Any study employing a cross-disciplinary methodology must address multiple issues with source materials. Varied disciplines approach source documentation in a way that is acceptable to that specific discipline but anathema to another. An example of this disparity regards the dating of sources. In the case of an archaeology thesis, it is expected that only the latest and most relevant research would be employed, and this approach has been implemented where appropriate. In the case of architectural history, and church architecture specifically, this approach can be
problematic. The World Wars demonstrated the ease at which cultural legacies could be destroyed. This provided the impetus to document existing churches in great detail. This documentation resulted in a series of technical works originating throughout Scandinavia consulted for this thesis. It is a truism that interpretations of church imagery, dating of fabric, historical analysis of inventory, etc., have been significantly refined over the course of the past decades. The architectural drawings though have not. As an example, the classical work on Nidaros Cathedral was written in the mid-1960’s (Fischer, Gerhard 1964 and Fischer, Gerhard 1965), and its architectural data is used in this thesis. Where interpretations that are more recent are available, that source is employed. As an extreme example, Appendix A employs extensive technical drawings, most a century or more old. If more recent architectural drawings or
interpretations exist, these are referenced. As a generalization, these older technical works are the most likely to be used, if only because the buildings no longer exist in their original form. This
approach, although not common in archaeology per se (which prefers the most recent research), is normal in studies related to architecture.
Similar dating issues relate to relevant literary works. With the expectation that the readers of this thesis will likely be English speakers, a special effort has been made to reference English translations of Old Norse/Icelandic works where possible. These are often quite old. The ‘Viking Revival’ of the 19th century resulted in numerous English translations. These are occasionally the only translations undertaken of these works. An example of this is the historically important (King) Sverris saga, whose only translation into English is from 1899 (Sephton 1899). The earliest Norwegian laws, the 12th century Gulathingslova (Helle 2001 details the various dating schemes) have only been fully translated into English once (Larson 1935). A similar situation exists with regards to the vitae. These works, often well known in Scandinavia, are only rarely translated into English in the modern world. The important Life of Ansgar (the Vita Anskarii) for example currently exists in only one complete English translation (Robinson 1921).
The use of Anglo-Saxon examples for illustrative purposes regarding pre-Christian practices in this paper is, with exceptions, generally avoided. While agreeing with Carver that ‘It would be more productive... to move the whole study of Anglo-Saxon paganism... into a region defined by the Northern Seas...’ (Carver 2010: 5), the unique history of cultural traditions in England in this time frame complicates analysis. These traditions included a mix of Brythonic influences and Roman influences not seen in Scandinavia (Blair 2005: 8-11). To the extent that we can trust Bede, he states that Eadbald the king of Kent (died c. AD 640), was the first king to destroy idols (Ecclesiastical History Book III, Chapter 8,1 Bede 2008: 122; doubts regarding Bede’s historicity can be found in Church 2008: 162-80). It is suggested that by the time of Bede, paganism or at least references to it, had died out (Blair 2005: 168). This is most definitely not the case in Scandinavia. The centuries between the English and the Scandinavian conversions limits their applicability for illustrative purposes.
1 Anno Dominicae incarnationis sexcentesimo quadragesimo, Eadbald rex Cantuariorum transiens ex hac vita...Hic
primus regum Anglorum in toto regno suo idola relinqui ac destrui... Historiam Ecclesiasticam Gentis Anglorum, Liber Tertius: viii (Henderson and King 1930: 360-2)
1.6. Terminology
This work involves the comparison of sources in seven different languages and embraces
evidence from across 1000 years of Scandinavian history. The varied language and terminology, even for commonly used words, presents significant barriers. For example, the use of the term ‘Germanic’ is generally held to be appropriate to anything based out of the Germanic language groups. This covers such a large swathe of time and place as to render the term effectively meaningless. The term ‘Norse’ or ‘Nordic’ is preferable and commonly used by in academic writing. Gro Steinsland’s Norrøn Religion: Myter, Riter, Samfunn (2005) is an example. The term ‘Scandinavian’ and ‘Norrøn’ have historically been associated with pan-Scandinavian cultural movements in Norway, Sweden, and Denmark. Both terms imply a cultural uniformity and political unity that never existed in reality. Certainly employing these terms in relation to the Roman Iron Age is factually incorrect. It has recently been argued that the concept of ‘Norse’ began in the 6th century (Hedeager 2011: 30-2). In order to create comparative clarity in this thesis ‘Norwegian’ is employed to denote traditions culturally distinct to Norway. ‘Norse’ is used as cognate for ‘Scandinavian’, refering to traditions seen across a broader geographic and cultural range, limited in this discussion to the core countries of Norway, Sweden, Denmark, and Iceland. Scandinavia is strictly used as a geographiacal term, also restricted to Norway, Sweden, Denmark and Iceland.
The term used by many sources for ancient, indigenous belief systems is ‘pagan’, or ‘heathen’. Caution must be applied because both terms are used in contemporary English differently and each has religious connotations. ‘Pagan’ in modern usage refers to earth-centred, recreated or ‘new age’ religions, and ‘heathen’ to modern recreations of what is perceived as Norse-based or Viking religions. Both words have roots in the countryside. Pagan comes from the Latin
pāgānus, meaning ‘rustic, peasant, citizen’ (Hoad 1991: 330). The term ‘heathen’ exists in
essentially unchanged form in ON (hieðan, Cleasby and Vigfusson 1957: 245), OE (hǣðen, Clark Hall and Meritt 1969: 166), OS (hēthin, Onions 1966: 433) and OHG (heidan, Köbler 2006: 210-11). Modern Nordic languages are also similar: Nor (hedning, Kunnskapsforlaget 2012: 726), Sw (hedning, Norstedts 2010: 609), Dan (hedning, Axelsen and Jones 2003: 406) and Ic (heiðingi, JPV 2006: 323). The term, meaning ‘dwelling on the heaths’ functioned as a derogatory term for non-Christians. It was also occasionally employed as a derogatory term for Christians that were not a member of the sect of the writer. The earliest record of the term is in
Ulfilas’ 4th century Gothic bible translation, and its use implies the non-Christian meaning (haiþno Mark 7:26, Snædal 2013a: 183; text at Snædal 2013: 28). The etymology is somewhat uncertain although it is suggested it may come from Armenian meaning ‘nation’ (Barnhart 2005: 471). In either case, the words pagan and heathen are cognate. These terms did not identify a specific brand of pre-Christian religion, but described a religious practice that was not Christian (Palmer 2007: 403-4). Christianity drew a distinction between its urban ‘civilized’ practices, and the rustic ‘uncivilized’ practices performed in the countryside. This has ramifications in
interpreting the ancient sources and to which persuasion of paganism they are referring. These terms do not represent what these peoples or religions called themselves. The use of a single term (pagan) is problematic in that it suggests a uniformity that is non-existent. There is no reason to believe that pagan practices in non-Norse regions were similar to the pagan traditions followed in Norway. In Norse sources, the term ‘heathen’ is commonly employed. Scholarly tradition translates this term into English as ‘pagan’ (Steinsland 2005: 13-4). Although this tradition is occasionally followed herein, to avoid confusion the term ‘pre-Christian’ is preferred. It is important to stress that this refers specifically to pre-Christian practices in the Norse
countries, when referring to pre-Christian practices in other regions this is demarcated. Also, to reiterate, these terms were occasionally used derogatorily against different Christian sects. Terminology related to timber building traditions is vital to this work. It also presents difficulties in that there is no universally agreed architectural vocabulary. Architectural and archaeological works that define the vocabulary of timber structures often contradict each other. Britain has been in the forefront of timber-building archaeology with its long-standing tradition of investigating barns, guildhalls, houses, and other medieval timber structures. The Council of British Archaeology has produced a more or less, standard terminology around timber
architecture (Alcock et al 1996). There are though limitations in the relevancy of this work in that England has only one timber church from the study period (The Church of St Andrew, Greensted-juxta-Ongar, Essex). The terminology, therefore, does not encompass the features seen in Norwegian stave-built churches. There are architectural works in Norwegian that have attempted to codify this vocabulary (for example Christie 1976; Anker and Havran 2005a: 338- 41, translated into English 2005: 338-41). These works though are not presented in a coherent standardized form. Works related to eastern Europe and Russia describing architectural forms and vocabulary exist but are similarly complicated by being regionally limited and in several
languages (for example: Дpaґaн [Dragan] 2014: 207-398; Cisowski et al 2005; Patterson 2001: 47-54; Opolovnikov and Opolovnikova 1989: 251-254; Żak and Piecuch 2011; among others). The standard work on the timber church building tradition (Ahrens 2001) is written in German, but tends to use Norwegian-based terminology when referring to stave church traditions. All of these works have suggested some of the terminology used in this thesis, and where this becomes incomprehensible, the terms are cross-referenced to each other to facilitate clarity. The larger issue specifically in the Norwegian stave church tradition is that Norwegian words and the features they describe are sometimes non-existent in English. For example, there is no easy translation of the term svalgang. This feature does not exist in other Christian building traditions. It is translated as walkway, pentice, ambulatory, galilee, or side-aisle, and although it has
elements of all of these, it cannot be characterised by any one of these terms (is it detailed in section 4.3). Kniplingskrave (literally Nor ‘lace collar’, the term was coined by Bugge 1981: 10) has been defined as ‘a sort of “lace collar” of stiffening spandrels, which in concert with the frame of staves, forms a raised baldachin’ (Norberg-Schulz 1996: 83). Although this is accurate, one is not the wiser by knowing it, and the likelihood of recognizing it is no better after reading the definition. As a rule, illustrative photographs are provided when a new term is introduced to the text.
Given that this thesis is a cross-disciplinary work, there are many definitions from different fields that are employed. The most important terms referenced herein are:
*(asterisk): following common usage in historical linguistics, an asterisk (*) is placed before a word that is not directly recorded in a text, but whose existence can be inferred and reconstructed based on other linguistic sources (usually via comparative linguistics).
Central Places: the modern countries of Norway, Sweden, and Denmark did not exist in the early Viking Age. This era was dominated by what can be described as a rural power centre called a central place. It should be noted that the Norse version of this operates somewhat differently than seen in other parts of Europe (Hamerow 2004: 157). These sites are demarcated from other agrarian sites in that they are artefact-rich sites whose remains can be interpreted in terms of centralized ‘functions of an ideological, political, judicial and economic nature’ (Lundqvist 1997: 179). From a practical perspective, the local kings/queens participated and controlled what were effectively state religions on a small scale (Slupecki 2009: 26). These sites
would reach their zenith between c. AD 400 to c. AD 900 (Nicklasson 2002: 121-2).
Archaeologically, these sites can be identified by a number of specialized features (Fabech 2006: 27):
• There is a central area usually focused on a hall. • The hall is often associated with royalty.
• Workshop and production areas are in this central area. • There are satellite communities around it.
• Ritual sites, such as cult-buildings, sacred trees and cemeteries are part of the central place (see below).
• Often (and controversially, see below) an assembly site is found at these sites. • Extensive and varied material artefacts, often the result of trading networks, are
discovered at these sites.
There is significant controversy regarding this concept and its applicability to Scandinavia
(summarised in Sundqvist 2015: 100-2). Among other issues, the most recent research challenges the traditional split between a few central places and numerous peripheral areas. Metal
detectorist finds in areas in Sweden not associated with central places, such as Halland
(Nicklasson 2002: 111-23) and Östergötland (Rundkvist 2011) suggest that the number of central places, and their respective halls, have been underestimated, The identification of assembly sites archaeologically, in an era with little or no written records, has also proven problematic. Recent work in England and Sweden suggests previous views of apparent shared similarities in form and function, as well as topographical associations, of assembly sites in Europe are overstated
(Sanmark and Semple 2008: 245-6). This research suggests that although it is accurate to say there is evidence for places of assembly for large groups, these sites have great variety (Sanmark and Semple 2008: 255-6). It is worth noting that theophoric toponymical elements (especially the term ve, vi and the later hov; detailed in section 5.3) are often associated with central places and this suggests a ritual element (Brink 2012: 63-4; examined in section 5.3).
Conversion: the concept of conversion, and indeed its very definition, is complicated and warrants its own detailed analysis (this is presented in section 1.1.3).
Cult-House: historically the term ‘temple’ is the commonly used translation for various Old Norse terms in Old Norse/Icelandic literature (detailed in section5.3). Archaeology has also
inconsistently employed over time, several different terms for similar structures (detailed in section 1.1.2). Following the lead of Olof Sundqvist, I employ the term ‘cult-house’ for a
dedicated cultic structure, and ‘hall’ for a multi-functional aristocratic structure (Sundqvist 2015: 5, 108).
Diplomatic Edition: this is a printed or online manuscript that replicates, as closely as possible, the original manuscript. There is no normalization of the spelling or letters, no stylistic changes (such as putting in paragraphs where none were in the original), nor are poetic elements spaced out according to modern sensibilities. Most of the commonly used Icelandic reference works are in fact not diplomatic editions. Spelling and formatting is normalized (for example in the Íslenzk
Fornrit editions, or Kuhn’s Edda). This approach and its critique is detailed in the section: Notes
on Translations.
External ambulatory (Nor svalgang): this is an enclosed, often narrow, walkway that goes around the entire outside of a church. This feature is described in detail in section 4.3, Fig. 4 provides an image for immediate reference.
Fabric: this term is used in the architectural sense, meaning the structural parts of a building. This excludes portable artefacts such as furniture (Curl 2006: 273).
Fortidsminneforeningen: this is the Society for the Preservation of Ancient (Norwegian) Monuments. This organization is responsible (among other structures and monuments) for conserving the existing stave churches.
Guldgubber: guldgubber (Sw), gullgubber (Nor) and guldgubbar (Dan): means ‘little man made of gold’. These are small stamped gold foil figures, although examples in silver are known. There is a long history of investigating these and they are often found in locations with historical reference to cultic activities, areas with theophoric names, deposition sites in land or water, and associated with central places (Lundqvist: 1997: 184). Although widely interpreted as cultic in nature, their meaning, imagery, and usage has been problematic. Various interpretations have been suggested. Royal imagery in the form of a hiero-gamos sacred marriage (Steinsland 1990 and 1991), temple money (Watt 1999: 174), votive offerings (Lundqvist 1997: 184), legal related usage and iconography (Ratke and Simek 2006) have all been proposed. The very heterogeneity
of this imagery suggests variant usages and supports regional or localized rationales. One of the Icelandic sagas appears to record a tradition that bears similarities to guldgubber usage. The
Saga of the People of Vatnsdal (Section 10), uses what appears to be a guldgubber as a plot
device, Igimund is given an amulet by the King. This amulet disappears from his purse. It reappears when he is digging postholes for his cult-house (an abridgement of the relevant parts of that saga are in Q:84).
Inventory: this term refers to the portable artefacts located within a structure. For example, chairs, movable fonts, etc. located in a church.
Medieval popular Bible: this term is widely used in literary disciplines. It signifies the entirety of biblical material as presented to the lay public, and what they would have been familiar with. The biblical text employed in medieval times was different in different places (and changed over time). The standard bible traditionally employed by scholars when discussing this time frame is the Douai Rheims English translation (1609/1582) of St Jerome’s Vulgate (DRB 2011). I have followed this tradition to maintain consistency in source quotes, and because this is as close to the Latin original that would have been recognizable at the time. It is recognized that different textual versions were employed and available in different temporal and geographic regions. Post Church (Nor stolpekirke): this term, rather counter-intuitively, is defined somewhat
differently in different countries (this is detailed in section 1.1.2). It is best thought of as a timber church with intermittent earth-bound posts (Nor stavs), with a sill between it to support the wall plate.
Raised central room (Nor indre reisning): a majority of stave churches possess this
characteristic. In a stone church this is similar to what would be termed a ‘clerestory’. This is a tall nave, with an upper story lined with windows. This is usually supported by an arcade in basilican structures. In these timber structures, the same idea is employed, but there are no windows (Fig. 5).
Riksantikvaren: (technically the Direktoratet for kulturminneforvaltning): this is the governing body responsible for Norwegian cultural heritage. In the context of this work, they maintain the stave church archives and provide funding for research and upkeep of these structures.
Sheela na gig (the term is Irish, its etymology is a topic of debate): these are graphic
representations, usually in stone, of a female exposing her genitals. Their meaning and purpose has been greatly debated. Scholars have interpreted them in various ways, but most commonly as pagan deities (Margaret Murray 1934: 93-100), symbols of luck for childbearing (Freitag, 2004: 68-105) and as graphic representations of the sin of lust (Weir and Jerman 1999: 11-22). Sheela
na gigs are not a universal feature of catholic imagery, they are geographically found mainly in
England, Wales, and Ireland (Fig. 6).
Stave Church: the definition of a stave church as defined by the Norwegian cultural heritage authorities is a standing wooden church built between c. 1130 (the date of the current Urnes Stave Church) and 1350 (the date of the black plague in Norway).These structures share many common defining structural elements, including: a port and beam construction situated on a rock foundation with corner posts that are notched together (details after Anker and Havran 2005. 27). I detail and critique this dating and definition in section 1.3.
Stemma: this is a document (often displayed graphically) that traces the evolution of surviving manuscripts, with the original manuscript and the relationships of various copies. It can be visualized as a family tree of a particular manuscript.
Linguistic and Translation Issues
Modern Norwegian names, terminology and especially place-names, can be confusing. Norway has two languages, Nynorsk (also called New Norwegian or Landsmål) and Bokmål (also called Dano-Norwegian or Riksmål). Additionally, ligatures and certain letters (like Å) can be spelt out using different vowels. Older documents are written in a Norwegian that is heavily based in Danish. This makes Norwegian spelling appear inconsistent. English does not tolerate variant spellings. Rather than change all these spellings to be internally consistent and thereby modify source document quotes, the tendency employed is to use the original spellings as they appear in the source documents. The kings in Scandinavia are known by various names in each country and often these names were nicknames and changed over time. There has been an attempt to employ the more commonly known name as recognizable to an English speaker. A cross-over list of the variant names is provided in the Notes on Translations. Also included is a summary of the methodology for choosing particular translations, as well as the linguistic rules imposed on source translations.
This work makes extensive and repeated use of a full range of written accounts. In order to