Introduction
The construction of permanent houses marked a significant departure from previous forms of non- or semi-sedentary living, providing physical protection from various threats, natural or otherwise, as well as serving as visible and effective claims to land. From this basis, houses often become symbols of social identity and economic prosperity (Blanton, Reference Blanton1994; Carsten & Hugh-Jones, Reference Carsten, Hugh-Jones, Carsten and Hugh-Jones1995), with architectural forms functioning as a medium for expressing the social dimension (Delitz, Reference Delitz, Trebsche, Müller-Scheeßel and Reinhold2010). In Scandinavian Bronze Age archaeology, this investment in houses and their diverse sizes is commonly argued to reflect the transformation to a clearly stratified society, characterized by the presence of chiefly farms (Brink, Reference Brink2013: 433). Many scholars interpret this as the establishment of a ‘house society’, where house sizes reflect individual status and socio-political power (e.g. Streiffert, Reference Streiffert2005: 25, 67; Artursson, Reference Artursson2009: 198; Austvoll, Reference Austvoll2020: 23; Reference Austvoll2021: 55–56).
Although numerous archaeological, historical, and ethnographic examples show a positive correlation between wealth and house sizes (David & Kramer, Reference David and Kramer2001; Kohler et al., Reference Kohler, Smith, Bogaard, Feinman, Peterson and Betzenhauser2017; Hofmann et al., Reference Hofmann, Müller-Scheeßel and Müller2024), making the concept of a ‘house society’ a fruitful line of enquiry, such models and their inherent structural inequalities need to be proven rather than simply assumed (Harding, Reference Harding2013). To this end, this article presents a newly collected corpus of 149 houses and other buildings from fifty-six sites in south-eastern Norway (Figure 1), dating to the Late Neolithic (2300–1700 bc) and the Bronze Age (1700–500 bc).
Figure 1. The south-eastern Norway study area within Scandinavia, with the extent of the Nordic Bronze Age overlaid following Prescott and Glørstad (Reference Prescott, Glørstad, Prescott and Glørstad2012), extended to encompass the Netherlands and northernmost Germany.
The primary objective of this article is to provide, for the first time, a comprehensive overview of the settlement evidence in south-eastern Norway, offering detailed supplementary data for further studies. A second aim is to present a comparative analysis, using data from across southern Scandinavia, and consider trends discernible in a broader corpus of settlements in the Netherlands and northern Germany. Given that south-eastern Norway is traditionally viewed as a geographical and cultural periphery in the Nordic Bronze Age world (Kristiansen, Reference Kristiansen1998: 70), and recognizing that models need to be proven rather than assumed, it is particularly important to examine how the houses and settlement patterns in our region align with the wider trends outlined for the Nordic Bronze Age. Further, this study aims to assess and discuss the settlement and social organization in south-eastern Norway, as evident from its houses. Chronologically, the focus is on developments during the Late Neolithic and the Early Bronze Age (1700–1100 bc), with the trends emerging in the Late Bronze Age (1100–500 bc) addressed more briefly.
Background
In Scandinavian archaeology post-built, aisled structures about twice as long as they are wide are referred to as houses or longhouses and believed to represent architecturally a supra-regionally shared way of dwelling, although variations in practice may be concealed by architectural similarity. Based on the idea that the size of permanent dwellings correlates with agricultural productivity and access to resources, as traditionally associated with Iron Age Scandinavia (e.g. Webley, Reference Webley2008: 52), clear differences in house sizes are also commonly interpreted as indicative of status, with larger structures demonstrating the existence of chiefdoms during the Nordic Bronze Age, as these are considered chiefly farms (e.g. Vandkilde, Reference Vandkilde1996; Earle, Reference Earle2002; Kristiansen & Larsson, Reference Kristiansen and Larsson2005; Kristiansen, Reference Kristiansen and Sjögren2006; Artursson, Reference Artursson2009; Egelund Poulsen, Reference Egelund Poulsen and Brattli2009). Regarding houses, such hierarchical manifestations are interpreted by several scholars as the emergence of a ‘house society’ (e.g. Artursson, Reference Artursson2009, Reference Artursson, Prieto Martinez and Salanova2015; Nordvall, Reference Nordvall and Wranning2019; Austvoll, Reference Austvoll2020; Earle et al., Reference Earle, Haack-Olsen, Eriksen, Henriksen and Kristensen2022). This concept draws on Claude Lévi-Strauss’ theoretical framework (Lévi-Strauss, Reference Lévi-Strauss1982), which has been successfully applied to numerous anthropological and archaeological studies (e.g. Joyce & Gillespie, Reference Joyce and Gillespie2000; González-Ruibal, Reference González-Ruibal2006; Beck, Reference Beck2007; Borić, Reference Borić and Jones2008; Sharples, Reference Sharples, Currás and Sastre2020). The archaeological application of the concept primarily revolves around houses reflecting social inequality owed to unequal access to capital (González-Ruibal & Ruiz-Gálvez, Reference González-Ruibal and Ruiz-Gálvez2016: 385), a notion that resonates with the dominant Marxist approach to political economies in Scandinavian Bronze Age research (Earle et al., Reference Earle, Ling, Uhnér, Stos Gale and Melheime2015).
Within this framework, the house served as a hierarchical symbol during the Nordic Bronze Age, the conceptual underpinnings of which are illustrated in Figure 2 (Kristiansen, Reference Kristiansen and Sjögren2006: 192, fig. 50). This model shows an increase in house sizes from the end of the Late Neolithic through the Early Bronze Age, in which 200 m2 could potentially mark a significant threshold (see Bech & Haack-Olsen, Reference Bech, Haack-Olsen, Bech, Valentin Eriksen and Kristiansen2018: 135, fig 4.3). Large houses often co-existed with smaller houses, indicating that it is not just the size of some houses that are at the basis of the chiefdom model, but rather that a relatively wide size range can be present. A prominent example is the Højgård houses in southern Jutland, where a 400–500 m2 house was contemporaneous with several smaller houses (Ethelberg, Reference Ethelberg, Ethelberg, Jørgensen, Meier, Robinson, Hardt and Poulsen2000).
Figure 2. Schematic model of the long-term change of house types and sizes from the Late Neolithic to the Bronze Age in southern Scandinavia (from Kristiansen, Reference Kristiansen and Sjögren2006: fig. 50). Reproduced by permission of K. Kristiansen .
From around 1300 bc, house sizes in southern Scandinavia decreased and continued to decline throughout the Late Bronze Age. Interestingly, this trend contrasts with northern Germany, where house sizes decrease from 1700 bc before starting to increase again from 1400 bc (Kneisel et al., Reference Kneisel, Dörfler, Dreibrodt, Schaefer-Di Maida and Feeser2019: 1611). During the Early Bronze Age, houses in northern Germany typically occupied a surface of 100–200 m2, although a few slightly large examples are known (Donat, Reference Donat2018: 184). From the transition to the Late Bronze Age, houses sharply decrease in size again in northern Germany (Kneisel et al., Reference Kneisel, Dörfler, Dreibrodt, Schaefer-Di Maida and Feeser2019: 1612), as also seen in southern Scandinavia. As evident in Figure 2, large farmhouses in southern Scandinavia reached 100–150 m2, and while the houses decreased in size, they increased in number. Hence, houses are more often organized in hamlets in the Late Bronze Age, although this form of living is known from earlier periods (Kristiansen, Reference Kristiansen2007: 166; Bech, Reference Bech, Bech, Valentin Eriksen and Kristiansen2018: 27–28; Løken, Reference Løken2020; Risch et al., Reference Risch, Friederich, Küssner and Meller2022). In such hamlets, houses are typically located 50–100 m apart (Horn et al., Reference Horn, Austvoll, Artursson and Ling2023: 16).
Against this background, it becomes evident that while a dominant model of house development in southern Scandinavia exists, a comparison with other regions both within and adjacent to southern Scandinavia in northernmost Europe reveals diverging developments. Suggesting that the outlined conceptual model is not universal. Looking outside the Nordic zone, Stijn Arnoldussen (Reference Arnoldussen2008) has argued that house size in the Dutch area is a direct reflection of the number of people living in the house, irrespective of social stratification. Within Scandinavian Bronze Age research, however, alternative explanations of size differences are rarely considered or given much attention (Brück & Fokkens, Reference Brück, Fokkens, Fokkens and Harding2013).
Material Presentation and Analysis
The south-eastern Norwegian settlement evidence consists of 119 post-built houses and thirty other buildings, the latter mostly interpreted as serving various economic functions (see Supplementary Material 1 for details). While the state of preservation varies (see Supplementary Material 2 for details), around eighty per cent of the houses had enough definable features of their ground plans to derive sufficient measurements for analysis. Although primarily relying on the information and interpretations of the settlements presented in excavation reports (links in Supplementary Material 1), all available GIS data and additional documentation have been reviewed to ensure a thorough examination of all excavated features, including those not included in the presentation of houses in the reports. Further details on this matter are available in Supplementary Material 2.
Temporal distribution
There are 269 radiocarbon dates available, ranging between 2873–2207 cal bc (4000 ± 110 bp, T-18022) and 775–400 cal bc (2415 ± 80 bp, T-19882), visualized in summed probability plots using the rcarbon package (Crema & Bevan, Reference Crema and Bevan2021) and calibrated with the InCal20 calibration curve (Reimer et al., Reference Reimer, Austin, Bard, Bayliss, Blackwell and Bronk Ramsey2020) using the rintcal package (Blaauw, Reference Blaauw2022). The summed probability distribution (Figure 3) shows a noticeable establishment of houses around 2200 bc, followed by growth until approximately 1800 bc. From this point on, there is a gradual decline until c. 1600 bc, when a significant drop occurs among the two-aisled material. Concurrently, three-aisled houses (although introduced earlier) become established, and two-aisled houses are constructed later (Figure 4). Considering the distribution of radiocarbon dates for each construction, Figure 4 reveals that some structures within the same sites overlap, suggesting that they may have been contemporary. This is observed at, for example, sites 8, 18, 23, 24, 47, 48, and 54. To further investigate this possibility, their spatial distribution needed to be analysed.
Figure 3. Summed probability plot of available radiocarbon dates from the settlements (n = 269) binned in 200-year long phases, differentiated in building types.
Figure 4. Start and end times of all individual radiocarbon-dated buildings, differentiated by building types.
Spatial distribution
Spatial analysis suggests that houses with overlapping dates must have been successive, as they are often constructed in the same location (Figure 5). On the other hand, sites situated very close to each other do not feature contemporary structures, resulting in houses located at least 1 km—but more commonly several kilometres—away from any contemporary buildings. From around 2200 bc to 1400/1300 bc, the distribution of sites shows a general distance of around 4–7 km, increasing to 10–12 km around 1200–1100 bc. From the Late Bronze Age onwards, the distances decrease, with contemporary houses occasionally as close as 1 km apart towards the transition to the Pre-Roman Iron Age (500 bc–ad 1). Overall, the spatial arrangement suggests a large living environment throughout the entire study period.
Figure 5. Spatial distribution of the settlement sites within south-eastern Norway. For details on the number of buildings and their dates, see Figure 4. The inset examples (A–D) illustrate sites with buildings that overlap in date, but which are too close to each other or directly on top of each other to have been contemporaneous.
There are, however, a few instances of contemporary constructions at the same site in terms of houses alongside a four-post structure, which is typically interpreted as storage buildings at the farmstead (Løken, Reference Løken and Løken1998). Other economical buildings, such as the wall-ditch buildings, are more typically suited further away from the houses, sometimes encountered in more isolated locations, possibly due to their use as outbuildings, a topic I will return to later.
Measurements and sizes
The post-built houses range from 5 to 25 m in length and from 3.5 to 10 m in width. Typically, three-aisled houses are wider, while two-aisled are slightly longer but have a narrower range of square metre (Figure 6, details available in Supplement 1). To quantify these measurements and assess changes in the house sizes to create a chronological distribution similar to that presented for the south Scandinavian material in Figure 2, weighted means of house sizes were calculated in 200-year intervals (see also Løvschal, Reference Løvschal2020).Footnote 1 The results show an increase from an average surface of 25 m2 during the Middle Neolithic transition to 86 m2 in the latter part of the Bronze Age, peaking at 114 m2 between 1000 and 800 bc (Figure 7). To test the normality in the size distribution (see e.g. Peterson & Drennan, Reference Peterson, Drennan, Kohler and Smith2018: 41), Gini coefficients were calculated for the same 200-year intervals using the DescTools package in R (Signorell, Reference Signorell2024; see also Hofmann et al., Reference Hofmann, Müller-Scheeßel and Müller2024). The results indicate low to moderate levels of unevenness (ranging between 0.09 and 0.31; Gini coefficient values are defined as ranging from 0 (perfect uniformity) to 1 (maximum disparity)), in which the period 900–800 bc marks a shift (Figure 7).
Figure 6. Violin plots of measurements grouped by house types.
Figure 7. House surfaces by weighted m2 following the 200-year intervals (blue line), with the maximum and minimum m2 represented as a shaded outline. The Gini coefficient of the same 200-year intervals is illustrated in the plot below.
House Features
Architectural forms and features
The most distinctive building type (Figure 8) is the curved-end rectangular house, common in Late Neolithic and Early Bronze Age houses, with clear parallels in Sweden, eastern Denmark, and eastern Germany (Petersson, Reference Petersson, Peterson and Toreld2010; Fokkens, Reference Fokkens, Meller, Friederich, Küssner, Stäuble and Risch2019b). Obround houses (i.e. with parallel sides and rounded ends) emerge as most typical for the Late Bronze Age. Similarities for this type are found across southern Scandinavia, typically dating to the Late Bronze Age (Artursson, Reference Artursson2009; Løken, Reference Løken2020). Rectangular houses are most numerous overall, but are more evenly spread throughout the study periods. In the Late Neolithic/Early Bronze Age, some resemble the distinctive, small, rectangular Fosie-type houses (e.g. Björhem & Säfvestad, Reference Björhem and Säfvestad1989).
Figure 8. Time interval of the settlement evidence, with examples of the different building types from the south-eastern Norwegian corpus.
All these houses used a trestle construction technique, grindverk in Norwegian, a technique associated with Scandinavian post-built houses (Oma, Reference Oma2018: 59). Another architectural feature in south-eastern Norway consists of possible instances of staveline construction (Gunnarsjaa, Reference Gunnarsjaa2007: 731; Godal et al., Reference Godal, Moldal, Oalann and Sandbakken2018: 158, 164). This technique involves building with timbers set lengthways, with the two rows of roof-bearing posts bound laterally by the top sill, instead of transversally in trestle pairs. Archaeologically, this feature is noticeable in the skewed relative position of the posts, giving these buildings a distinct trapezoidal form. While this could belong to a three-aisled construction, it is interpreted as one-aisled, built in a less robust fashion. What appears to be roof-bearing posts may be wall-posts supporting the roof, as known from historical sources. Historically, these ‘staveline’ buildings often served economic functions, such as boat-houses and hay barns (Godal et al., Reference Godal, Moldal, Oalann and Sandbakken2018: 159).
Exterior features
Identifying the exterior architecture from excavated materials can be challenging, but specific wall types can be suggested for about one-third of the structures (Table 1a). These wall types are primarily identified by examining the spacing and size of the wall-posts (see Austvoll, Reference Austvoll2021: 77, with further references). Closely spaced, smaller posts, often arranged in a curved exterior pattern, are associated with wattle-and-daub constructions. In contrast, larger posts spaced further and more evenly apart indicate bole walls, which can be either made of solid timber or planks, with the presence of wall-ditches suggested to indicate the latter. Based on these traits, bole walls are most common in two-aisled houses dating to the Late Neolithic and Early Bronze Age, while three-aisled structures more frequently consist of wattle-and-daub. This is further indicated by the presence of clay with imprints of latticework found in postholes.
Table 1. Overview of building features, including orientation. For details, see Supplementary Material.
Building orientation
A north–east to south–west orientation is generally most prevalent (Table 1b), with a relatively even distribution among other orientations and no apparent overarching pattern. There are, however, variations between different types of buildings; for example, an east–west orientation is more common among three-aisled houses compared to two-aisled houses.
Interior elements
Interior features are generally fragmentary, with entrances and hearths being the most common (Table 1c). Evidence of pens is only found in two Late Bronze Age buildings. The lack of indoor stalls does not necessarily exclude the possibility of animals being kept indoors; this could, for instance, be attested by the replacement of posts through more intensive rotting as well as by a smaller distance between the trestles (Petersson, Reference Petersson2006: 64–66). These features are found in eight houses, dated to the Late Bronze Age, predominantly towards the transition to the Pre-Roman Iron Age. Floors constitute another interior feature, and, while these mainly consist of rammed clay-like soils, there is one possible instance that a floor was covered in wood. Additionally, the revision of the data has also revealed instances of sunken-floor houses (Figure 8), previously unrecognized within the region.
Discussion
The south-eastern Norwegian settlement evidence aligns with overarching trends observed in Scandinavian Bronze Age houses, such as the adoption of three-aisled structures around 1600 bc, while the Late Neolithic and Early Bronze Age curved-end houses align closely with an eastward trend in house styles observed across northern Europe (see Fokkens, Reference Fokkens, Meller, Friederich, Küssner, Stäuble and Risch2019b: 928, fig. 13), particularly across the border and within south-western Sweden (see e.g. Petersson, Reference Petersson, Peterson and Toreld2010). Obround-shaped houses, on the other hand, are less common in this period, but well known across Jutland and northern Germany at this time (Ethelberg, Reference Ethelberg, Ethelberg, Jørgensen, Meier, Robinson, Hardt and Poulsen2000; Bech & Haack-Olsen, Reference Bech, Haack-Olsen, Bech, Valentin Eriksen and Kristiansen2018; Geschwinde & Heske, Reference Geschwinde, Heske, Meller, Friederich, Küssner, Stäuble and Risch2019). The eastward trend is supported by a similar trapezoidal ‘staveline’ type of construction being recognized in south-western Sweden (Lönn, Reference Lönn, Peterson and Toreld2010), and is further reinforced by the narrow, oblong wall-ditch constructions without posts, which, to my knowledge, only have parallels in Halland along the west coast of Sweden (Nicklasson, Reference Nicklasson, Johansson, Streiffert and Wranning2001; Häggström & Kalmar, Reference Häggström Aulin and Kalmar2003).
The settlement data also reveal distinct regional characteristics. While east–west oriented houses are predominant among settlements in southernmost Scandinavia (e.g. Bunbury et al., Reference Bunbury, Austvoll, Jørgensen, Vatsvåg Nielsen, Kneisel and Weinelt2023: 10), this orientation is only most common in three-aisled houses compared to two-aisled houses in south-eastern Norway. Considering that eighty-five per cent of east–west oriented houses in southernmost Scandinavia are found in Jutland (see Bunbury et al., Reference Bunbury, Austvoll, Jørgensen, Vatsvåg Nielsen, Kneisel and Weinelt2023: appendix D), the differing orientation of houses in south-eastern Norway could be a result of a more varied topography and a different landscape compared to further south. This varied topography, with different wind directions and exposure to sunlight, is likely to have influenced the orientation and siting of the houses. The most notable regional characteristic in the south-eastern Norwegian settlement evidence, however, is the tendency to construct smaller houses compared to those traditionally associated with the Late Neolithic and Bronze Age in Scandinavia (Figure 9). In the following, I will explore some possible reasons behind the emergence of this particular trend in the region.
Figure 9. Chronological size distribution in south-eastern Norway (blue) compared to the south Scandinavian development model (grey), illustrating differences in size variation. The results in Figure 7 were smoothed using spline interpolation within ggplot2 and then overlaid on the model in Figure 2 using Adobe Illustrator CS6.
Potential reasons for building small houses
One possible explanation for building small could be the need for more efficient heating in colder latitudes compared to further south (Porčić, Reference Porčić2012; Sand-Eriksen & Mjærum, Reference Sand-Eriksen and Mjærum2023). This trend can also be observed along the western Swedish coastline, where Bronze Age houses appear to decrease in size from Halland northwards into Bohuslän (Carlie, Reference Carlie1992; Streiffert, Reference Streiffert, Claesson and Munkeberg2004; Nordvall, Reference Nordvall2015, Reference Nordvall and Wranning2019). Furthermore, a similar pattern is attested to in western Norway, where the average length of the houses dating to the Late Neolithic and Early Bronze Age in southern Rogaland is 17 m compared to 10 m further north along the coast (Austvoll, Reference Austvoll2021). An interpretation linked to climate is thus plausible but insufficient since houses in northern Germany are generally smaller than those typically associated with the Scandinavian Bronze Age.
Another possibility is that variations in size reflect different subsistence practices. This is particularly relevant given that monumental houses in southernmost Scandinavia are considered to be directly proportionate to agricultural productivity. Although husbandry is regarded as a fundamental part of the domestic economy in southernmost Scandinavia, it has recently been argued that the monumentalization of houses has its roots in agricultural intensification beginning in the Late Neolithic (Johannsen, Reference Johannsen2024). Considering that south-eastern Norway has a shorter growing season compared to southernmost Scandinavia (Moen, Reference Moen1999: 21), arable farming may have been on a smaller scale due to climatic factors, requiring communities to rely on other forms of subsistence economy. Reduced house sizes may thus relate to both climatic and domestic factors. There is, however, little evidence of animals being accommodated indoors before the latter part of the Late Bronze Age. If we further consider that the shift from two- to three-aisled buildings is associated with the indoor stalling of livestock (e.g. Fokkens, Reference Fokkens, Fabech and Ringtved1999; Rasmussen, Reference Rasmussen, Fabech and Ringtved1999; Oma, Reference Oma2018), the overall small house sizes within south-eastern Norway seem less compatible with indoor stalling. Therefore, animal husbandry must have taken another form here, possibly oriented and adapted to outdoor conditions (Petersson, Reference Petersson2006).
Given that both the wall-ditch and one-aisled, trapezoidal ‘staveline’ buildings are less robust and less suited to permanent residence, they could have served as temporary shelters for human and animals, as well as for storage of equipment and fodder. On the other hand, it is equally possible that they were associated with other activities than pastoral practices, such as hunting and gathering. Although further archaeological data sources should be considered to confirm this, which is beyond the scope of this study, the ‘large’ outdoor living space between farmsteads indicates the possibility of a subsistence strategy reliant on high mobility.
This inference is further supported, I argue, by the low Gini coefficient observed among the houses in this region. Kohler et al. (Reference Kohler, Smith, Bogaard, Feinman, Peterson and Betzenhauser2017: 619) found a correlation between higher Gini coefficients and agricultural societies (greater than 0.3), compared to lower coefficients in hunter-gatherer societies (less than 0.2), with horticulturalist societies falling in between. These provides us with values to make inferences about subsistence practices based on the Gini coefficients in our own study area. The Gini coefficient calculated here shows that south-eastern Norwegian households only reach an agricultural level from c. 800 bc and remain below 0.2 until about 1500 bc (Figure 7). In comparison, when we apply the same calculation to material from Sweden and Denmark, as derived from Bunbury et al. (Reference Bunbury, Austvoll, Jørgensen, Vatsvåg Nielsen, Kneisel and Weinelt2023), the Gini coefficients for around 1100 houses dating to the Late Neolithic and Early Bronze Age reach an agricultural level of 0.33. The material from western Norway remains slightly below the 0.3 threshold, with an overall coefficient of 0.29. However, a regional analysis along the coast reveals a clear south-to-north gradient: the southernmost region of Rogaland exhibits coefficients reaching agricultural levels, while those further north stay below this threshold.
This pattern, combined with evidence of mixed subsistence practices in northern areas—including pastoralism, hunting, and fishing alongside arable farming (see Austvoll, Reference Austvoll2021: 143–75)—suggests that houses in these regions may not have been intended as year-round dwellings for everyone. Instead, they may have allowed for certain, possibly seasonal, groups to live away from the house. Consequently, the lower Gini coefficients may not only reflect subsistence practices but could also support the argument for more mobile lifestyles that, in turn, permitted a smaller overall house size.
A ‘house society’ or a society with houses?
Considering how architecture can serve as a medium for the social structure and that socio-economic factors can influence the size of houses, the variation in house sizes within western Norway—specifically between the Jæren region in Rogaland and areas further north—suggest differences in household organization. These range from coercive to cooperative models (Austvoll, Reference Austvoll2020, Reference Austvoll2021; Lund et al., Reference Lund, Furholt and Austvoll2022), with the former related to a hierarchical house-based form of societal organization, and the latter diverging from this structure. These findings align well with results obtained from the Gini coefficients in the present study. Against this, the smaller and more uniform house sizes in southeastern Norway suggest a social organization that differs from the dominating ‘house society’ model within the Nordic Bronze Age world.
However, significance could also be attributed to houses through different means (Lévi-Strauss, Reference Lévi-Strauss1982). Studies of ‘house societies’ in Southeast Asia show that ancestral practices are common. For example, the Atoni of Timor make offerings to the inner left post in the house, as this is the ‘head mother', while the Torajan in Indonesia put similar value to the centre or ‘navel-post’ (Waterson, Reference Waterson1997: 85, 124–27). The Kayapó or Mẽbêngôkre of central Brazil prioritize upkeeping, and it is the oldest house that is the most important (Lea, Reference Lea, Carsten and Hugh-Jones1995: 208). Relating to these examples, I explored features or finds suggesting significance attributed to houses beyond their size. Although forty-nine houses had finds in structural features, these were primarily isolated occurrences and mainly consisted of types typically associated with household refuse (see Supplementary Material 1). While a few instances of arrowheads and an axe placed in postholes have been interpreted as possible house offerings (for details see e.g. Valum, Reference Valum2011), no noticeable patterns were observed within the material. On the other hand, the frequent rebuilding of houses in the same locations across many sites (see Figure 5) could be linked to some form of ancestral reverence, signifying the enduring importance of place across generations (see e.g. Brink, Reference Brink2013; Fokkens, Reference Fokkens, Fokkens, van As and Jansen2019a), even when large spaces were available. Yet, the empirical evidence overall does not support the idea of a ‘house society’ in south-eastern Norway.
The uniform nature of the settlement evidence suggests an organizational structure resembling segmentary societies. These societies consist of smaller, autonomous domestic groups without central authority, indicating a more symmetrical power relation (Currás & Sastre, Reference Currás, Sastre, Currás and Sastre2020; Lund et al., Reference Lund, Furholt and Austvoll2022). Considering the region's connection to the Nordic Bronze Age networks, the results of this study suggest a coexistence of different social systems in proximity to each other (see also Fontijn, Reference Fontijn2003; Graeber & Wengrow, Reference Graeber and Wengrow2021; Hofmann et al., Reference Hofmann, Nikulka, Schumann, Hofmann, Nikulka and Schumann2022). While further studies using additional archaeological data are needed, I believe that the key lies in its spatial organization within an environment that afforded mobile activities.
The significant distance between contemporary houses suggests that households had extensive areas at their disposal. This does not imply that communities avoided each other, but rather indicates a socioeconomic organization reliant on and centred around movement. Collective efforts, as shown by parts of the western Norwegian evidence (Austvoll, Reference Austvoll2021), could have been a structuring feature within this, as often encountered among mixed farming communities in historical and ethnographic sources in these areas (Kaldal, Reference Kaldal, Kaldal, Johansson, Fritzbøger and Snellman2000). Within this, the house functioned as place of reference for people throughout the year, and not necessarily as a place of dwelling and activity for all, as farmsteads are more traditionally portrayed (Holst & Rasmussen, Reference Holst, Rasmussen, Bergerbrant and Sabatini2013). Stable isotopes studies from neighbouring south-western Sweden have shown increased mobility during the Late Neolithic and the transition to the Bronze Age that lend support to such an interpretation (Blank et al., Reference Blank, Sjögren, Knipper, Frei, Malmström and Fraser2021).
In the Late Bronze Age, however, changes took place. Contemporary settlements began to appear closer together, though they did not form hamlets like those found further south. From around 900–800 bc, there was greater variety in house sizes, with the first occurrence of a house surpassing 200 m2. The Late Bronze Age settlement evidence from south-eastern Norway hence diverges from southern Scandinavia during this time, where houses are expected to decrease in size, but also from the previous period within the same area (see Figure 9). Given the common interpretation that large houses indicate greater wealth, this shift may reflect an increased social stratification in south-eastern Norway at this time. Alternatively, the houses might be built larger to accommodate greater numbers of people (as argued by Arnoldussen, Reference Arnoldussen2008), as seen in the wider increase in the population at this time following the final breakthrough of farming towards the end of the Late Bronze Age (Solheim, Reference Solheim2024). To fully understand this, however, these results warrant further investigation.
Conclusion
The settlement corpus from south-eastern Norway displays clear connections to the wider Nordic Bronze Age world, with houses reflecting broader trends across southern Scandinavia and northern Europe, especially in their eastward orientation. However, significant differences also emerge when comparing regions, underscoring the need to critically assess existing models, even within traditionally unified cultural spheres like the Nordic Bronze Age.
The main result of constructing smaller house sizes in south-eastern Norway, compared to other parts of southernmost Scandinavia, indicate a different social organization. Considering this alongside the significant distances between contemporary settlements, implying that households had extensive landscapes or territories, suggests a continued reliance on mobile activities, like hunting, gathering, and pastoralism, even after settling in post-built houses. Thus, the farmstead served more as a fixed point in a regularly traversed landscape than a continuous dwelling place.
The uniformity among these houses, on the other hand, reflects a society organized around symmetrical power relations and cooperation rather than hierarchy. Living far apart may have encouraged collective efforts for various purposes, such as cooperative hunting or pastoralism, a pattern seen later among mixed farming communities in this region of southern Scandinavia, though further studies using additional archaeological data are needed to confirm this.
By the Late Bronze Age, increased size variation and the closer construction of contemporary houses marks a shift, possibly indicating social stratification, changed mobility patterns, and/or population growth. Even with this shift, south-eastern Norwegian settlement practices diverged from southern Scandinavian trends, highlighting the importance of recognizing regional variations. Moreover, these results remind us not to assume the validity of existing models without testing them first, while further illustrating how distinct social systems can coexist in close proximity.
Acknowledgements
I would like to express my gratitude to Mette Løvschal for invaluable comments on an early draft of this article and for engaging in later discussions on topics explored in this article. I extend my thanks to Steinar Solheim and Knut Ivar Austvoll for insightful comments on the article, and to Hendrik Raese and Linn Nordvall for discussions on settlement material from northern Germany and south-western Sweden, respectively. I thank the four reviewers and the editors for the constructive feedback, which significantly improved this article. Special thanks to Åshild Lappegård Lahn for a final read-through.
Supplementary Material
To view supplementary material for this article, please visit https://doi.org/10.1017/eaa.2024.50.
Open access
