No. 46 2025 Local and Regional Variation in the Emergence of the Shieling System Kristoffer Dahle Journal of the North Atlantic Journal of the North Atlantic The Journal of the North Atlantic (Online ISSN #1935-1933, Print ISSN #1935-1984), with an international editorial board, is a collaborative publishing effort of the Eagle Hill Institute, PO Box 9, 59 Eagle Hill Road, Steuben, ME 04680-0009 USA. Phone 207-546-2821. E-mail: office@eaglehill. us. Website: www.eaglehill.us/jona. Copyright © 2025, all rights reserved. On-line secure subscription ordering: rate per year is $40 for individuals, $32 for students, $300 for organizations. Authors: Instructions for authors are available at www.eaglehill.us/jona. The Eagle Hill Institute (Federal ID # 010379899) is a tax exempt 501(c)(3) nonprofit corporation of the State o f Maine, USA. The Journal of the North Atlantic (JONA) is a multi-disciplinary, peer-reviewed and edited archaeology and environmental history journal focusing on the peoples of the North Atlantic, their expansion into the region over time, and their interactions with their changing environments. The journal—published online in the BioOne.org database and on the JONA website, and indexed in a full range of journal databases—serves as a forum for researchers, and as an information resource for instructors, students, and the intellectually curious who would like to learn about the latest research and study opportunities within the region. The journal publishes a wide diversity of research papers, as well as research summaries and general interest articles in closely related disciplines, which, when considered together, help contribute to a comprehensive multi-disciplinary understanding of the historical interplay between cultural and environmental changes in the North Atlantic world. Specifically, the journal’s focus includes paleo-environmental reconstruction and modelling, historical ecology, archaeology, ecology of organisms important to humans, anthropology, human/ environment/climate interactions, climate history, ethnography, ethnohistory, historical analyses, discussions of cultural heritage, and place-name studies. The journal publishes individual papers on an article-by-article basis. Whenever a manuscript has completed its peer review process and the article galley has been approved by the author, it will be immediately published online in the BioOne database and on the JONA website. This publishing model is also available for special volumes such as conference and symposium proceedings or other collections of papers. In effect, this means that articles are grouped online over time, i.e., the table of contents of volumes will grow as articles are posted online, which has the advantage of rewarding prompt authors, while enabling tardier authors to retain the option of being included in a special volume without delaying its publication. The Journal of the North Atlantic’s publishing format is versatile enough that authors can include supplementary files with their articles. These supplements may include dataset, figure, and table files (e.g., files requiring a larger than normal journal page size, such as large maps), as well as text and protocol files, audio and video files (e.g., for ethnographic studies), and even Powerpoint files. The Journal of the North Atlantic is indexed in the Web of Science (clarivate.com), EBSCO.com, and by way of author entries in Google Scholar and Researchgate. It is included in full-text in BioOne.org and JSTOR.org. Board of Editors Jette Arneborg, National Museum of Denmark, Copenhagen, Denmark Gerald F. Bigelow, Archaeology Institute, University of the Highlands and Islands, Kirkwall, Orkney, UK Rosie Bishop, Museum of Archaeology, University of Stavanger, Norway Colin Breen, University of Ulster, Coleraine, Northern Ireland Alison Cathcart, Division of History, Heritage and Politics, University of Stirling, Stirling, UK Mike J. Church, Department of Archaeology, Durham University, Durham, England, UK Jane Downes, Orkney College UHI, Kirkwall, Orkney, Scotland, UK Andrew J. Dugmore, Institute of Geography, University of Edinburgh, Edinburgh, Scotland, UK Mark Gardiner, School Of History And Heritage, University of Lincoln, UK Erika Guttmann-Bond, Vrije Universiteit, Amsterdam, The Netherlands Ramona Harrison, Department of Archaeology, History, Cultural Studies and Religion, University of Bergen, Norway Agnar Helgason, deCODE Genetics, Reykjavík, Iceland Joerg-Henner Lotze, Eagle Hill Institute, Steuben, ME, USA • Publisher Niels Lynnerup, Laboratory of Biological Anthropology, The Panum Institute, Copenhagen, Denmark Christian Koch Madsen, National Museum and Archives of Greenland, Nuuk, Greenland Ingrid Mainland, University of the Highlands and Islands Archaeology Institute, Orkney College, Kirkwall, Orkney, UK Meriel McClatchie. School of Archaeology, University College Dublin, Dublin, Ireland Thomas H. McGovern, Dept. of Anthropology, Bioarchaeological Lab, Hunter College, New York, NY, USA Natascha Mehler, Department of Medieval Archaeology, University of Tübingen, Germany Helgi D. Michelsen, Tjóðsavnið, The Faroe Islands National Museum, Faroe Islands Jacqui A. Mulville, School of History and Archaeology, Cardiff University, Cardiff, Wales, UK Anthony Newton, School of Geosciences, University of Edinburgh, Edinburgh, Scotland, UK • Editor Astrid E.J. Ogilvie, Institute of Arctic and Alpine Research, University of Colorado, Boulder, CO, USA Alexandra Sanmark, Center for Nordic Studies, University of the Highlands and Islands, Kirkwall, Orkney, UK J. Edward Schofield, Dept. of Geography & Environment, School of Geosciences, University of Aberdeen, Scotland Niall Sharples, Cardiff Universty, Cardiff, Wales, UK Ian A. Simpson, Department of Archaeology, Durham University,(Hon Research Fellow), Durham, UK Przemyslaw Urbanczyk, Institute of Archaeology and Ethnology, Warsaw, Poland Eileen Tisdall, Biological and Environmental Sciences, University of Stirling, Stirling, UK Chase Uy, Eagle Hill Institute, Steuben, ME • Production Editor Orri Vésteinsson, Institute of Archaeology, Reykjavík, Iceland Alex Woolf, University of St Andrews, St Andrews, Scotland, UK James Woollett, Université Laval Québec, QC, Canada Cover Photograph: Høgsetra, Rauma, Møre and Romsdal. Photograph © Kristoffer Dahle. “Skálholt Map” courtesy of The Royal Library, Copenhagen, Denmark Journal of the North Atlantic K. Dahle 2025 No. 46 1 2025 JOURNAL OF THE NORTH ATLANTIC 46:1–22 Local and Regional Variation in the Emergence of the Shieling System Kristoffer Dahle1,2,3* Abstract - Transhumance is known across the world. This paper focuses on the Scandinavian and North Atlantic shieling system and transhumant practices across the county of Møre and Romsdal, Norway. By analysing aggregated radiocarbon dates from extensive archaeological surveys, I have identified four phases of intensified transhumant activity through the Iron and Middle Ages, each separated by a period of decline. The main focus is on Phase 3 (ca. AD 700–1300). During this period the number of dates escalates, and we find the first evidence of dwellings in outland grazing areas unsuitable for cultivation, interpreted as shielings. The first sites emerged during the late Merovingian Period and early Viking Age in the inner parts of Sunnmøre and Romsdal, with similarities to fjord areas further south. The late Viking Age and Early Middle Ages witnessed further expansion across the central fjord regions, but there is no evidence of shielings along the coast and in the northernmost shire until ca. AD 1100. In the inner areas of Nordmøre, there is still no decisive evidence of shielings prior to the Black Death, and the main expansion seemed to have occurred in the subsequent Phase 4. Introduction The Scandinavian shieling system is one of many transhumant practices across the world. For more than a century the origins, development, and explanations behind this system have been subjects of ongoing debate. Most of the early research was based on historical, ethnographic, and onomastic material, primarily describing the practice and its local and regional variations during the last centuries, but also theorising upon its emergence (Beito 1949; Reinton 1955, 1957, 1961; Solheim 1952). The archaeological material, however, was still sparse and mainly based on stray finds and Iron Age graves (Hougen 1947). From ca. 1960–1990 extensive archaeological projects were carried out in mountain areas because of the large-scale construction of hydroelectric dams. These projects included several shieling sites (Bjørgo et al. 1992, Gustafsson 1982, Kvamme 1988, Randers and Kvamme 1992). At this point, radiocarbon dates and pollen analyses offered a far better proxy for understanding variations in time and space within these remote outland settlements. Throughout the last few decades the number of archaeological shieling sites and dates has increased significantly. Yet, despite some comparative studies (Austrheim et al. 2015, Stene 2015), there is still a lack of synthesis on how the shieling system evolved in various areas and regions across Norway, with further consequences for how we perceive its spread across the North Atlantic. According to Askeladden (2004), the Norwegian Database for Cultural Heritage, prehistoric and medieval shieling sites are highly concentrated along Western Norway (Fig. 1). This may have various explanations, whether due to differing demographies and landscapes, varying economies and regional specialisation, biases regarding taphonomy and formation processes, or simply the result of differing emphasis, priorities, strategies, and methods within current archaeological research and cultural heritage management (Dahle 2024). 1Department of Culture, Møre and Romsdal County, Molde, Norway. 2Department of Archaeology and Cultural History, Norwegian University of Science and Technology, Trondheim, Norway. 3Department of Historical and Classical Studies, Norwegian University of Science and Technology, Trondheim, Norway. *Corresponding author: kristoffer.dahle@mrfylke.no. Associate Editor: Ingrid Mainland, University of the Highlands and Islands. Journal of the North Atlantic K. Dahle 2025 No. 46 2 The most common Norwegian term for a shieling is a seter/sæter. Ivar Aasen has defined a sæter as “a remote pasture with a summer hut where milk from the cows is refined and brought back to the farm” (Aasen 1873, my translation), stressing the concept of dwellings, dairy production, and the dependent relations between farm and shieling. Goats may also have been kept for milking, but otherwise this characterises the Scandinavian shieling system (cfr. Larsson 2003). Transhumance, on the other hand, is used as a much wider term that does not necessarily involve fixed buildings or milking and dairy production. It has been defined as “an animal production and land use strategy involving the movement of herds of domestic herbivores between altitudinally differentiated and complementary seasonal pastures” (Greenfield 1999, cfr. Costello and Svensson 2018, Geddes 1983, Larsson 2003). In the following analysis, shieling landscapes will refer to outland pastures in mountains, forests, and other “marginal” agrarian landscapes that contained shielings in recent history, regardless their existence in the past. The main aim of this article is to promote a better understanding for transhumant strategies in Møre and Romsdal during the Iron and Middle Ages (500 BC–AD 1536), and more specifically for local and regional variation as to when shieling system emerged. By following a post-phenomenological approach (Idhe 1990, Verbeek 2016), I will explore how spatial and temporal models, offered by the technologies of radiocarbon dating, Geographic Information Systems (GIS), and statistical modelling may contribute to new chronologies and narratives about the past. Further, I will discuss some of the factors behind variation and change across various time scales and examine how the spatial and temporal distribution patterns align with varying explanatory models. Figure 1. Heat map of prehistoric and medieval shieling sites, protected by law and listed in the Norwegian Database for Cultural Heritage (Askeladden) prior to this study. Møre and Romsdal county is situated on the boundary between Western and Central Norway. Journal of the North Atlantic K. Dahle 2025 No. 46 3 The study is based on several minor development-led archaeological surveys carried out by regional authorities throughout the county over the past decades. In 2021–2023 the results from these investigations were supplemented by additional fieldwork, employing the same strategies and methods, to create a more representative archaeological record across the county. As part of my Ph.D., this initial survey has been followed up with the use of non-invasive methods (Dahle et al. 2025) and geoarchaeological and palynological investigations at selected sites. My main objective here, however, is to study local chronological variation based on the extensive survey material across Møre and Romsdal, ranging from south to north, and from the outer coast to the interior highlands. Situated on the boundary between Western and Central Norway, and with close connection across the mountains to Eastern Norway, this county may also serve as a “laboratory” for studying regional patterns at a national scale (Fig. 1). This may further contribute to a better understanding of the shielings in Western Norway within the wider Scandinavian and North Atlantic context. Perspectives on time and space Before embarking on the analysis of the archaeological material, I would like to present some general theoretical perspectives on time and space, and how these two overarching categories themselves are constituted by the archaeological material (cfr. Lucas 2021). Post-phenomenologists, such as Don Ihde (1980) and Peter-Paul Verbeek (2016), have emphasised the role of technology and technological mediation in the experience and understanding of time, space, and human-world-relations. Technology is not simply part of the world, but also matters to our human perception and understanding of the world. Indeed, this may be particularly relevant for human-past-relations, as they never can be embodied and perceived directly, but only through past remnants and the technologies that help visualise them. To move beyond our senses and ethnographic accounts, the finitude of the human body and our own experienced time and space, we need certain tools. At the very basis there are fundamental frameworks, such as history and chronology, and the idea of a linear and synchronised time (Lucas 2021). History is often confused with the past, simply as something that “has been”, but rather it should be regarded as present narratives, explanations, and perceptions of the past (Harding 2005, Lucas 2021). Chronology could thus be regarded as a “technological” extension the body, and a way of experiencing time (Crystal 2018). Artefacts and ecofacts may have survived from the past, but our histories and chronologies are always constructed, maintained, and changed in the present. From simply thinking about how past remnants are ordered and placed along an already existing timeline, as representations from past events and processes, Gavin Lucas (2021) has emphasised how the very same objects also constitute or make time through their own materiality. For the shielings, often lacking datable objects and features, radiometric methods play a crucial role in our understanding of their prehistory. These methods not only provide insight into the chronological framework but could also shed light on some of the historical processes. Regarding time, it is impossible to bypass the work of the French Annales scholar Fernand Braudel and his three levels of structural duration: longue durée, conjunctures and événements (Braudel 1980). Several scholars have tried to cope with the multi-temporality of change and causation and how various processes seem to operate at different scales (Bailey 1981, Crellin 2020, Harding 2005, Hull 2005, Lucas 2021). Rather than different durations or tempos inferring timescales that cannot be presupposed, Lucas tries to de-scale Journal of the North Atlantic K. Dahle 2025 No. 46 4 Braudel and replace his three-scale concepts in terms of the type of change they evoke: persistence as an active power for explaining stasis, recurrence for explaining cycles of growth and decline, and severance for explaining unique turning points and structural disjunctions (Lucas 2021). Whereas absolute chronologies (radiometric dating and statistical modelling; (Bronk Ramsey 2017, Crema and Bevan 2021) have helped us conceptualise development through time, distribution maps, GIS-technology and various spatial models have provided the means of understanding past phenomena in space. In Paris: Ville Invisible, Bruno Latour and Emilie Hermant (1998) argue that it is impossible to capture Paris at a glance as a whole and with all its structures. In order to transform the “plasma” into an ordered reality, they argue we have to move away from reality itself and rather look at the map. Through the efforts of cartographers, technicians, and civil servants, Paris has been made visible (de Vries 2016). The same applies to the study of the past; by generating and visualising spatial and temporal patterns GIS technology offers a way of indirect and temporally abstract imagination, creating new histories, narratives and explanations that were never accessible to the ones being there-in-the-world (Crystal 2018). Explanatory models and factors Early studies on the origin and development of the shieling system had a largely functionalistic approach. They simply originated due to the lack of pastures in the vicinity of the farm (Grude 1891, Kleiven 1938, Olafsen 1907, Østberg 1930). At the end of the 1920’s, the Swedish ethnographer John Frödin (1929) postulated a new theory. Rather than being developed from the farm, transhumance was primary. From moving between several equal sites, one of them was intensified and established as a permanent farm settlement. Lars Reinton (1961) also argued that the shieling system originated from a semi-nomadic system, but due to international similarities he argued that this stage was already surpassed at the time agriculture was introduced to Norway in the Neolithic. Early ethnographers and cultural geographers often favoured a longue durée-perspective (persistence), explaining variation as mainly due to natural factors such as topography and geology or to inherent cultural traits or evolutionally stages (Erixon 1918, Frödin 1929, Reinton 1961). Archaeologists like Johs. Bøe (1944) and Bjørn Hougen (1947) were inspired by the same discourse, but also included archaeological remnants and factors such as demography and climate (recurrence) when explaining the origins of the shieling system. Functionalistic perspectives had a revival during the era of processual archaeology, mainly explaining the origins as internal change due to resource pressure (Bjørgo et al. 1992, Gustafson 1982, Randers and Kvamme 1992). Rather than simply being determined by either cultural or natural factors, archaeologists in the post-processual era emphasised the role of human actors and strategies (Dahle 2005, Emanuelson et al. 2003, Svensson 1998) making unique choices at the local scale that could reconfigure both material and mental structures (severance). From this short review, it appears that the explanatory models have developed in line with the technological development and the resolution of the archaeological record. Since the introduction of radiometric methods (14C), most archaeologists have left Reinton’s theory of diffusion. Yet, there have been various opinions on origins and chronology due to definitions, emphasis, and regional variation (Dahle 2007, Larsson 2003). However, through quantitative studies of radiocarbon dates in time and space, we may be able to refine our chronological frameworks and detect local and regional variation in spatial distribution patterns. Further, by comparing related patterns we may identify some of the factors Journal of the North Atlantic K. Dahle 2025 No. 46 5 explaining variation and change. However, rather than claiming a mono-causal explanation (cfr. Crellin 2020), I will present various related factors at various scales and show how these may relate to the origin and the development of the shieling system. Material and Methods Throughout the last few decades thousands of radiocarbon samples have been analysed from various archaeological sites and contexts across the county of Møre and Romsdal, either as part of development-led archaeological surveys or other small-scale projects. Yet, only a small portion of the thousands of shielings that are known historically in the region (cfr. Reinton 1961) has been investigated. Prior to the onset of this study more than 30 shielings in Møre and Romsdal were categorised as automatically protected by law, 26 of which were older than the Black Death. In accordance with the patterns observed at the national level, these were concentrated in the southern half of the county, i.e., Western Norway, and mainly in the central fjord region—between the outer coast and the mountains further inland (Figs. 2 and 3). Obviously, the number of radiocarbon dating samples and analyses are relevant to the number of sites finally listed in the national database (Dahle 2024). Before my investigations 72 radiocarbon dates had been determined from possible transhumant sites in shieling landscapes across Møre and Romsdal. These were concentrated along the same fjord districts, contrasting with the very few sites which had been investigated or dated along the coast. In this central region most of the surveys had been conducted in Sunnmøre Figure 2. Heat map of historically known shielings in Møre and Romsdal, displaying their concentration in central fjord areas and equally distributed across the three historical shires (Sunnmøre, Romsdal and Nordmøre). The specific boundaries between the sub-areas are based on municipalities and later administrative divisions. Journal of the North Atlantic K. Dahle 2025 No. 46 6 and Romsdal in the southern part of the county. In Nordmøre, the northernmost region, only 8 sites had been dated. A notable concentration of sites (n = 18) and dates (n = 29) in the central parts of Romsdal relates to a former M.A. study (Dahle 2005). These are generally excluded from the analyses, to avoid geographical bias, but serve as a reference material and are displayed in the maps (Fig. 2, Table 1). Sampling and selection of sites New surveys undertaken in 2021–2022 tried to correct some of this bias by prioritising sites in coastal areas and in the northern region. As development-led surveys may result in Figure 3. Investigated shielings across Møre and Romsdal and sub-areas, displaying the natural topography in the background. Notice the alpine fjord landscape in the central fjord region of Sunnmøre, in contrast to the gentler terrains of Nordmøre. Table 1. Number of sites and samples from various parts of the county, both as part of developmentled archaeology and the current Ph.D. project. Extra numbers from the M.A. thesis (Dahle 2005) are in parentheses (cfr. Fig. 2). North–South axis East–West axis D-LED PHD SUM SITES D-LED PHD SUM SITES Nordmøre 10 14 24 21 Inner Ares 16 13 29 22 Romsdal 16 10 26 (29) 20 (15) Central Areas 28 14 42 (29) 29 (15) Sunnmøre 21 22 42 31 Outer Areas 3 19 21 21 SUM 47 46 93 (29) 72 (15) SUM 47 46 93 (29) 72 (15) Journal of the North Atlantic K. Dahle 2025 No. 46 7 more than 1 date from each site, the number of samples in a region does not necessarily reflect the density of sites being sampled. Hence, the county was divided into 2 different zones; both from coast to inland and from north to south, aiming at a minimum of 20 dated sites from each of these subregions. In addition, the sampling strategy tried to fill other blank spots in the maps and was generally seen in relation to the size of each region and the density of historically known shieling sites. In the Middle Ages the southern shire of Sunnmøre was divided into two administrative areas, both equal to the shires of Nordmøre and Romsdal, in accordance with the lething system and how many ships they could mobilise (cfr. Dahle and Hill 2024). The large number of sites from Sunnmøre and the central fjord region also corresponds with areas where summer farming was more widespread in recent centuries (Fig. 2). Taking this into account, the uneven numbers are justified as being more representative for the total population (Table 1). It should be added that the additional surveys to a somewhat greater degree aimed at the lowermost stratigraphical deposits, trying to establish the earliest phase of human activity. In contrast to development-led investigations the selection of sites was not limited by areas of construction. This selection of representative sites could thus be more purely based on the archaeological potential. This includes factors such as the age and centrality of the farm, historical accounts, names, and local traditions, as well as judgements based on LiDAR and orthophotos. Further, the walk-over survey, soil coring, and test pits were not restricted by the boundary of modern development areas. Some sites were subject to more thorough investigations in 2023 (Dahle et al. 2025), but still only a few samples for radiocarbon dating have been determined from each site. In addition to geographical location, the survey was also aimed at sampling sites and landscapes with varying topography at the micro-level. In particular, I have been looking at the topographic profile between the farm and the shieling i.e., the difference in elevation divided by the direct distance, to see if there are significant statistical variations in age due to local topography, rather than pure geographical location (cfr. Dahle 2005). A high ratio would typically indicate a shieling further up the valley (horizontal profile), whereas a low ratio indicates a site located on an edge or plateau above the farm by the fjord or the valley floor (vertical profile). Aggregated radiocarbon dates Datasets of radiocarbon dates have been analysed by using Summed Probability Distribution plots (SPDs) within the RCarbon Package (Crema and Bevan 2021) and calibrated by the IntCal20 calibration curve (Reimer et al. 2020). The use of SPDs in summing up the calibrated radiocarbon dates within a defined chronological range has several limitations (Williams 2012). To avoid biases of well dated sites or misinterpretations due to the calibration curve, I have followed methods described by Shennan et al. 2013, Timpson et al. 2014, and Crema and Bevan 2021, comparing my results with a theoretical null hypothesis fitted to the empirical data using an exponential growth model. To avoid biases for well documented sites, this includes the use of bins with a cut-off value at 100 years (cal.) at site level. By employing Monte Carlo simulation, a null model (grey envelope) is predicted, based on randomisation of the input variables (Crema and Bevan 2021). To generate a 95% confidence envelope this process is repeated 1000 times (Crema et al. 2016, Crema and Bevan 2021). Deviations from the null model exhibit periods with significantly higher (red) or lower (blue) number of dates. In recent years this methodology has been employed as a proxy for growth and decline in demography, settlement, and general level of activity (Arthur et al. 2024, Loftsgarden and Solheim 2023, Solheim and Iversen 2019). In this study, the same Journal of the North Atlantic K. Dahle 2025 No. 46 8 procedure has been employed on varying subsets of the regional database for radiocarbon dates, for instance relating to activity in the shieling landscape or to settlement and cultivation in the lowlands (Figs. 4 and 5). This enables comparisons between the development of the shieling system and general demographic and socio-economic tendencies, although some biases require further discussion. To assess local variation in the various datasets, either between the shires (N–S) or various topographic zones (E–W), I have further divided the data into smaller geographical subsets or performed permutation tests that generate simulation envelopes for each SPD by shuffling the dates belonging to the different subsets (Fig. 6). The test also takes different sample sizes into consideration, as subsets with fewer samples will produce a wider envelope reflecting greater uncertainty. GIS and spatial distribution GIS and the use of various spatial models can create conceptual visualisations of the past that were never embodied by anyone inhabiting the same past lifeworld (Crystal 2018). As translations or technological mediations, they enable us to see, interpret, and interact with past worlds, and to facilitate comparisons with other datasets (de Boer et al. 2021, Latour and Hermant 1998). I have chosen to separate the sites from each chronological phase or cycle, according to the aggregated radiocarbon dataset, as local variation and spatial development within each of these phases may have followed various paths. In the maps I have further subdivided these phases into smaller timespans to visualise local, chronological variation. In order to place the radiocarbon dates into such a box, however, the symbolisation is based on Figure 4. Activity in the shieling landscape, based on SPD and Hypothesis testing of all radiocarbon dates from transhumant sites in Møre and Romsdal between 2500 BP and the present day (n= 89), analysed by using Rcarbon package (Crema and Bevan 2021) developed for R programming language (RStudio). Significant periods of growth (red) or decline (blue) are marked by the bars. Journal of the North Atlantic K. Dahle 2025 No. 46 9 median dates cal. BC/AD (Figs. 7 and 8). The sites are also symbolised according to my interpretation of the sites. The separation between shielings and possible farms is based on archaeological remnants (such as lynchets or agricultural fields), as well as oral and written sources. Sites with discernible house foundations or substantial deposits (>5cm), both indicating a dwelling, are categorised as shielings whereas more ephemeral activity areas are simply referred to as possible shielings. The latter could be pure open air milking sites or herder camps, but the existence of dwellings cannot be ruled out. Results Over the last 30 years more than 1700 samples from various sites and contexts across the county of Møre and Romsdal have been analysed. As part of developmentled archaeological surveys, budgets only allow a few dates from each site. The dataset material thus covers extensive areas, but some areas, such as the shieling landscapes, are less affected by development pressure and thus less represented. The new investigations make up for some of this, and despite being a keyhole study with limited material (n = 93, excluding the aforementioned M.A. study), it still enables quantitative and comparative studies and the detection of various spatial and temporal patterns. Temporal patterns In addition to some very early dates the SPD displays two periods with increased activity in the Early Iron Age (Phase 1, 500–100 BC; Phase 2, AD 200–600), one in the Late Iron Figure 5. Settlement and activity in the lowlands. SPD and hypothesis testing against a 0-model based on all radiocarbon dates (except outland sites) in Møre and Romsdal between 2500 BP and the present day (n = 1034, M and R County database 1990–2023). Analysed by using Rcarbon package (Crema and Bevan 2021) developed for R programming language (RStudio). Journal of the North Atlantic K. Dahle 2025 No. 46 10 and Middle Ages (Phase 3, AD 700–1300), and then again following the Late Medieval Agrarian Crisis (Phase 4, AD 1500–1900). The most significant booms appear around AD 1100 and in the 19th-century. It is also worth noting the gaps in between the four phases: in the 7th-century, in the 14th- and 15th-centuries, and most notably in the centuries around the beginning of the Common Era (cfr. Bjørgo et al. 1992, Dahle and Busengdal 2022). This general pattern corresponds with the former M.A. study from Romsdal (Dahle 2005, 2007). Although new dates may contribute to elaborate on this chronology, some overall tendencies are considered significant and visible as deviations from the null model (Fig. 4). The exclusion of the earliest dates is mainly due to the aim and scope of this article, but the limited number of dates (n = 4) over this great time span would also make local comparisons highly uncertain. I acknowledge that mountain pastures may also have been grazed in the Bronze Age, and that there may have been transhumant systems preceding the period in focus. However, the origin and development of the shieling system, as defined above, must be seen in relation to permanent farm settlements and the more stable fieldand- meadow system (Emanuelsson et al. 2003, Øye 2002). Although structural changes may have occurred later (Grønnesby 2019), most researchers advocate more stable settlements and more intensified field systems from the Pre-Roman or Roman Iron Age (Meling 2023, Mjærum 2020, Ystgaard 2019). Based on the archaeological remains of buildings and more substantial deposits from Møre and Romsdal, it appears as if the shieling system, according to our definition above, emerged in what I have called Phase 3. By removing all outland sites from the total dataset of radiocarbon dates, e.g., shieling sites, hunting sites, and iron and tar production sites that may be due to specific economic booms, we are left with remnants of what may reflect general agrarian settlement and Figure 6. Permutation tests of all radiocarbon dates from transhumant sites in Møre and Romsdal, 0–2500 BP (n = 89) across regions (left) and topographical zones (right). Analysed by using Rcarbon package (Crema and Bevan 2020) developed for R programming language (RStudio). Nordmøre displays a deviating pattern with a notable decline in the shieling landscape throughout the Late Iron Age. In inner areas the early Viking Age expansion is followed by decline in the Middle Ages, whereas there is a significant growth in the central fjord region. Journal of the North Atlantic K. Dahle 2025 No. 46 11 cultivation in the Iron and Middle Ages (n = 1034, Fig. 5). Even here, the three periods of decline appear quite significant. Thus, it seems as if the development in the shieling landscape (Fig. 4) follows the general level of agrarian settlement and activity in the lowlands. This will be further discussed below. Further models and permutation tests, some of which will be presented in the discussion, are all based on the material presented above. By adding the M.A. study (n = 29) or removing samples from possible farm sites (n = 23), the total population of transhumant sites has also been altered without having any notable effect on the overall results. In the maps, the most definite shielings are displayed as a separate group and with a different symbology than possible farms and other transhumant sites. Spatial patterns Being situated in space, the sites, samples and phases may also be analysed geographically. Until 2020, the earliest shielings documented in the northern part of the county (Nordmøre) were from the 15th- or 16th-century. Would this spatial pattern still remain when adding new extensive, systematic, and targeted surveys? The new surveys across Nordmøre resulted in radiocarbon dates spanning across all four phases. Currently, 9 sites from the region have dates prior to the Black Death (Fig. 2). Therefore, it seems historical shielings across the county could potentially be prehistoric, underscoring the necessity of radiocarbon sampling to assert their age and legal status during archaeological surveys (Dahle 2024). Yet, there is still significant chronological variation between various parts of the county (Fig. 6). In the first phase (ca. 500–100 BC, Phase 1) there are few sites, but still a tendency of a gradual spread, from the south to the north (Figs. 6 and 7). The earliest site is located in Figure 7. Sites from Phase 1 with the earliest date sorted by median age, displaying a gradual spread from south to north. Journal of the North Atlantic K. Dahle 2025 No. 46 12 Sunnmøre and dated to the 5th-century BC. Whereas the expansion in Romsdal appears to occur in the 4th- and the early 3rd-century BC, the dates from Nordmøre are mainly from the 2nd-century BC. Most of these deposits are quite ephemeral and it is hard to tell if they represent single events or rather initial clearings for subsequent grazing that cannot be confirmed in the archaeological record. The temporal gap left in the shieling landscape between ca. 100 BC–AD 200 is noted across the county, and there are no signs of further outland expansion until the Late Roman Iron Age (cfr. Bjørgo et al. 1992, Dahle and Busengdal 2022). In Phase 2, ca AD 200–600, there are new signs of expansion in the inner and central fjord region (Fig. 6). Still, the sites mainly consist of cooking pits, hearths, clearances, and thin deposits. There are few indications of dwellings, the only exception perhaps being Hammarset at Sunnmøre, although interpreted as a possible farm. In the 7th-century, however, there is a new decline. We can also discern spatial variation in Phase 3. In this period, we are better able to distinguish between shielings and possible farms (Fig. 8). In the late Merovingian Period and early Viking Age the shielings are concentrated in the inner parts of Sunnmøre and Romsdal. Between AD 850 and 1100, however, the growth is rather concentrated along the central fjord areas, the topographic zone with the highest concentration of historically known shielings (cfr. Fig. 2). In the outermost coastal areas, the same expansion mainly occurs after AD 1100 with no significant variation between the three shires. In central areas of Nordmøre, however, we do not see any signs of activity until the High Middle Ages, and none of the sites from the inner parts of the shire have yet been dated to Phase 3. In all of the three shires there is a gap in the 14th-century before the new expansion in the Late Middle Ages and Early Modern Period (Phase 4). Figure 8. Sites from Phase 3 with the earliest date sorted by median age, displaying a gradual spread from south to north. Journal of the North Atlantic K. Dahle 2025 No. 46 13 Discussion In the following I will discuss the spatial and temporal patterns presented above in terms of local and regional variation when it comes to early transhumant practices and the emergence of the shieling system. In the last section I will return to some of the traditional narratives and explanations and discuss the origin and development of the shieling system in light of the presented patterns. Local and regional variation Starting with the similarities, as displayed by the charts, some temporal patterns are more or less congruent across the county (Fig. 6) and in line with the general demographic and economic development in the lowlands (Fig. 5). This applies both to the periods of expansion and growth in the shieling landscape that have prompted the four phases presented above, as well as the crises and declines that are separating them. Yet, as displayed in Figures 6–8, some deviating tendencies deserve further discussion; do these deviatons represent significant local variation as to when the shieling system emerged, and can they also indicate regional variation at a national level? At first, we must acknowledge the limited material that barely qualifies as mass data. In the absence of large-scale excavations with numerous radiocarbon dates, we also have to acknowledge that we are not able to measure duration i.e., for how long each site was occupied and the time span represented by each sample (cfr. Lucas 2021). By focusing on the earliest bottom dates (TPQ) we can estimate when the activity at a site began, but we still do not know much about later continuity or change. Secondly, we have to look closer at the interpretation of the sites that have been sampled. The separation between farm and shieling remains a static interpretation, not considering temporal dynamics and how sites could have changed function through time. Hence, deposits and dates from the two initial phases are hard to classify as either farm or shieling. The majority may simply be characterised as milking sites or other kinds of transhumant herder sites. Phase 3 is considered as the formative period for the shieling system in the region, as we know it, and as something separate from both farms and other transhumant herder sites. Accordingly, this phase will be the main focus for analysis. With these premises taken into consideration, gradual spread from the south to the north is discernible in both Phases 1 and 3. Rather than simply reflecting diffusion, this could represent varying levels of resource pressure in the three shires. In Phase 3 there is no expansion in Nordmøre until the Middle Ages (Figs. 6 and 8), contrary to the other shires. The earliest site is quite ephemeral, with deposits less than 3 cm, and may simply represent a transhumant herder site rather than a shieling. Further, 2 of the 3 sites from the Early and High Middle Ages are dated by pine (Pinus) samples, suggesting an even later origin. This is not the case for any of the Phase 3 shielings from in Sunnmøre and Romsdal. Rather, several dates are sampled from hearths in abandoned house foundations, and the shielings may thus have been established even earlier. In summary, the spatial variation between the shires displayed by the maps and charts in the results section (cfr. Figs. 6 and 8) may rather be understated. Considering the topographic axis (east–west), we may discern some variation from the coast to inland, from the hill lands along the Atlantic seaboard to the gradually narrower fjords and the more alpine landscape of the interior. In Phase 3 the early Viking Age dates are mainly from the inner parts of Sunnmøre and Romsdal. In the central and outer parts the expansion peaks in the Early and High Middle Ages, possibly displaying a westward spread. The most striking variation, however, is the fact that the expansion in the northern Journal of the North Atlantic K. Dahle 2025 No. 46 14 shire seems to start in the outermost areas. The three earliest sites are both situated along the coast, whereas the fourth is located in the central fjord region. There are still no sites found in the inner parts of Nordmøre until after the Black Death. Based on the limited number of samples, this study is still of a hypothetical character. Hence, it remains to be seen if these spatial and temporal patterns hold when faced with new material evidence in the future. As noted above, the majority of the sites in the fjord landscapes of Sunnmøre and Romsdal are dated as part of development-led archaeology. By contrast, the new sites along the outer coast and in the northern part have been more carefully selected on the basis of potential and representativeness (Table 1, Fig. 2). This makes it possible to prioritise some of the most interesting sites according to both location, appearance, and local traditions. The presented results are thus strengthened by the fact that the recently investigated sites are biased in the opposite direction. The relatively low population of historically known shieling sites along the outer coast (cfr. Fig. 2) further contributes to strengthen the overall pattern, as a higher percentage of the total population has been sampled. Moreover, the new investigations have had a higher focus on tracking the earliest deposit, whereas development-led investigations often have sampled hearths or floor levels potentially reflecting later events. Hence, the validity of the spatial patterns in Møre and Romsdal, and as displayed by the National database (Fig. 1), seems to be strengthened. Questions remain if local variation within the county also may express regional variation at the national level; if the “watershed” by the Romsdal fjord in Phase 3 also represents a significant boundary between Western and Central Norway, and if any of the two expansions into the interior in the early Viking age (Phase 3) or in the Late Middle Ages (Phase 4) have parallels in the upper valleys of Eastern Norway. The current state of knowledge suggests that most of the expansion in Eastern and Central Norway happened after the Black Death (Amundsen 2007, Austrheim et al. 2015, Weber et al. 2007), but we still need more comparable radiocarbon dates to conclude (cfr. Dahle 2024). Explaining spatial and temporal patterns In this last section, I will return to some of the explanatory models and factors that were presented in the introduction. Are we able to distinguish between any of these, based on the spatial and temporal patterns offered by the new surveys in Møre and Romsdal, and could this lead to revised narratives on the origin and development of the shieling system? By arguing that the distribution of sites in time and space may relate to various explanatory models, I will describe and discuss the spatial and temporal patterns presented above in relation to other patterns and to factors such as topography, demography, economy, and politics. Further, I propose that causality and change may work at different temporal scales (Bailey 1981, Braudel 1980, Crellin 2020, Hull 2005) and in relation to Lucas’ concepts of persistence, recurrence, and severance (Lucas 2021). Persistence: topography and landscape. Regional variation is commonly explained by different topographies and landscapes (Erixon 1918, Reinton 1961). Could this also explain the different patterns between Sunnmøre and Romsdal on the one hand and Nordmøre on the other? Indeed, the Romsdal fjord represents a “watershed”, not only between cultural entities, but also between different natural environments from the alpine fjord landscape of Western Norway to the forested hill landscape of Central Norway. The shire of Romsdal has always been “in between”. Even though it culturally and legally has been adhering to Central Norway, the topographic boundary is the fjord basin at the heart of the shire. Out of the 13 dated shielings in Romsdal from Phase 3, 11 are from the southern side of the fjord (Fig. 8). According to early theories (Grude 1891, Olafsen 1907), the shieling system originated as a response to grazing pressure. From topography, it could thus be deduced that the available pasJournal of the North Atlantic K. Dahle 2025 No. 46 15 tures surrounding the farms in the steeper landscape in the southern part of the county were less sufficient than further north and perhaps reserved for winter fodder (cfr. Eriksson 2023). One approach to assessing the significance of topography involves examining the correlation between the age of the investigated sites and their associated landscapes, as the topographic profile may infer something about the availability of cultivable land and pastures in the lowlands. Generally, this quotient is lower in the northern and outer parts, but there are local variations across the county. In order to isolate this topographic factor, the sites were carefully selected to represent the span between horizontal and vertical profiles across all the subregions (Fig. 9). In Phase 1 there is no significant variation. In Phases 2 and 3, however, the general tendency is that most of the shielings are established on farms with a vertical profile. Thus, contrary to what one may expect (Cabouret 1989, Grude 1891), the distance between farm and shieling is less decisive than the height difference, or rather the gradient. There are some exceptions, but the expansion in Phase 3 mainly relates to farms with a short and steep topographic profile, whereas most shielings in distant valleys are established in the subsequent Phase 4 (Fig. 9, cfr. Kleiven 1938). The relative proportion of sites with the two profile types within in the total population of the three shires may thus indicate that the potential for finding Viking Age and Early Medieval shielings is higher in the southernmost and more alpine part of the county. Topographic and environmental explanations have, for long, been criticised for being deterministic (Crellin 2020), but recent perspectives highlight the reciprocity in human- Figure 9. Across all shires the expansion of shielings in the Viking age and Medieval Period starts at farms with a vertical topographic profile, perhaps an indication of less access to nearby pastures by the home farm. Journal of the North Atlantic K. Dahle 2025 No. 46 16 environment relations (Eriksson 2023). We may say that they offer a longue durée or persisting framework for understanding conjunctures or recurrences and more groundbreaking événements or severences (Braudel 1980, Lucas 2021), due to factors such as demography, climate, economy and politics. Recurrence: Demography and climate. It has become fairly common to look at the aggregated distribution of radiocarbon dates across time and space, as a proxy for human demography, settlement, and activity, and often in relation to climatic changes (Gundersen et al. 2020, Loftsgarden and Solheim 2023, Solheim and Iversen 2019). From the results above, we can see how the aggregated dates from transhumant contexts correspond well with the general booms and busts of demographic and socio-economic development, represented by the dates from lowland areas (Fig. 5). The question arises if the local variation discernible in the shieling landscape is visible at a more general level: Could the various chronologies be due to the demographic variation between the shires? If we exclude all outland sites and mainly look at dates from settlement and cultivation in the lowlands (n = 1034), the majority of radiocarbon dates are sampled from Sunnmøre (n = 455), which also is the largest and most populated region. The number of samples in Romsdal is also fairly high (n = 376) in relation to its size and current population. This may both be due to a high development rate as well as its vicinity to the regional administrative centre. The lowest number of samples are from Nordmøre (n = 203). In order to look at variation in demography and lowland activity, I have compared the chronological development within each of these three shires (Fig. 10). Figure 10. Variation in demography, settlement, and lowland activity between the three shires. Hypothesis testing against a 0-model in all three regions based on the radiocarbon dates from settlement and cultivation in the lowlands from 2500 BP until present (n = 1034, M and R County database 1990–2023). Journal of the North Atlantic K. Dahle 2025 No. 46 17 The number of dates from the Early Iron Age are fairly similar across all the shires, following similar paths. The expansion in the Pre-Roman Iron Age is followed by a decline around 100 BC and a new rise in the period AD 200–600. The 6th-century crisis appears to have hit all three areas, but in the northern shire this decline may have had more long-lasting consequences. In Sunnmøre and Romsdal new growth appears already by the Merovingian- Viking Age transition. Sunnmøre reaches a peak by the early 9th-century, whereas Romsdal displays a boom in the 12th-century. In Nordmøre, there is no significant period of expansion within Phase 3 (Fig. 10). This may be due to a general lack of dates, but the Late Iron Ages still seems like a period of general decline. From the Early Middle Ages there is gradual growth until the 13th- and 14th-century, corresponding well with the radiocarbon dates from the shieling landscape. Various spatial patterns may be caused by different methods and foci among archaeologists conducting the surveys in various parts of the county. Yet, combined with the topographic variation mentioned above, relative variations in demography, settlement, and agrarian activity in the lowlands may explain some of the local and regional variation as to when the shielings were established. Thus, from simply suggesting that resource pressure may have been a prime factor in the development of the shieling system (Cabouret 1989, Grude 1891, Olafsen 1907), we are now able to date and understand these processes more properly. Severances: Economic and political strategies. When explaining historical watersheds (severences), archaeologists often tend to focus on negative events such as plagues, climatic events, and natural disasters (Arthur et al. 2024, Gundersen 2020, Solheim and Iversen 2019). Rather, I have looked at some unique positive events. During Phase 3 four different (yet related) events may have contributed to favour, formalise, and manifest the shieling system across the county; unification, introduction of new technology, commoditisation, and literalisation. The unification of the kingdom may have erased regional boundaries but could at the same time have contributed to regional specialisation by strengthening transregional trade. Resource strategies were no longer confined to regional chiefdoms but subject to the same socio-political hierarchy and infrastructure. This did not only lead to a flow of successful economic strategies across the kingdom, but also better ways of exploiting comparative advantages (Dahle and Hill 2024). Between the Viking Age and the High Middle Ages regional variations were reinforced, exemplified by more intensive fishing along the coast, large-scale reindeer trapping in the high mountains, and massive production of bog iron in the interior valleys (Mikkelsen 1994, Narmo 1996, Wickler and Narmo 2014). Similarly, the exploitation of mountain pastures may have been considered a comparative advantage and a way of maximising surplus production in the fjords of Western Norway (Dahle and Hill 2024). The expansion of shielings could also have been facilitated by the introduction of the plunge churn. Based on a combination of archaeological, artistic and linguistic evidence across Europe, as well as the absence of this, Janken Myrdal (1988) has suggested this must have happened ca. AD 1000. This technological innovation may have acted as a catalyst, favouring outland expansion and increased butter production. After a minor decline from ca. AD 850, this may correlate with the rapid growth from the late 10th-century. Dairy products may have served as tributes and gifts in earlier phases (Ystgaard 2023), but due to contemporary processes of feudalisation and urbanisation, and the land rent system replacing the personal bands between patrons and clients, butter may soon have become subject to commoditisation. The permutation tests also suggest that the growth from the late 10th-century coincides with new centres of gravity (Figs. 4 and 6). The new Journal of the North Atlantic K. Dahle 2025 No. 46 18 shielings in central and outer fjord areas may thus have become more intimately related to dairying. In a recent article Kristoffer Dahle and David Hill (2024) have demonstrated the spatial and temporal correlation between urbanisation and the expansion of shielings in the uplands of Møre and Romsdal. From mainly being political and administrative strongholds, early medieval towns gradually developed into important markets for both consumption and trade. By the end of Phase 3 butter had thus become one of the most important export products as well as a main currency for landownership (Dahle and Hill 2024, Hill 2010, Skre 2017). The last turning point relevant for the development of the shieling system is the literalisation and the introduction of a written language in the 12th- and 13th-centuries. From simply being an opportunistic economic strategy in time and space, the shieling system became manifested and standardised by means of written words. Specific rights and obligations became constituted in laws, cadastres, and diplomas, as having existed since time immemorial. Hence, these historical sources perfectly exemplify how our perceptions of long-term patterns (persistence) may have been reconfigured by the writing of these documents as unique man-made events (severences). Conclusion By supplementing the results from development-led surveys of shielings sites in Møre and Romsdal with new investigations, and thus contributing to a more representative archaeological record, I have demonstrated local variation in transhumant practices and the emergence of the shieling system across the county. Based on clusters in the aggregated radiocarbon dates, 4 separate phases have been suggested—separated by periods of general decline. Both in the Pre-Roman Iron Age (Phase 1) and in the Late Roman Iron Age and Migration Period (Phase 2), there is evidence of increased activity in the shieling landscapes of the inner and central fjord region. As we still do not have evidence of dwellings, these sites are interpreted as transhumant herder or milking sites. Due to the sparse archaeological record, regarding both quantity and quality, the interpretation of spatial and temporal patterns during these two phases is still considered highly hypothetical. The main emphasis, however, is put on Phase 3, ca. AD 700–1300. In this period, the number of dates escalates, and we also have the first evidence for dwellings in grazing areas that are clearly not suitable for cultivation and permanent farm settlement. The first sites start to appear by the late Merovingian Period and early Viking Age in the inner parts of Sunnmøre and Romsdal, displaying similarities to sites and areas further south. By the late Viking and Early Middle Ages, there is a rapid increase in the number of sites across the central fjord regions, whereas the main expansion to the coast and to the northern shire appears to have occurred in the High Middle Ages. In the innermost parts of Nordmøre, however, we still do not have any evidence of definite shieling sites until after the Black Death and what we have characterised as a new expansion (Phase 4). As mentioned in the introduction, these spatial and temporal patterns are mainly representative for the county of Møre and Romsdal. As a hypothesis, however, it is suggested that they may express regional variation at a higher level, contributing to explain the uneven records of shieling sites between Western Norway, on the one hand, and Eastern and Central Norway, on the other (Fig. 1). In general, the origin and development of the shieling system is best explained as a result of pressure on grazing land surrounding the farm, as scholars have argued since the Journal of the North Atlantic K. Dahle 2025 No. 46 19 late 19th-century. There are however several factors involved in explaining emergence and geographical variation, including both pull and push factors that are working at varying spatial and temporal scales. Rather than singular causation, I have tried to describe various correlating patterns and events that may have had an effect on variation, stability, and change in the shieling landscape. Acknowledgements This study has been financed by Møre and Romsdal County Council and The Norwegian Research Council (grant no. 327209), and is based on unpublished material provided by county archaeologists (cfr. Supplemental tables, available online at https://doi.org/10.18710/I9C0ND). I would like to thank my supervisor Arne Stamnes, James Barrett, and Aaron Johnston for commenting on my draft. Literature Cited Aasen, I. 1873. Norsk ordbog: Med dansk forklaring. P.T. Mallings Boghandel, Christiania, Norway. 976 pp. Amundsen, T. 2007. Undersøkelser i seterområder: Rødseter og Deset Østseter. Pp. 137–286, In T. Amundsen (Ed.). Elgfangst og bosetning i Gråfjellområdet. Varia 64. Gråfjell-prosjektet (Vol. 2). University of Oslo, Oslo, Norway. 370 pp. Arthur, F., K. Hatlestad, K.-J.Lindholm, K. Loftsgarden, D. Löwenborg, S. Solheim et al. 2024. The impact of volcanism on Scandinavian climate and human societies during the Holocene: Insights into the Fimbulwinter eruptions (536/540 AD). The Holocene 34:619–633. Askeladden. 2004. The Norwegian database for cultural heritage. Available online at askeladden. ra.no. Accessed 16 October 2023. Austrheim, G., P.J.E. Sjøgren., K. Stene, K.L. Hjelle, J. Rosvold, and A.M. Tretvik. 2015. Fjellets kulturlandskapshistorie. Fjellets kulturlandskap. Arealbruk og landskap gjennom flere tusen år. DKNVS Skrifter 3. Museumsforlaget, Trondheim, Norway. 207 pp. Bailey, G. 1981. Concepts, time-scales, and explanations in economic prehistory. Pp. 97–118. In A. Sheridan and G. Bailey (Eds.) Economic Archaeology: Towards an integration of Ecological and Social Approaches. British Archaeological Reports International Series 96. Oxford, UK. 303 pp. Beito, O.T. 1949. Norske sæternamn. Instituttet for Sammenlignende Kulturforskning, Oslo, Norway. 339 pp. Bjørgo, T., S. Kristoffersen, and C. Prescott. 1992. Arkeologiske undersøkelser i Nyset- Steggjevassdragene 1981–87. Arkeologiske rapporter 16. University of Bergen, Bergen, Norway. 327 pp. Bøe, J. 1944. Høgfjellsristningar i Luster i Sogn. Viking 8:169–182. Braudel. F. 1980. On History. University of Chicago Press, Chicago, USA. 226 pp. Bronk Ramsey, C. 2017. Methods for summarizing radiocarbon datasets. Radiocarbon 59:1809–1833. Cabouret, M. 1989. Esquisse d’une chronologie des etapes de la formation de la vie pastorale en Norvège. Historisk tidsskrift 68:28–37. Costello, E., and E. Svensson (Eds.). 2018. Historical Archaeologies of Transhumance across Europe. Themes in Contemporary Archaeology (Vol. 6). Routledge, London/New York, UK/USA. 282 pp. Crellin, R. 2020. Change and Archaeology. Routledge, London, UK. 266 pp. Crema, E.R., J. Habu, K. Kobayashi, and M. Madella. 2016. Summed probability distribution of 14C dates suggests regional divergences in the population dynamics of the Jomon period in Eastern Japan. PLoS ONE 11, e0154809. Crema, E.R., and A. Bevan. 2021. Inference from large sets of radiocarbon dates: Software and methods. Radiocarbon 63:23–39. Crystal, D. 2018. Postphenomenology and archaeology: Towards a temporal methodology. Time and Mind. The Journal of Archaeology, Consciousness and Culture 11:297–304. Dahle, K. 2005. Norm og praksis. Bruk og forvaltning av utmark i midtre Romsdal i et langtidsperspektiv. M.A. Thesis. University of Bergen, Bergen, Norway. Journal of the North Atlantic K. Dahle 2025 No. 46 20 Dahle, K. 2007. Støl, sel og seterstøl. Seterdrift gjennom 2000 år i Romsdal. Heimen 44:343–358. Dahle, K. 2024. Tilblivelsen av setre som automatisk fredete kulturminner. Kvantitative og kvalitative tilnærminger til kulturminneforvaltningens håndtering av setre som arkeologiske kulturmiljø. Viking 89:29–59. Dahle, K., and S. Busengdal. 2022. Living on the edge. Patterns of agrarian settlement and land-use in the fjord landscape of Inner Sunnmøre. Pp. 25–44. In D.E. Mooney, L. Guðmundsdóttir, B. Dahle, H. Roberts, and M. Ramstad (Eds.). Expanding Horizons. Settlement Patterns and Outfield Land Use in the Norse North Atlantic. University of Bergen Archaeological Series (Vol. 13). University of Bergen, Bergen, Norway. 230 pp. Dahle, K., and D. Hill. 2024. Smjǫr-fra buðdeigja til byjar-monnum. Urbanisering og utmarksekspansjon i vikingtid og middelalder. Pp. 150–166. In K. Dahle, M. Langnes, A. Nytun, and B. Rundberget (Eds). Kjøpstaden på Veøya. Middelalderbyen som forsvant. Vitark (Vol. 12). Museumsforlaget, Trondheim, Norway. 186 pp. Dahle, K., D.-Ø.E. Solem, M.M. Gran, A.A. Stamnes. 2025. Making the invisible visible: The applicability and potential of non-invasive methods in pastoral mountain landscapes—New results from aerial surveys and geophysical prospection at Shielings Across Møre and Romsdal, Norway. Remote Sensing 17:1281. https://doi.org/10.3390/rs17071281 de Boer, B., H. te Molder, and P.P. Verbeek. 2021. Understanding science-in-the-making by letting scientific instruments speak: From semiotics to postphenomenology. Social Studies of Science 51:392–413. de Vries, G. 2016. Bruno Latour. Polity Press, Cambridge, UK. 224 pp. Emanuelsson, M., A. Johansson, S. Nilson, S. Pettersson, and E. Svensson. 2003. Settlement, Shieling and Landscape. The Local History of a Forest Hamlet. Lund Studies in Medieval Archaeology 32. Almqvist and Wiksell International, Stockholm, Sweden. 154 pp. Eriksson, O. 2023. Domesticated forest landscapes in Central Scandinavia during the Iron Age: Resource colonization for iron and subsistence strategies based on livestock. Journal of Field Archaeology 48:315–326. Erixon, S. 1918. Bebyggelsesundersøkingar. Øfversikt. Periodiska bebyggelsestyper. Fäbodvesen. Fataburen 1918:21–57. Frödin, J. 1929. Om fäbodbebyggelsens utbredning och olika typer i Europa. Meddelanden frå Uppsala Universitets Geografisk Institution, ser. A, no. 1. Svenska Geografins Årbok 1929. Geddes, D.S. 1983. Neolithic transhumance in the Mediterranean Pyrenees. World Archaeology 15:51–66. Greenfield, H., 1999. The advent of transhumant pastoralism in the temperate southeast Europe: A zooarchaeological perspective from the Central Balkans. Pp. 15–36, In L. Bartosiewicz and H.J. Greenfield (Eds.). Transhumant Pastoralism in Southern Europe. Recent Perspectives from Archaeology, History and Ethnology. Archaeolingua Series Minor 11. Amulett’98 KFT, Budapest, Hungary. 247 pp. Grønnesby, G. 2019. “…en pludselig og stærk omveltning»? Eldre jernalder og overgangen til yngre jernalder i Trøndelag. Praksis og overregionale nettverk. Ph.D. Thesis. NTNU University Museum. Trondheim, Norway. Grude, J. 1891. Stølsdriften på Vestlandet. Steenberg, Stavanger, Norway. 174 pp. Gundersen, I.M, C.L. Rødsrud, and J. Post Melby. 2020. Kokegroper som massemateriale. Regional variasjon i en kulturhistorisk brytningstid. Pp. 187–199, In C.L. Rødsrud and A. Mjærum (Eds.). Ingen vei utenom: Arkeologiske undersøkelser i forbindelse med etablering av ny rv. 3/25 i Løten og Elverum kommuner, Innlandet. Cappelen Damm Akademisk, Oslo, Norway. 354 pp. Gustafsson, L. 1982. Månavassdraget. Unpublished report. Vitenskapsmuseet, Trondheim, Norway. Harding, J. 2005. Rethinking the Great Divide: Long-term structural history and the temporality of the event. Norwegian Archaeological Review 38:88–101. Hill, D. 2010. Medieval Towns and the Rural Economy in Eastern Norway: Central Place Theory, Settlement and Taxation AD 1000–AD 1350. VDM Verlag, Saarbrücken, Germany. 143 pp. Hougen, B. 1947. Fra seter til gård. Studier i norsk bosetningshistorie. Norsk Arkeologisk Selskap, Oslo, Norway. 364 pp. Hull, K.L. 2005. Process, perception, and practice. Time perspectivism in Yosemite native demography. Journal of Anthropological Archaeology 24:354–377. Journal of the North Atlantic K. Dahle 2025 No. 46 21 Ihde, D. 1990. Technology and the Lifeworld. From Garden to Earth. Indiana University Press, Indiana, USA. 226 pp. Kleiven, I. 1938. Sæterbruke på Vaagaa. Norsk Geografisk Tidsskrift 7:193–202. Kvamme, M. 1988. Pollen analytical studies of mountain summer farming in western Norway. Pp. 349–367, In H.H. Birks, H.J.B. Birks, P.E. Kaland and D. Moe (Eds.). The Cultural Landscape— Past, Present and Future. Cambridge University Press, Cambridge, UK. 540 pp. Larsson, J. 2003. Vad är en fäbod. Ett försök att definiera ett begrepp. Bebyggelseshistorisk tidsskrift 45:131–142. Latour, B., and E. Hermant. 1998. Paris ville invisible. Les Emëcheurs de penser en rond. La Découverte, Paris, France. 160 pp. Loftsgarden, K., and S. Solheim. 2023. Uncovering population dynamics in southeast Norway from 1300 BC to 1500 AD using summed radiocarbon probability distributions. Pp. 99–112, In M. Ødegaard and I. Ystgaard (Eds.). Complexity and Dynamics: Settlement and Landscape from the Iron Age and Medieval Period in the Nordic Countries. Sidestone Press, Leiden, Netherlands. 224 pp. Lucas, G. 2021. Making Time. The Archaeology of Time Revisited. Routledge, London/New York, UK/USA. 154 pp. Meling, T. 2023. Settlement structure and landscape use in Southwest Norway in the last millennium BC. Pp. 41–52, In M. Ødegaard and I. Ystgaard (Eds.). Complexity and dynamics: Settlement and Landscape from the Iron Age and Medieval Period in the Nordic Countries. Sidestone Press, Leiden, Netherlands. 224 pp. Mikkelsen, E. 1994. Fangstprodukter i vikingtidens og middelalderens økonomi. Universitetets Oldsaksamling skrifter - ny rekke nr 18. University of Oslo, Oslo, Norway. 218 pp. Mjærum, A. 2020. The emergence of mixed farming in eastern Norway. Agricultural History Review 68:1–21. Myrdal, J. 1988. The plunge churn from Ireland to Tibet. Pp. 111–137, In A. Fenton and J. Myrdal (Eds.). Food and Drink and Travelling Accessories. Essays in Honour of Gösta Berg. John Donald Publishers, Ltd., Edinburg, UK. 216 pp. Narmo, L.E. 1996. Jernvinna i Valdres og Gausdal: et fragment av middelalderens økonomi. Varia 38. University of Oslo, Oslo, Norway. 239 pp. Olafsen, O. 1910. Ullensvang: en historisk-topografisk-statistisk Beskrivelse af Ullensvangs Herred. Nilsen and søns Bogtr. and Reprodukstionsanst, Bergen, Norway. 628 pp. Østberg, K. 1930. Norsk Bonderett 6. Seterbruket i Norge Med et Tillegg om årets—især landbrukets— merkedager. Webergs boktrykkeri A/S, Oslo, Norway. 120 pp. Øye, I. (Ed.). 2002. Vestlandsgården—Fire Arkeologiske Undersøkelser. Havrå-Grinde-Lee-Ormelid. Arkeologiske Avhandlinger og Rapporter 8. University of Bergen, Bergen, Norway. 80 pp. Randers, K., and M. Kvamme. 1992. Breheimenundersøkelsene 1982–1984 II: Stølsområdene. Arkeologiske Rapporter 15. University of Bergen, Bergen, Norway. 109 pp. Reimer, P.J., W.E.N. Austin, E. Bard, A. Bayliss, P.G. Blackwell, C. Bronk Ramsey et al. 2020. The IntCal20 Northern Hemisphere Radiocarbon Age Calibration Curve (0–55 cal kBP). Radiocarbon 62:725–757. Reinton, L. 1955. Sæterbruket i Noreg I. Sætertyper og driftsformer. Aschehoug and Co., Oslo, Norway. 481 pp. Reinton, L. 1957. Sæterbruket i Noreg II. Anna arbeid på sætra i haustingsbruket og matnøytsla elles. Aschehoug and Co., Oslo, Norway. 286 pp. Reinton, L. 1961. Sæterbruket i Noreg III. Aschehoug and Co., Oslo, Norway. 594 pp. Shennan, S., S.S. Downey, A. Timpson, K. Edinborough, S. College, T. Kerig et al. 2013. Regional population collapse followed initial agriculture booms in mid-Holocene Europe. Nature Communications 4:2486. Skre, D. 2017. Viking-Age Economic Transformations. The West-Scandinavian case. Pp. 1–27, In Z. Glørstad and K. Loftsgarden (Eds.). Viking-Age Transformations. Trade, Craft and Resources in Western Scandinavia. Routledge, London, UK. 302 pp. Solheim, S. 1952. Norsk sætertradisjon. Aschehoug and Co., Oslo, Norway. 708 pp. Journal of the North Atlantic K. Dahle 2025 No. 46 22 Solheim, S. and F. Iversen. 2019. The Mid-6th-century crises and their impacts on human activity and settlements in south-eastern Norway. Pp. 423–434, In N. Brady and C. Theune (Eds.). Settlement Change across Medieval Europe. Old Paradigms and New Vistas. Ruralia (Vol. 12). Sidestone Press, Leiden, Netherlands. 446 pp. Stene, K. 2015. Resource exploitation and settlement in mountain areas in Southern Norway during the Early Iron Age—an altered perception of landscape and landscape use? Pp. 191–202, In S. Indrelid, K.L. Hjelle, and K. Stene. (Eds.). Exploitation of Outfield Resources-Joint Research at the University Museums of Norway. University of Bergen, Bergen, Norway. 255 pp. Svensson, E. 1998. Människor i utmark. Lund Studies in Medieval Archaeology 21. Almqvist and Wiksell International, Stockholm, Sweden. 221 pp. Timpson. A., S. Colledge, E. Crema, K. Edinborough, T. Kerig, K. Manning et al. 2014. Reconstructing regional population fluctuations in the European Neolithic using radiocarbon dates: A new casestudy using an improved method. Journal of Archaeological Science 52:549–557. Verbeek, P.-P. 2016. Toward a theory of technological mediation: A Program for Postphenomenological Research. Pp.189–204, In J.K.B. Friis and R.C. Crease, Technoscience and Postphenomenology: The Manhattan Papers. Lexington Books, London, UK. 259 pp. Weber, B. et al. 2007. Vesle hjerkinn-Kongens Gård og Sælehus. Norske Oldfunn (Vol. 21). Universitetets Kulturhistoriske Museer, Oslo, Norway. 221 pp. Wickler, S., and L.E. Narmo. 2014. Tracing the development of fishing settlement from the Iron Age to the Modern Period in Northern Norway: A case study from Borgvær in the Lofoten Islands. Journal of Island and Coastal Archaeology 9:72–87. Williams, A.N. 2012. The use of summed radiocarbon probability distributions in archaeology: A review of methods. Journal of Archaeological Science 39:578–589. Ystgaard, I. 2019. Environment and Settlement: Ørland 600 BC–AD 1250. Archaeological Excavations at Vik, Ørland Main Air Base. Cappelen Damm Akademisk, Oslo, Norway. 426 pp. Ystgaard, I. 2023. Activities and community organization in Roman Iron Age Vik, Ørland, Central Norway. Pp. 143–156, In M. Ødegaard and I. Ystgaard (Eds.). Complexity and Dynamics: Settlement and Landscape from the Iron Age and Medieval Period in the Nordic Countries. Sidestone Press, Leiden, Netherlands. 224 pp.