With the introduction of Christianity to Norway in the 10th and 11th centuries, churches of different sizes and forms were built. Some may have been made partly of stones, some wooden buildings had earth-bound posts, and some had their lower construction set on a frame. Even though the wooden churches had structural differences, they make a recognizable general impression. Formal differences may hide common features of their planning; while apparently similar buildings may turn out to have their structural elements organized differently. Certain basic principles must have been common to all types of building. Basic geometrical figures, numbers that were easy to work with, one or just a few length units and simple ratios, and perhaps proportions as well, were among the theoretical aids all builders inherited. The specialist was the man who knew a particular type of building so well that he could systematize its elements in a slightly different way from what was the case in the buildings known hitherto, thus carrying development a stage further. My research has been concentrated on the planning and design of the early wooden churches, especially in Norway.

Summary of articles

Fra en omvisning i Gol stavkirke
(A Guided Tour in Gol Stave Church)
Middelalderforum, nr.2/ 1996: 15-23. Abstract in Norwegian. Full text in Norwegian.

This paper describes how the wooden church from Gol in Buskerud, bought by the king in 1884, was reconstructed in the Folk Museum in Oslo. In spite of all changes in the building through the centuries, the architect Waldemar Hansteen was able to reconstruct most parts. Where original parts were missing, he copied forms and sizes from the Borgund stave church. The paper discusses some of the problems arising when moving an old church away from its original setting into a museum.

Innvielsesrituale for trekirke
(Ordo for dedication of a wooden church)
Hikuin, vol. 24, 1997: 83-98. Abstract in Norwegian. Full text in Norwegian.

In the literature of medieval wooden churches it has been suggested that it was contrary to Canon Law to raise churches entirely of wood,  because the Roman Church demanded stone as the main material. If it has been an established practice to build in stone, there seems to be no documentation showing that such a rule was ever given as an order. An Irish Ordo from before 900 CE for the dedication of a wooden church is translated. The Ordo shows that the same ritual was performed for churches of wood and stone.

Var det krav om høye stavkirker i middelalderen?
(Was it a requirement to build lofty stone churches in the Middle Ages?)
Viking, vol. LX, 1997: 85-94. Abstract in Norwegian. Full text in Norwegian.

The paper discusses three interrelated questions pertaining to the building of churches in the Middle Ages. Firstly, were there ever any ecclesiastic regulations stipulating that churches should be built of stone? The answer given in this paper is no; there is nothing in the liturgy or Canonic Law that states this as a requirement. There were, however, traditions for modelling churches on Solomon’s Temple, and on Mediterranean building practices in general. The second question raised is whether churches had to be tall based on liturgical demands. The answer suggested here is also negative, as the liturgy was primarily focused on the celebration of the Eucharist at the altar. The extreme heights to which many cathedrals were built were condemned as an expression of vanity and greed by members of the Church who adhered to a purer and more apostolic view of Christianity. Thirdly, the more recent allegory of the nave as a lofty forest is dismissed as having no basis in medieval Christian thought.

Røldal stavkirke eller…?
(Røldal – Stave Church or what…?)
Viking, vol. LXI, 1998: 131-145. Abstract in Norwegian. Full text in Norwegian.

In this article the stave church of Røldal is discussed. In form and construction the church resembles other stave churches in Norway, but its details and overall construction have made it difficult to place typologically; it resembles the Finneloft building at Voss. The church is usually dated to the 14th century. The construction of the lower part of the church is discussed, and the author hypothetises that the church originally had corner posts dug into the ground. If so, at least two of these wooden posts survived in their original state until 1913. This would contradict the assumption that wooden posts set in the ground lose their capacity to bear the superstructure after a short period of time due to rot. The author concludes that the building represents a profane type of construction that was in use before, during and after the time of other stave churches.

Research in Medieval, Norwegian wooden Churches, relevance of Available Sources
Nordic Journal of Architectural Research, vol. 13, nr. 4, 2000: 7-23. Abstract in Norwegian. Full text in English.

This paper points out that there are several sources available to answer questions posed to a medieval building, and that the sources have different potentials. The paper seeks to identify the sources of interest for the study of medieval church planning and design. Of the sources available three main groups may be described, normative, descriptive and attributive texts, drawings and models. For the history of planning, design and building, however, the archaeological evidence and building remains are direct, original sources in the absence of the building proper. Finally, the standing buildings represent the richest source for the architect/building historian. If physical objects and texts are examined according to their innate characteristics then, the combination may broaden the knowledge of planning and design of the early wooden churches of Norway.

Hvorfor ble kirkegårder velsignet?
(Why were churchyards blessed?)
Hikuin, vol. 27, 2000: 57-66. Abstract in Norwegian. Full text in Norwegian.

The question asked in this article is why churches were dedicated and churchyards were solemnly blessed in the Middle Ages. Five possible reasons are proposed. Firstly because a possible pagan place of worship had to be purified by exorcism and blessing before it was suitable for Christian worship. Secondly, because a priest was not allowed to celebrate mass in a church before the altar was dedicated to God in the name of a Saint. By this action the whole place was reserved for Christian worship perpetually. Thirdly, because the dedication ceremony was meant to create awe and respect for the holy place so fugitives seeking sanctuary should be left unharmed. Fourthly, because the public and circumstantial rite was meant to make visible the transfer of property, and the new property boundaries. Finally, because the ritual should inculcate into the public respect for the property of the Church. Dedication of churches and solemnly blessing of churchyards were therefore both means and aim. The aim was to promote respect for the site and the people residing there. The means was to create awe by connecting it to holiness. In a predominantly oral culture this had to be done by actions and statements.

“Stavkirkeprekenen” som bygningshistorisk kilde
(The Stave-church Sermon as a source for building history)
Fortidsminneforeningen, Årbok 2001: 91-96. Abstract in Norwegian. Full text in Norwegian.

This paper discusses a 12th-century Icelandic/Norwegian homily for the day of dedication of a church. It is known as The Stave-church Sermonbecause the text gives a multilevel allegorical explanation of different parts and elements of a wooden church building. While the rite of dedication of a church was normative, the homily belongs to a different tradition. The impression is that the description of the various parts of a wooden church is not documentation of a specific edifice, perhaps not even a type of church. The church described cannot be placed topographically, nor can it be dated or placed typologically. Building elements known to a 12th century congregation are mentioned in order to illustrate the Scripture. The importance of the homily for the building historian is its use of terminology connected to parts of wooden buildings.

Trekirkene før stavkirkene. En undersøkelse av planlegging og design av kirker før ca. år 1100
Avhandling for graden Arkitekthøgskolen i Oslo.
(Wooden churches before the stave churches. An investigation into the planning and design of churches prior to c. 1100 CE. Dissertation for the Degree of Oslo School of Architecture.)
Abstract in English. Abstract in Norwegian.

From Vitruvius to Alberti: Systematic Thinking in medieval church design.
Paper read at the conference: (Theorising) History in Architecture and Design,  at Arkitekthøgskolen i Oslo, 25th-27th April 2003. Sammendrag på norsk. Full text in English.

Some 1500 years after Vitruvius the Italian architect Leon Battista Alberti wrote On The Art Of Building In Ten Books. Medieval scholars educated in literary traditions describe churches outwardly and superficially. The builders tested different hypotheses for solutions in design and construction without writing down and elaborating them methodologically. Therefore it could not be called a theory proper but rather an architectural sub-theory. The paper concludes that the builders’ praxis was rejected by scholarly discourse, but that craftsmen with their sub-theory did not require written discourses to initiate, fund, plan, design, construct and develop the buildings that were required.

The “Inverse Design problem” In Medieval Wooden Churches of Norway.
At the 6th Asian Design International Conference in Tsukuba, Japan, 14-17 October 2003. Sammendrag på norsk. Full text in English.

This paper discusses the “inverse design problem”: how to reconstruct the original design principles in medieval wooden churches of Norway. In Northern Europe, many of the early medieval churches were of wood and the principal load-bearing elements consisted of earth-bound posts. Norse traders, raiders and craftspeople became acquainted with Christianity in the 8th and 9th centuries, if not earlier. So, before the arrival of the first missionaries to Scandinavia, Norse craftsmen may have known how to plan, design and construct churches. In the early 12th century, the most common way of building churches in Norway was with the posts/staves on low foundations of stone. The medieval craftsman had his knowledge of form transferred mnemonically by words and actions from other carpenters; no design manual seems to have survived. The original designers may have considered many design problems, only one of which was established in plan and elevation. Attention to measurements and ratios was obviously essential to ensuring an easy and sound construction. The modern building archaeologist, however, has to solve the “inverse design problem”: how to extrapolate the lengths and ratios from the existing solution back to several possible options. The options are contingent on different materials, static necessities, economy, Biblical symbolism, sub-Euclidean geometry and vernacular building traditions. The paper proposes a way to circumvent the inverse design problem in a research situation, exemplified in the medieval wooden churches of Norway. Even if this problem may not have a solution in a strict sense, the probability of one option versus another has to be discussed and evaluated before one elaborates a hypothetical design system.

Grunnforskning på grunnstokker. For en nysgjerrighetsdrevet kirkeforskning
Fortidsminneforeningen, Årbok 2004, vol. 158: 137-140 Abstract in Norwegian. Full text in Norwegian.

The Department for the Environment through its Directorate for Cultural Heritage is responsible for the heritage management of churches, but there appears to be no authority with responsibility for the long-term development of competence and professional research on their structural archaeology. Building up and maintaining the research environment in this field should be a public undertaking, but neither architect training nor any university institute offers professional courses in building archaeology with regard to medieval churches. What can be done about this?

Preken fra en takstol, bokanmeldelse av: Ola Storsletten: Takene taler. Norske takstoler 1100-1350, klassifisering og opprinnelse. Del I, 444 s., del II, 314 s. CON.TEXT, avhandling 10, Arkitekthøgskolen i Oslo 2002.
Fortidsminneforeningen, Årbok 2004, vol. 158: 189-192 Abstract in Norwegian. Full text in Norwegian.

Ola Storsletten, civil architect,, and research officer at NIKU, has published a work on the known medieval roof structures in Norway. On the basis of an extensive systematization of the documentation he considers the origins of tie-beams in stone-built churches and stave churches. The reviewer is critical about the range and disposition of the material, the somewhat unclear presentation of the problems, and the subsequent vague conclusions. However, the many clear, reliable and informative drawings and the comprehensive descriptions make the work a valuable foundation for further research on the roof structures of medieval churches in Norway.

De tidlige trekirkenes planlegging og design
(Planning and Design of the Early Wooden Churches. A Contribution to a Theory.)
Hikuin, vol. 30, (2003), 2005: 95-101. Abstract in Norwegian. Full text in Norwegian.

For more than 160 years, researchers have conducted investigations into the churches which existed before the stave churches in Norway, and they have pointed out possible prototypes both at home and abroad. However, theories relating to origins often deal only with external form. In my dissertation (Jensenius 2001) the conclusion reveals that church building was part of an experience-based tradition wherein knowledge about the buildings’ visible and invisible order (measurements) was transferred from prototypes by a set procedures.

Hvor ble det av proporsjonsbegrepet i arkitekturen?
(Where are the proportions in architecture gone?)
Nordisk Arkitekturforskning, vol. 19, No. 2. 2006: 81-93. Sammendrag på norsk. Full text in Norwegian (PDF).

The article discusses the notion of proportional systems and asks if they have relevance for today. “Proportion” is traced back in time to the history of arithmetic and geometry. The numbers and ratios, fulfilling practical needs for reckoning quantity in planning and design, were idealized as “perfect” numbers and ratios by the theorists. Very seldom, craftspeople wrote down their constructive mathematics, based on their experiences. The axiomatic theoretical mathematics, often intermixed with speculative, cosmological variants, belong to a tradition calling itself universal, abstract and general. The paper concludes that systems of proportions may be used, with or without adhering to a neoplatonic worldview.

Lorentz Dietrichsons ”De norske stavkirker”: en klassiker.
Viking, bind LXIX – 2006: 275-288. Sammendrag på norsk. Full text in Norwegian (PDF).

This article discusses some of the underlying premises in Prof. Lorentz Dietrichson’s great classic “De Norske Stavkirker” (The Norwegian Stave Churches) from 1892. In this book, he made an invaluable contribution to the research on these buildings, as he collected and systematized the large material, seen as an important part of Norwegian cultural history. His investigation made it clear that stave churches were also in fact an independent section of the European architectural heritage. However, the author points out that Dietrichson had few models for his investigations, and that contemporary scholars abroad hardly knew about these buildings. Dietrichson’s conclusions therefore were both subjective and controversial. But his rhetorical power, his wide reading, and his extensive travels made his book difficult to discuss or contradict on a specialist basis. Everyone who wished to study these churches has always had to relate to Dietrichson’s presentations, and his interpretations of the buildings. This has led to a certain tired stagnation within the research activity. The article points out an alternative interpretation of the churches, which may lead to new conceptions.

Staver som brytes og kirker som faller Vedlikehold og rivning av kirker i middelalderen
(Broken columns and falling churches. Maintenance and demolition of churches in Medieval times.)
Kirke og Kultur, nr. 5/2007: 418-429. Norsk sammendrag. Full text in Norwegian (PDF).

In recent years there has been much discussion about how to maintain churches from the last century, when they are no longer in demand or are in a poor condition. Sale or alteration of the pattern of utilization of the building or change in purpose for which the building is used, has been proposed. Some have reacted to this, since it seems too dramatic just to tear a holy church down. The paper, with a design historical approach, asks how the Church in the Middle Ages practiced maintenance and demolition of consecrated buildings in Norway.

Designing wooden churches in Viking and Medieval Norway
(Prosjektering av trekirker i norsk vikingtid og middelalder.)
Viking, vol. 73, 2010: 157-172. Sammendrag på norsk. Full text in Norwegian (PDF).

This paper is a contribution in archaeology of building for design of wooden churches in Norway in the Middle Ages. The varied design of the known wooden churches is commonly recognized, yet the reasons for this eclecticism have received little study in scholarly literature in the last century. The dearth of written reflections on design heightens the importance of the buildings themselves as primary source “documents”. Is it possible to reconstruct the original process from idea to form, although it is not forwarded in writing? It has been assumed that craftsmen forwarded an extensive but unsystematic practice. This knowledge consisted of practical methods and concrete rules for planning, design and assembling, rules which led the builder from idea to completed building. In this paper the background for different rules for practice are discussed. The rules were transferred orally through craftsmen. These rules were parts of a sub-theory, the condition for the development of statics and different theories of architecture.

Structural and archaeological observations on wooden churches in Viking and Medieval Norway
(Bygningstekniske og arkeologiske bemerkninger om trekirker i Norge i vikingtid og middelalder) 
Collegium Medievale, vol. 23, 2010: 149-180. Sammendrag på norsk. Full text in Norwegian.

The main objective of this research is to identify the architectural consequences and possibilities of different ways of foundations. As a premise is shown a graph from an article written by the German researcher Haio Zimmermann, a graph I have retraced as a mean of clarity (Fig. 1). Here is shown a rough proportional relationship in percentage between post-frame buildings and timber-frame buildings, used in Northern and Central Europe. The important point is the arch of the curve around the year 1000, marked here with an arrow. This bending of the curve coincides with the introduction of Christianity to Norway and the subsequent church building projects. Zimmermann has no simple explanation on the fact that timber-frame buildings gradually were preferred to the post-framed. After the site had been levelled and the top soil had been removed, there were dug a series of individual holes down to the dense subsoil around the oriented perimeter of the design. For some buildings, the holes are just wider than the posts, like in this illustration, in other churches they may be 1,0- 1,5 meter in diameter. In the bottom of the more elaborate holes it may have been spread a layer of sand as drainage and levelling. On this, it may have been placed a stone footing, for the distribution of weight mainly (Fig.2). Around the lower end of the posts there were some times placed charcoal, or the charcoal found may be from the charring of the end of the post itself. It is a question if this way of treatment really could prevent decay. The end of the post was fixed by a bracing of stones to keep the post plumb after the erection, then packed tight with stones and tamped earth. An embedment depth of 0.3 to 0.9 m is far from being adequate to take the required lateral thrusts, so the stability of the edifice is not more than partly dependent on this fixing of the posts. The diaphragm effects of wall planks, however, would stiffen building connections, in addition to knee braces. The posts were placed in the deepest part of the hole and aligned two ways with the rest of the posts in the wall. All posts should be perfectly plumb and be placed in a straight line along a wall; all building corners should be perfectly square. A 4 to 5 m post may have had a weight of about 280 kg. To get all the posts aligned and with a regular on-centre spacing seems to have been of some difficulty, and the churches may also be out-of-square (Fig.3). It may be guessed that when this church was in use, the floors were slanting out of level and the walls were leaning out of vertical. Soundness is the least culturally encoded of the principles of design. The builders had to follow the law of gravity, the limitation of the materials and the demand of the climate, and design the buildings accordingly.

Development of controlled site drainage for water runoff must have been an important part of the site design; moist soils were a major threat to wooden constructions. The timber would absorb moisture and so produce fungi which decomposed the wood (Fig.4). So, as the illustration indicates, the rot in the lower part of any wooden wall may have several causes. How much of which type of humidity, or when the materials decayed may be difficult to describe. There is no doubt, however, that many builders still found the post-frame buildings suitable for their purpose. New post-frame buildings replaced older ones in Urnes, Kaupanger, Mære and Faret. Other churches were replaced by stone churches or timber-frame buildings.  The building of a new church therefore gives the date of the dismantling of the post-frame building. One may guess that new post-frame churches were erected as long as the advantages with the post-frame construction were greater than the drawbacks. This illustration shows two solutions of corner joints in the timber-frame churches (Fig.5). The connections between stave and sill in the uppermost drawing may resemble those in some of the post-frame buildings. When the plans of the excavated churches are examined, one thing becomes clear: all post-frame churches were succeeded by a new building with larger nave and chancel. In my opinion therefore, the post-frame buildings were dismantled, when and because the naves and chancels had become too narrow for practical use. They were then replaced, regardless of their over all condition.

There was no ecclesiastical demand that churches had to be tall from a liturgical point of view, as the liturgy was primarily focused on the celebration of the Eucharist at the altar. The will to build lofty must have been a cultural and a social one. Structural safety must have presented a special concern, and nothing indicates that the construction had to be of the greatest possible height. In any case, to plan for the structural integrity of a high-rise, one had to select an alternative foundation to the post-frame, for the building to be both horizontally and vertically located.

To stabilize the inner posts in the nave, therefore, one had to fix them uniformly, and this was done in the raft beams. The foundation consists primarily of four structural quoins, that is boulders set in the ground. To carry the weight of the superstructure, the upper faces of the four stones had to be exactly level, they then acted as a datum line. On the four boulders were placed the joints of the four raft beams of the nave, made to form a rectangle (Fig.6). The raft beams had three functions. Firstly, they diminished the risk of subsidence of the ground. Secondly, they kept the finished floor of the building higher than the surrounding site, thus reducing the possibility of a flooded building during heavy rainfall or snow melting. Thirdly, they supported the edifice structurally by keeping the floor level. The height of the superstructure was the outcome of the planning process, the accuracy employed made the height possible. The principal posts falled in the joints of the raft beams. Gravity was its greatest asset. The lower post to beam connection was not subjected to any critical forces, but its contact pressure to the raft beams should be as uniform as possible. A rounded or squared tenon in the end of the post was mortised into the full depth of the beams, which was sufficient to resist any lateral movement.

To sum up in three points. First, the greater number of the churches of Norway before the 12th century were probably post-frame buildings. In terms of reasonable durability, performance and cost, the low-rise post-frame church building may have been seen as a time-honored viable church concept. A weak point may have been the trouble of transferring the design to the site, with deviation from measured values as a result. Second, the known post-frame churches were, in all likelihood, disassembled when and because the naves and chancels had become too narrow for practical use. This may have been hastened if parts of the buildings needed an overhaul. However, no documentation shows buildings to be derelict or abandoned when replaced. The replacement of the post-frame church buildings seems to have been performed very slowly, over a period of maybe hundred years. Also, the act of improving, expanding, enlarging or refining the post-frame and the timber-frame churches must have been parallel developments through the centuries. Features held in common may have developed independently. Third, a building with an extended area combined with a heightened superstructure, would make great demands on a foundation. If the superstructure was changed, the foundation would also have to be changed. Or, said in another way, the accurately levelled and squared floors and plumb staves made the new height possible. The constructional safety generated by this accuracy was the precondition for the building of high-rise timber framed churches in Norway.

Form and use of the stave church door: Hof kirke in Solør as example
(Stavkirkedørenes form og bruk: Hof kirke i Solør som eksempel)
Viking, vol. LXXIV, 2011: 225-240. Abstract in Norwegian. Full text in Norwegian.

After defining a practical route in and out of the building with serviceable elements, the carpenter’s task for constructing doors in stave churches was fivefold. First he had to fit planks for the door together. Second, he had to determine the direction of opening. Third, he would make a rigid suspension for the door in the jamb. Fourth, make hinges which would rotate for a long period without maintenance. Fifth, make the door so it could be locked to the jamb. The door is strongly constructed. Strap hinges in the end of bandings made of flat stock extend around the sides of the door. The hinges are hanging by a loop of each hinge on an iron pintle, an L-shaped hook set into the door jamb. Hinges, banding straps, locks, lock plates, forged nails and pulls and handles constitute the hardware used on doors. Wrought iron nails are driven in from the face side of the door and clinched over or bifurcated on the battens and braces. One would think that accessibility is the single most vexing topic among makers of doors. However, most of the wooden doors in the stave churches are narrow and high, often in a ratio of 1:4. For opening a door one has to apply a larger force to make the door swing as one gets closer to the hinge side of the door. When pushing as far from the hinges as possible, one can apply a smaller force, albeit for a longer period of time. Therefore, a low and wide door is the easiest to pull open. The reason for the preference of high and narrow doors may be the limited area in the aisle inside the church.