólfr inn fróði ór Hvini. Ynglingatal. KVHAAH. 28:2. Red. och översättning A. Noreen.
Stockholm 1925.
Þrymsqviða. See Edda.
Translatio S. Alexandri. Monumenta Germaniae Historica. Scriptorum. Tomus II.
Edidit Georgius Heinricus Pertz. Hanover 1829.
U = Upplands runinskrifter. Granskade och tolkade av E. Wessén & S. B. F. Jansson.
Sveriges runinskrifter. 6–9. KVHAAH. Stockholm 1940–58.
Vafþrúðnismál. See Edda.
Vápnfirðinga saga. In Austfirðinga sǫgur: Þorsteins saga hvíta, Vápnfirðinga saga,
Þorsteins saga hvíta, Vápnfirðinga saga, Þorsteins þáttr stangarhǫggs, Ǫlkofra þáttr,
Hrafnkels saga Freysgoða, Droplaugarsona saga, Brandkrossa þáttr, Gunnars
þáttr Þiðrandabana, Fljótsdæla saga, Þorsteins saga Síðu-Hallssonar, Draumr
Þorsteins Síðu-Hallssonar, Þorsteins þáttr Austfirðings, Þorsteins þáttr Sǫgufróða,
Gull-Ásu-Þórðar þáttr. Íslenzk fornrit 11. (Ed.) Jón Jóhannesson. Reykjavík 1950.
Vatnsdæla saga. In Vatnsdæla saga: Hallfreðar saga. Kormáks saga. Hrómundar þáttr
halta. Hrafns þáttr Guðrúnarsonar. Íslenzk fornrit 8. (Ed.) Einar Ól. Sveinsson.
Reykjavík. 1939.
———. The Saga of the People of Vatnsdal. Trans. A. Wawn. In The Sagas of Icelanders.
New York.
Vellekla. See Skj and SkP.
Literature
Andersson, T. M. 2008. From Tradition to Literature in the Sagas”. In Oral Art Forms and
their Passage into Writing. (Eds.) E. Mundal & J. Wellendorf. Copenhagen, 7–17.
Andrén, A. 1991. Guld och makt—en tolkning av de skandiniaviska guldbrakteaternas
funktion. In Samfundsorganisation og Regional Variation—Norden i romersk jernal-
der og folkevandringstid. Beretning fra 1. nordiske jernaldersymposium på Sandbjerg
Slot 11–15 april 1989. (Eds.) C. Fabech, & J. Ringtved. Jysk Arkæologisk Selskabs
Skrifter XXVII. Århus, 245– 256.
———. 2002. Platsernas betydelse. Norrön ritual och kultplatskontinuitet. In Plats och
praxis: Studier av nordisk förkristen ritual. Vägar till Midgård 2. (Eds.) K. Jennbert,
A. Andrén & C. Raudvere. Lund, 299–342.
———. 2004. I skuggan av Yggdrasil: Trädet mellan idé realitet i nordisk tradition. In
Ordning mot kaos—studier av nordisk förkristen kosmologi. Vägar till Midgård 4.
(Eds.) A. Andrén, K. Jennbert & C. Raudvere. Lund, 389–430.
———. 2006. Warrior Culture, Hybridity, and Old Norse Cosmology. In Old Norse
Religion in Long- Term Perspectives: Origins, Changes, and Interactions. Vägar till
Midgård 8. (Eds.) Andrén, A. & K. Jennbert & C. Raudvere. Lund, 33–38.
———. 2014. Tracing Old Norse Cosmology. The world tree, middle earth, and the sun in
archaeological perspectives. Vägar till Midgård 16. Nordic Acadeic Press, Lund.
Andrén, A., K. Jennbert & C. Raudvere. 2006. Old Norse Religion. In Old Norse Religion
in Long-Term Perspectives: Origins, Changes, and Interactions. Vägar till Midgård 8.
(Eds.) A. Andrén, K. Jennbert & C. Raudvere. Lund, 11–14.
Anjou, S. 1935. Uppsalatemplets gudabeläten avbildade på bonaden från Skog.
Fornvännen (30), 257–264.
Antonsen, E. H. 2002. Runes and Germanic Linguistics. Trends in Linguistics. Studies
and Monographs 149. Berlin/New York.
Åqvist, C. 1996. Hall och harg—det rituella rummet. In Religion från stenålder till
medeltid. Riksantikvarieämbetet. Arkeologiska undersökningar. Skrifter 19. (Eds.)
K. Engdahl & A. Kaliff. Linköping, 105–120.
Arne, T. J. 1932. Vendel före vendeltiden. Fornvännen 27, 1–22.
Arrhenius, B. 1961. Vikingatida miniatyrer. Tor (7), 139–164.
———. 1990. Utgrävningen av den östligaste storhögen på gravfältet Ormknös, raä 111,
Björkö,Adelsö sn, Uppland. Laborativ Arkeologi (4). Stockholm, 65–80.
———. 1995. 65–80. Regalia in Svealand in Early Medieval Times. Tor (27), 311–335.
———. 1998. Bebyggelsestrukturen i det förhistoriska Vendel och järnålderns mång-
mannagårdar. In Centrala platser—Centrala frågor: Samhällsstrukturen under
Järnåldern. En Vänbok till Berta Stjernquist. Uppåkrastudier. 1. (Eds.) L. Larsson &
B. Hårdh. Stockholm, 179–187.
———. 2000. Tuna och Husby i Vendel. In En bok om Husbyar. Riksantikvarieämbetet.
Avdelningen för arkeologiska undersökningar. Skrifter 33. (Ed.) Michael Olausson.
Örebro, 93–100.
Blomkvist, T. & P. Jackson. 1999. Alt ir baugum bundit: Skaldic Poetry on Gotland in a Pan-
Scandinavian and Indo-European Context. Arkiv för nordisk filologi (114), 17– 29.
Bø, O. 1960. Gravøl (Norge). In Kulturhistoriskt lexikon för nordisk medeltid. Vol. 5.
Malmö, cols. 450–452.
———. 1965. Källa. Noreg. In Kulturhistoriskt lexikon för nordisk medeltid. Vol. 10.
Malmö, cols. 57–59.
Bøe, A. 1964. Konge (Innleiing og Noreg). In Kulturhistoriskt lexikon för nordisk medel-
tid. Vol. 9. Malmö, cols. 1–4.
Bolle, K. W. 1987. Cosmology. An Overview. In Encyclopedia of Religion. Vol. 4. (Ed.)
M. Eliade. New York, 100–107.
Boor, H. de 1930. Die religiöse Sprache der Völuspá und verwandter Denkmäler. In
Deutsche Islandforschung 1930. (Ed.) W. H. Vogt. Breslau, 68–142.
Boudriot, W. 1928. Die altgermanische Religion in der amtliches kirchlichen Literatur des
Abendlandes vom 5. bis 11. Jahrhundert. Untersuchungen zur allgemeinen Religions
geschichte, Heft 2. Bonn.
Braarvig, J. 1997. The Function of Religious Rites in Creating and Sustaining the
‘Meaning’ of Symbols: A Commentary on Aristotle’s De interpretatione 1. In
In Search of Symbols. (Eds.) J. Braarvig & T. Krogh. Oslo, 44–67.
Bratt, P. 2008. Makt uttryckt i jord och sten. Stora högar och maktstrukturer i Mälardalen
under järnåldern, Stockholm Studies in Archaeology 46. Stockholm.
Braune, W. 1889. Vingolf. Beitrage zur Geschichte der deutschen Sprache und Literatur
14, 369–376.
Brendalsmo, J. 2006. Kirkebygg og kirkebyggere: Byggherrer i Trøndelag ca. 1000–1600. Oslo.
Brereton, J. P. 1987. Sacred Space. In Encyclopedia of Religion. Vol. 12. (Ed.) M. Eliade.
New York, 526–535.
Brink, S. 1990a. Sockenbildning och sockennamn: Studier i äldre territoriell indelning i
Norden. Acta Academiae regiae Gustavi Adolphi. LVII. Studier till en svensk ort-
namnsatlas utgivna av Thorsten Andersson. 14. Uppsala – Stockholm.
———. 1990b. Cult Sites in Northern Sweden. In Old Norse and Finnish Religions and
Cultic Place-Names. Scripta Instituti Donneriani Aboensis. XIII. (Ed.) T. Ahlbäck.
Stockholm, 458–489.
———. 1992. Har vi haft ett kultiskt *al i Norden? In Sakrale Navne. NORNA-rapporter.
48. (Eds.) G. Fellows-Jensen & B. Holmberg. Uppsala, 107–121.
———. 1994. En vikingatida storbonde i södra Norrland. Tor (26), 145–162.
———. 1996a. Political and Social Structures in Early Scandinavia. A Settlement-
Historical Pre-Study of the Central Place. Tor (28), 235–281.
———. 1996b. Forsaringen—Nordens äldsta lagbud. In Beretning fra femtende tvær-
faglige vikingsymposium. (Eds.) E. Roesdahl & P. Meulengracht Sørensen. Århus,
27–55.
———. 1997. Political and Social Structures in Early Scandinavia II. Aspects of Space
and Territoriality—The Settlement District. Tor (29), 389–437.
———. 1999a. Social Order in the Early Scandinavian Landscape. In Settlement and
Landscape: Proceedings of a Conference in Århus, Denmark, May 4–7 1998. Jutland
Archaeological Society. (Eds.) C. Fabech & J. Ringtved. Højbjerg, 423–438.
———. 1999b. Fornskandinavisk religion—förhistoriskt samhälle: En bosättningshis-
torisk studie av centralorter i Norden. In Religion och samhälle i det förkristna
Norden: Et symposium. (Eds.) U. Drobin et al. Odense, 11–55.
———. 1999c. Nordens husabyar—unga eller gamla. In Et hus med mange rom:
Vennebok til Bjørn Myhre på 60-årsdagen. Bind B. (Eds.) I. Fuglestvedt et al.
Stavanger., 283–291
———. 2000a. Husby. In Reallexikon der Germanischen Altertumskunde. Band 15. (Eds.)
H. Beck et al. Berlin– New York, 274–278.
———. 2000b. Nordens husabyar—unga eller gamla? In En bok om Husbyar.
Riksantikvarieämbetet. Avdelningen för arkeologiska undersökningar. Skrifter 33.
(Ed.) Michael Olausson. Örebro, 65–73.
———. 2001. Mythologizing Landscape: Place and Space of Cult and Myth. In
Kontinuitäten und Brüche in der Religionsgeschichte: Festschrift für Anders Hultgård
zu seinem 65. Geburtstag am 23.12.2001. (Eds.) M. Stausberg, O. Sundqvist & A. van
Nahl. Berlin – New York, 76–112.
———. 2003. ambátt, seta, deigja—þræll, þjónn, bryti: Termer för trälar belyser träldo-
mens äldre historia. In Trälar: Ofria i agrarsamhället från vikingatid till medeltid.
Skrifter om skogs- och lantrukshistoria 17. (Eds.) T. Lindkvist & J. Myrdal. Stockhom,
103–117.
———. 2004. Mytologiska rum och eskatologiska föreställningar i det vikingatida
Norden. In Ordning mot kaos—studier av nordisk förkristen kosmologi. Vägar till
Midgård 4. (Eds.) A. Andrén, K. Jennbert & C. Raudvere. Lund, 291–316.
———. 2005. Verba Volant, scripta manent? Aspects of Early Scandinavian Oral Society.
In Literacy in Medieval and Early Modern Scandinavian Culture. (Ed.) P. Hermann.
Odense, 77–135.
———. 2007. Skiringssal, Kaupang, Tjølling—the Toponymic Evidence. In Kaupang in
Skiringssal. Kaupang Excavation Project. Publication Series. Vol. 1. Norske Oldfunn
XXII. (Ed.) D. Skre. Oslo, 53–64.
———. 2008. Law and society—polities and legal customs in Viking Scandinavia. In
The Viking World. (Ed.) S. Brink. London, 23–31.
———. 2010. Är Forsaringen medeltida? In Hälsingerunor (2010), 109–117.
———. 2011. Gudhem—the toponymic evidence (or rather challenge). In The Gudme/
Gudhem Phenomenon. (Eds.) O. Grimm & A. Pesch. Neumünster, –.
Broberg, A. 1991. Religionsskifte och sockenbildning i Norduppland. Kyrka och socken i
medeltidens Sverige. Studier till det medeltida Sverige 5. Riksantikvarieämbetet.
(Ed.) O. Ferm. Stockholm, 49–79
Brøgger, A. W. 1916. Borrefundet og Vestfold-kongernes graver. Det norske Videns
kapsselskapets Skrifter. II. Hist. Fil. Kl. No. 1. Kristiania.
Carp, R. M. 2014. Material Culture. In The Routledge Handbook of Rearch Methods in the
Study of Religion. (Eds.) M. Stausberg & S. Engler. London, 474–490.
Carsten, L. 2012. Posts of Power. The Sagas’ Use of the Past Exemplified by an Iron Age
Building. In The 15th International Saga Conference: Sagas and the Use of the Past.
Aarhus, 97–98.
———. 2014. Lordship, Symbols and Memory: the Great House at Runsa from a
Philological Point of View. In Runsa Borg. Representative Life on a Migration Period
Hilltop Site—a Scandiavian Perspective. Papers from the project Runsa borg,
Uppland. No 2. (Ed.) M. Olausson. Östersund, 145–166.
Celander, H. 1955. Förkristen jul: Enligt norröna källor. Göteborgs universitets årsskrift.
Vol. LXI. 1955. 3. Stockholm.
Chadwick, N. 1950. Þorgerðr Hölgabuðr and the Trolla þing. A note on the sources In
The Early Cultures of North-West Europe. Cambridge, 395–417.
Christensen, T. 1991. Lejre Beyond Legend—The Archaeological Evidence. Journal of
Danish Archaeology (10), 163–185.
———. 1997. Lejrehallen. In “. . . gick Grendel att söka det höga huset . . .”: Arkeologiska
källor till aristokratiska miljöer i Skandinavien under yngre järnålder. Rapport från ett
seminarium i Falkenberg 16–17 november 1995. Hallands länsmuseers Skriftserie
No. 9/GOTARC C. Arkeologiska skrifter. No. 17. (Eds.) J. Callmer & E. Rosengren.
Halmstad, 47–54.
———. 2001. Lejre. §2. In Reallexikon der Germanischen Altertumskunde. Band. 18.
(Eds.) H. Beck et al. Berlin – New York, 249–254.
———. 2009: Odin fra Lejre. I Årskrift fra Roskilde Museum, 6–25.
———. 2010: Odin fra Lejre. I Danefæ—Skatte fra den danske muld. (Eds.) P. O. Nielsen
& M. Andersen. Kopenhagen, 183–186.
Clark Hall, J. R. 1916. A Concise Anglo-Saxon Dictionary for the Use of Students. Cambridge.
Clarke H. & B. Ambrosiani. 1993 (1991). Vikingastäder. Org. title Towns in the Viking Age.
Höganäs.
Clauss, M. 1990. Mithras: Kult und Mysterien. München.
———. 1992. Cultores Mithrae: Die Anhängerschaft des Mithras-Kultus. Heidelberger
Althistorische Beiträge und Epigraphische Studien 10. Stuttgart.
Cleasby R. & Guðbrandur Vigfusson. 1957 (1874). An Icelandic-English Dictionary. Oxford.
Clunies Ross, M. 1994. Prolonged Echoes: Old Norse Myths in Medieval Northern Society.
Vol. 1. Odense.
———. 1998. Prolonged Echoes: Old Norse Myths in Medieval Northern Society. Vol. 2.
Odense.
———. 2006. The measures of Old Norse religion in long-term perspective. In Old
Norse Religion in Long-Term Perspectives: Origins, Changes, and Interactions. Vägar
till Midgård 8. (Eds.) A. Andrén, K. Jennbert & C. Raudvere. Lund, 412–416.
———. 2008. The creation of Old Norse mythology. In The Viking World. (Ed.) S. Brink.
London, 231–234.
Cöllen, S. 2011. Der Rätselhafte Gott: Heimdallr im Licht altnordischer Vorstellungen
von Ahnen und Ordnung. Uppsala.
Damell, D. 1980. Om en fosfatkarta över fornminnesområdet vid Gamla Uppsala. In
Inventori in honorem: En vänbok till Folke Hallberg. (Ed.) Å. Hyenstrand. Stockholm,
64–67.
———. 1991. Fornsigtuna: En kungsgårds historia. Upplands Bro.
———. 1993. About Royal Manors from the Late Iron Age in Middle Sweden. Current
Swedish Archaeology 1. (Eds.) M. Burström & A. Carlsson. Stockholm, 39–47.
Davidson, H. R. E. 1943. The Road to Hel: A Study of the Conception of the Dead in Old
Norse Literature. Cambridge.
———. 1975. Scandinavian Cosmology. In Ancient Cosmologies. (Eds.) C. Blacker &
M. Loewe. London, 172–197.
———. 1994 (1993). The Lost Beliefs of Northern Europe. London – New York.
Dillmann, F.-X. 1997. Kring de rituella gästabuden i fornskandinavisk religion. In
Uppsalakulten och Adam av Bremen. (Ed.) A. Hultgård. Nora, 51–73.
———. 2000a. Chronique des études nordiques. Proxima Thulé. Revue d’études nordi-
ques. Vol. 4. (Ed.) F.-X. Dillmann. Paris, 101–112.
———. 2000b. Ad noen geografiske navn i Heimskringla. In Oluf Rygh: Rapport fra
symposium på Stiklestad 13.–15. Mai 1999. NORNA-rapporter. 70B. (Eds.) B. Sandnes
et al. Uppsala, 43–54.
———. 2000c. See Snorri Sturluson in Sources.
———. 2006. Les magiciens dans l’Islande ancienne: Études sur la représentation de la
magie islandaise et de ses agents dans les sources littéraires norroises, Acta Academiae
regiae Gustavi Adolphi. XCII. Uppsala.
———. 2007. Valhǫll. In Reallexikon der Germanischen Altertumskunde. Band 35. (Eds.)
H. Beck et al. Berlin – New York, 341–366.
———. 2008. Omkring Halvdan Svartes gravferd. Høvdingen: Om Snorre Sturlasons liv
og virke. (Eds.) J. O. Askedal & K. J. Myrvoll. Oslo, 73–87.
DMS. Det medeltida Sverige. 1960–. Stockholm.
Douglas, M. 1966. Purity and Danger: An Analysis of Concepts of Pollution and Taboo.
London.
Drescher, H. & K. Hauck. 1982. Götterthrone des heidnischen Nordens. Frühmittel
alterliche Studien. Jahrbuch des Instituts für Frühmittelalterforschung der Uni
versität Münster. 1982. Band 16, 237–301.
Driscoll, M. J. 2008. See Ágrip above.
Drobin, U. 1991. Mjödet och offersymboliken i fornnordisk religion. In Studier i religion-
shistoria tillägnade Åke Hultkrantz professor emeritus den 1 juli 1986. (Eds.)
L. Bäckman et al. Löberöd, 97–141.
Drobin, U. & M.-L. Keinänen. 2001. Frey, Veralden olmai och Sampo. In Kontinuitäten
und Brüche in der Religionsgeschichte: Festschrift für Anders Hultgård zu seinem 65.
Geburtstag am 23.12.2001. (Eds.) M. Stausberg, O. Sundqvist & A. van Nahl. Berlin –
New York, 136–169.
Dronke, U. 1997. See The Poetic Edda above.
DuBois, T. A. 1999. Nordic Religions in the Viking Age. Philadelphia.
Duczko, W. 1993a. Bibliografi över Gamla Uppsala. In Arkeologi och miljöekologi i Gamla
Uppsala: Studier och rapporter. OPIA. 7. Societas Archaeologica Upsaliensis. (Ed.)
W. Duczko. Uppsala, 71–78.
———. 1993b. Introduktion till Gamla Uppsala. In Arkeologi och miljöekologi i Gamla
Uppsala: Studier och rapporter. OPIA. 7. Societas Archaeologica Upsaliensis. (Ed.)
W. Duczko. Uppsala, 9–37.
———. 1996a. Gamla Uppsala. In Reallexikon der Germanischen Altertumskunde.
Band 10. (Eds.) H. Beck et al. Berlin – New York, 409–418.
———. 1996b. De litterära och arkeologiska kungarna. In Arkeologi och miljöekologi i
Gamla Uppsala: Studier och rapporter. Vol. II. OPIA. 11. Department of Archaeology.
Uppsala University. (Eds.) W. Duczko. Uppsala, 9–19.
———. 1996c. Kungsgården. In Arkeologi och miljöekologi i Gamla Uppsala: Studier och
rapporter. Vol. II. OPIA. 11. Department of Archaeology. Uppsala University. (Ed.)
W. Duczko. Uppsala, 37–51.
———. 1996d. Uppsalahögarna som symboler och arkologiska källor. In Arkeologi och
miljöekologi i Gamla Uppsala: Studier och rapporter. Vol. II. OPIA. 11. Department of
Archaeology. Uppsala University. (Ed.) W. Duczko. Uppsala, 59–94.
———. 1997. Gamla Uppsala—svearnas maktcentrum i äldre och nyare forskning.
In “. . . gick Grendel att söka det höga huset . . .”: Arkeologiska källor till aristokratiska
miljöer i Skandinavien under yngre järnålder. Rapport från ett seminarium i
Falkenberg 16–17 november 1995. Hallands länsmuseers Skriftserie No. 9/GOTARC C.
Arkeologiska skrifter. No. 17. (Eds.) J. Callmer & E. Rosengren. Halmstad,
71–81.
Dumézil, G. 1958. La Rígsþula et la structure sociale indo-européenne. Revue de
l’histoire des religions (154), 1–9.
———. 1959. Les dieux des Germains. Essai sur la formation de la religion scandinave.
Paris.
———. 1973. Gods of the Ancient Northmen. Berkeley.
———. 1986 (1948). Loki. Paris.
———. 2000. Mythes et dieux de la Scandinavie ancienne. Édition établie et préfacée
par F.-X. Dillmann. Paris.
Dumville, D. N. 1979 (1977). Kingship, Genealogies and Regnal Lists. In Early Medieval
Kingship. The School of History. University of Leeds. (Eds.) P. H. Sawyer &
I. N. Wood.Leeds, 72–104.
Marriage and Property Devolution in the Middle Ages. (Ed.) L. I. Hansen. Tromsø
2000, 133–165.
van Gennep. A. 1960. The Rites of Passsage. Orig. title Les rites de passage. Trans.
M. B. Vizedom & G. L. Caffee. Chicago.
Georges, K. E. 1951 (1913). Ausführliches Lateinisch-Deutsches Handwörterbuch. I–II.
9. Auflage. Nachdruck der achten verbesserten und vermehrten Auflage von
H. Georges. Basel.
Gilhus, I. S. & L. Mikaelsson. 2003. Nya perspektiv på religion. Orig. title Nytt blikk på
religion: Studiet av religion idag. Trans. U.-S. Rask. Stockholm.
Gilhus, I. S. & E. Thomassen 2010. Oldtidens religioner. Midtøstens og middelhavsområ-
dets religioner. Oslo.
Girard, R. 1977. Violence and the Sacred. Trans. P. Gregory. Baltimore.
———. 1986. The Scapegoat. Trans. Y. Freccero. Baltimore.
Gisli Sigurðsson. 2004. The Medieval Icelandic Saga and Oral Tradition. Cambridge, MA.
Golzio, K.-H. 2002. Tempel. Religionsgeschichtliche. In Theologische Realenzyklopädie,
42–46.
Göthberg, H. 2008. Bosättning och kyrkogård vid Gamla Uppsala kyrka. Uppsala.
Göthberg H., C. Lovén & G. Dahlbäck. 2010. Domkyrkan i Gamla Uppsala. In Uppsala
Domkyrka. II. Domkyrkan i Gamla Uppsala. Nuvarande domkyrkans omgivningar.
Sveriges kyrkor 228. (Eds.) R. Carlsson et al. Uppsala, 11–64.
Götherström, A. 2002. The Value of Stallions and Mares during the Medieval Time in
Upper Class Svealand: Molecular Sex Identifications on Horse Remains from Vendel
and Eketorp. Journal of Nordic Archaeological Science 2002, 75–78.
Graber, G. 1919. Der Einritt des Herzogs von Kärnten am Fürstenstein zu Karnburg: Eine
Studie zur kärntischen Volkskunde. Akad. d. Wiss. Wien, Phil-hist. Kl., Sitz.-ber. 190.
Band 95. Wien.
Gräslund, A.-S. 1983–84. Tor eller Vite Krist? Några reflektioner med anledning av
Lugnås-hammaren. Västergötlands fornminnesförenings tidskrift, 229–235.
———. 1985. Den tidiga missionen i arkeologisk belysning—problem och synpunkter.
Tor (20), 291–313.
———. 1987. Runstenar, bygd och gravar. Tor (21), 241–262.
———. 1992. Kultkontinuitet—myt eller verklighet? Om arkeologins möjligheter att
belysa problemet. In Kontinuitet i kult och tro från vikingatid till medeltid. Projektet
Sveriges Kristnande. Publikationer 1. (Ed.) B. Nilsson. Uppsala, 129–150.
———. 1997. Adams Uppsala—och arkeologins. In Uppsalakulten och Adam av
Bremen. (Ed.) A. Hultgård. Nora, 101–115.
———. 1999. Is there any Evidence of Powerful Women in Late Iron Age Svealand? In
Völker an Nord- und Ostsee und die Franken. Akten des 48. Sachsensymposiums
in Mannheim vom 7. bis 1. September 1997. Kolloquien zur Vor- und Frühgeschichte.
Band 3. Mannheimer Geschichtsblätter, Neue Folge, Beiheft 2. (Eds.) U. von Freeden
et al. Bonn, 91–98.
———. 1981 (1954). Det eldste kronogodset i Trøndelag. In Bosetning og økonmi. Norske
historikere i utvalg 5. Nye middelalderstudier. (Eds.) J. Sandnes & C. Krag. Oslo.
Hållans Stenholm, A.-M. 2010. Lilla Ullevi—en kultplats. In Makt, kult och plats.
Rapport från arkeologiska seminarier vid Stockholms läns museum. (Eds.) P. Bratt
& R. Grönwall. Stockholm, 31–38.
———. 2012. Fornminnen: det förflutnas roll i det förkristna och kristna Mälardalen.
Vägar till Midgård 15. Lund.
Hallberg, P. 1966. Medeltidslatin och sagaprosa: Några kommentarer till Lars Lönnroths
studier i den isländska sagalitteraturen. Arkiv för nordisk filologi (81), 258–276.
———. 1973. The Concept of gipta—gæfa—hamingja in Old Norse Literature. In
Proceedings of the First International Saga Conference 1971. (Ed.) P. Foote. London,
143–183.
———. 1979. Den isländska sagan. Verdandis skriftserie. 6. Lund.
Hallencreutz, C. F. 1984. Adam, Sverige och trosskiftet: Det missionsvetenskapliga per-
spektivet. In Adam av Bremen: Historien om Hamburgstiftet och dess biskopar. Trans.
E. Svenberg. Comm. by C. F. Hallencreutz, K. Johannesson, T. Nyberg, A. Piltz. (Ed.)
B. Sahlin. Stockholm, 355–378.
———. 1993. När Sverige blev europeiskt: Till frågan om Sveriges kristnande. Stockholm.
———. 1997. Missionsstrategi och religionstolkning: Till frågan om Adam av Bremen
och Uppsalatemplet. In Uppsalakulten och Adam av Bremen. (Ed.) A. Hultgård.
Nora, 117–130.
Hallinder, P. 1970. Iron. Excavation at Helgö III. (Ed.) W. Holmqvist. Stockholm.
Hansson, A.-M. 2011. Bread offerings at Foundation IV, Helgö—further influence
from the classical world? In Excavations at Helgö XVIII. (Eds.) B. Arrhenius &
U. O’Meadhra. Stockholm, 53–77.
Haraldsen, T. H. 1998. Gullgubben fra Vingom—på Frigg og Odins Hov? In Fåberg og
Lillehammer: En lokalskrift utgitt av Fåberg Historielag. Fåberg, 136–146.
Hårdh, B. 1996. Silver in the Viking Age: A regional-economic study. Lund.
Häringe, K. 1991. Gödåker: Ett uppländskt gravfält från romersk järnålder. Uppsats
C/20p. Institutionen för arkeologi. Uppsala universitet. Uppsala.
Hasenfratz, H.-P. 1997 (1992). Die religiöse Welt der Germanen: Ritual, Magie, Kult,
Mythus. Freiburg – Basel – Wien.
Hastrup, K. 1985. Culture and History in Medieval Iceland: An Anthropological Analysiss
of Structure and Change. Oxford.
Hauck, K. 1950. Geblütsheiligkeit. In Liber Floridus Mittellateinische Studien, Paul
Lehmann zum 65. Geburtstag. (Eds.) B. Bischoff & S. Brechter. St. Ottilien, 187–240.
———. 1954. Halsring und Ahnenstab als herrscherliche Würdezeichen. In
Herrschaftszeichen und Staatssymbolik I. (Ed.) P. E. Schramm. Stuttgart, 145–212.
———. 1985–86. Die Goldbrakteaten der Völkerwanderungszeit. Mit Beiträgen von
M. Axboe, K. Düwel, L. von Padberg, U. Smyra und C. Wypior. Münstersche
Mittelalterschriften. Vols. 1–3. München.
———. 2001. Journey of Civilisation: The Late Iron Age View of the Human World.
OPIA. 24. Institutionen för arkeologi och antik historia. Uppsala universitet.
Uppsala.
———. 2009. The Early Iron Age in South Scandinavia: Social Order in Settlement and
Landscape. OPIA 46. Uppsala.
Herschend, F. & D. Kaldal Mikkelsen. 2003. The main building at Borg (I:1). In Borg
in Lofoten: A chieftain’s farm in North Norway. (Eds.) Munch, G. Stamsø et al.
Trondheim, 41–76.
Heusler, A. 1911. Das Strafrecht der Isländersaga. Leipzig.
Hewitt, M. A. 1996. How New is the ‘New Comparativism’? Difference, Dialectics, and
World-Making. Method & Theory in the Study of Religion (8–1), 15–20.
Heywood, A. 1998. Political ideologies. An Introduction. London.
Hinnells, J. R. 1995. Temples (Ancient Near Eastern). In A New Dictionary of Religions.
London, 517–518.
Hjärne, E. 1951. Från Alsnö och Sko: Två undersökningar i svensk medeltidshistoria.
Uppsala.
Hjulström, B. 2008. Patterns in diversity: Geochemical analyses and settlement changes
during the Iron Age—Early Medieval time in the Lake Mälaren region, Sweden. Theses
and papers in scientific archaeology 11. Stockholm.
Hoffmann, E. 1976. Königserhebung und Thronfolgeordnung in Dänemark bis zum
Ausgang des Mittelalters. Beiträge zur Geschichte und Quellenkunde des
Mittelalters. Band 5. Berlin – New York.
Höfler, O. 1952. Germanisches Sakralkönigtum. Band 1. Der Runenstein von Rök und die
Germanische Individualweihe. Tübingen/Münster/Köln.
———.1959. Der Sakralcharakter des germanischen Königtums. In La Regalità Sacra.
Leiden, 664–701.
Holmberg, B. 1986. Den hedenske gud Tyr i danske stednavne. In Mange bække små: Til
John Kousgård Sørensen på tresårsdagen 6.12.1985. (Eds.) V. Dalberg & G. Fellows-
Jensen. København, 109– 126.
Holmberg, K. A. 1969. De svenska tuna-namnen. Acta Academiae regiae Gustavi
Adolphi. XLV. Studier till en svensk ortnamnsatlas utgivna av Jöran Sahlgren. 12.
Uppsala.
Holmberg-[Harva], U. 1917. Valhall och världsträdet. Finsk tidskrift för vitterhet, veten-
skap, konst och politik (48), 349–377.
———. 1922. Der Baum des Lebens. Annales Academiae Scientiarum Fennicae B 16:3.
Helsinki.
Holmbäck, Å & E. Wessén. 1979 (1933–46). See Svenska landskapslagar.
Holmgren, G. 1937a. Gamla Uppsala och Mora Äng i medeltidslagarnas valföreskrifter.
Uppsala.
———. 1987. Runes in Sweden. Orig. title Runinskrifter i Sverige. Trans. P. Foote.
Stockholm.
Jensen, J.S. 2014. What is religion? Durham.
Jesch, J. 1994. In Praise of Ástríðr Óláfsdóttir. Saga-Book. Vol. XXIV. Part 1, 1–18.
———. 1996 (1991). Women in the Viking Age. Woodbridge.
Jocobsen, L. & E. Moltke. 1941–42. See DR.
Johanson O. S. & G. Stamsø Munch 2003. Introduction and summary. In Borg in Lofoten:
A chieftain’s farm in North Norway. (Eds.) Munch, G. Stamsø et al. Trondheim, 11–18.
Johnsen, S. 1968. Stuttruner i vikingtidens innskrifter. Oslo.
Jón Hnefill Aðalsteinsson. 1992. A Piece of Horse-Liver and the Ratification of Law. In
Snorrastefna: 25.–27. júlí 1990. Rit Stofnunar Sigurðar Nordals 1. (Ed.) Úlfar Bragason.
Reykjavík, 81–98.
———. 1998. A Piece of Horse Liver: Myth, Ritual and Folklore in Old Icelandic Sources.
Reykjavík.
Jón Jóhannesson. 1956. Íslendinga saga I. Reykjavík.
Jón Viðar Sigurðsson. 1994. Forholdet mellom verdslig og religiøs makt på Island i
fristatsperioden. In Myte og ritual i det førkristne Norden: Et Symposium. (Eds.)
J. P. Schjødt et al. Odense, 129–139.
———. 1999. Chieftains and Power in the Icelandic Commonwealth, The Viking
Collection, 12. Odense: Odense University Press.
———. 2003. Kristninga i Norden 750–1200. Oslo.
———. 2008a. Det norrøne samfunnet: Vikingen, kongen erkebiskopen, og bonden. Oslo.
———. 2008b. Iceland. In The Viking World. (Ed.) S.Brink. London & New York, 571–578.
———. 2010. Den vennlige vikingen: Vennskapets makt i Norge og på Island 900–1300.
Oslo.
———. 2011. Kings, Earls, and Chieftains: Rulers in Norway, Orkney and Iceland
c. 900–1300. In Ideology and Power in the Viking and Middle Ages: Scandinavia,
Iceland, Ireland, Orkney and the Faeroes. (Eds.) G. Steinsland et al. Leiden, 69–108.
——— et al. 2005. Ingólfr: Norsk–Islandsk Hopehav 870–1536. Førde.
Jónas Kristjánsson. 1988. Eddas and Sagas: Iceland’s Medieval Literature. Trans. P. Foote.
Reykjavík.
Jonsson, E. 2006. Eldstålsformade ringar: En studie av när de tillverkades, och var och hur
de hittas inom gränserna för dagens Sverige. Kandidatuppsats (C) i arkeologi.
Institutionen för arkeologi och antikens kultur, Stockholms universitet. Stockholm.
Jørgensen, L. 1998. En storgård fra vikingetid ved Tissø, Sjælland—en foreløbig præsen-
tation. In Centrala platser—Centrala frågor: Samhällsstrukturen under Järnåldern.
En Vänbok till Berta Stjernquist. Uppåkrastudier. 1. (Eds.) L. Larsson & B. Hårdh.
Stockholm, 233–248.
———. 2002. Kongsgård—kusted—marked: Overvejelser omkring Tissøkompleksets
struktur og funktion. In Plats och praxis—studier av nordisk förkristen ritual. Vägar
till Midgård 2. (Eds.) K. Jennbert, A. Andrén & C. Raudvere, Lund, 215–247.
Kiil, V. 1960. Hliðsljálf og seiðhjallr. Arikiv för nordisk filologi (75), 84–112.
———. 1962. The Norse Prophetess and the Ritually Induced Prostitution. Norveg (9),
159–174.
Klos, L. 2007. Lady of the Rings: Järnålderns kvinnor mellan makt och kult. In Kult, guld
och makt: Ett tvärvetenskapligt symposium i Götene. Historieforum Västra Götaland.
Serie B. Vetenskapliga rapporter och Småskrifter. No. 4. (Ed.) I. Nordgren. Skara,
70–86.
Kluge, F. 1957 (1883). Etymologisches Wörterbuch der deutschen Sprache. Berlin.
Knirk, J. E. 2006. Tune: Runological. In Reallexikon der Germanischen Altertumskunde.
Band 31. (Eds.) H. Beck et al. Berlin – New York, 332–335.
Knott, K. 1998. Issues in the study of religion and locality. Method & Theory in the Study
of Religion (10), 279–290.
———. 2009. From locality to location and back again: A spatial journey in the study
of religion. Religion (39/2), 54–60.
———. 2014. Spatial Methods”. In The Routledge Handbook of Research Methods in the
Study of Religion. (Eds.) M. Stausberg & S. Engler. London & New York, 491–501.
Köbler, G. 1986. Eid. § 3. Rechtshistorisches. In Reallexikon der Germanischen
Altertumskunde. Band 6. (Eds.) H. Beck et al. Berlin – New York, 539–242.
Köhler, U. 2004. Sakralkönigtum. §2. Ethnolog. In Reallexikon der Germanischen
Altertumskunde, Band 26. (Eds.) Heinrich Beck et al. Berlin – New York, 181–183.
Kolbrún Haraldsdóttir, 1994. Fagrskinna. In Reallexikon der Germanischen Alter
tumskunde. Band 8. (Eds.) H. Beck et al. Berlin – New York, 142–151.
Kousgård Sørensen, J. 1985. Gudhem. Frühmittelalterliche Studien (19), 131–138.
———. 1989. Om personnavne på -vi/-væ og den før-kristne præstestand: Med nogle
overvejelser over en omstridt passage i Glavendrup-stenens indskrift. Danske stu-
dier. København, 5–33.
———. 1992. Haupttypen sakraler Ortsnamen Sudskandinaviens. In Der historische
Horizont der Gotterbild-Amulette aus der Ubergangsepoche von der Spatantike zum
Fruhmittelalter. (Ed.) K. Hauck. Göttingen, 228–240.
Krag, C. 1991. Ynglingatal og Ynglingesaga: En studie i historiske kilder. Studia humani-
ora. 2 (Oslo: Universitetsforlaget, 1991).
———. 2000. Norges historie frem til 1319. Oslo.
———.2001. Trosskiftet og teorien om sakralkongedømmet. Collegium Medievale (14),
211–225.
———. 2007. Trøndelag. In Reallexikon der Germanischen Altertumskunde. Band 35.
(Eds.) H. Beck et al. Berlin – New York, 268–272.
———. 2008. The creation of Norway. In The Viking World. (Ed.) S. Brink. London &
New York, 645–651.
———. 2009. Ynglingatal str. 4—omskrivningen viljes borg som dateringskriterium.
Maal og minne (02/2009), 129–146.
Krause, W. 1930. Die Kenning als typische Stilfigur der germanischen und keltischen
Dichtersprache. Schriften der Königsberger gelehrten Gesellschaft. 7. Jahr. Geistes
wissenschaftliche Klasse. Heft 1. Halle (Saale).
———. 1944. Ing. Nachrichten von der Akademie der Wissenschaften in Göttingen aus
dem Jahre 1944. Philol.-Hist. Klasse, Nr. 10. Göttingen.
———. 1966. Die Runeninschriften im älteren Futhark 1. Text. Mit Beiträgen von
Herbert Jankuhn. Abhandlungen der Akademie der Wissenschaften in Göttingen,
Philologisch-historische Klasse, 3. Folge 65. Göttingen.
Kristján Eldjárn. 1956. Kuml og Haugfé: Úr heiðnum sið á Íslandi. Reykjavík.
Kristján Eldjárn & Adolf Friðriksson. 2000. Kuml og Haugfé: Úr heiðnum sið á Íslandi.
New edition. Reykjavík.
Kristoffersen, S. 2000. Sverd og spenne: Dyreornamentik og sosial kontekst. Studia huma-
nitatis Bergensis 13. Bergen.
———. 2006. Tu. In Reallexikon der Germanischen Altertumskunde. Band 31. (Eds.)
H. Beck et al. Berlin-New York, 307–309.
Kuhn, H. 1971. Kleine Schriften II. Aufsätze und Rezensionen aus den Gebieten der ger-
manischen und nordischen Sprach-, Literatur- und Kulturgeschichte. Zweiter Band.
Literaturgeschichte, Heldensage und Heldendichtung, Religions- und Sittengeschichte,
Recht und Gesellschaft. Berlin – New York.
———. 1978. Kleine Schriften IV. Aufsätze und Rezensionen aus den Gebieten der germa-
nischen und nordischen Sprach-, Literatur- und Kulturgeschichte. Vierter Band.
Aufsätze aus den Jahren 1968–1976. Berlin – New York.
La Farge, B. 1994. Review Gro Steinsland 1991. Skandinavistik (24:1), 55–59.
La Farge, B. & J. Tucker. 1992. Glossary to the Edda. Heidelberg.
Läffler, Fr. L. 1905. Ännu några ord om de hedniska edsformulären i Norden. Antiqvarisk
tidskrift för Sverige (13:3), 1–11.
Läffler, L. Fr. 1911. Det evigt grönskande trädet vid Uppsala hednatämpel. In Fästskrift
till H. F. Feilberg. Svenska landsmål och svenskt folkliv. (Ed.) J. A. Lundell. Stockholm,
617–196.
Lamm, J. P. 2004. Figural Gold Foils Found in Sweden: A Study Based on the Discoveries
from Helgö. In Excavations at Helgö XVI. Exotic and Sacral Finds from Helgö. Kungl.
Vitterhets Historie och Antikvitets Akademien. Stockholm, 41–142.
Lamm J. P., H. Hydman & M. Axboe. 1999. “ ‘Århundradet s brakteat’—kring fyndet av
en unik tionde brakteat från Söderby i Danmarks socken, Uppland. Fornvännen
(94), 225–243.
Lamm, K. 1999. Helgö. §§ 2–5. In Reallexikon der Germanischen Altertumskunde. Band.
14. (Eds.) H. Beck et al. Berlin – New York, 286–291.
———. 2004. In J. P. Lamm 2004 (see above).
Landau, Peter 1982. Eigenkirchenwesen. In Theologische Realenzyklopädie 9. Berlin –
New York, 399–404.
Lange, Wolfgang. 1958. Studien zur christlichen Dichtung der Nordgermanen: 1000–1200.
Palaestra. Untersuchungen aus der Deutschen und Englischen Philologie und
Literaturgeschichte. Band 222, Göttingen: Vandenhoeck und Ruprecht.
Larsson, G. 2007. Ship and Society: Martime Ideology in Late Iron Age Sweden. Aun 37.
Uppsala.
Larsson, L. 2006a. Ritual building and ritual space: Aspects of investigations at the Iron
Age central site Uppåkra, Scandia, Sweden. In Old Norse Religion in Long-Term
Perspectives. Origins, Changes, and Interactions. Vägar till Midgård 8. (Eds.)
A. Andrén, K. Jennbert & C. Raudvere. Lund, 248–253.
———. 2006b. Hall, harg eller hof. Ett kulthus i Uppåkra. Kulthus & gravhus. Det ritua-
liserade rummets teori och praktik. (Eds.) M. Anglert et al. Stockholm, 143–152
———. 2006c. Rum, rymd och areal. Ett kulthus och dess närmiljö ur ett ceremoniellt
perspektiv. Bebyggelsehistorisk tidskrift (52), 7–19.
———. 2011. A ceremonial building as a ‘home of the gods’? In The Gudme/Gudhem
Phenomenon. (Eds.) O. Grimm & A. Pesch. Neumünster, 189–206.
Larsson, L. & K.-M. Lenntorp. 2004. The Enigmatic House. In Continuity for Centuries:
A Ceremonial Building and its Context at Uppåkra, Southern Sweden. Uppåkrastudier.
10. (Ed.) L. Larsson. Stockholm, 3–48.
Larsson, M. G. 1990. Runstenar och utlandsfärder: Aspekter på det senvikingatida sam-
hället medutgångspunkt i de fasta fornlämningarna. Acta Archaeologica Lundensia,
series in 8o, 18. Lund.
———. 1993. Kung Eriks nyordning? Spår av samhällsorganisation i vikingatidens
Uppland. In Kärnhuset i riksäpplet. Upplands fornminnesförening och hembygds-
förbunds årsbok. Uppland 1993. (Eds.) K. Blent et al. Uppsala, 105–134.
———. 1997. Vikingar i österled. En samlingsutgåva av: Ett ödesdigert vikingatåg, Ingvar
den vittfarnes resa 1036–1041. Väringar, Nordbor hos kejsaren i Miklagård. Rusernas
rike, Nordborna och rysslands födelse. Stockholm.
Lassen, A. 2011. Odin på kristent pergament: En teksthistorisk studie. Københavns
Universitet.
Laur, W. 2001. Germanische Heiligtümer und Religion im Spiegel der Ortsnamen:
Schleswig-Holstein, nördliches Niedersachsen und Dänemark. Kieler Beiträge zur
deutschen Sprachgeschichte. Band 21. (Eds.) F. Debus & W. Laur. Neumünster.
Leeuw, G. van der. 1956 (1933). Phänomenologie der Religion. Neue Theologische
Grundrisse. Tübingen.
Lehmann, F. R. 1930. Die Polynesischen Tabusitten: Eine ethnosoziolog. u. religionswis-
senschaftl. Untersuchung. Leipzig.
Lerche Nielsen, M. 1998. Glavendrup. § 2 Runologisches. In Reallexikon der Germani
schen Altertumskunde. Band 12. (Eds.) H. Beck et al. Berlin – New York, 196–200.
———. 2003. Från prästgården till drottning Disas Samna: Två problematiska runste-
nar från Gamla Uppsala. Fornvännen (98), 195–205.
———. 2009. Klerkernas tid 800–1520. In Det svenska samhället 800–1720: Klerkernas
och adelns tid. (Eds.) T. Lindkvist & M. Sjöberg. Lund, 31–202.
Lindkvist, T. & K. Ågren. 1985. Sveriges medeltid. Stockholm.
Lindquist, I. 1944. Kungadömet i hednatidens Sverige. Arkiv för Nordisk Filologi (58),
221–234.
———. 1963. Två vikingatida gudabeläten. In Kulturen (1962), 70–78.
Lindqvist, S. 1934. En förnäm kvinnas grav på Valsgärde gravbacke. Finska fornminnes-
föreningens tidskrift (40), 138–142.
———. 1936. Uppsala högar och Ottarshögen. KVHAAH. Stockholm.
———. 1941–1942. Gotlands Bildsteine. I–II. KVHAAH. Stockholm.
———. 1962–63. Forngutniska altaren och därtill knutna studier. Kungl. Humanistiska
Vetenskaps- Samfundet. Årsbok. 1962–1963. Stockholm.
Lindow, J. 1988. Scandinavian Mythology: An Annotated Bibliography. New York –
London.
———. 2001. Norse Mythology. A Guide to the Gods, Heroes, Rituals, and Beliefs. Oxford.
Ljungberg, H. 1938. Den nordiska religionen och kristendomen: Studier över det
nordiska religionsskiftet under vikingatiden. Nordiska texter och undersökningar 11.
Stockholm – København.
———. 1947. Tor: Undersökningar i indoeuropeisk och nordisk religionshistoria. Uppsala.
Ljungkvist, J. 2005. Uppsala högars datering. Fornvännen (100), 245–259.
———. 2006. En hiar atti rikR: Om elit, struktur och ekonomi kring Uppsala och Mälaren
under yngre järnålder. Aun 32. Uppsala.
———. 2008a. Dating two royal mounds of Old Uppsala: Evaluating the elite of
6th to 7th century Middle Sweden. Korrespondenzblatt: Römisch-Germanisches
Zentralmuseum. Mainz.
———. 2008b. Vendel—Development and Change on a Burial Ground over 1300 Years.
In Valsgärde studies: The place and its people, past and present, OPIA 42. Uppsala,
13–55.
———. 2013. Monumentaliseringen av Gamla Uppsala. In Gamla Uppsala i ny belys-
ning. (Eds.) O. Sundqvist and P. Vikstrand with J. Ljungkvist. Religionsvetenskapliga
studier från Gävle 9. Gävle, 33‒67.
——— et al. 2000. I maktens närhet: Två boplatsundersökningar i Gamla Uppsala.
Raä 285, Norra Gärdet. Raä 547, Mattsgården. Gamla Uppsala socken, Uppland.
Societas Archaeologica. Upsaliensis. Skrifter nr 1. Uppsala.
Löfving, C. 2005. Forsaringen åter till medeltiden. In Hälsingerunor: En hembygdsbok.
Bollnäs, 111–118.
———. 2010. Forsaringen är medeltida. Fornvännen (105), 48–53.
Løken, T. 1992. Ullandhaug sett i lys av Forsandundersøkelsene. AmS-Småtrykk 26.
Stavanger.
———. 2001. Oppkomsten av den germanske hallen: Hall og sal i eldre jernalder i
Rogaland. Viking (LXIV), 49–86.
———. 1869. Die Quellenzeugnisse über das erste Landrecht und über die Ordnung der
Bezirksverfassung des isländchen Freistaates. Abhandlungen der philosophisch-
philologischen Classe der Königlich Bayerischen Akademie der Wissenschaften.
12. Bd., 1. Abth. München.
———. 1873. Zur Urgeschichte der Godenwürde. Zeitschrift für deutsche Philologie (4),
125–130.
Mauss, M. 1997 (1950). The Gift: The Form and Reason for Exchange in Archaic Societies.
Orig. title Essai sur le Don. Trans. W. D. Halls. London.
McCutcheon, R. T. 1997. Manufacturing Religion: The Discourse on Sui Generis Religion
and the Politics of Nostalgia. New York – Oxford.
McDonnel & Maclean 2009. See Lucas 2009.
McKinnell, J. 2010. Ynglingatal—A Minimalist Interpretation. Scripta Islandica—
Isländska sällskapets Årsbok 60/2009, 23–48.
McKinnell, J., R. Simek & K. Düwel. 2004. Runes, Magic and Religion: A Sourcebook.
Studia Medievalia Septentrionalia. 10. Wien.
McTurk, R. W. 1975–76. Sacral Kingship in Ancient Scandinavia. A Review of Some
Recent Writings. Saga-Book. Vol. XIX. Parts II–III, 139–169.
———. 1994. Scandinavian Sacral Kingship Revisited. Saga-Book. Vol. XXIV. Part I,
19–32.
Meissner, R. 1921. Die Kenningar der Skalden. Ein Beitrag zur skaldischen Poetik.
Bonn – Leipzig.
Melefors, E. 2014. Gutnal þing och Gutnalia än en gång. In Gutagåtor. Historiska pro-
blem och tolkningar. (Ed.) P. Stobaeus. Visby, 211–230.
Meletinskij, E. 1977. Scandinavian Mythology as a System of Oppositions. In Patterns in
Oral Literature. (Eds.) H. Jason & D. Segal. Paris, 251–260.
Meskell, L. 2005. Objects in the Mirror Appear Closer Than They Are”. In Materiality.
(Ed.) D. Miller. Durham – London, 51–71.
Meulengracht Sørensen, P. 1986. Thor’s Fisching Expedition. In Words and Objects:
Towards a Dialogue between Archaeology and History of Religion. (Ed.) G. Steinsland.
Oslo, 257–278.
———. 1991a. Om eddadigtenes alder. In Nordisk hedendom: Et Symposium. (Eds.) Gro
Steinsland et al. Odense, 303–319.
———. 1991b. Håkon den gode og guderne: Nogle bemærkninger om religion og
centralmagt i det tiende århundrede– og om religionshistorie og kildekritik. In Fra
Stamme til Stat i Danmark II. Høvdingesamfund of kongemagt, pp. 235–244. (Eds.)
P. Mortenssen & B. M. Rasmussen. Århus, 235–244.
———. 1993. Saga and Society: An Introduction to Old Norse Literature. Odense.
———. 2003. The hall in Norse Literature. In Borg in Lofoten: A chieftain’s farm in North
Norway. (Eds.) Stamsø Munch, G. et al. Trondheim, 265–272.
Meyer, B. et al. 2010. The origin and mission of Material Religion. Religion (40:3),
207–211.
———. 2006. The Edges of the Old Norse World-View. A Bestiary Concept. In Old-
Norse Religion in Long-Term Perspective: Origins, Changes, and Interactions. Vägar till
Midgard 8. (Eds.) A. Andren, K. Jennbert & C. Raudvere, Lund, 63–67.
Nielsen, A.-L. 1996. Hedniska kult- och offerhandlingar i Borg. In Religion från stenålder
till medeltid. Riksantikvarieämbetet. Arkeologiska undersökningar. Skrifter nr 19.
(Eds.) K. Engdahl & A. Kaliff. Linköping, 89–104.
———. 2006. Rituals and power. About small buildings and animal bones from late
Iron Age. In Old Norse Religion in Long-Term Perspectives: Origins, Changes, and
Interactions: an international conference in Lund, Sweden, June 3–7, 2004, (eds.)
A. Andrén, K. Jennbert, C. Raudvere, Lund, 243–247.
Nikolaysen, N. 1881–1891. Kunst og Haandværk. Kristiania.
Nilsson, B. 1989. De Sepulturis: Gravrätten i Corpus Iuris Canonici och i medeltida nordisk
lagstiftning. Bibliotheca Theologiae Practicae. Kyrkovetenskapliga studier 44.
Stockholm.
———. 1991. Frids- och asylföreskrifter rörande den medeltida sockenkyrkan. In
Kyrka och socken i medeltidens Sverige. Studier till det medeltida Sverige 5.
Riksantikvarieämbetet. (Ed.) O. Ferm. Stockholm, 473–504.
———. 1992. Till frågan om kyrkans hållning till icke-kristna kultfenomen: Attityder
under tidig medeltid i Europa och Norden. In Kontinuitet i kult och tro från vikinga-
tid till medeltid. Projektet Sveriges Kristnande. Publikationer. 1. (Ed.) B. Nilsson.
Uppsala, 9–47.
———. 2000. Recension: Uppsalatemplet granskat på nytt. Henrik Janson, Templum
Nobilissimum. Adam av Bremen, Uppsalatemplet och konfliktlinjerna i Europa
kring år 1075. (Avhandlingar från Historiska institutionen i Göteborg. 21.) Göteborg
1998. Historisk tidskrift (1/2000), 105–111.
Ninck, M. 1935. Wodan und germanischer Schicksalsglaube. Darmstadt.
Nordahl, E. 1993. Södra Kungsgårdsplatån. Utgrävningarna 1988–1991. In Arkeologi och
miljöekologi i Gamla Uppsala: Studier och rapporter. OPIA. 7. Societas Archaeologica
Upsaliensis. (Ed.) W. Duczko. Uppsala, 59–63.
———. 1996. . . . templum quod Ubsola dicitur . . . i arkeologisk belysning. Aun 22. Med
bidrag av Lars Gezelius och Henrik Klackenberg. Uppsala.
———. 2001. Båtgravar i Gamla Uppsala: Spår av en vikingatida högreståndsmiljö.
Aun 29. Med bidrag av Anita Malmius, Petra Molnar, Anna Kjellström och Bengt
Schönbäck. Uppsala.
Nordberg, A. 2003. Krigarna i Odins sal: Dödsföreställningar och krigarkult i fornnordisk
religion. Stockholm.
———. 2006a. Ull und Ullin. In Reallexikon der Germanischen Altertumskunde. Band
31. (Eds.) H. Beck et al. Berlin – New York, 405–408.
———. 2006b. Jul, disting och förkyrklig tideräkning: Kalendrar och kalendariska riter i
det förkristna Norden. Acta Academiae regiae Gustavi Adolphi. XCI. Uppsala.
Price, N. S. 2002. The Viking Way: Religion and War in Late Iron Age Scandinavia. Aun. 31.
Uppsala universitet. Uppsala.
Price, S. R. F. 1993 (1985). Delphi and divination. In Greek Religion and Society. (Eds.)
P. E. Easterling & J. V. Muir. Cambridge, 128–154.
———. 1994 (1984). Rituals and Power: The Roman Imperial Cult in Asia Minor. Cambridge.
Puhvel, J. 1975. Remus et frater. History of Religions (15), 146–157.
Quinn, J. 2000. From Orality to Literacy in Medieval Iceland”. In Old Icelandic Literature
and Society. (Ed.) M. Clunies Ross. Cambridge, 30–60.
———. 2005. Women in Old Norse Poetry and Sagas. In A Companion to Old Norse-
Icelandic Literature and Culture. (Ed.) R. McTurk. Oxford, 518–535.
Radcliffe-Brown, A. R. 1952. Structure and Function in Primitive Society: Essays and
Adresses. London.
———. 1979. Taboo. In Reader in comparative religion: An anthropological approach
(fourth ed.). (Eds.) W. A. Lessa & E. Z. Vogt. ??, 46–56.
Rahmqvist, S. 1986. Gamla Uppsala by—Upplands största. In Från Östra Aros till
Uppsala: En samling uppsatser kring det medeltida Uppsala. Uppsala stads historia
7. Uppsala, 254–271.
———. 2000. Gamla Uppsala under medeltiden. In Myt, makt och människor: Tio
uppsatser om Gamla Uppsala. (Ed.) G. Friberg. Stockholm, 47–49.
Ramqvist, P. H. 1987. Mellannorrland under äldre järnålder: Några aspekter under äldre
järnålder. Bebyggelsehistorisk tidskrift (14), 105–126.
———. 1991. Perspektiv på regional variation och samhälle i Nordens folkvandringstid.
In Samfundsorganisation og Regional Variation: Norden i romersk jernalder og folkev-
andringstid. Beretning fra 1. nordiske jernaldersymposium på Sandbjerg Slot 11–15
april 1989. Jysk Arkæologisk Selskabs Skrifter. XXVII. (Eds.) C. Fabech & J. Ringtved.
Aarhus, 305–318.
———. 1995. Resources, Population and Power in the Migration Period: An Analysis of
the Archaeological Remains in Central Norrland. In Archaeology East and West of
the Baltic: Papers from the Second Estonian-Swedish Archaeological Symposium.
Sigtuna, May 1991. Estonian-Swedish Archaeological Symposium. 2. Theses and
Papers in Archaeology N.S.A. 7. Department of Archaeology, Stockholm University.
(Ed.) I. Jansson. Stockholm, 33–38.
Rappaport, R. A. 2000 (1999). Ritual and Religion in the Making of Humanity. Cambridge
Studies in Social and Cultural Anthropology. 110. Cambridge.
Ratke S. 2009. Guldgubber—Einblicke in die Völkerwanderungszeit. Ph.D. dissertation,
University of Bonn.
Ratke S. & R. Simek. 2006. Guldgubber: Relics of Pre-Christian law rituals? In Old Norse
Religion in Long- Term Perspectives: Origins, Changes, and Interactions: an interna-
tional conference in Lund, Sweden, June 3–7, 2004. (Eds.) A. Andrén, K. Jennbert,
C. Raudvere, Lund, 259–266.
Raudvere, C. 2002. Mellan liv och text: Ordets makt som litterärt motiv och rituell prak-
tik i norrön litteratur. In Plats och praxis: Studier av nordisk förkristen ritual. Vägar
till Midgård 2. (Eds.) K. Jennbert, A. Andrén & C. Raudvere. Lund, 31–68.
———. 2003. Kunskap och insikt i norrön tradition: Mytologi, ritualer och Troll
domsanklagelser. Vägar till Midgård 3. Lund.
Raudvere, C. & J. P. Schjødt (eds.). 2012. More than Mythology: Narratives, Ritual
Practices and Regional Distribution in Pre-Christian Scandinavian Religion. Lund.
Reichert, H. 2003. Vier Miszellen zum Urgermanischen und ‘Altrunischen’. Zeitschrift
für deutsches Altertum 132, 355–356.
Reifegerste, E. M. 1989. Die Hervarar Saga: Eine kommentierte Übersetzung und
Untersuchungen zur Herkunft und Integration ihrer Überlieferungsschichten.
Altnordische Bibliothek. 6. Köln – Leverkusen.
Riesebrodt, M. 2007. Cultus und Heilsversprechen. Eine Theorie der Religion. München.
Ringgren, H. 1989. Israels religion. Stockholm.
Ringgren, H. & Å. V. Ström. 1984 (1957). Religionerna i historia och nutid. Stockholm.
Ringstad, B. 1987. De store gravminnene—et maktideologisk symbol? Viking (50), 65–78.
———. 1991. Graver og ideologi: Implikasjoner fra vestnorsk folkevandringstid. In
Samfundsorganisation og Regional Variation: Norden i romersk jernalder og folkevan-
dringstid. Beretning fra 1. nordiske jernaldersymposium på Sandbjerg Slot 11–15 april
1989. Jysk Arkæologisk Selskabs Skrifter. XXVII. (Eds.) C. Fabech & J. Ringtved.
Aarhus, 141–150.
Ringsted, N. 2005. Karsvik platåhus—ett märkligt fornminne. Brommaboken: Bromma
hembygdsförenings Årsskrift. Bromma.
Robberstad, K. 1950. Frå gama log ny rett. Oslo.
———. 1967. Odelsrett. Norge. In Kulturhistoriskt lexikon för nordisk medeltid. Vol. 12.
Malmö, cols. 493–499.
Roesdal, E. 1982. Viking Age Denmark. London.
Rønne, P. 2002. Hus på hus i Melhus. SPOR (17), 21–23.
———. 2011. Horg, hov and ve—a pre-Christian cult place at Ranheim in Trøndelag,
Norway, in the 4rd–10th centuries AD. Adoranten (2011), 1–14.
Rosén, J. 1939. Striden mellan Birger Magnusson och hans bröder. Studier i nordisk poli-
tisk historia 1302–1319. Lund.
———. 1964. Krongods. In Kulturhistoriskt lexikon för nordisk medeltid. Vol. 9. Malmö,
cols. 434–438.
Røskaft, M. 1997. Kultkontinuitet og tidlig kirkebygging i Trøndelag. In Helgonet i
Nidaros: Olavskult och kristnande i Norden. (Ed.) L. Rumar. Stockholm, 225–239.
———. 2003. Maktens landskap: Sentralgårdar i Trøndelag ved overgangen fra vikingtid
til kristen middelalder, ca. 800–1200, No. 39 Skriftserie fra Historisk institutt,
Trondheim: Norges teknisk- naturvitenskaplige universitet.
Rostvik, A. 1967. Har och harg. Acta Academiae Regiae Gustavi Adolphi. XLIV. Studier
till en svensk ortnamnsatlas utgivna av Jöran Sahlgren. 11. Uppsala.
Røthe, G. 1997. Fra Olav Geirstadalv til Olav den helige. En religionshistorisk kommen-
tar til tåten om Olav Geirstadalv. In Gokstadhøvdingen og hans tid. (Ed.) T. Frost.
Sandefjord, 25–42.
———. 2007. Þorgerðr Hölgabrúðr—the fylgja of the Háleygjar family. Scripta
Islandica (57), 33–55.
Rousell, A. 1943. Islands gudehove. In Forntida gårdar i Island. (Ed.) M. Stenberger.
København, 215–223.
Rüpke, J. 1996. Controllers and Professionals: Analyzing Religious Specialists. Numen.
Vol. XLIII. (Eds.) H. G. Kippenberg & E. T. Lawson. Leiden, 241–262.
———. 2001. Die Religion der Römer: Eine Einführung. München.
Ruthström, B. 1988. Oklunda-ristningen i rättslig belysning. Arkiv för Nordisk Filologi
(103), 64–75.
———. 1990. Forsaringen—vikingatida vi-rätt? Arkiv för Nordisk Filologi (105), 41–55.
Rydh, H. 1920. Den magiska ringen från Hamrånge. Meddelanden av Gästriklands
Kulturhistoriska förening, 3–6.
———. 1936. Förhistoriska undersökningar på Adelsö. KVHAA. Stockholm.
Rydving, H. 1990. Scandinavian–Saami Religious Connections in the History of
Research. In Old Norse and Finnish Religions and Cultic Place-Names, Scripta insti-
tuti Donneriani Aboensis, XIII. (Ed.) Tore Ahlbäck (Stockholm: Almqvist & Wiksell,
1990), 358–373.
———. 1993. The End of the Drum-Time: Religious Change among the Lule Saami,
1640s–1740s. Acta Universitatis Upsaliensis. Historia Religionum 12. Uppsala.
———. 1996. Suttungamjödet, vishetens källa: Öl och mjöd i Skandinavien under
vikingatid och medeltid. In Rus och rit: Om droger och religioner. Religionshistoriska
forskningsrapporter från Uppsala. 9. (Ed.) H. Rydving. Uppsala, 155–162.
———. 2005. Review-article on Sundqvist 2002. Namn och bygd (93), 180–182
———. 2006. Constructing Religious Pasts: Summary Reflections. In The Making
of Christian Myths in the Periphery of Latin Christendom (c. 1000–1300). (Ed.) L. B.
Mortensen. Copenhagen, 315– 321.
———. 2011a. The ‘Bear Ceremonial’ and Bear Rituals among the Khanty and the Sami.
Temenos (46,1) (2010), 31–52.
———. 2011b. Le chamanisme aujord’hui: Constructions et deconstructions d’une illu-
sion scientifique. Études mongoles et sibériennes, centrasiatiques et tibétaines 42
(1911), 1–13.
Rydving, H. & R. Kristoffersson. 1993. Några samiska offerplatser. Fornvännen (88),
195–210.
Sabourin, L. S. J. 1973. Priesthood: A Comparative Study. Leiden.
Særheim, I. 1989. Litt om eldre busetjingsnamn i Klepp. Frá haug ok heiðni (2),
213–224.
———. 2015. Stadnamn og sentralitet i eit jærsk jordbrukssamfunn frå jenralderen. In
Et Akropolis på Jæren? Tinghaugplatået gjennom jernalderen. AmS-Varia 55. Eds
E.S. Kristoffersen, M. Nitter og E. S. Pedersen. Stavanger, 49–62.
Sahlgren, J. 1953. Från Morasten till Håga. Namn och Bygd (41), 1–45.
Salberger, E. 1980. Oklunda-hällens runristningar. Östgötska runstudier: Scripta runica.
(Ed.) E. Salberger. Göteborg 1, 2–23.
Saler, B. 1993. Conceptualizing Religion: Immanent Anthropologists, Transcendent
Natives, & Unbounded Categories. Studies in the History of Religions (Numen
Bookseries), Vol. LVI. Leiden – New York – Köln.
Van der Sanden, W. & T. Capelle (2001). Mosens Guder: Antropomorfe træfigurer fra
Nord- og Nordvesteuropas fortid/Immortal Images: Ancient anthropomorphic wood
carvings from northern and northwest Europe. Silkeborg.
Sandnes, J. 1964. Gno. hörgr i gardsnavn. Maal og minne, 113–120.
———. 1967. Trøndelags eldste politiske historie. Historisk Tidsskrift (46), 1–20.
———. 1971. Ødetid og gjenreisning: Trøndsk busetningshistorie ca 1200–1600. Oslo.
———. 1987. Hedensk kulthus og kristen kirke. In Festskrift til Alfred Jakobsen. (Eds.)
J. R. Hagland, J. T. Faarlund & J. Rønhovd. Trondheim, 144–152.
———. 1992. Haupttypen sakraler Ortsnamen Westskandinaviens Anhang: Karte der
sakraler Ortnamen in Rogaland (Stavanger Amt nach M. Olsen). In Der historische
Horizont der Götterbild-Amulette aus der Übergangsepoche von der Spätantike zum
Frühmittelalter. (Ed.) K. Hauck. Göttingen.
———. 2001. Lade. In Reallexikon der Germanischen Altertumskunde. Band 17. (Eds.)
H. Beck et al. Berlin – New York, 541–542.
Sandnes, J. & O. Stemshaug. 2007. Norsk stadnamnleksikon. Oslo.
Sanmark, A. 2004. Power and Conversion: A Comparative Study of Christianization in
Scandinavia. OPIA. 34. Uppsala.
———. 2005. Dietary Regulations in Early Christian Norway. Viking and Medieval
Scandinavia (1). Turnhout, 203–224.
Sanmark, A. & S. Semple. 2008. Places of assembly : new discoveries in Sweden and
England. In Fornvännen (103), 245–259.
Santesson, L. 1989. En blekingsk blotinskrift: En nytolkning av inledningsraderna på
Stentoftenstenen. Fornvännen (84), 221–229.
———. 1993. Eine Blutopferinschrift aus dem südschewedischen Blekinge: Eine
Neudeutung der einleitenden Zeilen des stentofter Steines. Frühmittelalterliche
Studien (27), 241–252.
Sawyer, B. 2000 (1992). Kvinnor och familj i det forn- och medeltida Skandinavien.
Occasional Papers on Medieval Topics. 6. Göteborg.
———. 1975. The Conversion of Iceland: A Survey. Viking Society for Northern Research.
Text Series. Vol. VI. Trans. P. Foote. London.
Stutz, U. & H. E. Feine 1989. Forschungen zu Recht und Geschichte der Eigenkirche:
Gesammelte Abhandlungen. Aalen.
Sundkvist, A. 2001. Hästarnas land: Aristokratisk hästhållning och ridkonst i Svealands
yngrejärnålder. OPIA 28. Uppsala: Institutionen för arkeologi och antik historia.
Sundqvist, O. 1995. Recension—Claus Krag, Ynglingatal og Ynglingesaga: En studie i
historiske kilder, Oslo: Universitetsforlaget, 1991. Svensk Religionshistorisk Årsskrift
(5). (Ed.) D. Westerlund. Stockholm, 158–162.
———. 1997. Runology and History of Religions: Some Critical Implications of the
Debate on the Stentoften Inscription. In Blandade runstudier II. Runrön. 11.
Runologiska bidrag utgivna av Institutionen för nordiska språk vid Uppsala univer-
sitet. (Eds.) L. Elmevik & L. Peterson. Uppsala, 135–174.
———. 1998. Kultledare och kultfunktionärer i det forntida Skandinavien. Svensk reli-
gionshistorisk årsskrift (7). (Ed.) D. Westerlund. Stockholm, 76–104.
———. 2002. Freyr’s Offspring: Rulers and Religion in Ancient Svea Society. Acta
Universitatis Upsaliensis. Historia Religionum. 21. Uppsala.
———. 2003a. The Problem of Religious Specialists and Cult Performers in Early
Scandinavia. Zeitschrift für Religionswissenschaft (1/2), 107–131.
———. 2003b. Priester und Priesterinnen. In Reallexikon der Germanischen
Altertumskunde. Band 23. (Eds.) H. Beck et al. Berlin – New York, 424–435.
———. 2004a. Uppsala och Asgård: Makt, offer och kosmos i forntida Skandinavien. In
Ordning mot kaos—studier av nordisk förkristen kosmologi. Vägar till Midgård 4.
(Eds.) A. Andrén, K. Jennbert & C. Raudvere. Lund, 145–179.
———. 2004b. Sakralkönigtum (Skandinavische Quellen). In Reallexikon der Germa
nischen Altertumskunde. Band 26. (Eds.) H. Beck et al. Berlin – New York, 279–293.
———. 2005a. Kvinnliga kultledares religiösa och sociala position i forntida
Skandinavien. Chaos. Dansk-Norsk tidsskrift for religionshistoriske studier (43), 9–29.
———. 2005b. Aspects of Rulership Ideology in Early Scandinavia—with Particular
References to the Skaldic poem Ynglingatal. In Das frühmittelalterlische Königtum.
Ergänzungsbände zum Reallexikon der Germanischen Altertumskunde. Band 49.
(Ed.) F. R. Erkens. Berlin – New York, 87–124.
———. 2005c. Tempel. In Reallexikon der Germanischen Altertumskunde. Band 30.
(Eds.) H. Beck et al. Berlin – New York, 327–338.
———. 2005d. Tabu. In Reallexikon der Germanischen Altertumskunde. Band 30. (Eds.)
H. Beck et al. Berlin – New York, 249–260.
———. 2005e. Sagas, Religion and Rulership: The Credibility of the Description of
Ritual in Hákonar saga goða. In Viking and Medieval Scandinavia 1. (Eds.) J. Quinn
et al. Turnhout, 225–250.
———. 2006. Från vikingatida aristokratiska hallar till medeltida stormannakyrkor.
Bebyggelsehistorisk tidskrift (52), 20–32.
———. 2007. Kultledare i fornskandinavisk religion. OPIA 41. Institutionen för arkeo-
logi och antik historia. Uppsala universitet. Uppsala.
———. 2009a. The question of ancient Scandinavian cultic buildings: with particular
reference to Old Norse hof. Temenos (45/1), 275–294.
———. 2009b. The Hanging, the Nine Nights and the ‘Precious Knowledge’ in Hávamál
138–145: The Cultic Context In Analecta Septentrionalia. RGA-E-Band 65. Berlin &
New York, 649–668.
———. 2010a. Om hängningen, de nio nätterna och den dyrköpta kunskapen i
Hávamál 138–145: Den kultiska kontexten. Scripta Islandica—Isländska Sällskapets
Årsbok (61), 68–97.
———. 2010b. Vi-platsernas väktare: Till frågan om kultledare i det förkristna
Mälarområdet. In Makt, kult och plats. Rapport från arkeologiska seminarier vid
Stockholms läns museum. (Eds.) P. Bratt & R. Grönwall. Stockholm, 90–95.
———. 2011a. An Arena for Higher Powers: Cultic Buildings and Rulers in the Late Iron
Age and the Early Medieval Period in the Mälar Region. In Ideology and Power in the
Viking and Middle Ages: Scandinavia, Iceland, Ireland, Orkney and the Faeroes. (Eds.)
G. Steinsland et al. Leiden, 163–211.
———. 2011b. Gudme on Funen: A central sanctuary with cosmic symbolism? In The
Gudme/Gudhem Phenomenon. (Eds.) O. Grimm & A. Pesch. Neumünster, 63–76.
———. 2012a. Var sejdhjällen (fvn seiðhjallr, hjallr) en permanent konstruktion vid
kultplatser och i kultbyggnader? Fornvännen (107), 280–285.
———. 2012b. Religious ruler ideology in pre-Christian Scandinavia. A contextual
perspective. In More than Mythology. Narratives, ritual practices and regional distri-
bution in pre-Christian candinavian religions. Eds. C. Raudvere & J. P. Schjødt,
J. P. Lund, 225–262.
———.2013. Gamla Uppsala som förkristen kultplats: en översikt och en hypotes. In
Gamla Uppsala i ny belysning. (Eds.) O. Sundqvist & P. Vikstrand with J. Ljungkvist.
Gävle, 69–112.
———. 2013b. Snorri Sturluson as a historian of religion—The credibility of the
descriptions of pre-Christian cultic leadership and rituals in Hákonar saga góða. In
Snorri Sturluson—Historiker, Dichter, Politiker. Ergänzungsbände zum Reallexikon
der Germanischen Altertumskunde. Band 85. Red. H. Beck et al. Berlin/New York:
de Gruyter, 71–92.
———. 2014. Religious and Ideological Aspects of Hall Interiors in Late Iron Age”. In
Runsa Borg. Representative Life on a Migration Period Hilltop Site—a Scandiavian
Perspective. Papers from the project Runsa borg, Uppland. No 2. (Ed.) M. Olausson.
Östersund, 109–144.
———. 2015. Gårdarna Hauge och Tu i Klepp. En kultplats med kvinnliga kultledare?
In Et Akropolis på Jæren? Tinghaugplatået gjennom jernalderen. AmS-Varia 55. Eds
Ellen Siv Kristoffersen, Marianne Nitter og Einar Solheim Pedersen. Stavanger,
89–105.
———. 2015. M.S. Vapen, våld och vi-platser: Skändande av helgedomar som makt-
strategi i det vikingatida Skandinavien. In Krig och fred i vendel- och vikingatida tra-
ditioner. (Eds.) S. Olsson & H. Rydving. Bergen.
Sundqvist, O. & A. Hultgård. 2004. The Lycophoric Names of the 6th to 7th Century
Blekinge Rune Stones and the Problem of their Ideological Background. In
Namenwelten. Orts- und Personennamen in historischer Sicht. Ergänzungsbände
zum Reallexikon der Germanischen Altertumskunde. Band 44. (Eds.) A. van Nahl
et al. Berlin – New York, 583–602.
Sundqvist, O. & P. Vikstrand. 2014. Disevid och Distingen—spår av östnordisk diskult?
In Den heliga platsen. Handlingar från symposiet Den heliga platsen Härnösand 15–18
september 2011. (Eds.) E. Nyman et al. Sundsvall, 153–178.
Svanhildur Óskarsdóttir. 1994. Daið þér Ynglinga? In Sagnaþing helgað Jónasi
Kristjánssyni sjötugum 10. April 1994. 2 vols. Ed Gísli Sigurðsson et al., 761–768.
Svavar Sigmundsson. 1992. Áttrúnaður og örnefni. In Snorrastefna 25.–27. Júlí 1990.
Ritstjóri: Úlfar Bragason. Reykjavík.
Sveaas Andersen, P. 1977. Samlingen av Norge og kristningen av landet. 800–1130.
Handbok i Norges historie. Bind 2. Oslo.
Sveinbjörn Rafnsson. 1974. Studier i Landnámabók: Kritiska bidrag till den isländska
fristatstidens historia. Lund.
———. 1983. Frásögur um Fornaldarleifar 1817–1823. Reykjavík.
Svensson, K. 2010. Götavi—en vikingatida kultplats i Närke. In Makt, kult och plats.
Rapport från arkeologiska seminarier vid Stockholms läns museum. (Eds.) P. Bratt
& R. Grönwall. Stockholm, 58–68.
Tambiah, S. J. 1979. A performative approach to ritual. Proceedings of the British
Academy 65, 113–169.
Taranger, A. 1890. Den angelsaksiske kirkes indflydelse paa den norske. Kristiania.
———. 1913. The Meanings of the Words Óðal and Skeyting in the Old Laws of Norway.
In Essays in Legal History Read before the International Congress of Historical Studies
Held in London in 1913. (Ed.) P. Vinogradoff. London, 159–73.
———. 1934–1936. Om Kongevalg i Norge i Sagatiden. Norsk Historisk Tidsskrift (30),
110–166 / 273–311.
Tesch, S. 1990a. Bakgrund och metod. In Makt och människor i kungens Sigtuna:
Sigtunautgrävningen 1988–90. 28 artiklar om de preliminära resultaten. (Ed.) S. Tesch.
Sigtuna, 7–19.
———. 1990b. Stad och stadsgård. In Makt och människor i kungens Sigtuna:
Sigtunautgrävningen 1988–90. 28 artiklar om de preliminära resultaten. (Ed.) S. Tesch.
Sigtuna, 23–50.
———. 2001. Olof Palme, S:ta Gertrud och Sigtunas medeltida kyrkotopografi. In
Biskopen i museets trädgård. En arkeologisk gåta. Sigtuna Museers skriftserie 9.
(Eds.) S. Tesch & R. Edber. Sigtuna, 9–44.
Thompson, C. W. 1975. Studies in Upplandic Runography. Austin, Texas – London.
Vikstrand, P. 1993. Förkristna sakrala ortnamn i Jämtland. Namn och bygd (81), 49–84.
———. 1996. Kan ortnamnselementet skälv, fsv. skialf, skiælf, ha haft en sakral
innebörd? In Från götarna till Noreens kor: Hyllningsskrift till Lennart Elmevik på
60-årsdagen 2 februari 1996. Skrifter utgivna genom Ortnamnsarkivet i Uppsala.
Serie B. Meddelande. 11. (Eds.) E. Brylla et al. Uppsala, 201–208.
———. 2000. Konungen och helgedomen. In Oluf Rygh: Rapport fra symposium på
Stiklestad 13.–15. Mai 1999. NORNA-rapporter. 70B. (Eds.) B. Sandnes et al. Uppsala,
43–54.
———. 2001. Gudarnas platser: Förkristna sakrala ortnamn i Mälarlandskapen. Acta
Academiae regiae Gustavi Adolphi. LXXVII. Studier till en svensk ortnamnsatlas
utgivna av Thorsten Andersson. 17. Uppsala.
———. 2002. Sacral place-names in Scandinavia. ONOMA (2002), 121–143.
———. 2004a. Sakrale Namen. In Reallexikon der Germanischen Altertumskunde. Band
26. (Eds.) H. Beck et al. Berlin – New York, 167–179.
———. 2004b. Skutá and Vendil. Two Place Names in Ynglingatal. In Namenwelt. Orts-
und Personennamen in historischer Sicht. Ergänzungsbände zum Reallexikon der
Germanischen Altertumskunde. Band 44. (Eds.) A. van Nahl, L. Elmevik & S. Brink.
Berlin – New York, 372–387.
———. 2006. Ásgarðr, Miðgarðr, and Útgarðr: A Linguistic Approach to a Classical
Problem. In Old- Norse Religion in Long-Term Perspective: Origins, Changes, and
Interactions. Vägar till Midgård 8. (Eds.) A. Andrén, K. Jennbert & C. Raudvere.
Lund, 354–357.
———. 2007. Bebyggelsenamnen i Mörbylånga kommun. Sveriges ortnamn. Ortnamnen
i Kalmar län 7. Uppsala.
———. 2009. Förkristna sakrala personnamn. Studia anthroponymica Scandinavica 27
(2009), 5–31.
———. 2010. Ortnamn och den äldre järnålderns högstatusmiljöer. In Makt, kult och
plats. Rapport från arkeologiska seminarier vid Stockholms läns museum. (Eds.)
P. Bratt & R. Grönwall. Stockholm, 23–30.
———. 2011. Sakrala ortnamn—och centrala: Om den religiösa sfärens ställning i
dimensionen centrum- pereferi under förkristen tid. In Navnemiljøer og samfund i
jernalder og vikingetid: Rapport fra NORNAs 38. Symposium i Ryslinge 12.–15. Maj 2009.
(Eds.) L. Eilersgaard Christensen & B. Jørgensen. Uppsala, 315–346.
———. 2013. Namnet Uppsala. In Gamla Uppsala i ny belysning. (Eds.) O. Sundqvist &
P. Vikstrand with J. Ljungkvist. Religionsvetenskapliga studier från Gävle 9. Gävle,
135–160.
Vries, J. de. 1932. Über Sigvats Álfablót-Strophen. Acta Philologica Scandinavica (7),
169–180.
———. 1956–57. Altgermanische Religionsgeschichte I–II. Grundriss der germanischen
Philologie 12/1–2. Berlin.
———. 1977. Altnordisches etymologisches Wörterbuch. Leiden.
———. 2011a. Arkeologin bakom Rimbert: Om Hergeiirs och Gautberts kyrkor och
borgen i Birka. Fornvännen (106), 100–112.
———. 2011b. Exotiska föremål och nya seder under mellersta järnåldern. In Förmodern
globalitet: Essäer om rörelse, möten och fjärran ting under 10 000 år. (Ed.) A. Andrén.
Lund, 109–130.
———. 2013. Gamla Uppsala—på nytt. In Gamla Uppsala i ny belysning. (Eds.)
O. Sundqvist & P. Vikstrand with J. Ljungkvist. Religionsvetenskapliga studier från
Gävle 9. Gävle, 161–204.
Zachrisson, T. & M. Kempff Östlind. 2006. Ceremoniella rum. Bebyggelsehistorisk tid-
skrift (52), 5–6.
Zernack, J. 1998. Fulltrúi. In Reallexikon der Germanischen Altertumskunde, Band 10.
(Eds.) H. Beck et al. Berlin – New York, 243–245.
Zinser, Ha. 1993. Kosmologie. In Handbuch religionswissenschaftlicher Grundbegriffe.
Band III. (Eds.) H. Cancik, B. Gladigow & K.-H. Kohl. Stuttgart – Berlin – Köln,
446–448.
Žižek, S. 2012 (1994). Introduction: The Spectre Ideology. In Mapping Ideology.
London – New York, 1–33.
——— (ed.) 2012 (1994). Mapping Ideology. London – New York.
Frostuþingslǫg 300 goðar/goði
Frotho III (king) 467–468 in general 56–57
full (memorial drinking rites) 481–482 chieftains as 168–169
fulltrúi 88, 519 cultic leadership of 57, 177–178, 306,
funeral feasts 330–331
drinking rituals at 481–488 and aðalból 452–453
for Ǫnundr 3 functions of 148, 164–165, 174–176
rituals performed at 3–4 gift-giving ceremonies by 359–360
royal 476–492 hereditary nature of 176–179
fylkiskirkja/fylkiskirkjur (shire churches) judical function of 398
51, 52, 141, 147, 210, 440, 441, 442 and oath-rings 376, 387–390, 398
fylki (shire) 47–48, 171, 300, 308, 440–441 and power control 532–533
Fyrby stone (Sö 56) 262 as protector of cultic sites 148
recruited from lower Norwegian elite 57
Gådi stone (U 739) 362 relations with gods 87, 89, 539
garðr 262 religious origin of 60
Gautrekr (king) 494 as sacrificers 330–331
Gautreks saga 494 selection of 175–176
Geertz, Clifford 289 sharing of goðorð by 179
Geirmundar þáttr heljarskinns 257 women as 371–373
Geirmundr Hjǫrsson 55 see also *gudhar/*gudhi; gyðjur; hofgoðar/
genealogical knowledge hofgoði; Þórsnesingar family
existance of 69 Godøy (Saltfjorden) 278
importance for the elite 93 n156 gods
and land rights 448–453 images of
and royal inaugurations 492–493 aniconic 280
in Ynglingatal 65–66, 70–71 possessed by the devil 277–279
Gennep, Arnold van 498 personal devotion to 88–89, 517–518
geography rings carried by 427
of Svetjud 35–36 worship of many 516–518
of Trøndelag area 46–47 worshipped
Gerðr 73–74, 79, 83, 412, 415–416 in Iceland 88–89
Gesta Hammaburgensis Ecclesiae Pontificium by Svear 72, 121–122
(Adam of Bremen) see also under specific gods
evaluations of 113–115 god’s nails (reginnaglar) 236, 237, 257
extracts from 110–113 gœstning 528
people contesting 114–115 gold foils
purpose of 114–115 in general 236–237, 407–408
see also Adam of Bremen from Bergagård (Slöinge) 417
gift-giving ceremonies 359–363, 364 from Borg (Lofoten/Vestervågøy)
Girard, René 343 related to hearth 244
Gísla saga Súrssonar 286, 287, 323, 331, 359 related to post-holes 142, 145–146,
Glaðsheimr 201, 202, 217, 521 229, 236, 295–296, 411–412, 417
Glælognskviða (Þórarinn loftunga) 358 as temple coins 194
Glavendrup stone (DR 209) 184, 386 from Eskilstuna 194, 409
Glitnir 218 gesture language on 414–415
goð place-names 260–261, 278 from Gudme 417
goðahús 95, 104 n27, 155–156, 284, 298 from Hauge-Tu 255–256, 410, 412
Goðakelda 260 from Helgö 128