The excavations

The first known archaeological investigations of Raknehaugen were undertaken in 1869 and 1870 by antiquarian Anders Lorange, who attempted to reach its centre through trenches and a shaft sunk from its summit (Figure 2). This allowed him to reach the mound’s base, where the anticipated central burial failed to materialize (Lorange, Reference Lorange1870). The explorations nevertheless provided valuable insights into the mound’s construction, revealing one of its more unusual features: a series of timber layers that apparently corroborated a local myth describing the mound as the final resting place of a king buried in a stone chamber under layers of timber (Lorange, Reference Lorange1871: 478–80).

Figure 2.

Lorange’s (Reference Lorange1871) drawing of Raknehaugen.

Sigurd Grieg directed a second investigation between 1939 and 1940. This involved the excavation of trenches towards the mound’s centre, where a larger area was opened. No cairn or burial chamber was found, and the recovered artefacts consisted of broken and discarded objects unrelated to any grave goods. Furthermore, the few cremated bones recovered were too fragmented to establish whether they were human, forcing Grieg to question whether Raknehaugen constituted a burial at all. He eventually concluded that it was most likely a cenotaph built in honour of a high-ranking man (Grieg, Reference Grieg1941: 25–26).

While these excavations yielded little evidence of a burial, they confirmed Lorange’s earlier findings and offered a more detailed view of the mound’s interior. They revealed that the mound had been raised on top of a formerly cultivated field, in which features such as pits, stakeholes and ard marks were identified (Skre, Reference Skre1997: 15–17). The base of the mound structure itself consisted of a turf layer covered by alternating deposits of clay and sand that included a small burnt layer containing fragments of cremated bone. Above these lay a series of timber constructions, beginning with a pyramid of thin, unbarked pine and deciduous branches packed with moss and sandy clay (Grieg, Reference Grieg1941: 12–13). This was followed by yet another sequence of sand and clay layers, topped by a solid conical arrangement of heavier pine and deciduous logs stacked vertically around a central post and horizontally near the edges; budding branches on the logs indicated a spring or early summer deposition (Grieg, Reference Grieg1941: 11–12). Above this lay a metre-thick deposit of sand, covered by the uppermost timber layer. Comprising some 25,000 logs and branches, this was stacked into a tent-like structure, with split timbers and clay used to create a compact, near-impenetrable layer (Grieg, Reference Grieg1941: 8–11). The entire structure was then sealed by layers of sand, including a layer of humic sand several metres thick, and a final layer of topsoil.

Dating and reassessment

A much later reappraisal by Dagfinn Skre redated one of the timbers from the mound, narrowing the time it was erected to the mid-sixth century ad. Samples from its ten outer growth rings returned a radiocarbon date of 1508±23 bp, calibrated in IntCal20 to ad 539–636 (at 2σ). Skre also calculated its volume to be c. 26,200 m³, suggesting that 450–600 people had been involved in its construction, with 30–60 individuals providing the timbers (Skre, Reference Skre1998: 314–19). A further reassessment of the cremated bone fragments found in 1940 identified these as belonging to a human skull from an individual between 20 and 40 years old (Sellevold, Reference Sellevold1992), and thus the mound’s interpretation shifted yet again from cenotaph to burial mound, and its apparent similarity to other large mounds of the period was taken as evidence of conforming to a burial rite characterized by poorly furnished cremations capped by large mound constructions (Myhre, Reference Myhre1992). A subsequent radiocarbon analysis, however, dated the bone to 3020±30 bp (Gaut, Reference Gaut and Nyberg2016), calibrated to 1391–1130 bc at 2σ (IntCal20), showing that the bones probably represent prehistoric material redeposited within the mound construction, rather than a primary burial.

The curious timbers

Following the extensive excavations of 1939 and 1940, analyses of the timbers from the mound were undertaken to establish a relative dendrochronology and to gain an understanding of the local vegetation history. A first visual inspection of the timbers by forest historian Asbjørn Ording noted their rough qualities, consisting mainly of young, stubby pine trees with a pronounced taper. He further noted that the trees had been felled unusually high on the trunk—between 1 and 1.5 m from the ground—while some were found to have been broken rather than cut (Figure 3A). The timbers had then been left unbarked and crudely limbed, with long branch stumps protruding from the stems. Most of the timbers had been bucked into shorter logs, two to three metres long, and some had been riven, i.e. halved or quartered lengthways to create interlocking layers in the mound. In some cases, holes had been cut, and the ends cut to ease towing (Figure 3B), but otherwise the timbers showed no further signs of having been worked or reshaped. The layers also included loose branches, as well as the stump ends of trees with parts of roots still attached, leading Ording (Reference Ording1941: 93–95) to describe the assemblage as ‘unusually ugly’ (Figure 4).

Figure 3.

A) Timbers extracted for dendrochronological analyses. Note how some of the timbers have been broken rather than cut (photograph: S. Grieg, 1940, Museum of Cultural History, Oslo). B) Sketch by S. Grieg showing cut timbers with ‘eyes’, presumably fashioned to ease transport (archives of the Museum of Cultural History, Oslo). A) Reproduced under licence: CC BY-SA 4.0. B) Archival material in the Museum of Cultural History, Oslo.

Figure 4.

A) Timbers extracted from Raknehaugen during the 1939–1940 excavations. (photograph: S. Grieg, 1940, Museum of Cultural History, Oslo). B) One of the stump ends found in the mound. The rucksack on the left is c. 50 cm tall (photograph: E. Mork in Ording, Reference Ording1941: 106). A) Reproduced under licence: CC BY-SA 4.0. B) Reproduced by permission of the Norwegian Institute of Bioeconomy Research.

A more refined analysis of 100 pine timbers selected from the assemblage established that the sampled trees varied in age between sixteen and sixty years, with an average age of thirty-three years. Their diameters ranged from 20 to 30 cm, and calculations suggested their heights would have varied between c. 7 and 15 m. Modern local examples, in comparison, were considerably slimmer and taller than those in the mound. Ording further discovered that ninety-seven of the sampled pines had been felled in the same year, the remaining three having been felled two to five years earlier. Furthermore, the lack of significant cracking or rot indicated that the timbers had not been stored outside for an extended period, which in turn was taken as evidence for a relatively rapid construction phase, probably within two or three seasons of the trees being felled. These analyses allowed for a reconstruction of the pines used in erecting the mound, suggesting that these had grown in an open landscape with ample access to light and nutrients throughout their lives, similar to present-day pastures. Finally, Ording (Reference Ording1941: 127) established that the timber came from an area of c. 1 km², cleared either through logging or forest fire a few decades before the mound was built, and subsequently turned into pasture.

A second analysis undertaken by John Johnsen in 1943 broadened the study to include timbers from birch and alder as well as pine. These findings largely corroborated Ording’s studies, showing that most timbers had been felled at the same time, although Johnsen also found that two birch timbers may have been felled a year later than the main body of trees. Additionally, Johnsen (Reference Johnsen1943: 26–27) found that, although the timbers were generally well-preserved, several showed signs of rot, perhaps revealing a sample bias in Ording’s earlier analyses.

A much-emphasized aspect of Ording and Johnsen’s work is an anomalous growth ring observed in most of the samples analysed (Ording, Reference Ording1941: 111–12; Johnsen, Reference Johnsen1943: 28–30). This ring was formed fifteen years before the trees in Raknehaugen were felled and appears unusually narrow and faint, suggesting a sudden interruption in tree growth (Figure 5). Although the cause of the interruption was never identified in the original study, recent research has connected the phenomenon with the ‘Dust Veil’ event of ad 536, the geological event triggering a widespread climatic downturn around the time the mound was built (see e.g. Iversen, Reference Iversen, Iversen and Petterson2016: 45–46; Skre, Reference Skre and Nyberg2016: 140–41). This connection permits a more precise dating of the felling of the mound’s timbers to ad 551 (Skre, Reference Skre and Nyberg2016), but it also situates the mound in time and place, allowing for new interpretations and for the exploration of the relationship between Raknehaugen and its immediate landscape.

Figure 5.

A slice of one of the timbers investigated by Ording. The white tabs on the slice mark the fifteenth, almost invisible growth ring (photograph: H. Roll-Hansen in Ording (Reference Ording1941: 113). Reproduced by permission of the Norwegian Institute of Bioeconomy Research.

Static and fluid landscape forms

Raknehaugen is located centrally in a glacially formed landscape that can generally be divided into two separate landscape types (Figure 6). To its north lies a 130 km² gravel-rich, sandy plain formed by meltwater deposited during a hiatus in the glacial retreat around 9500 bp. A characteristic of such landscapes is that their soils are generally nutrient-poor and more suited to woodland than agriculture. It can thus be inferred that these plains provided abundant timber and firewood. Excavations have indeed demonstrated that the area was primarily used for outfield activities such as charcoal production and hunting in the Iron Age and Middle Ages (Helliksen, Reference Helliksen1997: 121–37) (Figure 7). In contrast, to the south and west of Raknehaugen, the geology is dominated by heavy marine clays deposited beneath the seawater that abutted the retreating glacier. These clays have subsequently been shaped by fluvial activity, with streams and rivers creating deep ravines, while the many fossilized fault scarps (i.e. the preserved traces of past landslides) in the area reveal a history of frequent landslides. Here, the clay-rich soils are highly fertile, laying the foundation for some of the largest farms of the region in the Middle Ages (Skre, Reference Skre1997: 17–19).

Figure 6.

Quaternary sediment map of the area surrounding Raknehaugen (contains data under the Norwegian Licence for Open Government Data made available by the Geological Survey of Norway.

Figure 7.

Extract from the map ‘Akershus amt nr 35-2’ from 1852 (© Kartverket, 2024).

Raknehaugen is thus strikingly located at the boundary between two distinct landscape types, and, while attributing emotional responses to these environments is inherently speculative, it is plausible that these landscape types evoked different experiences. Given the mound’s position, it is tempting to suggest that the contrasting landscape forms were recognized and imbued with significance, and that its location was deliberately chosen to reflect this.

The LiDAR evidence

Recent advances in LiDAR visualization techniques, effectively blending different models to enhance subtle landscape details, have proven particularly useful for seeing large but shallow features of the palaeolandscape (Kokalj & Somrak, Reference Kokalj and Somrak2019; Gustavsen, Reference Gustavsen2024). Its effectiveness is particularly evident in the LiDAR data from the Raknehaugen area, again revealing a clear division in the landscape: the terrain to its north-west is relatively even, interrupted only by a few glacially formed circular lakes, whereas the terrain to its south and west is marked by deep ravines and river channels cutting into the underlying marine clays (Figure 8). Further, the visualization technique applied to the LiDAR data provides details of the terrain that may otherwise be overlooked.

Figure 8.

LiDAR images from the areas surrounding Raknehaugen (centre). A) Visualization of the LiDAR data by overlying a simple local relief model with a narrow histogram stretch onto a hillshade model. B) Interpretation map of the LiDAR data showing the fault scarp immediately SW of Raknehaugen (background data: © Kartverket, 2025). Reproduced under licence CC BY 4.0.

The southern part of the dataset reveals a distinct, curvilinear ridge immediately south and west of the mound, some 3800 m long, up to 20 m wide but only about 40 cm high, encircling an irregular depression extending some 1000 m south-west of the mound, where its terminus is obscured by later ravine activity. Along its course, the terrain drops—first sharply by about 5 m near the mound, then gradually by another 18 m near its terminus. Overall, this depression describes a near-rectangular scar in the terrain, covering approximately 1 km², aligning with an undated landslide previously mapped by the Geological Survey of Norway (Løken et al., Reference Løken, Jørstad and Heiberg1970). This leads to the hypothesis that the mound itself was built on the fault scarp left by one of the most significant landslide events recorded in the region, a possibility that invites a reassessment of Raknehaugen in light of the archaeological record from its excavations.

Mound construction

As detailed above, Raknehaugen constitutes a highly complex and unusual structure, with soils consisting largely of clays and sand of different hues and thickness, interspersed by three layers of roughly worked timbers and branches. Previous interpretations of the different mound materials tended to see these in terms of rational practicality, either as introduced to facilitate transport, to hold other materials in place, or to stabilize the mound. This line of reasoning echoes Ingold’s ‘building perspective’ (Ingold, Reference Ingold2000: 213–33), where form takes precedence over process, and where the building of the mound follows a preconceived plan with the final form being the purpose.

More recent research, however, argues that the complexity observed in mound architecture is more a reflection of the social and symbolic dimensions of their construction than merely the result of practical or engineering considerations. Accordingly, the sourcing and collecting of materials for a mound, as well as the act of erecting it, served as a social unifying and identifying activity, while the materials used are likely to have had a symbolic, ritual, and cosmological significance (Gansum & Oestigaard, Reference Gansum and Oestigaard2004: 69–72; Leary & Field, Reference Leary, Field, Leary, Field and Campbell2013: 213–15). Building on these arguments, Cannell (Reference Cannell2021) found that, in several instances, materials were purposefully sourced and selected from the areas surrounding the mounds, and efforts were made to keep them separate from one another. By referencing and incorporating elements of the local environment, the very materiality of the mounds, as well as their settings, emerges as significant to the mound builders, and it suggests that they sought to establish a tangible connection between the monument, the people, and the surrounding landscape (Cannell, Reference Cannell2021).

Turning to Raknehaugen, the soils in Raknehaugen were likely to have been sourced fairly locally (Skre, Reference Skre1997: 25), although their exact origins have yet to be established. Furthermore, as seen in the documentation from the 1939–1940 excavations, the general homogeneity of the layers and the sharp interfaces between them suggest that the mound builders strived to keep the soils separate, depositing these in distinct phases as work progressed, perhaps pointing to the soils being imbued with symbolic or ritual significance (Figure 9).

Figure 9.

View of the 1939–1940 excavations, showing the western part of the main trench where the second timber layer has been exposed. Note the distinct layers in the background section (photograph: G. Sand, 1940). Reproduced under licence CC BY-SA 4.0.

The timber layers consisted of well-preserved pine logs, but also included deciduous species such as birch, alder, and rowan. Moreover, pieces of timber with rot, as well as rooted stumps, branches, and twigs—some with leaves or needles still attached—were added to the assemblage. The timbers had been carefully stacked around a central post, but little effort seems to have gone into reworking them before inclusion in the mound, apart from splitting and bucking.

The analyses further demonstrated that the main body of sampled trees was largely felled within a single season. This, as has been argued, would have required a considerable workforce, which in turn has been cited as evidence for a high degree of social organization (e.g. Hagen, Reference Hagen1997: 198–99; Skre, Reference Skre1997: 38). Yet, certain often overlooked peculiarities in the description of the timbers point to an alternative interpretation. Ording (Reference Ording1941: 95) noted that many of the trees had been felled unusually high along the trunk, a practice generally discouraged as it can inhibit regrowth. While this may be explained by deep snow at the time of felling (Hagen, Reference Hagen1997: 195), it does not explain why several timbers had been broken rather than cut with an axe. Similarly, the inclusion of tree roots implies that trees had been pulled from the ground, both unusually cumbersome and inefficient ways of felling, particularly in winter.

Ording’s analyses further suggested that the timbers were sourced from an area of 1 km², cleared several decades before the construction of the mound. This was later supported by pollen analyses from the nearby lake (Høeg, Reference Høeg1997: 58–59). The area was subsequently turned to pasture, as indicated by moss analyses from within the mound itself (Størmer, Reference Størmer1949) and a more recent multi-proxy study by Bajard et al. (Reference Bajard, Ballo, Høeg, Bakke, Støren and Loftsgarden2022), which shows a shift in agricultural practices from crop cultivation to animal husbandry in the area during the sixth century. As a consequence, grazing would have prevented the re-establishment of a dense forest, forming instead an exposed landscape with trees growing in small, scattered clusters, and developing relatively stubby and tapered forms (Ording, Reference Ording1941: 127). Given that denser forests probably grew to the north of Raknehaugen, it seems probable that the trees in the mound were sourced from the more open, pasture-like areas to its south.

The inclusion of these particular timbers was clearly important to the builders of the Raknehaugen mound. It is notable that pre-Christian societies regarded trees as sacred or imbued with auspicious qualities (Näsström, Reference Näsström2001: 79–81; Davies & Robb, Reference Davies and Robb2004: 145) but their significance seems neither connected to specific tree species nor to their condition or degree of preparation before burial. Instead, it seems that their origin—the particular place from which the trees were obtained—was significant.

The importance of landscape

When focus is restricted to appearance and content it reduces the mound to a static, two-dimensional form, detached from its immediate surroundings, and downplays or neglects the undoubtedly complex relationships that existed between the mounds, their builders, and their immediate environs. Thus, while also depriving the mound builders of agency, the landscape becomes secondary to the mound, a backdrop against which human activity took place and onto which human symbolism was projected. Such a view contrasts with current relational approaches to landscape archaeology, in which the landscape is seen as active, agential, and constantly negotiated (e.g. Tilley, Reference Tilley1994: 24; Knapp & Ashmore, Reference Knapp, Ashmore, Ashmore and Knapp1999: 2; Bradley, Reference Bradley2000; Bender, Reference Bender2002: 104; Brück, Reference Brück2005). These perspectives have generally been underexplored in Scandinavian archaeology, where research has tended towards studies that consider landscape forms as significant on account of the presence of deposited artefacts (e.g. Rødsrud, Reference Rødsrud2005; Lund, Reference Lund, Carver, Semple and Sanmark2010; Fredengren, Reference Fredengren2018), or where the terrain is used as a means to enhance or elevate the monument itself rather than as imbued with any numinous quality (e.g. Gansum & Oestigaard, Reference Gansum and Oestigaard2004: 64; Thäte, Reference Thäte2007: 132–42; Maher, Reference Maher, Harrison and Maher2014).

This interpretative tendency may, in part, stem from the assumption that veneration in and of nature had largely ceased in Scandinavia around the mid-first millennium ad, giving way to more organized forms of religion practised within buildings (Fabech, Reference Fabech, Dickinson and Griffiths1999: 38). Late Antique and early medieval written sources, however, attest to the sustained ritual significance of the natural landscape for northern European peoples, as do medieval sources for our understanding of Norse religious beliefs (Näsström, Reference Näsström2001; Rødsrud, Reference Rødsrud2005). Furthermore, early Christian law texts and moral guidelines explicitly condemn the worship of natural elements (Andrén, Reference Andrén, Jennbert, Andrén and Raudvere2002; Brink, Reference Brink and Nordeide2013). Thus, we must assume that such practices persisted beyond the introduction of Christianity, and even into modern times, according to ethnographic evidence (Lönnroth, Reference Lönnroth, Almqvist and Gröning2021).

Such indications invite us to look beyond isolated natural features, such as stones, trees, and wetlands, to borders and transitional zones as loci of sacred significance. This is supported by archaeological evidence for ritual deposits or the deliberate placement of monuments in liminal landscapes (Lund, Reference Lund, Carver, Semple and Sanmark2010; Hedeager, Reference Hedeager2011: 172). Moreover, in Icelandic written sources, transitional landscapes serve as boundaries between the earthly realm and the otherworld (Leslie, Reference Leslie and Sävborg2010). Likewise, ethnographic studies suggest that areas marked by abrupt transition, whether it be in geology, hydrology, or vegetation, are likely to invoke a sense of the sublime (Taçon, Reference Taçon, Ashmore and Knapp1999: 36–37). In short, these transitional zones in the landscape were places where the sacred manifested itself through sudden and transformative revelation, bridging the otherworldly and the earthly, and setting it apart from the ordinary (Eliade, Reference Eliade1990: 107; Fabech & Näsman, Reference Fabech, Näsman, Nordeide and Brink2013: 66).

This notion of liminality can help us frame the contrasting landscapes on either side of Raknehaugen, revealing how environmental qualities may have shaped past perceptions of ritual significance. Thus, the dense vegetation and featureless terrain of the forests to the north of the mound afforded limited light and restricted sightlines, perhaps creating an enclosed atmosphere that may have been perceived as both threatening and protective—a liminal space where the familiar meets the unknown (Lönnroth, Reference Lönnroth, Almqvist and Gröning2021: 119). In contrast, the southern landscapes were more open, with ample light and broader vistas. Yet, their undulating terrain and winding watercourses constrained movement through the landscape, while their streams, unstable soils, and frequent landslides may have contributed to a sense of fluidity and unpredictability. These landscapes were not static backdrops; they were part of a dynamic and sometimes volatile environment which, in the mid-sixth century, was affected by severe climatic disturbances.

The long sixth century

The sixth century witnessed a climatic deterioration likely to have been triggered by a series of volcanic eruptions between ad 536 and 660, which injected aerosols into the stratosphere, obscuring the sunlight and causing considerable cooling of the atmosphere (Büntgen et al., Reference Büntgen, Myglan, Ljungqvist, McCormick, Di Cosmo and Sigl2016; Bajard et al., Reference Bajard, Ballo, Høeg, Bakke, Støren and Loftsgarden2022: 2), leading to widespread crop failure, famine, disease, and a general population decline. In Scandinavian archaeology, the event of ad 536 has received particular attention, as it appears to coincide with a period of extensive socio-cultural change and may be linked to the Fimbulwinter myth, a portent of Ragnarǫkr, the end times in Norse mythology (Axboe, Reference Axboe1999; Gräslund & Price, Reference Gräslund and Price2012; Iversen, Reference Iversen, Iversen and Petterson2016; for a critique, see Gundersen, Reference Gundersen2019 and Gjerpe, Reference Gjerpe2021).

It has been argued that the emergence of large mounds in the mid-sixth century should be seen as a societal response to these crises, either as a physical affirmation of land ownership and authority, or as collective religious acts (e.g. Löwenborg, Reference Löwenborg2012; Gjerpe, Reference Gjerpe2021: 43; Loftsgarden & Iversen, Reference Loftsgarden and Iversen2024: 116–17), with some researchers making an explicit link between the construction of Raknehaugen and these events (Price & Gräslund, Reference Price, Gräslund and Riede2015: 12; Kaliff & Oestigaard, Reference Kaliff and Oestigaard2020: 242–46; Gjerpe, Reference Gjerpe2021: 42–43; Sæbø, Reference Sæbø2025). This connection is largely based on the fifteenth growth ring observed by Ording and Johnsen, which Skre (Reference Skre and Nyberg2016), in combination with radiocarbon dating of the timbers, used to date the felling of the trees to ad 551, and the construction of the mound to the following year. It should, however, be noted that the growth rings (see Figure 5) on either side of the fifteenth appear completely normal (Ording, Reference Ording1941: 112), and that tree growth seems to have been entirely unaffected by other negative climatic events in the period, introducing some uncertainty as to the suggested dates. Nonetheless, it is reasonable to assume that the mound was built in the mid-sixth century, amid worsening climatic conditions, which included a cooling of the atmosphere and increased levels of precipitation (Gräslund & Price, Reference Gräslund and Price2012: 438; Bajard et al., Reference Bajard, Ballo, Høeg, Bakke, Støren and Loftsgarden2022).

Such conditions would have a considerable impact on the physical landscape around Raknehaugen. As Hedeager (Reference Hedeager2011: 172) and Bajard et al. (Reference Bajard, Ballo, Høeg, Bakke, Støren and Loftsgarden2022) demonstrate, the sixth century saw a local shift from crop cultivation to pastoralism, coinciding with Ording’s claim that the mound timbers came from nearby land cleared some decades before the trees were felled. This transformation would have altered the established drainage patterns, as animal trampling and the removal of water-absorbing vegetation increased surface runoff and destabilized the soil. Moreover, the combination of increased precipitation and the lowering of the temperatures would also have reduced evapotranspiration levels, increasing the frequency of soil saturation and advancing the leaching of stabilizing agents in the soil. As a result, the conditions were in place for the formation of ‘quick clay’, in which the clay temporarily acquires a liquid consistency, making it highly susceptible to landslides (Torrance, Reference Torrance, Clague and Stead2012).

A ritual response to a cataclysmic event?

The combined effects of human-induced landscape changes and increased precipitation in the sixth century created the ideal conditions for a catastrophic event, which is potentially reflected in the massive landslide scar discussed above. If indeed there is a temporal relationship between the two, then the period—already marked by social instability due to food shortages, epidemics, and high mortality—was further punctuated by a massive, cataclysmic landslide. The full impact of such an event is naturally difficult to grasp, but it is likely to have been traumatic, as the entire landscape moved and transformed into a violent liquid that devastated homes and fields, claiming lives and livestock. Likewise, the long-term consequences may have exacerbated hunger and disease in an already vulnerable population, while the immense scar left on the landscape would have stood as an ominous and constant reminder, perhaps of human insignificance and unearthly powers.

In the wake of such devastation, it is not uncommon for communities to come together to make sense of their situation and take action to prevent further calamity. An intriguing example in this context is the large mound, Ngundeng’s Mound or Pyramid, constructed by the Nuer of Sudan in the late nineteenth century ad following a twin outbreak of smallpox and rinderpest in a period of drought. Honouring a deity seen as the vanquisher of these diseases, Ngundeng’s Mound was built to prevent their return by symbolically confining evil thoughts and actions to the mound (Whittle, Reference Whittle1997: 148–49). It represents a good example of a societal effort to neutralize external forces that threatened a community. Furthermore, it serves as an apposite example of how a decentralized society with no prior experience of cooperative labour or monument building could suddenly unite to build an impressive and robust structure (McGregor, Reference McGregor2001: 201). In a similar vein, the construction of megalithic monuments in France and Spain has been linked to earthquake events, where these structures were perhaps built to appease destructive forces (Bradley & García Sanjuán, Reference Bradley, García Sanjuán, Bickle, Cummings, Hofmann and Pollard2017: 190) and, as mentioned above, recent research has already made a direct connection between Raknehaugen and the effects of the ‘Dust Veil’ event of ad 536.

Applying this perspective to Raknehaugen, its construction can be read as a response by the affected communities to what was perceived as an external threat, whereby this not only served as a means of restoring the physical landscape but also of re-establishing the sacred and social order. The materials incorporated in Raknehaugen may thus have served a propitiatory role or been an apotropaic device, acting as a protective barrier against malevolent forces or harmful spirits. And what better location than where the gentle landscape of the forests met the volatile terrain overlooking the vast scar left by the landslide, a liminal ‘thin place’ where interaction with the divine was possible (Hooke, Reference Hooke, Hooke and Hyer2017: 108).

While a direct connection between the mound and the landslide has not been determined, this could, for instance, be explored through detailed soil analyses of the landslide zone to determine its type, erosional force, and volume, and by dating the landslide using traditional radiocarbon or luminescence dating, or more innovative techniques such as cosmic ray exposure (Pánek, Reference Pánek2015). In the absence of these investigations, there are still elements of the already existent archaeological material that lend credibility to the idea. For example, the gentle appearance of the landslide scar in the landscape suggests some antiquity, which is further supported by the absence of any historical references to the event. Furthermore, several aspects of the mound’s construction align well with the suggested interpretation. In particular, the timbers, which included a variety of species originating from the open landscape south of the mound (an area prone to landslides) and the size of the scar coincide well with the estimates of the area cleared to provide material for the mound (Figure 10). Further, the timbers themselves included stumps that had been pulled out of the ground, while others had been broken rather than cut with an axe. Others again had been cut unusually high up on the trunk, perhaps suggesting partial burial when this occurred. Combined with the narrow timeframe in which the timbers were felled, this raises the possibility that their felling resulted from a landslide rather than by deliberate human action.

Figure 10.

Landslide scar lying immediately south of Raknehaugen, measuring just over 1 km²(background data: © Kartverket, 2025). Reproduced under licence CC BY 4.0.