In the eye of the beholder: key architectural elements in 25 years of visual analysis of Danish megalithic tombs
Abstract
A number of construction features in passage graves appear to be repeated at multiple sites in Denmark. These constructional choices are likely to be reflections of certain functional and/or symbolic demands, and are thus considered architectural. Passage graves should be experienced as three-dimensional spaces and their physical impact should be felt on the body. The passage itself has several remarkable features: the antechamber, thresholds and doorframes, and an intriguing use of white burnt flint. A principle of duality may be reflected both in double chambered and twin chambered passage graves as well as in the use of twin stones and the layout of the individual chambers. In some areas, passage grave chambers are extended upwards by means of a so-called intermediate layer, virtually forming a corbelled roof. The aspiration to create chambers much taller than actually needed for the burials as such can be seen as the pinnacle of megalithic architecture in this region.
Keywords: passage grave, keystone, capstone, twin stone, crushed white flint, dualism
Introduction
A number of constructional features in passage graves appear to be anything but coincidental or opportunistic. On the contrary, these features are repeated at multiple sites, some throughout Denmark and others only within certain regions.1 The interpretation offered in this paper is that the constructional choices made by the builders are likely to reflect certain functional and/or symbolic demands – not the constraints imposed by the unworked blocks.
No matter how hard we try, we can never extract the same amount and quality of meaning from the megaliths as did the Neolithic societies who built and used them. When investigating and analysing data from passage graves, it is all too easy to understand and interpret observations within our own experiences from modern day life. In fact, this is all we can do, no matter what sophisticated theories on Neolithic society and religion we filter that data through. I will do my best to avoid that and offer only a set of categorised observations open to comparison with the results and interpretations of other researchers.
Megaliths are structures to be experienced in three dimensions – not by studying plans and drawings. One has to crawl into them, sit there for a long time, contemplate what one is seeing, and discuss it with colleagues. Having done that myself with two of my own colleagues, Svend Illum Hansen and Torben Dehn (Dehn et al. 2008, 274–280), many times throughout the last two decades, I am glad to have the opportunity to present some of our observations.
Architecture of the passage and the antechamber
In the Danish language, passage graves are still termed “jættestuer”, literally meaning “giants’ halls”. Despite the fact that, as early as 1744, an excavator concluded that the passage graves were indeed erected by normal human beings (Kaul 2010, 39), the Danish language has not adopted a term similar to the ones used in the neighbouring countries: “passage grave”, “gånggrift”, “ganggrab” etc. Although this erroneous and outdated term may appear charming, it draws less attention to the passage than the more accurate terms mentioned before. This is regrettable, because most visitors are already prone to crawl hastily through the passage in order to get into the chamber as quickly as possible. This is human nature: the uncomfortably low and narrow passage will, at best, cause a slight physical discomfort and may even trigger a feeling of claustrophobia. Furthermore, the well-prepared visitor anticipates that the chamber will reward him or her with a more agreeable ceiling height and a more ample space to move around. However crude, this feature of the passage is indeed an example of a very basic property of architecture: the ability of a structure to impose certain feelings on a visitor. The builders could quite easily have made the passage higher, wider, or shorter, but obviously chose not to do so.
Fig. 9.1: Bogø passage grave (listed monument number 4227:13) showing the position of thresholds within the passage (Hansen 1993, 105) A single or double set of thresholds and doorframes are still preserved in most Danish passage graves
Fig. 9.2: Hjortegårdene passage grave (listed monument number 2826:4): elevation of the passage side wall (Dehn et al. 2000, 224). A uniform feature of most Danish passage graves is that the passage is lower and narrower at the entrance but becomes wider and higher towards the chamber
Having recognised this architectural principle of the passage, one begins to notice other typical features. Most passages have one, two, or in rare cases three sets of doorframes and thresholds, which are relatively easy to identify (Fig. 9.1). Furthermore, the typical passage is not evenly low and narrow. Usually it is lowest at the entrance, with a slight but significant increase of the ceiling height towards the chamber (Fig. 9.2). Subtler is a feature of the innermost one or two metres of the passage, immediately before the access to the chamber proper. Here, the walls usually bulge a little just before the passage meets the two cornerstones (the pair of orthostats that, together with the so-called keystone, form the entry to the chamber). This widening of the passage immediately before the transition to the chamber clearly constitutes an antechamber (Fig. 9.3). The effect may be enhanced by a slight angling of the two innermost pairs of orthostats and by placing the cornerstones close to each other. Very often, cornerstones are selected and placed with flat surfaces carefully angled in order to accentuate the relative width of the antechamber (Fig. 9.4).
Fig. 9.3: Græse passage grave (listed monument number 2927:2) (Hansen 1993, 119): the slightly wider space between the corner stones and the innermost pair of orthostats in the passage constitutes an antechamber in most if not all passage graves
Fig. 9.4: Antechamber of the Tjæreby passage grave (listed monument number 3621:14). (Photo: Jørgen Westphal)
One might speculate over the purpose of this architectural element. Did it have a role in the ritual use of the passage graves? Did people, material objects, or bodies of the dead pause at this place in their movement in or out of the chamber? Or was it intended to divert the visitor’s experience from the otherwise monotonous feeling of constraint when moving through the passage?
The passage-chamber transition
Harnessing the forces of gravity that work on the several hundreds or thousands of tons of mass in a building is a challenge facing any constructor, thus also the builders of the passage graves. A modern constructional engineer would apply the science of statics. Statics is the branch of mechanics that deals with the analysis of loads on physical systems in static equilibrium. An experienced engineer can calculate the necessary dimensions of the building components, making the building able to withstand any projected impact without excessive use of materials. The builders of the passage grave probably did not have such deep mathematical insight but had to rely on their practical experience, just as had builders of Medieval and Renaissance churches and castles. But how close did the Neolithic builders dare to go to the maximum? The materials they chose, various natural stones, are known to have very good compression strength: there is hardly any limit as to how many pieces of stone one can stack upon each other. But stone has rather poor tensile strength and it breaks easily when pulled. Moreover, a pull is exactly the force that occurs in the lower part of a capstone that, in technical terms, acts as a horizontal load-bearing beam (Fig. 9.5), both under its own weight and certainly when further materials are placed on top of it. The capstones of passage graves always seem to be of appropriate thickness. Yet there is another constructional element where stones used as horizontal load-bearing beams sometimes fail. This is at the point where the passage joins the chamber. Here two orthostats of the chamber, the cornerstones, are placed at a sufficient distance to allow access from the passage to the chamber. Usually this access is relatively wide – 70cm is typical – but sometimes it so narrow that even a thin person has to squeeze through to get into the chamber. However, no matter what, there has to be a stone bridging the two cornerstones – the so-called keystone (Fig. 9.6).
The keystone has multiple functions. As mentioned above, it straddles the gap between the two cornerstones, thus forming a roof for the access to the chamber, but also ensures that the two cornerstones are locked in their respective positions and are not pushed inwards towards each other, blocking the access. The keystone is also the support for the immense weight of the overlying capstones. It appears that particularly robust types of stone were selected for this purpose. Given that there is not always a great difference between the height of the antechamber and the height of the chamber, there is often a limit to how thick the keystone can be. As a result, broken keystones can be seen in some passage graves. It is not known whether they broke during the construction process or at a much later stage. A prehistoric repair of such damage would have been challenging: given that the keystone is locked under the weight of the capstones, it would be difficult to remove and replace. During the recent years of megalith restoration in Denmark, no prehistoric repair of a broken keystone has been observed.
Fig. 9.5: The forces of compression and pull acting on a keystone serving as a load-bearing beam. (Diagram: Jørgen Westphal)
Fig. 9.6: Straight keystone in the Troldstuerne passage grave (listed monument number 2823:5). (Photo: Jørgen Westphal)
The problem of a keystone that is too long and too thin can be overcome by making it shorter and placing less weight on it. Instead of having a relatively long keystone bridge the gap between the highest points of the cornerstones, one can let a much shorter keystone rest on the shoulders of the cornerstones. By doing so, the cornerstones are still kept from being pressed towards each other; the keystone still forms a roof covering the access to the chamber; and most of the load from the chamber capstones is conducted through the cornerstones instead of the keystone (Fig. 9.7). This solution is often seen in western Denmark, where such an arrangement has a pointed stone serving as both the innermost capstone of the passage and as a keystone, with the pointed end mounted between the shoulders of the cornerstones (Fig. 9.8). This method is very efficient and suitable when the capstones of the chamber are intended to lie more or less directly on top of the orthostats (Fig. 9.9). In such a case, a keystone lying on the highest points of the cornerstones would be in the way.
Fig. 9.7: The principle of a pointed keystone. (Diagram: Jørgen Westphal)
Fig. 9.8: Approximate distribution of pointed and straight keystones versus straight keystones in Denmark. (Map: Jørgen Westphal)
Fig. 9.9: Pointed keystone in the Grønhøj passage grave, listed monument number 2912:8. (Photo: Jørgen Westphal)
Double chambers, twin chambers and twin stones – duality as principle
Approximately 10% of the passage graves in Denmark have two chambers, isolated from each other and each with a separate passage (Fig. 9.10). One of the chambers might be thought to be a later attachment to the other, but in Denmark such rebuilding or extension of the mound surrounding the two chambers is not observed. Thus, there does not seem to be any considerable time span between the construction of the two chambers. It is, however, risky to draw conclusions from negative data, and since other indications of rebuilding or extension are sometimes observed, the building history of double-chambered passage graves may not yet be fully understood. Denmark has a subset of double-chambered passage graves, however, where the structures of the two chambers are integrated in a manner that practically excludes an asynchronous building process: the so-called twin passage graves. In a twin passage grave, the two chambers actually form a single structure, only separated by one or two shared orthostats (Fig. 9.11). This variant, to our knowledge, is not found elsewhere in Europe.
Fig. 9.10: Ground plan of a typical double passage grave. (Diagram: Jørgen Westphal)
Fig. 9.11: Examples of twin passage graves: the shared orthostats are indicated in red (Diagram: Jørgen Westphal)
Fig. 9.12: Twin stones at the end of the Kong Svends Høj passage grave (listed monument number 4322:1). (Photo: Jørgen Westphal)
Fig. 9.13: Two halves of a large split stone found in the mound behind the chamber of the Måneshøj passage grave (listed monument number 4025:1). Combined weight 1.5 tons. (Photo: Jørgen Westphal)
In about 50% of all passage graves and great dolmens in Denmark, one or more pairs of so-called twin stones can be seen. ‘Twin stones’ is a Danish term for two halves of a single stone, incorporated in a megalithic tomb with the split surfaces facing inwards. The two halves might be orthostats standing next to each other or opposite each other, always with some kind of recognition of their relationship (Fig. 9.12), or they might, in rarer cases, be used as capstones. In Denmark, no clear tool marks have been found to indicate that the stones have been intentionally split by human action. Generally, they are assumed to have been split by natural forces, perhaps through direct pressure imposed by an Ice Age glacier or from the effects of frost action. In one case, the Månehøj twin passage grave, two halves of a large split stone were found deeply buried in the mound behind the chamber (Fig. 9.13). Between the two halves was abundant charcoal and ash, indicating that fire may have been used to split the stone. Whether this stone was originally intended to be used as a pair of twin stones in the structure of the tomb is of course difficult to determine.
Fig. 9.14: Capstone arrangement in Danish passage graves. Where more than one capstone is present, the capstones are always placed more or less parallel to the passage and transversely to the horizontal axis of the chamber. regardless of chamber length. (Diagram: Jørgen Westphal)
Fig. 9.15: Structural and architectural emphasis on the left-hand chamber in the twin passage grave Rævebakken (listed monument number 3017:37), where the dividing orthostats deliberately lean inwards to this chamber. (Photo: Jørgen Westphal)
Fig. 9.16: The principle of an intermediate layer between the orthostats and the capstones. (Diagram: Jørgen Westphal)
In fact, the general layout of a TRB passage grave in southern Scandinavia is based on a dualistic principle. The passage is attached at some point on the long side of the chamber: thus, when entering the chamber from the passage, there is always a right-hand side and a left-hand side. Capstones are always placed more or less perpendicular to the long axis of the chamber, no matter its length (Fig. 9.14). Symmetry, however, which might be seen as a higher order of duality, does not appear to be paramount. Very few passage graves have strictly symmetrical layouts, and more often several architectural elements point towards a deliberate deviation from symmetry. Most chambers have some kind of emphasis of the left side of the chamber: it is often either larger, taller, or built of more suitable or regular stones. This preference for the left side is usually also present in both double and twin passage graves. In one twin passage grave, Rævebakken on the island of Samsø, it is clear that the dividing orthostats between the two tombs were placed leaning inward towards the left-hand chamber with the same inclination as the other orthostats – giving the left-hand chamber a more regular shape than the right (Fig. 9.15).
The pinnacle of chamber architecture – intermediate layers and corbelled roofs
Almost all stones in Denmark are a result of glacial deposits, thus leaving the Neolithic constructors to work with whatever material past geomorphological processes brought to the area. As a general rule, few boulders have width-length ratios more than 1:2 and almost none approach ratios in the region of 1:3. Ratios between 1:1 and 1:1.5 are much more common. This phenomenon leads to a dearth of suitable building materials, which is further exacerbated by the fact that few stones also have a flat surface that can be placed inwards. Having reserved the longest stones for capstones, builders would only have a selection of relatively short stones for orthostats. One way to work around this problem is, of course, to add an extra layer of orthostats, the so-called intermediate layer, in order to increase ceiling height (Fig. 9.16). Simple as it sounds, this is an extremely difficult and dangerous constructional principle. Nevertheless, the method was used in a large number of passage graves, predominantly in western and northern Zealand (Fig. 9.17). In addition to creating taller chambers, this method also gives them the appearance of corbelled vaults. Since the stones of the intermediate layer are much smaller than the underlying orthostats, an optical illusion is created that further increases the visitor’s experience of space. Usually there is only a single intermediate layer, but in one case, Rævehøj at Dalby, three intermediate layers are present. The length of the chamber is, of course, a matter of available resources and how much time and effort the builders wished to invest. The construction of intermediate layers is, however, a genuine display of technical skill. Anybody with sufficient resources could build a passage grave with a very long chamber, but only those with access to the best builders and their superior skills and knowledge could build a chamber such as that of Troldstuerne (Fig. 9.18).
Fig. 9.17: Approximate distribution of passage graves with intermediate layers in Denmark. (Map: Jørgen Westphal)
The island of Samsø stands at the periphery of this area of megalithic brilliance. Megalithic tombs on the western and northern part of the island may have looked towards the western part of Denmark: Jutland. The eastern part of Samsø, on the other hand, shows architectural evidence of contact with western Zealand. This part of Samsø has two twin passage graves. The first one, the Pillemark passage grave (regrettably in a very ruinous state), has an intermediate layer that looks very similar to that at Troldstuerne on Zealand (Fig. 9.19). The second one, the Rævebakken passage grave (Fig. 9.20), has a much lower standard of construction than Pillemark. Shorter and thinner orthostats were used, and the remains show that it had a relatively low intermediate layer. Perhaps Rævebakken was an inferior copy of Pillemark, made by the locals, after the skilled craftsmen who engineered Pillemark had returned home to Zealand.
Fig. 9.18: One of the chambers at the Troldstuerne twin passage grave (listed monument number 2823:5), illustrating the very distinct intermediate layer. (Photo: Jørgen Westphal)
Fig. 9.19: The ruinous Pillemark twin passage grave (listed monument number 2917:25), with an intermediate layer of a quality resembling that of the Troldstuerne passage grave. (Image: Jørgen Westphal)
Fig. 9.20: Restored twin passage grave of Rævebakken (listed monument number 3017:37). (Photo: Jørgen Westphal)
Fig. 9.21: Deposit of white burnt flint in the Strids Mølle passage grave (listed monument number 3724:18). (Photo: Jørgen Westphal)
Fig. 9.22: Typical position of a narrow band of white burnt flint in front of one of the thresholds within the passage. (Diagram: Jørgen Westphal)
White burnt flint
White burnt flint is a common sight in passage graves in Denmark. It usually appears in large quantities in the chambers, either more or less evenly distributed, or in small piles (Fig. 9.21). In fact, burnt flint is often the marker that offers hints to field-walking archaeologists that they have encountered the last remains of a long vanished megalith. It is not straightforward to determine whether the very first users of the passage grave deposited this flint, or if it derives from rituals performed much later. White burnt flint also occurs in contexts more securely tied to the construction of the tombs, namely as a part of the drainage and sealing systems behind the dry walling in many passage graves and great dolmens in eastern and northern Jutland. Therefore, we do at least know that this material was not unfamiliar to or undesired by the builders. In two cases, Jordhøj and Svebølle, the mound surfaces may have been partly or even completely covered with white burnt flint.
In the course of restoration projects during recent decades, white burnt flint has been observed in a new context: as a narrow band across the floor of the passage, usually immediately in front of one of the threshold stones (Fig. 9.22). When sectioned, these bands appear as shallow ditches, 1–5cm deep and 10–20cm wide, filled with finely crushed white burnt flint. The purpose of the brightly reflective white flint could have been as both a ritual and a practical marker of a boundary that should not be crossed. Moreover, when found in front of a threshold, it could also be the remains of material used for sealing a door stone tightly in place. A recent discovery supports this interpretation. At the Mogenstrup passage grave, two smaller stones that were tossed aside when the chamber was plundered in the 19th century have been identified as door stones. They fitted remarkably well when mounted in the opening between the corner stones that gave access from the passage to the chamber. Patches of clay still remained on the door stones and on the sides of the corner stones, as well as on the floor underneath. Within the clay on the floor there was a large amount of white burnt flint – much more than seen in flint bands elsewhere (Fig. 9.23). Whether the clay and the white burnt flint indeed are sealing materials and how they relate to the original use of the chamber is, of course, uncertain. Nevertheless, ceramics from the site point to a rather short history of usage with no material from the Single Grave period (2800–2350 BC) or the Late Neolithic (2350–1700 BC). The elaborate doorframe and threshold systems may indicate that chambers were actually meant to be closed tightly between rituals or burials, efficiently isolating the chamber and the antechamber from the outside world.
Fig. 9.23: White burnt flint in front of one of the possible door stones in the inner part of the passage at the Mogenstrup passage grave (listed monument number 2017:54). (Photo: Jørgen Westphal)
Conclusion
The overall layout of a passage grave is by no means coincidentally guided by the shapes of the unworked blocks. The passage itself has distinctive features, such as variations in height and width, an antechamber, thresholds, and doorframes. The solution to the technically challenging transition between the passage and the chamber shows regional differences, as does the solution to the desire for an increased ceiling height. This may point to a geographical core area of passage grave construction where the builders were particularly skilled in comparison to neighbouring areas. The construction of the chamber seems to revolve around a principle of duality that is reflected both in the overall design and in the frequent use of split stones. In some cases, this duality is even expressed as double chambered and twin chambered passage graves. The use of coloured materials such as white burnt flint complements the architectural expression of the passage grave structure itself. Besides the visual properties of this material, this flint along with clay may also have served the purpose of sealing doorstones in place to isolate the chamber and the antechamber from the outside world. These architectural features that have so far been noted may yet be joined by others, as researchers spend more time observing the passage graves from the inside, both as visitors and as participants in restoration projects. It is by this means, and not by studying only maps, drawings, and artefacts, that new ideas about the intentions of the Neolithic builders may be generated.
Note
1 Further information on monuments mentioned in the text can be found in the “finds & monuments” database at the Danish Agency for Culture: www.kulturarv.dk/fundogfortidsminder/
References
Dehn, T., Hansen, S. and Kaul, F. 2000. Klekkendehøj og Jordehøj, Restaureringer og undersøgelser 1985-1990. Stenaldergrave i Danmark 2. Nationalmuseet & Skov- og Naturstyrelsen: København.
Dehn, T., Engberg, N., Etting, V., Frandsen, J., Hansen, S. I., Vesth, K. B. and Westphal, J. 2008. Bevaring og restaurering af fortidsminder. Nationalmuseets Arbejdsmark 1807–2007.
Nationalmuseet: København Hansen, S. 1993. Jættestuer i Danmark, Konstruktion og restaurering. Skov- og Naturstyrelsen: København.
Kaul, F. 2010. Erik Pontoppidans og Ove Høegh-Guldbergs udgravninger af jættestuer – Jægerspris 1744 og 1776. Aarbøger for Nordisk Oldkyndighed og Historie 2007, pp. 37–56.