THE AGENCY OF ROCK: CHARISMA OR ESSENCE?

Meaning is contextually and socially dependent, fluid, and multi-layered, but also formed and framed by ontological perspective or orientation. Thus, aiming to understand what rock meant to people in past societies, beyond its pragmatic qualities for tool production, is a major challenge for archaeologists. Ethnographically one can find examples of lithic procurement sites, quarries or outcrops, being regarded as arenas for communication and interaction with certain powers or spirits by some historic hunter-gatherer societies (cf. Robinson Reference Robinson, Boivin and Owoc2004, 97). Certain rock deposits have been perceived as places of ancestral presence, linking land, mythical characters or spirits, and people (eg, Taçon Reference Taçon1991; Hampton Reference Hampton1999; Dean Reference Dean2010). Rock procured from such a socially significant place can thus embed value, empowering the objects made from it. Ethnographic or anthropological examples from around the world demonstrate how objects and rock have been ascribed specific qualities, sometimes being regarded and valued as living entities themselves (Dean Reference Dean2010; Pétrequin & Pétrequin Reference Pétrequin, Pétrequin, Davies and Edmonds2011).

Endeavours have also been made to move beyond the perhaps arbitrary symbolic meaning of rocks. Instead, human–material encounters and engagement are considered in an attempt to elucidate how materials, things – rock – affect human practices as a result of their embedded and often non-quantifiable qualities. In the 1960s, Max Weber (Reference Weber1968) characterised the features, activities, symbols, and personalities of people holding positions in religious or political institutions as ‘charismatic’. The socially situated and acknowledged charisma aided in turn the legitimation and consolidation of authority. Since then, charismatic qualities have been extended to objects in an attempt to explain how objects affect human emotions, memories, and associations. Worn objects can thus be considered as extended parts of the wearer’s body, a type of ‘social skin’; they can be powerful and charismatic sources of expressed religious identity, for example (eg, Nødseth Reference Nødseth, Vedeler, Røstad and Kristoffersen2018). In cognitive psychology objects have similarly been acknowledged as contagions: that is, objects embedding qualities that affect the identity or status of the owners (Newman & Smith Reference Newman and Smith2016). In the 1990s, as one increasingly acknowledged that objects could embed the qualities of humans, theories of object-agency developed (Gell Reference Gell, Coote and Shelton1992; Reference Gell1998). Gell (Reference Gell1998, 17–21) pointed out that agency is relational, existing in the relationship between people and things. Hence, owning, using, or wearing an object which is recognised as significant can legitimise authority or a person or place’s enigmatic power or social significance. Thus, although originally perceived as a human quality, charisma can be transferred between people and things and be embedded by objects.

This recognition of the autonomy and status of things increased in the 2000s. Any worldly object, animate or inanimate, can be ascribed the ability to act, to have goals, and the power to influence people and society (Olsen Reference Olsen2003, 89; Boivin Reference Boivin, Demarrais, Gosden and Renfrew2004b; Gosden Reference Gosden2005, 196). Related arguments emphasising the capacities of objects to both emotionally and influentially affect humans, regarding objects as bodies capable of influencing and moving people, is part of post-anthropocentric debates (eg, Harris & Sørensen Reference Harris and Sørensen2010; Hamilakis Reference Hamilakis2013). Hence, objects can have the ability to arouse awe and to appeal to people’s aesthetic senses. Still, their impact also depends on things having a form, style, or origin that is recognised and acknowledged by the surrounding community. As pertains in this paper, to better understand the particularities of the procurement and use of certain rock types, I consider whether they could have possessed or been endowed with intangible qualities.

Our understanding of identified patterns of lithic distribution and exploitation may be inhibited by our Western European perspectives, if practices were initiated and anchored in an ontological perspective diverging from our Western one. An example of a diverging ontology can be found within the pre-Columbian and Amerindian communities. To them, the quality of light infused morality, represented life, and energised the world (Saunders Reference Saunders1999, 246). Light, bright colours and lustre indicated the presence of supernatural beings. Moreover, materials that Western European researchers regard as elementally different, such as metals, shells, pearls, wood, or rock that shone and shimmered would, for the pre-Colombians, be regarded as essentially the same, yet ‘actively transformational in appearance and spiritual essence’ (Saunders Reference Saunders, Boivin and Owoc2004, 136). Hence, the main quality of a raw material could be its transformability (Gosden Reference Gosden2005, 209; Conneller Reference Conneller2011, 93), that is, different substances, such as different rock types, may when transformed take on or realise forms and qualities which transgress Western classifications and boundaries between types (Conneller Reference Conneller2011). This more fluid and relational perspective on the qualities of the bodies of the world – its objects, people, animals, etc – has also been advocated by Eduardo Viveiros de Castro (Reference Viveiros De Castro2012, 125). Referring to Amerindian ontologies, he argues that understanding what things are is a matter of phenomenological perspective, in the sense that ‘body and soul, just like nature and culture, do not correspond to substantives, self-subsistent entities or ontological provinces’.

Examples from around the world can inform how different raw materials can be imbued with sacredness by virtue of their origin at locations conceived of as spiritually or ritually significant, such as a specific mountain. This can also mean origin in a wider sense, such as mountains in general. In the case of the pre-Colombians described by Saunders, the powerful energy of light was converted into solid objects, a transformative process that empowered or reinforced these objects with mythic identity. The empowered objects then merged with the owner, legitimising their social position or power (Saunders Reference Saunders1999, 246). Consequently, and in light of the outlined discussion of object-agency, transformation can make objects or, indeed, lithic raw materials charismatic. The challenge, though, is to anchor any claims of such intangible, fluid, and differentiated meaning of rock in the archaeological material. Based on the examination of lithic quarries, rock deposits, and lithic assemblages at open-air sites in two case study areas in South Norway, I will attempt to do just that.

USING ROCK IN MONTANE AREAS – TWO CASE STUDIES

The case studies are compiled from archaeological reports on surveyed and excavated hunting camps in two case study areas in the mountains of western South Norway. Both include rock quarries and seasonal, short-term camp sites located around lakes and rivers. Surveyed and excavated lithic materials from these sites, stored at the university museums in Bergen, Oslo, and Stavanger, have partly been re-examined. The museum catalogues, available onlineFootnote ¹ have also been consulted.

In this paper, I focus on the activity in the Late Mesolithic and Early/Middle Neolithic, yet many of the sites demonstrate activity in later prehistoric periods too. The material in both case studies concurs with general typological trends in western South Norway. Most relevant here is how, in the Late Mesolithic, the dominating technology is microblade production on single platform conical or bipolar cores. From the Early Neolithic blades are struck from cylindrical cores. Around 4000 bc characteristic transverse and tanged points occur, while tanged points continued to the end of the Middle Neolithic becoming more elaborate and curated with time. The latter is allegedly a cultural impulse from south Scandinavian Battle Axe Cultures, whereas, in the Early Neolithic, ground slate technology is thought to have been introduced to South Norway from northern Norway. Hence, during this period, the regions include and respond to external impulses but the predominantly hunter-gatherer-fisher semi-mobile society continues into the Late Neolithic.

The use of quartz crystal in the mountainous regions of Rogaland County

The first case study covers an area of about 600 km² in the mountainous interior of south-west Norway in the counties of Rogaland, Aust-Agder, and Telemark. The area is about 100 km east of the coastal zone that was relatively densely settled during both the Mesolithic and Neolithic. Steep valleys and lakes cut across the terrain of mostly treeless mountains with peaks between 1000 m and 1400 m asl. Compared to other regions, the variety of rock types suitable for tool production is low, with the bedrock consisting mostly of gneiss and phyllite. There is, however, one exception: large surface deposits of quartz with sizeable quartz crystals (colloquially known as rock crystals).

In the 1960s and ’70s, large scale archaeological surveys were initiated in response to massive hydro-power developments in this region. Consequently, prehistoric activity in parts of the montane region was thoroughly charted and resulted in archaeological excavations of the most exposed sites (eg, Odner Reference Odner1962; Reference Odner1963; Bang-Andersen Reference Bang-Andersen1975; Reference Bang-Andersen1983; Reference Bang-Andersen2008). During these surveys, six large sources of quartz crystal were also recorded (Bang-Andersen Reference Bang-Andersen1975) (Fig. 1).

Fig. 1. Map illustrating the large sources of quartz crystal within the case study areas and examined sites referred to in Table 1 (illustration: Astrid J. Nyland)

The quartz crystal sources comprise large, scattered blocks as well as broad, visible veins in the bedrock (Fig. 2). The search for crystals has left blocks smashed into smaller pieces and there are hollows from quarrying to access the pure rock crystal prisms embedded in cavities in the rock. The sources are not protected sites. Consequently, many of the signs of quarrying also stem from visits and crystal extraction in both historic and modern times as established hiking routes pass by the sites. There is evidence in the region of crystals being used for making beads in the Viking Age (Myhre Reference Myhre2005). Thus, it is difficult – if not impossible – to distinguish prehistoric from recent debris. In order to date and evaluate a potential exploitation of the sources and utilisation during the Stone Age, the lithic assemblages at 29 recorded sites in the vicinity of the sources were examined instead (Table 1, see also Fig. 1). The sites are dated based on chronological indicators in the lithic assemblages, especially points and blade technology. Quantified variation in rock composition in the assemblages indicates preferences in rock types and procurement strategies, for example, whether or not it was solely the local sources of rock that were exploited. Not all the examined sites are scientifically excavated, some comprise collected surface-exposed material only. Nevertheless, together they give an impression of which rock types were used in the mountains, compared to the coastal zone.

Fig. 2. The site comprises smashed quartz after the extraction of crystals embedded in cavities in a broad vein of quartz in the bedrock (photo: Astrid J. Nyland)

TABLE 1: THE RESULTS FROM THE EXAMINATION OF RAW MATERIAL VARIATION AT 29 SITES IN CASE STUDY AREA 1

The letters in column 1 correspond to the map in Fig. 1. The rock types found are: quartz crystal (qc), flint (f), quartzite (q), rhyolite (r) and slate (s). Small x indicates use and large X the period where the rock types dominates in the assemblage. Mus. no is the assemblage’s identification number in the stored collections at the Museum of Archaeology, UiS (S. no) and the Museum of Cultural History in Oslo, UiO, (C. no) (see also endnote 1). The sites are located in the counties of Rogaland (R), Hordaland (H), Aust-Agder (AA), and Telemark (T).

Some sites are evidential used repeatedly, demonstrating material that is from two (or more) periods. In the date column italic indicates the dominant period; parenthesis indicates the most likely date, where more precise dating is not possible.

The glacial withdrawal in this region is said to have concluded at the onset of the Middle Mesolithic (Blystad & Selsing Reference Blystad and Selsing1988). The oldest sites in the case study area date back to this period (Sites i, w and possibly æ). The majority of sites date to the end of the Late Mesolithic and the Early Neolithic. In the Late Mesolithic microblades are made of both flint and quartz crystals. Nine sites (u–å) have been radiocarbon dated (see Bang-Andersen Reference Bang-Andersen2008, 94–5, 136–9). At the coast and in the mountains sites demonstrate that quartz crystals were part of the known ‘lithic landscape’ from the Early Mesolithic (9500 bc onwards) into the Bronze Age (1700–400 bc) (eg, Bang-Andersen Reference Bang-Andersen, Larsson, Knutsson, Loeffler and Åkerlund2003; Reference Bang-Andersen2012; Skjelstad Reference Skjelstad2011). In this case study area the investigations demonstrate that flint and quartz crystal dominate as raw materials for tool production in the mountains at least from 5000 bc to 2700 bc.

The exploitation of quartzites in the mountains of Lærdal–Hemsedal

The second case study is located in the mountains of Lærdal and Hemsedal, in Sogn og Fjordane, Oppland, and Buskerud Counties. The study area is approximately 1000 km², above the present tree line, with lakes and rivers cutting across the rocky landscape and peaks around 1200–1400 m asl. The area constitutes a watershed between the valleys of inland eastern Norway and the inner fjord regions of western Norway. There are broad glaciated valleys to the east and steep river valleys to the west. As in the previous case study, most sites were discovered and charted during intense surveys associated with hydro-power developments in the 1960–70s (eg, Espedal Reference Espedal1965a; Reference Espedal1965b; Martens Reference Martens1965; Johansen Reference Johansen1967; Reference Johansen1978).

During the surveys, several exploited outcrops and boulders of quartzite were observed, including three definite quartzite quarries (Fig. 3). The quartzite in this area is particularly fine-grained, pure, and dense, splitting almost like flint (colloquially called ‘Lærdal quartzite’ in Norwegian Stone Age archaeology).

Fig. 3. Map of the three definite quartzite quarries and the location of examined sites referred to in Table 2 (illustration: Astrid J. Nyland)

The largest quarry is Kjølskarvet. It comprises two main focal points of quarried crags and outcrops (Sites I & II), separated by about 500 m (Espedal Reference Espedal1965a; Johansen Reference Johansen1967; Reference Johansen1968). Surrounding Sites I and II, and Lake Øljusjøen, there are numerous exploited quartzite erratics and outcrops but also workshop and camp sites (Fig. 4). This area thus constitutes the largest known mountainous extraction site of quartzite in Norway. The extent and thickness of the waste layers, the many workshops, and the surrounding sites imply that at least 100 m³ quartzite has been quarried in the area (Nyland Reference Nyland2016a).

Fig. 4. Surrounding the quartzite quarry comprising Kjølskarvet site I & II are numerous exploited erratics and debitage from initial reduction of blanks (photo: Astrid J. Nyland)

A smaller quarry, Stongeskaret, lies 3–4 km north-east of Kjølskarvet and just south-east of Lake Øljusjøen (Fig. 3). The quality and colour of the quartzite extracted is similar to that from Kjølskarvet but it is a much smaller quarry; the estimated extraction is only around 6 m³ (Nyland Reference Nyland2016a). About 12 km south-east of Kjølskarvet and Stongeskaret a third quarry, Halsane, is located on a protruding mountain ridge. Here, an outcrop of white, translucent cryptocrystalline quartzite was quarried. The volume of extraction is estimated to be a minimum of c. 10 m³ (Nyland Reference Nyland2016a).

Neither charcoal nor other organic materials have been discovered at any of the three quarries. Fire was most likely not used as a quarrying technique at these sites since large and sudden temperature fluctuations could have destroyed too much of the silica-rich rock in the process. Hammering and wedging were probably the more common techniques (Nyland Reference Nyland2016a). Cone-shaped scars from direct hammering were identified at most of the exposed outcrops as were flake scars and ‘steps’ in the rock from loosened disc-shaped blanks (Ballin Reference Ballin2008; Nyland Reference Nyland2016a). Because of the lack of organic material there are no direct radiocarbon dates from the quarries. The quarrying activity is therefore dated typologically and technologically by artefacts in the waste piles as well as at camp sites in their vicinity, some which are radiocarbon dated (eg, Gjerland Reference Gjerland1980; Ballin Reference Ballin1998; Matsumoto & Uleberg Reference Matsumoto, Uleberg and Hofseth2002; Uleberg Reference Uleberg2003; Årskog & Åstveit Reference Årskog and Åstveit2014a; Reference Årskog and Åstveit2014b; Reference Årskog and Åstveit2014c). Based on this, it is clear that the quarries were in use seasonally from the 8th millennium bc to the pre-Roman Iron Age (Nyland Reference Nyland2016a).

As in the Rogaland area the lithic raw material composition at sites within the demarcated study area have been identified and quantified (Table 2; Fig. 3). The number of examined finds per site in the 1960s varies. In the archaeological reports, the selected material is referred to as being ‘from representative excavation units’, and only c. 10% of the collected material from many of the sites was catalogued. The rest was weighed and an estimate of the various rock types given (eg, Espedal Reference Espedal1965a; Reference Espedal1965b; Johansen Reference Johansen1978; Ballin Reference Ballin1998; Table 2, column 4). Despite the variation in the number of finds and the degree of excavation of the selected sites, the material still sufficiently demonstrates a tendency in the dominant lithic procurement practice. Several of the sites have been re-examined since the 1960s (Ballin Reference Ballin1998; Matsumoto & Uleberg Reference Matsumoto, Uleberg and Hofseth2002; Uleberg Reference Uleberg2003; Nyland Reference Nyland2016a) and recent excavations complement the earlier information and work (eg, Sites 20–22 in Table 2, excavated in 2012–2014 with reports that provide complete find lists, as well as radiocarbon dates (Årskog & Åstveit Reference Årskog and Åstveit2014a; Reference Årskog and Åstveit2014b; Reference Årskog and Åstveit2014c).

TABLE 2: LITHIC RAW MATERIAL COMPOSITION AT 29 SITES IN CASE STUDY AREA 2

The numbers in column one correspond to the map in Fig. 3. The different rock types found are: quartz crystal (qc), flint (f), quartzite (q), slate (s) and other (o). Small x indicates use and large X the period where the rock types dominate in the assemblage. Mus. no is the assemblage’s identification number in the stored collections at the University Museum of Bergen, UiB (B. no) and the Museum of Cultural History in Oslo, UiO, (C. no) (see also endnote 1). The sites are located in the counties of Hordaland (H), Sogn og Fjordane (S&F), Oppland (O), and Buskerud (B).

Some sites are evidential used repeatedly, demonstrating material that is from two (or more) periods. In the date column italic indicates the dominant period; parenthesis indicates the most likely date, where more precise dating is not possible.

As expected, raw material from the three quarries dominates at the examined sites in the vicinity of each of them. However, it is also evident that a multitude of quartzite sources have been utilised, even at sites close to the quarries. At Lok.1 Smiu (No. 21 on the map), the excavators noticed rock collected at sources found within a radius of around 15 km (Årskog & Åstveit Reference Årskog and Åstveit2014c). In sum, the case study shows regular quarrying combined with an opportunistic exploitation of outcrops and erratics scattered across the terrain. In addition, people sporadically used quartz crystal and slate in the Early Neolithic. Another common tendency in both the Mesolithic and Neolithic is that flint was used at most of the sites. From the study area, it is 200 km or more to the nearest source of flint at the coast demonstrating that flint was transported rather far into an area abundant in quartzites.

DISTRIBUTION AS INDICATORS OF PREFERENCES IN THE CASE STUDY AREAS

Based on the production waste at the examined sites in the Rogaland case study, there seems to have been a deliberate targeting of the purest quartz and crystals (Fig. 5). Nevertheless, despite domination at some of the sites, the frequency is relatively low. The character of use is not compatible with intense and regular quarrying (cf. Nyland Reference Nyland2016a). Instead, hunters seem to have collected rock crystals from any source they encountered.

Fig. 5. Crystal clear rock was targeted. Late Mesolithic type conical core and microblades (photo: Astrid J. Nyland)

It is hard to determine the exact origin of the quartz crystals. Attempts to determine the provenance of quartz crystal distribution can be thwarted by the fact that these comprise a chemically homogeneous mineral consisting of silicon and oxygen and so provide poor objects for geochemical provenance investigations. Their crystal structure also disturbs the use of non-intrusive portable X-ray fluorescence spectrometers, which measure trace elements through light reflection. Hence, testing the same crystal twice to get the same result might prove difficult since light can be defracted differently if the x-ray hits at only a slightly different spot or angle. There are promising results from testing quartz using laser ablation inductively coupled mass spectrometry (LA-IPC-MS; Müller Reference Müller2008) but this has not been tested on the crystals examined for this study.

Considering the use of quartz crystal and flint in a wider perspective, flint is clearly the preferred rock type at the coast (Fig. 6). Beach flint is, as the name implies, collected along beaches, but from the Middle Mesolithic quarrying became an established practice too, intensifying around 4000 bc (cf. Nyland Reference Nyland2016a; Reference Nyland2016b). Contact between the landscape zones is evident through people bringing flint to the mountains but very few other rock types are demonstrably transported between the zones. One exception is slate where, often, complete or fragmentary tools are found at mountain sites, but little production waste. Another is the particular rhyolite found at Site e. This is an igneous fine-grained rock, dark bluish-grey with characteristic white zig-zag lines. It was quarried atop Mt Siggjo, a low mountain on a coastal island, from the onset of the Early to the first half of the Middle Neolithic. Previous distribution studies demonstrate it making up to 90% of lithics at coastal sites close to the quarry (Alsaker Reference Alsaker1987; Bergsvik Reference Bergsvik2006). It was also distributed up to 200 km north and south along the coast. However, it is only infrequently found further afield at mountain sites. It is recorded at only eight sites on the Hardangervidda plateau and never as more than a handful of pieces at each site (eg, Indrelid Reference Indrelid1994; Nyland Reference Nyland2016b). In the Rogaland study area, it was only found at a single site (e), where a handful of black, fine-grained quartzite fragments was also found. A similar looking rock has been found at contemporary Early Neolithic coastal sites (eg, Eilertsen Reference Eilertsen2009; Reference Eilertsen2010), pointing to this quartzite most likely originating at the coast too.

Fig. 6. Illustrating the dominance of settlements at the coast vs the mountains during the Stone Age in Rogaland, as well as the dominance of flint use vs quartz crystal. Case study one is in the upper right corner. Circle sizes range from small (1) to large (>10) sites per circle (maps by Krister Scheie Eilertsen)

In the Lærdal–Hemsedal case study further north, the situation is both different and similar. Quartzite is not brought from the coast to the mountain, nor is mylonite commonly used on the coast. Instead, in the mountains, there is a wide internal distribution of the greenish quartzite known from Kjølskarvet and Stongeskaret at sites across a wide area. However, provenancing quartzite securely is also challenging. There are similar looking greenish, fine-grained quartzite surfaces in several places within this part of the montane region (eg, Sjøvold & Martens Reference Sjøvold and Martens1971; Gustafson Reference Gustafson1983; Nyland Reference Nyland2016a). There is also great internal colour variation at the quarries Kjølskarvet and Stongeskaret, the quartzite ranging from vivid green in various shades to white, as well as coloured banding/stripes. This makes distribution studies based on visual identification challenging. Provenance studies based on geochemical or mineralogical composition analysis may also be difficult since most quartzites are made of nearly pure silica (eg, Pitblado et al. Reference Pitblado, Cannon, Neff, Dehler and Nelson2013). Some variation in geochemical composition between the quartzite sources is still detectable however, as shown in a pilot test utilising XRF on glass tablets (Table 3).Footnote ² The test did not aim to describe the geochemical signature of the whole deposit but shows variations in trace elements between identical-looking green quartzites from Stongeskaret. It also demonstrates internal variation in samples of white and green quartzite within the large Kjølskarvet quarry. In a study from the late 1970s the geochemical content of green quartzite flakes found at two Late Mesolithic and Early/Middle Neolithic coastal sites, Flatøy III and Ramsvikneset, south of the mouth of the Sognefjord, were compared with samples from the Kjølskarvet quarry using XRF. They were found to be a probable match (Bakka Reference Bakka1976; Bjørgo Reference Bjørgo1981). One may question the validity of limited provenance studies but, if valid, this strengthens the interpretation of a western coastal population seasonally or sporadically visiting the mountains. The fact that there is so little green quartzite at the coast is nevertheless puzzling.

TABLE 3: RESULTS OF XRF ANALYSES OF VISUALLY DIFFERENT ROCK SAMPLES FROM THE KJØLSKARVET AND STONGESKARET QUARRIES

¹quartzite quarry ID155736 (green); ²quartzite quarry II ID114851-I (green); ³quartzite quarry II ID114851-II (white)

Measurements undertaken by the Geological Survey of Norway (NGU) on 25 April 2013 (on request of the author)

Despite not being able to pinpoint the exact origin of the quartz crystal or green quartzite, movement of certain rock types into other areas implies relations between people and landscapes. Also, in the Lærdal–Hemsedal case study, the beach flint used at mountain sites implies movement between the mountainous and the coastal regions. The composition of raw materials and their use leave a lingering impression that there was a desire or preference for specific types of rock. Only certain rock types were in use in each region, whereas a selected few were used in both. Was this due to knapping properties or was there something else affecting choices and traditions?

HOW TO DETERMINE THE VALUE OF ROCK?

Both case study areas were evidently in continual use in the Late Mesolithic and Early/Middle Neolithic. Thus, one would expect that members of the societies concerned knew of the availability of lithic resources in each area. Nevertheless, flint was transported 200 km or more into the Lærdal–Hemsedal area where quartzite was easily available (Fig. 3) and half that distance into an area with available quartz crystals in Rogaland (Fig. 1). In both areas most of the distance could be travelled by boat on the fjords. Safe journeys would be weather dependant and the final part of the journey would nevertheless be a steep climb up to around 1300 m asl. It makes sense to bring raw material for tools into areas where there is otherwise little available suitable rock but why did people bring it into areas where rock was sufficient for their needs? As discussed earlier, ethnographic work suggests that rock procurement may have been structured or initiated based on spiritual ideas, knowledge, or stories embedded in the land and local traditions. The composition and frequency of certain rock types at sites can thus signpost the influence of aspects beyond pragmatics and optimal resource exploitation.

A common approach in archaeological research has been to ascribe value to rock based on its provenance, availability, or the degree of human endeavour required to procure or transport it across vast distances. Focusing on provenance, rock found beyond the region where it naturally occurs is often categorised as exotic as opposed to locally available rock (eg, Olausson Reference Olausson1983; Gould & Saggers Reference Gould and Saggers1985; Taffinder Reference Taffinder1998; Bergsvik Reference Bergsvik2002; Eriksen Reference Eriksen, Fischer and Eriksen2002). Hence, exotic rock indicates long-distance mobility or bartering and is most often considered prestigious and valuable. An exceptional example of long-distance distribution demonstrating this is the spread of jadeitites from the Alps more than 1000 km across Europe in the Early Neolithic. Jadeitite was made into everyday artefacts but also into highly polished axeheads and disc-rings – interpreted as prestige objects – and transported across Western Europe (Sheridan et al. Reference Sheridan, Pétrequin, Pétrequin, Cassen, Errera, Gauthier, Prodéo, Rodríguez-Rellán, Nelson and Fábregas Valcarce2019). In the case study areas on the west coast of Norway, neither beach flint, quartz crystals, nor quartzite are normally considered as prestigious. The artefacts discussed are those commonly found at the hunter-gatherer-fisher sites across South Norway. What then, determines the raw material composition or indeed preferences for certain rock types? Is it distance from origin that determined social significance, making it desirable and costly?

Comparing studies where lithic variation and production waste at settlement sites are assessed, it becomes apparent that using distance to define rock as exotic or local is highly relative and a matter of perspective. For example, in the dry deserts of central Australia rock was considered local when it originated within a radius of 40 km from a site, while rock from beyond this limit was considered exotic (Gould Reference Gould1978, 262; Gould & Saggers Reference Gould and Saggers1985, 119). In a northern European context, lithics from within a site’s predefined catchment area have been defined as local when occurring within the annual territory <10 km, as regional when occurring 10–50 km away, and as exotic when coming from beyond 50 km (Eriksen Reference Eriksen, Fischer and Eriksen2002, 31). In western Portugal the distance from specific sites to known sources of rock found at a collection of sites has been classified by Aubry et al. (Reference Aubry, Luís, Llach and Matias2012, 535, table 3) into three categories: long distance (later characterised as exotic) at 130–298 km/21–31 hours’ walk); intermediate (89–106 km/14–15 hours); and local. In the case studies in South Norway, presented here, people carried coastal flint pebbles and cores of other rock types into the mountains up to 200 km and perhaps even further. But, when found in the mountain regions of Norway, flint in particular is normally regarded as neither exotic nor extraordinary in archaeological interpretations. However, at one site in Rogaland the coastal beach flint is supplied in the form of a deliberate re-use of ground flint axes imported from South Scandinavia. This is identified by flint flakes with traces of grinding on their dorsal side from at least two mountain sites (Sites m and n). Flint from South Scandinavia is generally of better quality than beach flint and the size of the beach pebbles made them unsuited for producing axes of the type imported from the south. It has also been common to regard the imported flint tools as having added social value because of their association with farming cultures such as the Funnel Beaker and Battle Axe Cultures (eg, Bergsvik Reference Bergsvik2006; Nyland Reference Nyland2016b). We can question whether it was the axe itself or flint as a raw material that had meaning – and how we should interpret their deliberate destruction or transformation. It is clear that evaluations of distance and interpretation of value are contextually dependent and relational, not only today but also probably so in prehistory.

Besides measured distances to quarries, specific attributes of the lithic debitage have also been considered decisive in determining whether rock should be characterised as local or not. During a large excavation project of inland Mesolithic and Neolithic sites in south-east Norway (Damlien Reference Damlien and Stene2010, 64–5) rock was classified as either local, semi-local, or exotic, based on the amount of primary and secondary flakes in the lithic assemblages. A dominance of primary flakes arguably provides an indication of local origin with rocks having been picked up and initially shaped on the spot. When no primary flakes were found the cores, preforms, blades, or complete tools were assumed to have been transported to the site after initial shaping at other places, the raw material thereby being semi-local from sources not necessarily too far away. Only rock geologically unavailable in this region, the county of Hedmark, was deemed exotic. In this example distance travelled was also seen as a primary attribute that embued rock with significance.

However, this type of categorisation encounters potential problems too. In terms of the occurrence of primary flakes and cortex, different types of rock could require different treatment before transportation. In the Rogaland study a relatively high proportion of the flint flakes and fragments at sites examined displayed areas of cortex. At some, flakes from up to 11 different nodules could be identified (Bang-Andersen Reference Bang-Andersen2008, 85). Perhaps the raw material kept better as pebbles or blocks? Size of the required blank for specific tools could also have been an influential factor in how much initial reduction was needed before transportation. Socio-cultural codes as to where one could perform initial preparation is yet another factor which could have influenced the degree of initial knapping being done on site (Nyland Reference Nyland2016a). For example, during the Middle and Late Mesolithic in western Norway, the initial transformation of blanks into adze preforms was often separated from both quarries and settlements and undertaken at dedicated workshop sites (Olsen & Alsaker Reference Olsen and Alsaker1984), whereas the final completion of the adzes took place at settlement sites.

In order to define a particular rock as valuable, quantification has indicated the uniqueness of certain rock types at sites or in demarcated areas. Archaeological researchers are inclined to determine the most commonly used rock types, such as flint, rhyolite, or recycled stone axes, as markers of identity. In contrast, infrequently found rock types are considered rare. Certain rock types are autochthonous, meaning that they are geologically available only in specific areas or regions. Rare and exotic rock has thus been linked to notions of value or cost, linking again back to the presumed expense and effort made to procure it, such as travelling long distances (eg, Olausson Reference Olausson1983; Gould and Saggers Reference Gould and Saggers1985; Taffinder Reference Taffinder1998; Bergsvik Reference Bergsvik2002; Eriksen Reference Eriksen, Fischer and Eriksen2002). Hence, interpretations of the value of specific rocks vary with the distance removed from the source. However, considering the case studies described above, and the mobility of Late Mesolithic and Early Neolithic communities, how do we regard rock that is ‘rare’ for only parts of the year? In mountainous contexts, the Siggjo rhyolite in the Rogaland study area is, strictly speaking, rare but where it originated (at the coast about 100 km away), the source provided the area with unlimited amounts of it. Its distribution has been used to support discussions of the existence of ethnic groups, tribes, and territoriality on the west coast (Bergsvik Reference Bergsvik2006; Reference Bergsvik and Olofsson2011; Nyland Reference Nyland2016b; Reference Nyland, Teather, Topping and Baczkowski2019a). Considering how, in the Lærdal–Hemsedal study area, flint was brought more than 200 km to the mountains, surely this material should be considered ‘rare’, perhaps even costly? The montane quartzite dominated as a raw material within the mountain plains but was rare at the coast. Hence, definitions of rare, costly, or exotic rock are also dependent on context as well as experienced and interpreted relations. Perhaps distance removed should not be regarded as the most important aspect.

Another issue to consider is whether our modern definition of distance is compatible with that of a prehistoric mobile or semi-sedentary groups. Perhaps was it the concept of a journey – no matter how long – that made rock socially significant (eg, Taçon Reference Taçon1991; Boivin Reference Boivin, Boivin and Owoc2004a)? In areas where quarrying had become entangled in social systems and traditions over time, the practice of quarrying rock could, itself, have been key. Those quarrying loosened rock from a site where the activities of their ancestors were clear and present reinforced social knowledge and relations with each other by maintaining the shared social practice or tradition, as well as relations with the land and ancestors (Nyland Reference Nyland2016b; Reference Nyland, Teather, Topping and Baczkowski2019a). Through quarrying and the meaning associated with the source or site, rock could have gained social value, significance, or, indeed, charisma.

In the case studies presented above there is a tendency for formal tools (points) to be primarily made of flint, whereas flint, quartz crystal, and quartzites were used interchangeably to make more expedient tools such as knives and scrapers. The excavators of site Lok. 1 Smiu (No. 21 in Table 2) noted that flint was more exhaustively exploited than the locally available quartzite (Årskog & Åstveit Reference Årskog and Åstveit2014c). The ground slate points represent the only other lithic technology at the sites but were arguably brought into the mountains ready-made. Thus, except for these, there does not seem to be a rock type exclusively preserved for specific flake and blade tools. Indeed, even presumed prestigious objects, such as ground flint axes imported from South Scandinavia, were transformed into everyday blade and flake tools in the Rogaland region. This does not necessarily mean that these kinds of tools were of less value (cf. Nyland Reference Nyland2016b). Instead we might consider that making everyday tools out of presumed prestigious artefacts could indicate value being linked to transformability, that is, the fact that certain rock could cross categories as well as landscape zones. The Incas, for example, regarded rocks as living beings, ontologically on the same level as humans and animals, and as having transubstantial essence. Thus, significance was independent of form (cf. Dean Reference Dean2010, 5). To understand the use of rock an alternative approach is to move beyond quantification, distance, or predefined categories of value.

HUMAN–ROCK RELATIONS

Following on from the discussion above, we can suggest that the identified patterns of how people used and procured rock should also be considered as demonstrations of ascribed qualities, aspects of social value, or diverging ontological perspectives.

The mountain region in western Norway is a very different landscapes to the coast where people lived for the majority of the year during the later Mesolithic and earlier Neolithic. In the mountains there is an abundance of specific resources but sudden changes of weather and temperature would have made it a risky and hazardous territory to venture into in search of resources. The weather can be acutely felt and fog or clouds often cover peaks and the wider terrain. The rocky landscape has little vegetation, openly revealing geological formations or layers of rocks in different colours. In Rogaland, quartz crystal deposits can be seen from afar, gleaming white in an otherwise barren landscape in shades of grey (Fig. 7). To find good raw material sources required specific local knowledge, as demonstrated by the fact that the quartz crystal sources were largely reported by local hunters and hikers, not by surveying archaeologists (Bang-Andersen Reference Bang-Andersen1975). To this landscape, people transported flint – their familiar rock of choice for most of the year.

Fig. 7. The clouds touch the earth in the mornings in the mountains. Blocks of quartz are scattered and smashed by people searching for crystals (photo: Astrid J. Nyland)

Landscapes have been considered social products created by the actions of people (Gosden Reference Gosden1994, 81). The loosening of rock from a source is a very tangible action, leaving visible and persisting scars. The quarries may have functioned as nodal points in montane landscapes, as predictable and much-used sources of rock, as in Lærdal–Hemsedal. In Britain, researchers have also considered the striking features of landscapes and scenery as explanations for why certain quarries were utilised again and again (Bradley & Watson Reference Bradley, Watson, Rodríguez-Rellán, Nelson and Fábregas Valcarce2019, 151–6). However, this does not offer insight into the use of rock that was not necessarily quarried but perhaps randomly collected. Limited exploitation does not necessarily equal limited importance, the collection of crystal might have been a special-purpose trip made when visiting the mountains in search of what I have termed ‘cloudstones’ (see above).

A quartz crystal prism appears in quartz in its ready-made geometrical form; it catches the light and can reflect a rainbow of colours. It is not uncommon to find unused prisms at Stone Age sites in the mountains close to the sources (Such as site d and h; Table 1) but one also finds them at coastal sites.Footnote ³ Whether these are unused raw material or crystals collected for other reasons is hard to determine. The value of quartz crystals may have been anchored in their association with water, snow, or indeed the landscapes in which they were found, for example where clouds touch the mountains in the morning. In turn, through human engagement and interaction, these potent crystals could become tools that retained their charismatic qualities. In both Mesolithic and Neolithic lithic assemblages one can observe tools, cores, blades, and flakes with and without flat or angular prism sides. Hence, it may not be relevant whether the crystals are worked or not, their value, and perhaps social significance, had already been established. Off course, specific meaning could have changed over time, but the best indication of people finding quartz crystal socially or spiritually important is then perhaps the persistent and widespread use.

In Finland, in a similar manner to the pre-Colombians, the light-capturing qualities of crystals are argued as imbuing the use of quartz with cosmological significance (Mökkönen et al. Reference Mökkönen, Nordqvist and Herva2017). Emphasis is given to the fact that crystals were extracted from underground cavities inside granite and quartz has been suggested by some authors to be central in narratives of the creation of cultural identity relevant to the process of Neolithisation (Mökkönen et al. Reference Mökkönen, Nordqvist and Herva2017). Considering and acknowledging aspects of rock that was used every day, beyond its mechanical properties for knapping (be it quartz crystal or beach flint), can offer a shift in focus for archaeological interpretations. Moreover, and side-stepping our Western bias, acknowledging the possibility of past ontological perspectives affecting practices is vital if we are to gain a broader insight into patterns of use and procurement of lithic raw materials.

FINAL REMARKS

Although we cannot necessarily fully understand what an object meant to its producer and/or user, through identifying varying patterns in use some insights can be gained into the effect that objects had on people (Gosden Reference Gosden2005, 199). If we interpret the varying distribution patterns in a purely pragmatic way the varying sources may indicate an embedded procurement strategy, as described by Binford (Reference Binford1979). That is, hunters would collect rock whenever needed and when opportunity came along during the execution of a subsistence task. As one of Binford’s informants said: ‘Only a fool comes home empty handed’ (Binford Reference Binford1979, 259). Binford argued that only in extreme situations would one set out ‘for the express and exclusive purpose of obtaining raw material’. The present study could be said to correspond with this to some degree. In the Norwegian Late Mesolithic and Early Neolithic it seems that ‘only a fool’ left the coast for the mountains without flint – hunters arrived in the mountains with one set of tools and supplemented them with local rocks as needed during repeated hunting trips in the course of a seasonal stay. However, acknowledging the fact that rocks could affect people in more esoteric ways offers a new angle for exploring past communities in order to explain sometimes puzzling distribution patterns.

Lithic tools and production waste are more than traces of specific past activities discernible as hunting, fishing or gathering, long or short occupation, or snapshots evidencing mobility patterns. They are assemblages and collections of knowledge, social memories, and identity. Lithic assemblages express knowledge of resources in the mountains, of rewarding routes or passages to follow, of people knowing when to visit a certain landscape in order to ensure their safe return to the coast. They also comprise technological knowledge of how to find, collect, quarry, and knap a variety of rock types. Specific rock types, such as rhyolite or flint available as ground axes, were probably associated with groups of people from specific regions (eg, Nyland Reference Nyland2016b) and stones can be regarded as important because of magical or mythological associations (eg, Taçon Reference Taçon1991). Such qualities cannot necessarily be empirically grounded but to gain a fuller understanding of people’s lives in the past, archaeological studies must also include explorations of multi-sensorial aspects of the mineral world (cf. Boivin Reference Boivin, Demarrais, Gosden and Renfrew2004b; Conneller Reference Conneller2011). More often than not such aspects of rock are lost when transformed into archaeological reports: properties are presented before qualities and the objective surpasses the subjective (Nyland Reference Nyland, Day and Skeates2019b). Nevertheless, when the agency or influence of elusive and intangible qualities on human practices is acknowledged, this can offer new explanations of patterns demonstrating the use of specific rock types in certain areas beyond availability and pragmatics. The way humans meet and engage with the mineral world, bundling together memories, knowledge, and experience to make sense of it, has been described elsewhere as a process of appropriation; that is, rock is ‘thought-thickened’ through engagement that in turn imbues it with meaning (Ferguson Reference Ferguson2009, 183; Ferraby Reference Ferraby2015, 15). Hence, fleeting qualities such as beauty, light, or embedded spiritual sentiments can aid us in understanding past practices and human–rock relations.