From Coastal Sites to Elevated Hinterland Locations in the Mesolithic – Discussing Human–Woodland Interaction in the Oslo Fjord Region, Southeast Norway

5.1 General Remarks

All of the nine sites (see Table A1) have yielded solely lithic artefacts, often in high numbers (between 735 and 16,227 items). These were catalogued, following the same system, in the museum database (Gjenstandsbase, Museum of Cultural History, University of Oslo) in connection with the post-excavation report writing and were typologically and technologically dated; in most cases, the material was considered to be homogeneous (for references see Table A1). The occupation phases of the sites were dated to the EM with five sites (Table A1, nos. 1–5), the MM with three sites (nos. 6, 7, 9) and the LM with one site (no. 8). This covers the timespan between 9200 and 6400 calBC. Combining artefact chronology and shoreline reconstruction showed that the sites were coast based at the time of this artefact material, often located on smaller islands in skerries (Table 1, nos. 2–7; see references there). Only in some cases, it was stated that the material potentially might have very few later – Mesolithic – elements (see below).

All of these sites have also yielded one or more structures, mainly hearths, which were radiocarbon dated to later periods, hundreds, sometimes thousands of years later than the use of the site suggested by artefact material. Altogether 16 radiocarbon dates from the later Mesolithic exist from the nine sites: dates from eight sites pertain to the LM, between 6400 and 4400 BC (Table A2, nos. 2–9), and one to the MM (Table A2, no. 1). Even though an old wood effect (see e.g. Loftsgården, Rundberget, Larsen, & Mikkelsen, 2013; Persson, 2014) might have affected some of the datings, this would, in a Mesolithic context where one deals with rough dating frames, not change the general conclusion about that the absolute dating is later than the artefact material.

5.2 The Character of the Locations at the Times of the Later Mesolithic Radiocarbon Dates

The analysis showed that at the times of the later Mesolithic radiocarbon dates, the locations were placed on top of or only a few metres away from precipices, the former deep seabed with steep sides now turned into steep slopes along long, deep and narrow valleys which still today are typical for the region (Table 1). The topographic modelling further showed^[2] that they were placed

1. between approximately 25 and 87 m a.s.l. (Table 1 and Figure 3);

2. at or very close to potentially excellent viewpoints. Most locations were then overlooking one (Table 1, nos. 2, 5, 7) or more valleys (nos. 3, 4, 8), sometimes with a lake or watercourse, or a sound (no. 1). One slightly elevated location would have overlooked a valley with a little fjord (no. 9), or two slopes from a col (no. 6) (Figure 4);

3. The distance to the contemporary shoreline was now between approximately 105 and 3,000 m (Table 1). Three locations were around 300 m (Table 1, nos. 6, 7, 9), one ca. 500 m (no. 8) and four more than 1,000 m away from the shoreline (nos. 2–5) (Figure 5). They were overlooking valleys which in most of the cases lead directly from the coast into the hinterland, and surely were important communication corridors for humans (on foot/by boat) and animals. Some of these valleys are very marked up to this day (nos. 2, 3, 4, 7, 9).

5.3 General Context of the Samples on the Sites

For three of the sites, more than one later Mesolithic date exists (Table A2, nos. 1, 3, 4), either from one and the same structure (nos. 1, 4) or from two structures (no. 3). In two cases, the only hearth on the site (nos. 1, 9) gave a later Mesolithic radiocarbon date. On six of the sites (nos. 2, 4, 5, 6, 7, 8), there were other hearths which partly were not dated/could not be dated, but some of them yielded even younger radiocarbon dates (nos. 2, 6, 7), in two cases from the same structure as one of the later Mesolithic radiocarbon dates (nos. 4, 5). In these two cases (nos. 4, 5), the hearths are interpreted as belonging to the EM as the artefact material, due to their character and sometimes their relation to the artefact distribution.

5.4 The Quality of the Sample Material for Radiocarbon Dating

Plants which extract carbon from the atmosphere, such as hazelnut shell and wood, are reliable sources for radiocarbon dating (Persson, 2014). Two of the datings are conducted on charred hazelnut shells (Table A2, nos. 1 (1×), 4 (1×)), and the others are on charcoal including the following tree species: pinus/pine (eight times: nos. 1 (1×), 2 (1×), 3 (4×), 4 (1×), 7 (1×)), salix/willow (three times: nos. 1 (2×), 6 (1×) and betula/birch (two times: nos. 5 (1×), 8 (1×)), not recorded (one time: no. 9). An encompassing reconstruction of forest fires in historic times in the region showed that these were mostly of low/moderate severity and that pine trees as the species that mainly were affected by forest fires usually survived a fire by repairing itself leaving scars in the tree rings (Storaunet, Rolstad, Toeneiet, & Blanck, 2013). Only three of the sites have yielded later Mesolithic datings exclusively on pine (nos. 2, 3, 7), while two sites had several datings on different materials, amongst them pine, hazelnut shell and salix (nos. 1, 4). Five samples from three sites have yielded datings only on deciduous tree species (nos. 5, 6, 8). Deciduous tree contains much more water than coniferous trees and are thus lesser prone to forest fires.

5.5 The Quality of the Direct Archaeological Context of the Radiocarbon Sample

The quality of the archaeological context of a sample is the decisive factor for assessing if radiocarbon dates can point to human activity (e.g. Loftsgården et al., 2013). Most of the radiocarbon dates discussed here stem from layers with charcoal and burnt stones deposited in built hearths (Table A2, nos. 1, 3, 4, 5, 7, 8), some of them with kerb stones (nos. 4, 5). In some cases, they are rather monumental in extension and depths (nos. 3, 4, 5). The two hearths at Dørdal (no. 3), both with large amounts of burnt stones (34 and 150 kg), have yielded two later Mesolithic dates each and no other absolute dates. From the hearth at Anvik (no. 1), altogether four later Mesolithic dates exist (as only absolute dates), pointing to minimum two incidents which left (different types of) charred material. The two hearths with dates from several periods are Pauler 2 (no. 4), where the later Mesolithic radiocarbon dates stem from the same part of the hearth. At Bakke (no. 5), it was observed that the burnt gravel in the hearth might witness of long-running activities of fire (Nyland & Amundsen, 2012, p. 172). Structure A 15 at Pauler 1 (no. 2), which was characterized by charcoal, sooty earth, and some burnt stones, was under excavation dismissed as residue after a mouldering tree which was burnt at its outsides (Schaller Åhrberg, 2012, pp. 56–57). This structure is only published with some unclear photographs; therefore, this interpretation is difficult to substantiate. More doubtful however seems the context at Hegna Vest 1 (no. 6), where artefact material dated to the MM was deposited in the same, deep pit as the charcoal sample that was dated to the Late Mesolithic.

5.6 Find Material Related to the Later Mesolithic Radiocarbon Dates

In case of four of the sites, the excavators do not rule out that there are (minor amounts of) artefacts which might be deposited at the time of the later Mesolithic radiocarbon datings (Table A2, nos. 1, 7, 8, 9); these are however not specified. A restudy of the artefact material from these sites to identify potential later artefacts is desirable but would clearly go beyond the scope of this article.

5.7 Looking Closer at Two Examples

Pauler 2 (no. 4), was, on the backdrop of lithic artefact typology/technology (3708 lithic artefacts) and shoreline dating dated to 9150–8850 cal BC, interpreted as traces of one, maybe to visits within this period (Nyland, 2012). The site was shore based on a western sloping terrace towards a little bay, protruding into the southern coast of an island (Nyland, 2012, Figure 2) (see Figure 6a). Find concentrations indicated four activity areas, together with three stone accumulations and three hearths. One of the hearths (S1) yielded two overlapping LM radiocarbon dates, dated on hazelnut and on pine (Beta-234403: 6910 ± 40BP, 5880–5720 calBC and Beta-234404: 6990 ± 40BP, 5980–5760 calBC) from the same context (layer 2 – under the top layer). The circular hearth, 1.3 m × 1.1 m in diameter and a few centimetres deep, with kerb stones and a packing of burnt stones, was placed centrally on the site in an area with an artefact concentration. This and the structure’s similarity to hearths from two other EM sites strengthened an interpretation as an EM hearth; the later radiocarbon dates were interpreted as impact of later forest fires (Nyland, 2012). However, the context and the two later Mesolithic radiocarbon dates, one on pine and one on hazelnut shell, which stem from the same part of the hearth can indicate an LM (re-?)use of the hearth. The topographic situation at the time of these later Mesolithic dates, based on the shoreline curve from Sørensen et al. (2014), shows the significant changes which the area underwent due to land upheaval (Figure 6b): now the location was placed around 2 km away from the coast to the east, located about 85 m a.s.l., on a plateau at the northern side of a narrow and steep east–west going valley, which lead directly to the sea to the east. It was placed in a sheltered position above a little side valley (the former bay), with very good overview of the narrow valley to the south from higher points right beside the site, and about 1,770 m from the shoreline at the time. Six later dates from S1 might point to repeated use of this hearth through ages.

Figure 6

The site of Pauler 2 (red x) and its topographic changes through time. (a) At the time of its Early Mesolithic use (with the sea-level modelled to 120 m a.s.l.), the site was shore based at the eastern side of a little bay at the southern shore of an island. (b) At the time of the Late Mesolithic radiocarbon dates (hearth S1), the site was located on a precipice above an east-west going steep valley with very good views (orange arrows), and the contemporary shore ca. 2,000 m to the east (blue arrow). Map: Almut Schülke based on a landscape model by Gjermund Steinskog, MCH, UiO.

At Dørdal (no. 3), 735 lithic artefacts were excavated; they were typologically and technologically dated to the EM; the site was shoreline dated to the later EM, to 8600–8400 BC (Solheim et al., 2017). Then, it was shore based on a plateau at the southwest edge of a little island, sheltered in a little bay, with a slight elevation to its right (Solheim et al., 2017, Figure 13.3) (see Figure 7a). Two hearths were excavated in 4–5 m distance to each other. Hearth A792 was placed centrally on the offset, sharply delimited from the surroundings. It had a packing with 34 kg of burnt stones, an elongated shape, with a size of 1.35 m × 0.75 m and up to 0.2 m deep. Two charcoal samples (pine) were radiocarbon dated to respectively 6045–5925 calBC (95.4%) (Beta-417123, 7120 ± 30 BP) and 6010–5875 calBC (95.4%) (7050 ± 31 BP, Ua-53183). Hearth A1134, an almost circular packing with 150 kg of burnt stones, with a diameter of 1.6 m yielded two radiocarbon samples. At least one was taken from layer 3 in the middle of the hearth. They were dated to 6010–5895 calBC (95.4%) (Beta-417122, 7070 ± 30 BP) and 5970–5740 calBC (95.4%) (Ua-53184, 6956 ± 31 BP) (Solheim et al., 2017). All dates from the two hearths overlap, falling into the period 6045 to 5740 calBC – the earlier part of the LM phase 3. Dørdal is interpreted as one or two EM visits at a location with good possibilities to scout marine resources; the hearths were interpreted as maybe being of EM origin, the later dates as results of natural processes (Solheim et al., 2017). However, the context, the number of the samples and the contemporaneity of the datings speak for a later Mesolithic use (and even building?) of the hearth. Modelling the topographic situation of the location at the time of the later Mesolithic radiocarbon dates around 5900 calBC (using the shoreline curve from Sørensen, Høeg, & Gälman, 2015) shows that it then was placed around 63 m a.s.l. at the eastern end of an ENE–WSW-going mountain ridge that delimited a steeper valley to the south (Figure 7b). From the elevation west to the site, there were excellent views into several valleys meeting to the west of the mountain ridge. The linear distance to the coast was at the time around 1,400 m.

Figure 7

The site of Dørdal (red x) and its topographic changes through time. (a) At the time of its Early Mesolithic use, the site was placed at the western end of an elongated island. (b) At the time of the Late Mesolithic radiocarbon dates from the two hearths, the site was located at the west end of an elongated mountain ridge, with very good views into a junction of several valleys to the east, and approximately 1,400 km from the coast. Map: Almut Schülke based on a landscape model by Gjermund Steinskog, MCH, UiO.

5.8 Discussion

In terms of data quality regarding sample material and context much speaks for that at least five of the sites, which have either good archaeological context and/or sample material not only composed of pine, but of deciduous tree species or hazelnut shells (cf. Table A2, nos. 1, 3, 4, 5, 7), might witness of later Mesolithic human activity at a location that earlier was a coastal site. This argument is put forward on the background of that, given one would deal with sites from which artefact material chronologically would “match” the radiocarbon dates, the dating would without doubt be accepted as valuable absolute chronological indicator of human activity, and not as potential remains of a forest fire – as numerous publications on Mesolithic sites in the region attest. It is therefore important to consider the possibility that these – radiocarbon dated – situations can point to later Mesolithic human activity related to the burning of fire. The following points further support this suggestion:

1. When excavating Svingen boligfelt (no. 7) elk and elk faeces were repeatedly observed on site; locals confirmed that on elk track was running across the site. It was therefore suggested that the LM dating of charcoal (pinus) from one of the hearths might be connected to a short visit related to elk-hunting in later Mesolithic times and after the site had turned a hinterland location, a visit which would have left no significant material traces (Eymundsson, 2013, p. 73). To get there people would have followed the elk/elk track into the woods.

2. Recent studies have pointed out that natural forest fires in the Mesolithic period in Southeast Norway were of low severity. It is suggested that charcoal or other signs of burning from this period instead would point to human management of fires to manipulate the surroundings (Selsing, 2016, pp. 44–48). Burning isolated trees to produce smoke signals is e.g. pointed out as one method (Selsing, 2016, p. 46). Hearth A 15 at Pauler 1 (no. 2) could thus potentially be interpreted as remains of an intentional fire at the elevated hinterland site, for which also the location of the hearth at the highest point of the site could speak.

3. In his discussion of the later radiocarbon dates from hearths on EM sites excavated in the Pauler-project, Jaksland (2014, p. 31) makes the point that the respective locations would not have been directly accessible by boat in the later times of the radiocarbon dates. This argument is based on the idea of the vital role of boat transport in the EM, which he uses to support the interpretation of the later dates as remains of natural events. The possibility of movement on foot, vital for e.g. hunting and gathering (see below), should however not be dismissed as important practice also in the Mesolithic of the region. Charcoal kiln from historic times documented at Bakke (no. 5) illustrate that the location was accessible by foot and was used for fire management in later times. The charcoal particles which were observed during the Pauler excavations in the soils outside the archaeological sites (Jaksland, 2014, p. 32) could be remnants of charcoal which was spread (wind, animals, erosion, trampling) in the course of historic charcoal production/fire management.

4. Furthermore, the insight that human occupation does not always leave behind (similar) traces, not even when talking about repeated visits of one and the same site, is important. Grøn and Peeters (2021) discuss the topic of visibility and dateability of Mesolithic activities and the general difficulty of assessing the Mesolithic importance of what today is archaeological material, not least on the backdrop of ethnographic knowledge about extensive use of organic material in recent hunter-fisher-gatherer communities which often is not preserved in archaeological contexts. They discuss so-called “ghost sites” from different parts of Northern Europe, which have yielded Mesolithic absolute dates based on organic material, partly with remnants of organic dwelling constructions, but surprisingly little/no lithic material. Grøn and Peeters show that different activities were conducted at different times at the same location – in environments that changed through time. In Southeast Norway the assessment of activities is unavoidably based on the study of lithic artefacts due to poor preservation conditions of organic material in acid soils. Other activities, e.g. involving organic material, would seldom/not leave traces in the record, not to speak of many other prehistoric activities that do not leave traces at all. Diagnostic artefacts (e.g. sandstone knives) were used over long time spans and several periods and do therefore not help to date activities to more than to a long time span. Furthermore, not all of the lithic material might be visually dated that easily (here e.g. in case of no. 3 and no. 9). For some of the sites discussed here, it is mentioned that single artefacts might stem from a later occupation, later than the time of the larger part of the artefact material (Table A2, nos. 1, 7, 8, 9).

5. A most important point is that the topographic and experiential character of the locations discussed here changed drastically from earlier to later Mesolithic times – from being coastal sites directly at a deep seabed, to being an elevated and prominent hinterland locations with good views. These changing settings offered different possibilities of use at different times. They would have attracted people in different ways and for different reasons, what was easy access to sea water at first was later excellent views into valleys. Therefore, it is also reasonable to expect that not necessarily the same type of artefacts or traces of activities were left at these locations at different periods. For the later Mesolithic periods, this would be mainly the evidence of fire/burning either by re-using an old or building a new hearth – different to the deposition of lithic artefact material in earlier times.

5.9 To Sum Up

Even if some of the later radiocarbon dates, observed individually, can be questioned regarding the here forwarded suggestion, in their entity, the described observations support the assumption that at least some of the later Mesolithic radiocarbon dates from the here presented sites indicate human activities at elevated coastal hinterland locations which in earlier times were used as coastal settlement sites. Due to the drastic change of topography with a growing distance to the shoreline, the locations offered different possibilities of use at different times: in the earlier Mesolithic, they provided easy access to the sea with deep water from smaller island in archipelagos and were therefore valuable as places of settlement and, in the later Mesolithic, they offered good views into valleys (partly with waterbodies) which might have served as communication corridors for humans and animals (Figure 8). Eight of the sites provide possible evidence of the lighting of fire in a hearth (with charcoal as fuel or through the roasting of hazelnuts) in later Mesolithic times, while on dating seems questionable (no. 6). One dates to the later MM (no. 1). Seven date to the LM, phase 3 (6300–4500 calBC). Five of these are located on the western side of the Oslo fjord (South Vestfold and Bamble) and date to the earlier part of the LM (Table A2, nos. 2, 3, 4, 5, 8), while two, both from the eastern side of the fjord (former Akershus county), date to the later part of the LM (Appendix nos. 7, 9) (cf. Figure 1). The absence or only faint potential traces of diagnostic archaeological material (lithic artefacts) from these later periods need not speak against later visits of the locations. Instead, this might witness of short-term activities on these elevated and not always easily accessible locations, where objects or baggage were taking along (for not leaving traces?), or of activities that did involve organic material of which no traces are preserved.

Figure 8

Sketch of the shifting topographic contexts at the same location throughout the Mesolithic period. (a) The situation in the EM with the settlement placed directly at the coast on bedrock and with a strongly sloping seabed. (b) The situation in later Mesolithic times, after the shoreline, had retreated and the same placed would have been placed on top or very close to a rocky precipice with good overview into a valley. Illustrations: Almut Schülke.

These indications of Mesolithic people’s movement beyond coast and waterways into the wooded hinterlands in the Oslo Fjord area will now be discussed in a wider archaeological context.

6.1 Engaging with Hinterland Topography and Vegetation

The land behind the mainly rocky coast of the Oslo fjord region was also in later Mesolithic times characterized by pronounced bedrock formations, cut by rivers and numerous valleys with lakes and bogs. The vegetation cover with deciduous forests on thin and acid soils evolved throughout millennia after the last glacial, with land uplift and the separation of basins from the sea (e.g. Høeg, Henningsmoen, & Sørensen 2019; Sørensen et al., 2014; Wiekowska-Lüth, Solheim, Schülke, & Kirleis, 2018). The latitude of Southeast Norway is exposed to marked seasons with cold and dark winters, short spring periods, summers with only a few hours of darkness at night and autumns with early frost. Thus, the environment traverses enormous seasonal changes, where snow and ice in winter have the biggest impact on vegetation, animal life and on the physical communication potential, generally seen as a period of stress (Frislid & Jensen, 1994; Selsing, 2021).

The vegetation of the region in the Atlantic period, which can be paralleled with later Mesolithic times, was dominated by a mixed oak forest in which also linden occurred; the presence of species, such as mistletoe and ivy, indicates a warmer climate and longer growing seasons (Hoeg et al., 2019). Local pollen analysis explore the vegetation history of more delimited areas. A laminated lake core from Lake Skogstjern, Bamble, Telemark shows the changing vegetation through time. In the LM, Lake Skogstjern was placed about 500 m away from the coast, in the hinterland. Around the lake the vegetation in the LM was characterized by deciduous forest with elm, linden and hazel and indicated human manipulation of vegetation in the coast-near hinterland, which is interpreted as being intentional, at several occasions, to attract game, or even to form certain types of woods for other purposes such as weaving of wickerwork for shelters, fences, fish traps or baskets (Wiekowska-Lüth et al., 2018).

Archaeological evidence indicates the intentional use of wood for fire (Selsing, 2016, 2018), for construction and equipment, as e.g. indirectly manifested in the Nøstvet axes (Jaksland, 2005; Wiekowska-Lüth et al., 2018), and the collecting of hazelnuts. Other resources that potentially were collected in the woods, but that are difficult to grasp archaeologically are: fibres, berries, roots, diverse parts of plants and mushrooms as food, for medical purposes or for dying etc., with a high season in summer/autumn.

6.2 Encounters with Animals

Almost all of the animal bones preserved from Mesolithic coastal sites in the Oslo fjord region encompass fish/sea mammals and terrestrial animals (Glørstad, 2010, pp. 70–88; Mansrud & Persson, 2017; Solheim, 2020, Table 3.1). These, relatively seen, few bone-finds are all burned and highly fragmented; often they only can indicate the presence or absence of certain species (Mansrud & Persson, 2017). The species represented (Table A3) include ungulates (elk, red deer, roe deer or rein deer), rodents (beaver, squirrel), hares and predators (brown bear, red fox, mustelids and otter). While elk, squirrel and brown bear only are known from MM sites and roe deer, deer/reindeer, hare, red fox, mustelids and otter only from LM sites, red deer, wild boar and beaver occur in both periods. Wild boar occurs on most of the sites. This picture might however be biased by a number of source critical factors including poor preservation conditions.

To expand information the habitats of the represented animals were collected for this article (according to Frislid & Jensen, 1994; see Table A3). The animals’ habitats allow insights into human engagement with terrestrial animals in a spatial perspective as the surroundings in which the respective animal species dwell and move indicate which environments had to be frequented by humans to meet and hunt these animals.

Snow conditions in winter are the factor that affects animal movement most. Ungulates mainly tarry in wooded areas, but in the winter season, they move away from areas with deep snow as it limits possibilities of grazing (e.g. heather, blueberry leaves) (Frislid & Jensen, 1994, pp. 274–317). Some of the ungulate species move closer to the coast with milder climate and lesser snow, and red deer and elk travel across long distances, using the same tracks (Frislid & Jensen, 1994, pp. 280–286, 293–298). Thus, there were good chances for Mesolithic hunters to meet ungulates at the coast or along bigger watercourses. But, there are also species represented which only live in the woods (Table A3). Beaver, for example, depend on freshwater and deciduous trees (Frislid & Jensen, 1994, pp. 106–112). Seldomly and unpredictable they occur in coastal waters (Hardeng, 1998). Bear lives in coniferous woods or sometimes higher up in the mountains, preferably in areas that are difficult to access (Frislid & Jensen, 1994, pp. 200–206). Squirrels and mustelids mainly dwell in the woods (Frislid & Jensen, 1994, pp. 100–106, 230–234). Wild boar also prefers deciduous, slightly humid forests, but they can, as the other listed species (Table A3), have more varied habitats, which do not include woods. To encounter and hunt some of these animals, Mesolithic people had to access deciduous forests with fresh water sources (beaver) (Figure 9) or rather impassable coniferous forests (bear).

The animals’ different patterns of movement, sometimes explicitly seasonal, is decisive for human-animal encounters (Elliott, 2019; Fuglestvedt, 2018, p. 114). It must have been important for structuring human movements in the woods. Compared to the smaller terrestrial animals with more restricted territories of movement, which would have afforded less human mobility to follow them, the high mobility of the ungulates would have had impact on humans as hunting big game was difficult (Selsing, 2021). Following their paths was surely important. Big game was an important symbol and reference for Mesolithic people (Glørstad, 2010, pp. 227–236, 252–256; Fuglestvedt, 2018). Elk or deer are the dominating motives on the 10 LM rock art sites known from Southeast-Norway, all of which were situated directly at waterways in their time, three of them at the coast (Fuglestvedt, 2018, Table 2.1; Glørstad, 2010, pp. 216–223; Mikkelsen, 1977). A close communication and metaphysical interrelation between animals and humans might have been expressed e.g. in depictions of hunting situations on rock art, but also in eating meat and processing other parts of the animal (e.g. hides and bones into clothes and tools) (Fuglestvedt, 2018, p. 117 ff.). The close relationship of elk and water is also observed expressed in symbolic terms in the archaeological material (Glørstad, 2010, pp. 227–228, 252–256), which might suggest the hunting elk at well-known tracks or posts at open water, as it e.g. still is practiced in some parts of the large Northern Eurasian taiga (Willerslev, 2011). Scandinavian rock art fields with big game depictions often are placed close to seasonal gathering areas of larger groups of animals, and close to elk tracks in the woods, sometimes even placed close to historic elk-hunting pits (Gjerde, 2010, pp. 426–434). It was proposed that hunting pits are depicted on the SE-Norwegian rock art-sites Ekeberg and Skogerveien (Gjerde, 2010, pp. 433–434); the sites were both coast based in the Mesolithic. This might indicate that hunters tracked down elk in the hinterlands/woods, as hunting pits usually are placed along elk tracks between grazing areas. It required different senses and practices of movement for the hunter to detect and finally meet these animals, such as the ability to move quiet and undetected, including to prevent the animals from smelling the hunter (Willerslev, 2011), or waiting – in cases of hunting along known animal tracks. As the animals are very mobile, most likely different types of hunting had to be practiced, adjusted to the situation (Selsing, 2021).

These diverse sensual qualities and practices of movement were most likely also important for other aspects of social life taking place in the forest, such as taking control of situations, e.g. by conducting activities that should stay unseen for others, or observing other humans’ activities and movements, for examples from places with good views (see below).

6.3 Wayfinding Practices

To reach hinterland locations, as some of the ones presented above, rocky wooded hinterland areas would have need to be crossed on foot. One would have been confronted with significant height differences due to a splitting up of the landscape by basins, wetlands and valleys cutting the steep bedrock formations. This would have required other skills than navigating along the shores with a boat. Hunter-fisher-gatherers in the Boreal zone, especially in areas with a low density of people, would usually not have moved alone, as the forest was full of dangers and a safe return had to be guaranteed (Lovis, 2016). Wayfinding practices are movements of the actively perceiving and sensing practitioner, who avoided obstacles and who would have preferred the following of human and animal traces (Takada, 2016). Identifying safe pathways in hinterland areas would have included the crossing, overcoming (upwards and downwards) or circumventing of steep and often slippery slopes – to the elevated hinterland locations – as well as valley floors with bogs and wetlands, including the knowledge of which vegetation would indicate which type of underground at what time of the year (dry, wet, frozen, fluid, solid). Seasonal differences with different advantages/disadvantages of movement would be important, so that the cost of movement also would have played a role (Kelly, 2013, p. 101). In winter, snow and ice are challenges, but the crossing of wetlands and lakes on foot is possible; further, movement and transport can be made easier by using technical help such as ski or pulling aids. In frost-free seasons, lakes and watercourses can be overcome by boat/canoe or swimming. Reading animal tracks, spoors and excrements would help to identify paths and tracking animals. Senses such as smelling (e.g. the fox) and listening (e.g. to the song, calls or silence of the birds) would have given signs of safeness or danger, or of closeness of animals.

Furthermore, the memorizing of paths and places would have been crucial for orientation. Specific points in the perceived landscape would have been needed as important landmarks, to be connected in a memory sequence of strings/lines with different characteristics, which would pointed into the right direction (Takada, 2016). In Southeast Norway, these could for example include rock formations, wetland basins, characteristic trees or large anthills. Cumulative knowledge of an area would have increased the better the groups using it were familiar with the area (Lovis, 2016; see below). For areas in the Northern Boreal forest which are more regularly frequented by hunter-fisher-gatherer groups, landscape marking or the marking of tracks with e.g. cut off branches, cairns or animal skeletons in trees, is reported (Lovis, 2016).