Department of Archaeology and Ancient History

Multiple Expressions of the Wheel Cross Motif in South Scandinavian Rock Carvings: Case studies of Tanum and Enköping in

Yani Ma

MA thesis 45 credits in Archaeology Spring term 2020 Supervisor: Ing-Marie Back Danielsson Campus Engelska parken

Abstract

Ma, Y. 2020. Multiple expressions of the wheel cross motif in South Scandinavian rock carvings: case studies of Tanum and Enköping in Sweden.

Scandinavian rock carvings can be described as the special rock art languages that were written by prehistoric humans to express their ideas, beliefs and thoughts. Each piece of figurative motif language might tell a prehistoric story about for instance domestic life, social practice, ritual or cosmology. Among different motifs such as ship, human and animal, the wheel cross has received comparatively less attention. The wheel cross has many different variations, where the motif’s ambiguity and possible changing relations to other motifs as well as its relation to the rock itself and landscape over time, have not been studied in greater detail. To fill this gap, this work is aimed at investigating the multiple expressions and possible meanings of the wheel cross motif over time in South Scandinavian rock carvings. Two case study areas with rich rock carvings, Tanum and Enköping, located in the southwest and southeast of Sweden, respectively, are selected. The rock carving materials where the wheel cross motif is present are processed by chorological and chronological studies. A simple and general method is proposed to reconstruct the prehistoric shoreline of Scandinavia. The chronology of the wheel cross motif is analysed with stylistic and shoreline dating methods. The results of the chronological study are then analysed in a cultural-historical- geographical context using an intercontextual interpretation and comparative analysis method. It is suggested that the various visual expressions of the wheel cross, such as the wheel of a wagon or as the shield covering a human torso, have unique symbolic meanings beyond their practical and physical counterparts. With a cosmological meaning as the sun, and a religious meaning as a god, the wheel cross becomes the symbol of moving, fertility, power and life, which has been integrated into not only the rituals but also the domestic life of South Scandinavian society. Although the wheel cross motif is limited in number, its various forms, relating to other motifs, are argued to reflect how the Bronze Age society of Scandinavia imported and developed ideas, artefacts, stories, etc., from foreign cultures in Continental Europe and the Mediterranean area. Like the spoke physically supporting the cart, the wheel cross acts as the lifeline that closely links the landscape, ritual, artefact, human, and other materialities, to mentally support the South Scandinavian society.

Keywords: Rock carvings, Southern Scandinavia, wheel cross, intercontextual, materiality, lifeline, comparative analysis, Tanum, Enköping

Cover image: photo of Tanum 12 panel, adapted from Swedish Rock Art Research Archives.

Master thesis in Archaeology, 45 hp. Supervisor: Ing-Marie Back Danielsson. Defended and approved 2020-06-15. © Yani Ma Department of Archaeology and Ancient History, Uppsala University, Box 626, 75126 Uppsala, Sweden ii

Abstrakt

Ma, Y. 2020. Hjulkorsets mångtydigheter i Södra Skandinaviens hällristningar: fallstudier av Tanum och Enköping i Sverige.

Skandinaviens hällristningar kan på ett sätt beskrivas som ett slags språk, som ristats i berg av förhistoriska människor för att uttrycka till exempel idéer, social praxis, övertygelser och tankar. Varje figurativt motivspråk kan sägas framföra en förhistorisk berättelse om exempelvis det vardagliga livet, ritualer och/eller kosmologier. Bland olika motiv som skepp, människor och djur har det motiv som oftast kallas hjulkors jämförelsevis fått mindre uppmärksamhet. Hjulkorset finns i flera olika utföranden, och just motivets flertydighet i kombination med en studie av motivets relation till andra motiv, själva berghällarna samt plats i landskapet, har inte studerats detaljerat i någon större utsträckning. För att råda bot på denna kunskapsbrist syftar detta arbete till att undersöka motivets flerfaldiga uttryckssätt och möjliga betydelser i södra Skandinavien. För syftet har två fallstudieområden valts i Sverige, vilka bägge har rikligt med hällristningar. Den ena området är Tanum, beläget i sydvästra delen av Sydskandinavien, och Enköping, beläget i dess sydöstra del. Hällristningslokaler där hjulkors förekommer analyseras på flera sätt. En korologisk och kronologisk görs. En enkel och allmängiltig metod utvecklas vidare för att rekonstruera Skandinaviens förhistoriska strandlinjer. Kronologin för hjulkors utförs bland annat utifrån en stilistisk metod och en strandlinjedateringsmetod. Resultaten från den kronologiska studien analyseras sedan i ett kulturhistoriskt- geografiskt sammanhang med användande av en interkontextuell tolkningsmetod och en jämförande analysmetod. Det föreslås att de olika visuella uttryck i vilka hjulkors förekom, såsom exempelvis i form av vagnshjulet, eller i form av en sköld som täcker en mänsklig överkropp eller torso, har unika symboliska betydelser, utöver sina praktiska och fysiska motsvarigheter. Med en kosmologisk betydelse som solen och en religiös betydelse som gud, blir hjulkors-motivet en möjlig synonym för rörelse, fruktbarhet, makt och liv. Som sådan har den inte bara integrerats i ritualerna utan också i det vardagliga livet i de sydskandinaviska samhällena. Även om hjulkorsen är begränsade i antal, kan deras olika former och relation till andra motiv visa aspekter av hur bronsåldersamhället i Skandinavien importerade och utvecklade idéer, artefakter, berättelser och annat från främmande kulturer i Kontinentaleuropa och Medelhavsområdet. Liksom korset i betydelsen ekrar fysiskt stödjer ett vagnshjul fungerar hjulkorset även som en typ av livslinje som kopplade samman en mängd relationella företeelser, som var nödvändiga för det sydskandinaviska samhällets funktion på en mängd nivåer, exempelvis landskapet, ritualerna, artefakterna, människorna och förstås alla dessas samskapande och samverkande materialiteter.

Nyckelord: Hällristningar, södra Skandinavien, hjulkors, interkontextuell, materialitet, livslinje, jämförande analys, Tanum, Enköping

iii

Acknowledgements

I still remember my original motivation of applying for the master study, which was to further expand my archaeological knowledge, strengthen my scientific research ability and improve my English skills. When choosing a topic for the thesis, some of my friends suggested that I extend my bachelor’s thesis on the blue and white porcelain of the Ming Dynasty of China. I knew that would be a much easier way to go not only because I was familiar with that topic but also because most literature was published in my mother tongue. However, I decided to jump out of my comfort zone. As I came to Sweden for study, why not chose a local archaeological topic? Fortunately, I found my supervisor, Ing-Marie Back Danielsson, who encouraged me to take this challenge, and then guided me in the study and research. This thesis would never be completed without her constant help. I am also grateful for having the opportunity to study at the Uppsala University and in Sweden, an amazing university in a beautiful country. Uppsala University provides me not only a great platform to study but also gives me chances to explore more based on my own interests with an individual study plan. Thanks to all the courses, field work and internship, the journey of master study is meaningful and memorable. Finally, I give my deep and sincere gratitude to my family for their continuous and unparalleled love and support.

Yani Ma 2020-05-01

iv

Abbreviations

LN Late Neolithic, 2350–1700 BC BA Bronze Age (Nordic), 1700–500 BC EBA Early Bronze Age, 1700–1100 BC Period I 1700–1500 BC Period II 1500–1300 BC Period III 1300–1100 BC LBA Late Bronze Age, 1100–500 BC Period IV 1100–900 BC Period V 900–700 BC Period VI 700–500 BC PRIA Pre-Roman Iron Age, 500–1 BC SHFA1 Swedish Rock Art Research Archives FMIS2 Swedish National Heritage Board’s Archaeological Sites and Monuments Database and Associated Applications RAÄ Swedish National Heritage Board’s Archaeological Sites and Monuments Register GIS Geographic Information System CRS Coordinates Reference System GET3 Geodata Extraction Tool BP Before Present (AD 2000) m.a.s.l. Meters Above (present) Sea Level LUV Land Uplift Velocity

1 http://www.shfa.se/ 2 http://www.raa.se/tag/fmis/ 3 http://maps.slu.se/get v

Terminology

Rock carving A prehistoric image created by removing part of a rock surface through carving, also named as rock engraving or petroglyph (Swedish: Hällristning). Motif A figure created on the rock that has symbolic significance. There are many different kinds of rock carving motifs, such as ship, human, animal, circle and so on. When referring to a motif, for instance a ship motif, it is common to just use ship instead of ship motif. Wheel cross A motif generally consisting of an equilateral cross inside a circle ( ). However, it has many variations with different meanings and does not necessarily represent a wheel. Depending on the context, it is also named as sun cross, sun wheel or ring cross. Lifeline Depending on the context, in the thesis it means either the connecting lines in a closed motif that indicate the structure for moving, or the analytical tool for studying the connections between different motifs and materialities. Tanum A municipality4 of Sweden, one of the two case study areas of this work. This also applies to Enköping5. A site in Tanum parish, without the prefix RAÄ. Tanum 121:1 A panel of the Tanum 121 site. There could be several panels in one site, for instance, Tanum 121:2 is another panel. ‘Tanum 121:1’ sometimes is abbreviated as ‘Tanum 121’. Tanum 121:1(1)6 A unit of the Tanum 121:1 panel. There are usually several units in one panel, for instance, Tanum 121:1(2) is another unit.

4 Tanum is a municipality (Swedish: kommun) of Västra Götaland county (Swedish: län). Västra Götaland is one of the 21 . Sweden is also divided into 25 historical and cultural regions (provinces) that have no administrative function. In the thesis, when ‘Tanum’ is followed by a number, it means a parish (Swedish: socken) of Bohuslän province (Swedish: landskap). The size of is much larger than that of Tanum parish. 5 Enköping is a municipality of Uppsala county. Enköping parish is located in southwest of Uppland province. 6 The RAÄ number of a unit in the nil layer of FMIS. vi

Contents

Abstract ...... ii Abstrakt ...... iii Acknowledgements ...... iv Abbreviations ...... v Terminology ...... vi 1. Introduction ...... 1 1.1. Aim ...... 3 1.2. Theory ...... 5 1.2.1. Materiality ...... 5 1.2.2. Lifeline ...... 6 1.3. Method ...... 6 1.4. Material ...... 8 2. Background ...... 9 2.1. Chorology and chronology of Scandinavian rock carvings ...... 9 2.1.1. Stylistic and iconographic study ...... 9 2.1.2. Prehistoric landscape and shoreline of Scandinavia ...... 12 2.2. Interpretation of Scandinavian rock carvings ...... 16 2.2.1. Multiple perspectives of interpretation ...... 16 2.2.2. Previous understanding of the wheel cross motif ...... 18 2.3. Previous comparative studies of Scandinavian rock carvings...... 19 3. Historical and geographical background ...... 20 3.1. Landscape and historical remains...... 20 3.2. Location of rock carvings ...... 23 4. Case study of Tanum ...... 28 4.1. Typical rock carving motifs ...... 28 4.1.1. Cup mark ...... 35 4.1.2. Ship ...... 35 4.1.3. Human ...... 36 4.1.4. Animal ...... 36 4.1.5. Foot soles ...... 36 4.2. Multiple expressions of wheel cross motifs in Tanum ...... 37 5. Case study of Enköping ...... 42 5.1. Typical rock carving motifs ...... 42 5.1.1. Cup mark ...... 49 5.1.2. Ship ...... 49 5.1.3. Foot soles ...... 49 5.1.4. Animal ...... 49 5.1.5. Human ...... 50 5.2. Multiple expressions of wheel cross motifs in Enköping ...... 50 6. Discussion ...... 56 vii

6.1. Wheel cross motifs in a cultural-historical-geographical context ...... 56 6.2. The wheel cross beyond expressions and representation ...... 60 6.3. Comparative analysis of rock carvings of Tanum and Enköping ...... 61 7. Conclusion ...... 67 8. Summary ...... 68 Bibliography ...... 69 List of figures ...... 74 List of tables ...... 76 Appendix 1: Working with FMIS database ...... 77 i. Introduction ...... 77 ii. Automatic statistics ...... 77 iii. VBA code ...... 78 Appendix 2: Reconstructing historic shoreline ...... 80 i. Required data ...... 80 ii. Steps ...... 80

viii

1. Introduction

Most of the present coastal terrain of Southern Scandinavia was under water in the prehistoric period. Due to post-glacial rebound, the land uplifted with a velocity of a few meters per thousand years. The glacial striation made the surface of new coastal rock smooth and flat, ideal for the prehistoric humans to carve figures or images in (Bertilsson 1987: 32; Coles and Gräslund 2000: 11; Coles 2005: 9). Rock carvings are known throughout Scandinavia, from the Stone Age and onwards, but are perhaps most known from the Bronze Age (1700–500 BC). The northern traditions of rock carvings are generally older than those of Southern Scandinavia. Similar figures of the rock carvings have been found in different countries or even different continents (Goldhahn et al. 2010; Nimura 2012, 2015; Valdez-Tullett 2019). It indicates, to some extent, that the prehistoric humans in different regions interacted through communication, immigration or vast networks. There was no written language in Scandinavia during the BA. Carving figurative motifs on the rock was a special form of expression that had its own meaning. Thus, compared to other historical remains such as settlements and graves, rock carvings have their unique archaeological value. Rock carvings have been considered as symbolic signs (Coles and Gräslund 2000: 78; Vogt 2011: 17) and their value is not limited to the superficial meaning but associated with communication, symbolism, competitive and social practice, etc. For instance, Gell (1998: 29) views rock carvings as an agency for the communication between the society and individuals, whereas Fahlander (2012: 100) approaches rock carvings as a meaningful practice, such as a competitive practice, rather than a communication or narrative device. Moreover, Coles and Gräslund (2000: 99) point out that it is impossible to read rock carvings as a story or a set of meanings, but rather that rock carvings can be seen as symbols to trace the concerns of the society. Therefore, the interpretation of figurative rock carvings is not only important to archaeologists but also helpful to ethnographers, anthropologists, and museum curators from many different disciplines (Nimura 2015: 14).

(a) (b) Figure 1. (a) Photo of an original rock carving panel (Tanum 311); (b) Photo of a renewed rock carving panel (Bottna 43). Credit: (a) in the public domain via Wikimedia Commons; (b) Originator: Torsten Högberg. Equally, from the perspective of heritage protection, it is urgent to interpret rock carvings since the action of frost and plant roots as well as acid rain and other pollutants could cause damages to rock carvings (O. Walderhaug and Walderhaug 1998: 119). The damages could make the images on the rock unrecognizable, for instance, as shown in Fig.1a. Although some field work has renewed the images using chalk or paint to make them clear (e.g. see Fig.1b), it could mislead the interpretation of the original images to some degree (Coles and Gräslund 1

2000: 10). For this reason, it is of importance to document rock carvings in a high qualitative and professional way, to ensure that interpretations of them are of equal high quality. In the past half century, a lot of research work has continuously discussed the Scandinavian rock carvings (e.g. Kjellén 1976; Milstreu 1977; Malmer 1981; Bertilsson 1987; Sognnes 1995; Coles and Gräslund 2000; Sognnes 2001; Ling 2008; Goldhahn et al. 2010; Wessman 2010; Vogt 2011; Ling 2013; Nimura 2015; Petersson and Christina 2015; Skoglund 2016; Goldhahn 2019). Most of the previous works focus on the predominant motifs such as ship ( ), whereas other motifs such as the wheel cross generally have received much less attention. Fig.2 illustrates some examples of the wheel cross motif ( ). It shows that the wheel cross has a close connection to other motifs such as animal ( ), ship ( ), human ( ) and it can also exist in clusters individually. Although the wheel cross motif has been partially discussed in the literature (e.g. Malmer 1981; Winter 2001; Kaliff 2007; Lahelma 2017), the interpretations were made only for specific cases, lacking a systematic and dynamic overview picture.

(a) (b)

(c) (d) Figure 2. Examples of wheel cross motifs: (a) in Tanum 465; (b) in Tanum 12; (c) in Boglösa 131; and (d) in Tanum 470. Credit: all figures are adapted from SHFA. Originator: (a) Sven-Gunnar Broström; (b) anonymous; (c) Eva Kjellén; and (d) Ellen Meijer. To bridge this gap, this thesis focuses on specific sets of motifs that commonly are grouped together under the label ‘wheel cross motif’. In its different shapes as a circle ( ), a wheel ( ), a human ( ), two foot soles together ( ) etc., this multi-expressive motif offers a great potential to obtain a deeper understanding of its possible shifting meanings and agency throughout time and space. In extension, new understandings or facets of the BA society, with special regard to figurative rock carvings, can be obtained.

2

1.1. Aim The thesis is aimed at investigating multiple expressions and shifting meanings of the wheel cross motif throughout space and time in South Scandinavian rock carvings. Two typical regions with extremely rich rock carvings are selected as case study areas; Tanum and Enköping municipalities, as shown in Fig.3.

Figure 3. Overview of the location of study areas, Enköping and Tanum municipalities, in the map of Scandinavia. To be consistent, 21 counties, instead of 25 provinces of Sweden are shown. Made by the author. Tanum is located in southwest Sweden and it is characterized by fissure valleys and granite hills as well as the interconnected lowland area, as shown in Fig.4 (a,c). Enköping is located in the southeast of Sweden and it is characterized by the shoal and low-altitude bedrock, as shown in Fig.4 (b,d). The differences between the terrains and geographical positions make the comparison especially fruitful since it contributes to the understanding of how local natural landscapes affect the rock carvings. In addition, these two regions were located on the coastal route of cultural transmission of Scandinavia (Hjarthner-Holdar and Risberg 2001: 37–38; Kristiansen and Larsson 2005: 129), which implies their regional interaction with each other and with the outside world. Tanum is adjacent to the North Sea, and Enköping to Lake Mälaren that in turn is connected to the Baltic Sea. Consequently, the 3

similarities shared in these regions might show a common or similar social tradition, ritual practice and religion belief. Variations in the design of the same figure among different areas, on the other hand, may indicate the local characters.

(a) (b)

(c) (d) Figure 4. Elevation and coastal view of (a,c) Tanum and (b,d) Enköping. (a, b) are made by the author. (c, d) are taken from Google Street 2019.

To be more specific, the basic questions of this thesis work are summarized as below. These questions are formulated in a way that is in line with the point of departure of rock carvings study proposed by Malmer (1981: 102). 1. Did different expressions of the wheel cross in South Scandinavian rock carvings share the same meaning? If Yes/No, what might their meanings have been? 2. Did the meanings of the wheel cross change in different BA periods, and how? This question thus focuses on how the wheel cross motif has varied its expression throughout time. 3. Did different regions share the same answers to the above questions? In other words, what are the differences and similarities about the expressions of the wheel cross in different study areas? This question explores the motifs’ variation in space. The outline of the thesis is summarized as follows. In the next subchapters (1.2, 1.3, and 1.4), the theory, method and material which this study is based on, are described. Chapter 2 introduces the previous related work in the literature as well as the detailed research gap that motivated the specific work. Chapter 3 is dedicated to introduce the historical and geographical background of rock carvings in the studied areas. Other detailed results are presented in Chapters 4 and 5, followed by Chapter 6, discussing possible answers to the research questions. Finally, Chapters 7 and 8 summarize the main conclusions and content of this study.

4

1.2. Theory The theories of materiality and lifeline provide the theoretical basis through which to study the shifting meanings of the wheel cross motif and its shifting relations to other aspects. The two theories are described as follows.

1.2.1. Materiality Rock carvings are prehistoric images created by removing part of a rock surface. In one way, it can be seen as a kind of writing, like any other activity with marks and imprints, such as the dream (Derrida 1978: 209). Once carved on the rock, the meanings of the images were gone with the carver who made them. The only thing remaining is the physical and meaningless modified rock surface, i.e., the material. The interpretation by humans, here referred to as material culture or rock art, adds a reproduced meaning to the material. Naturally, there are numerous interpretations or different ways of seeing, resulting in different art and styles (Back Danielsson et al. 2012: 3). For instance, an animal-like motif ( ) could mean a horse, an elk, a cow, or a sheep, or allude to another feature, etc. Such interpretations are made by humans and they reflect the thoughts and practice of the representor (Fahlander 2004; Gibson 2006: 172). The interpretation should also take into account the background and relation to other aspects. For example, the only animal motif found in EBA Denmark is the horse and it is naturally interpreted as the symbol of a warrior, whereas the same horse motif found in Northern Scandinavia has a different meaning since it is among many different species of hunted animals (Malmer 1981: 108). In other words, the meaning of rock carvings shifts with the change of locations. It is the environment (landscape, culture, society, economy, religion, etc.) that enriches the interpretation. Such multidimensional interpretation is referred to as the materiality, which focuses on the processual and relational aspects between the material and other features (e.g. Demarrais et al. 2004; Fahlander 2004; Tilley and Bennett 2004; Miller 2005: 4; Hodder 2006; Tilley et al. 2006: 3; Ingold 2007; Meskell 2008; Sofaer 2008). The connection among the concepts of material, material culture and materiality is illustrated in Fig.5.

Society

Data Interpretation Landscape Relation Human (rock carvings) (meaning)

Artefact

Material Material Culture Materiality

Figure 5. Schematic diagram showing the connection among material, material culture and materiality. Made by the author. Materiality includes not only the practice of the human but also the relationship between the human and the society (e.g. Demarrais et al. 2004: 2; Tilley and Bennett 2004; Miller 5

2005: 4; Ingold 2006; Tilley et al. 2006: 3; Ingold 2007). Materiality expresses the relational aspects between humans and things, and it has been described as co-presence (Meskell 2008), bundling (Tilley et al. 2006: 3) or things as co-producers of art and styles (Latour 1992: 165). It is claimed that artefacts, landscape, settlements, vegetation, nature material, etc., are all involved with materiality (Fahlander 2008b: 135).

1.2.2. Lifeline The wheel cross is a motif generally consisting of an equilateral cross inside a circle ( ). The two crossed lines make the motif special and different from other motifs, such as ship ( ), cup mark ( ), animal ( ), weapon ( ), etc. The wheel cross may originally be carved as a circle, and then be recut inside the circle. The addition or change to the old symbols creates a new symbol, which could probably be a tradition (Milstreu 2017). The pattern of connecting lines in a closed motif is similar to those lines marked in the internal of human and animal bodies (e.g. elk ), referred to as the lifeline, which are found in large quantities in north Scandinavian rock carvings (e.g. Hallström 1938; Vastokas and Vastokas 1973; Mikkelsen 1983; Ramqvist et al. 1985; Hultkrantz 1986; Mikkelsen 1986; Whitley 2001; Goldhahn 2002: 46). The tradition of the internal marking depicted on the rock carvings continues to 500 BC (Malmer 1981; Sognnes 1995). Lifeline is also found on other objects such as animal bones (Olsen 2003; Helskog 2014) and in other countries such as Russia (Goldhahn 2002: 47). The lifeline reflects multiple expressions, not only in the active life such as the blood, the skeleton, the organ, etc., but also in life’s interaction with the society, the natural circumstances and the global world. Apart from the visual perspective, it is also natural and interesting to extend the concept of lifeline to study the non-visual connection. For example, Goldhahn (2002: 49) extends the interpretation of lifeline to include the non-visual aspect. It is found that quite a number of the animal motifs marked with the lifeline are located near rapids where the river roars, making the visitors feel the live audio-visual effect (Goldhahn 2002: 54). Meanwhile, the audio-visual sense gives a good inspiration about how different motifs – the human, ship, and animal etc. – communicate with each other. To this end, the lifeline itself is an agency that links different materialities. In this work, the wheel cross itself can also be viewed as a lifeline that connects different aspects. It can be a physical lifeline that relates different motifs to one another; it can be a visual lifeline that links humans and animals, landscapes and the environment. It can also be a virtual lifeline that integrates the society with culture, ritual and cosmology, etc. Consequently, the concept of lifeline enables the wheel cross to work as a button with which to open the door to a deeper understanding of rock carvings and societies of Scandinavia.

1.3. Method While the theoretical concepts accounted for above provide tools through which analyses are made, further methods used in this thesis consist of the chorological/chronological analysis, ideological analysis, and comparative analysis. The chorological analysis, i.e., spatial analysis, is the first step of a chronological analysis (Malmer 1981: 1). The chorological analysis is to solve the question where the material was located and the chronological analysis deals with the problem when the material was made, i.e., dating. The dating of rock carvings is one of the fundamental questions in rock carvings study. The traditional way of dating rock carvings is based on the typology or style (Sanz and Fiore 2014: 7105). Different from the typological dating, the stylistic dating requires the scholar to form an intuitive interpretation of the stylistic intentions of the artist who made it (Malmer 1981: 18). However, both of them are based on identifying similar characteristics. For instance, similar motifs on the rock were also found in other datable bronze pieces such as 6

razors, tweezers and knives (Kaul 1998: 74; Coles and Gräslund 2000: 26). However, such dating can be problematic if the artwork cannot be assigned to a specific stylistic set or unit (Sanz and Fiore 2014: 7108). This method, referred to as the informed method, is dependent on the experience and knowledge of the interpreter (Nimura 2015: 2). In contrast to the informed method, the dating of rock carvings can also be done with a more objective method, i.e. the formal method (Nimura 2015: 2). The method is based on the information that is independent of the experience of the interpreter, for instance, the popular radiocarbon (14C) dating method. Unfortunately, it is impossible to date a rock carving motif directly using the 14C dating method since the rock is an inorganic matter. However, rock carvings have their own special feature – a kind of historical remains that cannot be moved easily. It implies that the geographical information of rock carvings has significant values. Previous studies have found that Scandinavian rock carvings were located close to BA water regions. With the land uplifting, the prehistoric shoreline of Southern Scandinavia was also ‘elevated’ as the contour line of present elevation. Thus, the altitude of the rock panels (with respect to the present sea level) determines the upper bound date of when the figure was made because it was not possible to carve the rock below the sea level. Therefore, the altitude of the rock carvings provides another dating method, as a complementary to the traditional stylistic dating method. The use of shoreline displacement method can be dated back to the beginning of the last century and has been used by many rock carvings scholars (for a brief review, see Ling 2013: 15–21). The chorological analysis and ideological analysis are two general ways of interpreting rock carvings, as summarized by Bertilsson (1987: 3–5). The main difference is that chorological analysis focuses more on the material, whereas the ideological analysis focuses more on the material culture, as shown in Fig.5. There are many variants of the ideological analysis, such as the comparative study and holistic interpretation (Hultkrantz 1986: 42), the contemporary cultural context analysis (Malmer 1981: 109) and the intercontextual strategy of interpretation (Kristiansen and Larsson 2005: 12). The intercontextual strategy of interpretation suggests putting archaeology in a holistic cultural historical framework (Kristiansen and Larsson 2005: 372). Based on this framework, the theoretical and interpretive studies are done from the perspective of comparative, contextualised culture history. In addition to the cultural and historical perspective, this study also includes the geographical perspective to shed more light on the locations of rock carvings as well as the locations of case study areas. The comparative analysis of rock carvings in different regions, also named as regional study, has been used successfully by a number of researchers (e.g. Malmer 1981; Wessman 2010; Nimura 2015; Valdez-Tullett 2019). The regional study is based on the data from localized studies and is aimed at finding possible inter-regional connections. For instance, Nimura (2012; 2015: 14) performed a comprehensive study of different rock carvings sites in Denmark, Norway and Sweden. It is shown that the shoreline displacement dating method can be applied to the whole Scandinavian area rather than a specific location. The study gives a full picture of the distribution of Scandinavian rock carvings, i.e., the close connection with the sea. A similar regional study has also been done recently to study the rock carvings in the Atlantic areas (Valdez-Tullett 2019: 6). In the study, the landscape analysis and graphic analysis methods are used to find regional connections and regional characteristics. Usually, the comparison of different regions focuses more on similarities and it contributes to the general understanding of cultural transmission such as the Indo-European culture diffusion (Kristiansen and Larsson 2005: 129; Kaliff 2007: 28). Wessman (2010: 3) argues that differences have received much less attention than the similarities in regional studies of rock carvings. However, no matter what the foci are, similarities or differences, the approaches are essentially the same since once the similarities are well summarised, what is left would be the differences. In general, the similarities indicate the shared values, whereas the differences reflect the local adaptation.

7

1.4. Material The material data of rock carvings, including geographic coordinates, motif types and quantities, is downloaded from the Swedish National Heritage Board’s Archaeological Sites and Monuments Database and Associated Applications (FMIS). It includes 1489 and 1847 rock carving panels from Tanum and Enköping, respectively. Using the data from FMIS is not straightforward and the procedure used in this work is summarized in Appendix 1: Working with FMIS database. The images of the motifs are taken from the related references and the Swedish Rock Art Research Archive (SHFA). The landscape maps of Sweden as well as the elevation data with 2m×2m resolution are acquired from the GET portal. The landscape maps of Norway and Denmark are downloaded from EuroStat7. The data of Scandinavian land uplift velocity is retrieved from the NKG working group8. The software package used in this study is the QGIS 3.49 and Microsoft Excel 2010. It is necessary to keep in mind the limitations of material. First, due to limited time, the case study areas of this thesis do not include other parts of Southern Scandinavia, such as Skåne in the south of Sweden, and the related regions in Norway and Denmark. However, examples from those regions might be given in the discussion even though the detailed data are not included. Therefore, this thesis serves as the first step to investigate expressions of the wheel cross in Southern Scandinavia. In an extended study, other data, especially those from Denmark should be included in order to get a full picture. The second limitation comes from the FMIS database. The rock carving data documented in the FMIS database is not complete and the descriptions for some records are missing. The data in FMIS have been documented by different persons, and they registered the data based on their own understanding, sometimes with quite different categories. For instance, the wheel cross is categorized into the circle (Swedish: cirkel) and there are a few records having the same RAÄ number and category but different number of motifs. It is difficult to know if it is an error or if there are different meanings for the same type of motif unless you check them one by one with the original documentation. However, such different classification will have little effect on the result of relative statistics and analysis (Bertilsson 1987: 87). When analysing the data in FMIS database, it has been found that some motif records deviate from the original documentation. For example, Bottna 31:1 panel was registered with two wagon and 34 cup mark motifs in FMIS, whereas it actually has four wheel cross motifs and 38 cup mark motifs; Bottna 33:1 has 13 ship motifs in FMIS, whereas it has 12 ship motifs in the original documentation. However, it is believed that such defects are limited and will have negligible effect on the overall statistics. Another limitation of the FMIS database is that the data of geographical location is only available for each individual panel, but not for each unit. This limitation prevents this study from doing spatial analysis for each group of motifs. However, it is still valid to estimate the average age of a rock carving panel. The limitation also comes from the resolution of the elevation data. The highest resolution that can be accessed is 2m×2m from the economic map, which might cause large errors when investigating a specific rock carving panel (Ling 2006: 5). Such errors would be more significant when the rock is located on a steep terrain. However, the 2m×2m resolution is generally enough when performing an overall analysis.

7 https://ec.europa.eu/eurostat/web/gisco/geodata/reference-data/administrative-units-statistical-units/countries 8 https://www.lantmateriet.se/sv/Kartor-och-geografisk-information/gps-geodesi-och- swepos/Referenssystem/Landhojning/ 9 http://qgis.org/ 8

2. Background

This chapter accounts for the background for several aspects of rock carvings that are of relevance for the thesis. It includes the chorology and chronology of Scandinavian rock carvings, earlier interpretations of Scandinavian rock art and lastly an account of previous comparative studies of Scandinavian rock carvings. It is worth mentioning that a new general method of reconstructing the shoreline is presented in Subchapter 2.1.2, whereas other subchapters are devoted to present the history of the related research.

2.1. Chorology and chronology of Scandinavian rock carvings

2.1.1. Stylistic and iconographic study The systematic and pioneering work in the chorological and chronological study of rock carvings in Northern Europe including Scandinavia are conducted by Malmer (1981). Based on the registration and classification of numerous rock carvings data by himself, Malmer performed a detailed statistical calculation and chorological analysis for various motifs, i.e., ship, cart, weapons, feet, circle, human, animal, etc. The wheel cross motif was categorized into either cart or circle groups in Malmer’s study (Malmer 1981: 45, 67). As mentioned in Subchapter 1.3, rock is a natural inorganic matter which cannot be dated with 14C dating method. The first step of the traditional dating method is to perform a stylistic analysis (relative dating). In all different types of motifs, the ship is the most common and complex one. As shown in Fig.6, the complex structure of the ship makes it possible to classify and identify different ship figures on different rocks and objects. Prow

Gunwale Fore stem Aft stem Hull

Keel extension Keel Stabilizer

Figure 6. Structure and components of a BA ship. Remade by the author after Ling (2008: 187). The second step is to link, compare and refer to other historical remains such as metal that can be dated with the absolute dating method. For instance, some ship motifs were also found in the Rørby sword (Malmer 1981: 14), the Wismar trumpet (Malmer 1981: 15), the Kivik grave (Malmer 1981: 17) and the Sagaholm barrow (Malmer 1981: 36) that were dated to the EBA. In general, although with some exceptions, the older a ship is, the simpler its style is. Some examples of ship images and their features from Tanum and Enköping are shown in Table 1 and Table 2, respectively. As the images show, ships with stem raised and slight keel extension belong to Period I and II of the EBA. Ships with higher keel extensions and stems topped by horsehead or birdhead design appeared in Period III and IV. Ships from Period V and VI showed more inward stems and more elaborations and decorations such as complex

9

stems and curved keel extension (Coles 2005: 22). During the PRIA, ships were fully outlined but carved with parallel keel and gunwale line, narrow hulls without crew strokes (Coles 2005: 22). It is worth mentioning that some ships from the EBA were remade by adding details to have characteristics belonging to the LBA, which makes it difficult to perform a chronology study (Milstreu 2017: 46). It is found that those remade ship motifs usually have deeper depth and higher aesthetic norms (Ling 2008: 38). In a similar way, other motifs could also be dated. For example, humans with accentuated calves ( ) were introduced in Period II (Malmer 1981: 83). Scabbard equipped on humans with winged-chape form ( ) are dated to the LBA (Coles 2005: 42, 80). Huge curved battle axes, usually held by humans ( ), are dated to EBA Period I–II (Coles 2005: 79–80). Humans carrying a round shield ( ) are dated from Period III to V, whereas a human carrying a square shield and sitting on a horse ( ) are dated to PRIA (Coles 2005: 47–48). Four wheeled wagons ( ) are usually dated to the LBA (Johannsen 2011: 96; Kveiborg 2019: 256). Table 1. Example of ship motifs in different periods from rock carvings of Tanum. Interpretation and dating are mainly based on the work by Coles (2005).

Period, RAÄ Images Features Credit BC number

High raised stems, I, II Kville 157 horizontal and slight 1700–1300 keel extension

Slight higher keel III, IV extension, horsehead Kville 149 1300–900 or birdhead on stem Ellen top Meijer

Complex stem with I, VI curved lines, Tanum 192 900–500 prominent keel extension

10

Period, RAÄ Images Features Credit BC number

Bertil Tanum 75 Almgren Fully outlined, parallel keel and gunwale line, narrow PRIA hulls and shield-like 500–1 design, always unmanned (lacking crews) Tanum 208 Gerhard Milstreu

Table 2. Example of ship motifs in different periods from rock carvings of Enköping. Interpretation and dating are mainly based on the work by Coles and Gräslund (2000).

Period, RAÄ Images Features Credit BC number

Flat keel extension, I, II Boglösa Peter out turned prows, 131 Skoglund 1700–1300 slight raised stems

Animal heads on the III, IV Vårfrukyrk prow, inward keel a 176 1300–900 extension and prow

Einar Kjellen

Slightly upturned keel V, VI Boglösa extension with more 109 900–500 elaboration

Different from the ship motif, the wheel cross is lesser not only in quantity but also in stylistic variations. Therefore, it is difficult to perform chorological and chronological study for the wheel cross directly. However, since the wheel cross usually connects to other motifs, one way is to date the wheel cross by referring to the nearby datable motif, such as the ship. 11

Nonetheless, such a method might sometimes cause misdating since, for instance, the rock carvings for large panels (e.g. Tanum 12, 311; Boglösa 138) could be made on different occasions, as in different BA periods (Rédei et al. 2019: 3). Yet, it could be said that it is always difficult to know if the wheel cross, as any other motif, was made at the same time as other nearby motifs. In this case, an ambiguous dating is suggested based on other available information. When dating of other motifs is not possible or the wheel cross appears alone, the shoreline displacement dating method is employed to provide a reference and this is introduced in the next subchapter.

2.1.2. Prehistoric landscape and shoreline of Scandinavia The land of Scandinavia was loaded with kilometre-thick ice during the last ice age. As the ice began to melt about 20 000 years ago, the pressure on the Earth's crust eased and the land of Scandinavia began to rise (rebound). About 10 000 years ago, the last part of ice from Scandinavia disappeared. However, due to the Earth's slow-moving interior, it takes a very long time before it returns to its equilibrium position (Milne and Shennan 2013; Vestøl et al. 2016). Until now, the land of Scandinavia has risen several hundred meters and a few tens of meters are expected to remain. As the land approaches its equilibrium position, the land uplift velocity (LUV) decreases gradually. The LUV varies significantly from the north to the south of Scandinavia. At present, the land arises fastest in the northeast of Sweden with a velocity of 10 mm/year, whereas it is less than 1 mm/year in the south of Denmark (Hünicke et al. 2015: 160), as shown in Fig.7, where the distribution of present LUV relative to the geoid10 is shown. The coastal landscape has been changed significantly since the BA due to the effect of land uplift. There is a long history on the use of shoreline displacement as a tool to analyse BA rock carvings. With gradually increased exchange of knowledge between archaeology and geoscience in the last century, many important results were published. For a comprehensive review of the related work, the reader is referred to Ling (2008: 15–58). The BA shoreline (m.a.s.l.) as a function of time for the study areas is available in the literature, for instance, in Tanum (Ling 2008: 54) and in Enköping (Ling 2013: 20). However, the previous studies used a uniform shoreline displacement data for a region while it did not consider the spatial variations. As can be seen from Fig.7, the difference of LUV between the north and south of Tanum or Enköping is roughly 0.5 mm/year, corresponding to a difference of 2 meters in BA shoreline altitude. To take this effect into account, a more general approach has to be made to reconstruct the BA shoreline, especially for large areas. Another difficulty encountered in the previous studies, is that the historic shoreline in some cases requires correction. For instance, Ling (2008: 54, 75, 79, 82, 103, 114) has used the study of Berntsson (2006) to correct the results by adding 1.8 m extra to the shoreline altitude computed from Påsse’s model (Påsse 2001). The correction is critical to the chronological study based on shoreline displacement. Therefore, a general approach which does not require any correction is required. According to the definition made in (Vestøl et al. 2016: p2, presentation), three different LUV have to be distinguished. The absolute LUV is relative to the Earth’s centre of mass and the levelled LUV is relative to the geoid. The apparent LUV is relative to Mean Sea Level at a specific place over a time period and it changes randomly due to unpredictable contemporary climate factors such as ice melting, temperature increase and wind, etc. The levelled LUV can be converted to apparent LUV by applying a tide gauge. The tide gauge can measure the eustatic sea level, i.e., the relative change of Mean Sea Level. The differences of the three LUV are illustrated in Fig.8. In this study, the levelled LUV is used and the tide variations are neglected for several reasons. First, the tide gauge data varies significantly in space and time and the prehistoric data is generally unavailable. Second, tide variation is relatively much

10 The geoid is the sea surface with the influence of the gravity and rotation of Earth and without taking into account other influences such as winds and tides. The surface is extended through the continents. 12

smaller than shoreline variations caused by land uplift about 7000 years ago (Clemmensen et al. 2001: 329; Eronen et al. 2001: 28). Moreover, tide variation changes randomly even during several days, which is dramatically shorter than the time window of BA.

Figure 7. Land uplift velocity of Sweden with the study areas marked. Made by the author. The LUV data is based on the work of Vestøl, O., et al (2016) and the elevation data are retrieved from GET.

13

Rock

Sea Land Absolute land uplift (relative to the Earth’s centre of mass ) Time

Figure 8. Schematic diagram showing sea level variations during land uplift. Made by the author. It is not possible to carve on the rock which is below the sea level as shown in the left figure. Therefore, the altitude of a rock carving determines the upper bound of its age. When reconstructing the BA shoreline, it is of importance to take into account the change of LUV with time. In other words, the BA LUV is larger than the present LUV. The accurate calculation of LUV variation is out of scope of this study. However, there are some simplified models for instance by using a linear function to approximate the LUV variation in time, resulting in a quadratic function of the shoreline change with time. By defining the present time as zero, the shoreline is expressed as 푆(푡) = 푎 ∙ 푡2 + 푏 ∙ 푡 (1) 퐿푈푉(푡) = 푆′(푡) = 2푎 ∙ 푡 + 푏 (2) where 푆(푡) is the shoreline, 푡 the time11, a and b the coefficients. Assuming LUV 3000 years ago is ¼ larger in magnitude than that of today (Whitehouse 2018), 푆(푡) can be derived as 퐿푈푉(0) 푆(푡) = ∙ 푡2 − 퐿푈푉(0) ∙ 푡 (3) 24000 where 퐿푈푉(0) is the present LUV (positive number). With expression (3), the historic shoreline (m.a.s.l.) of Tanum and Enköping can be plotted in Fig.9. The average shoreline altitude corresponding to different periods is listed in Table 3. It is worth mentioning that the calculated averaged BA shoreline is in good agreement with previous studies (Ling 2008: 54; Plikk 2010; Sund 2010; Ling 2013: 20). However, the approach proposed here is more general and accurate since it takes into account the local position of each rock carving panel and it only depends on the present LUV. With the above method, it is not difficult to reconstruct the prehistoric shoreline and landscape, see Appendix 2: Reconstructing historic shoreline. As shown in Fig.8, the present elevation of a rock panel where a motif is located, can be used to roughly estimate the oldest age when the motif was made. The most famous rock panel that was dated using the shoreline displacement method is Tanum 311 (Ling 2008: 91, 237; Nimura 2015: 4).

11 t = -4000 for 4000 BP and t = 0 for present. 14

60

Enköping 50 Tanum

40

30 m.a.s.l.

20

10

0 7000 6000 5000 4000 3000 2000 1000 0 |<- Neolithic ->|<- Bronze Age ->|<- Iron Age ->|<- Historic time Years, BP

Figure 9. The historic shoreline of Enköping and Tanum. Made by the author. The dot line and solid line are computed with the averaged LUV of Enköping and Tanum, i.e., 5.75 mm/year and 3.75 mm/year, respectively. The dash lines show the upper and lower limits within the corresponding regions. Table 3. The computed average shoreline altitude of Enköping and Tanum during different periods.

Average shoreline altitude, m.a.s.l. Age/period Years Years, BP Enköping Tanum Neolithic 5000–1700 BC 7000–3700 52–24.6 m 34–16 m I 1700–1500 BC 3700–3500 24.6–23 m 16–15 m Early II 1500–1300 BC 3500–3300 23–21.6 m 15–14 m III 1300–1100 BC 3300–3100 21.6–20 m 14–13 m BA IV 1100–900 BC 3100–2900 20–18.7 m 13–12 m Late V 900–700 BC 2900–2700 18.7–17.3 m 12–11 m VI 700–500 BC 2700–2500 17.3–16 m 11–10 m Iron Age 500 BC–AD 1000 2500–1000 16–6 m 10–4 m

However, one should be always careful when using the shoreline dating method since it only suggests the upper bound of the age. A rock carving located in a lower area could not be made in earlier times but not vice versa, i.e., the one located at a higher altitude can be made at a later time. For example, Fig.10 shows a photo of the Boglösa 138 panel where some motifs and the shorelines of 1700 BC, 1500 BC and 1300 BC are illustrated. It is clear to see that the ship in the upper right corner was made at the very beginning of BA, i.e., Period I. The right middle ship has a vertical raised keel extension dated to LBA Period V (Ling 2013: 32), as also suggested by previous chronological study summarized in Subchapter 2.1.1. However, this ship was carved at a place corresponding to an EBA shoreline. For this reason, the shoreline displacement dating method could not be used alone but has to be cooperated with the stylistic dating method. As shown in Fig.10, the wheel cross and foot soles’ motifs spread in a quite large area of the panel and they were made not earlier than Period III and probably during the LBA considering the nearby ship motifs. Therefore, although shoreline 15

dating method is useful, it cannot provide any overall solution to the dating of rock carvings, as has also been noticed by Sognnes (2003: 206), who have used shoreline dating method to study the rock carvings in Norway.

Figure 10. The picture of rock carving panel Boglösa 138 in Enköping. The estimated shoreline is denoted by dash lines. The shoreline and the foot soles motif are added by the author. The elevation data (numbers in black) and the ship motifs are taken from Ling (2013: 36). Credit of the background photo: Einar Kjellén.

2.2. Interpretation of Scandinavian rock carvings

2.2.1. Multiple perspectives of interpretation Before getting into the details of previous interpretations, it is meaningful, as a first practice, to assume that one has never seen any interpretations about rock carvings before. In this way, one can build a step-by-step improved understanding, just as how the understanding was advanced year-after-year by the scientific community of rock carving research. For example, the ship ( ) can be said to show how the prehistoric transportation was made while the wagon ( ) shows the land transportation. This might be our first interpretation. A lot of questions might rise as more details are revealed. For instance, how the figures were carved into the rock surface when hard metal tools had not been invented? If it was not an easy task, why were so many ships so frequently carved? And even on rocks, at times, that were located far from the shoreline? The research of rock carvings are actually motivated by 16

these questions. To answer them, different scholars have had different perspectives, among which some have intersections and some are complementary. The typical perspectives are summarized below:  Archaeological vs. religious perspective. This was summarized by Burenhult (1980: 11). The archaeological perspective focuses on ideology and material culture, for instance, by linking to other BA grave goods or settlements (Kaliff 2007: 28). Whereas from the perspective of religion, the craftsmen carved similar pictures frequently on the rock to indicate their fear, wish and belief (Burenhult 1980; Andrén 2014). What they feared could also be used as a power to protect themselves (Goldhahn 2019: 9). It should be noted that making rock carvings during BA was very difficult since the available tools were probably just stones (Ling 2008: 21). One can imagine how difficult and time consuming it is to carve a figure into a rock that can last for thousands of years. If it was not for especially important reasons, what was it for?  Local vs. regional study. The local study views rock carvings as figures created from local events and religious rituals that developed independently (Winter 2001: 9), whereas the regional study emphasizes more the importance of cultural diffusion and regional interaction (Kristiansen and Larsson 2005: 129; Kaliff 2007: 28). For instance, as the most important materials of the BA, the raw metal of bronze, tin and copper had to be imported outside Scandinavia. It is argued that the supplies of bronze eventually led into the burial consumption, ritual displays, and overall dynamics of the BA society (Kristiansen 1984: 85).  Landscape vs. seascape. The rock carvings were found not only close to shoreline but also inland. There was a long debate whether rock carvings were more influenced and determined by the seascape or by the landscape (Ling 2008: 26). If rock carvings were located very close to the sea, they were probably linked to sea journeys, which were full of uncertainty and danger as well as hope and possibilities (Ling 2008: 255). The rock carvings located inland, however, might have nothing to do with the sea journey. In this context, ships and other motifs should have more symbolic meanings than their practical representation.  Ritual vs. domestic life. The view of interpreting rock carvings as rituals or religious remains was well accepted by many scholars (Kaliff 2007: 194; Kristiansen 2012: 69). However, based on a lot of excavation finds, it has also been argued that most rock carvings could probably be made in connection to domestic life and without religious reasons whatsoever (Petersson and Christina 2015: 279, 282).  Cultural vs. social perspective. The cultural perspective views rock carvings as communication or narrative devices (Tilley and Bennett 2008; Ljunge 2013), whereas the social perspective pays more attention to social expressions (Bertilsson 1987: 191; Wessman 2010), for instance, ‘social logic’ (Cornell and Fahlander 2009) and ‘Ego culture’ (Rédei et al. 2019: 9). Fahlander (2012: 101) approaches rock carvings as a meaningful practice such as competitive practice. It was found that some large ships take over the whole surface of a rock, which is viewed as a competition scenario (Sognnes 2001: 125; Fahlander 2012: 101). In short, the former interprets rock carvings as written language and focuses on the result, i.e., the content and manifestations, while the latter views them as practice or activity and puts more attention on the process; why and how they were made. Critical thinking is always necessary when interpreting rock carvings that were made several thousands of years ago. Some critical views have also been proposed regarding how to interpret the rock carvings in a better way. For example, the picture of rock carvings might be misleading since it differs from the original carvings (Ljunge 2016: 167) and the motifs on the stone might have been changed significantly over time (Jones et al. 2016: 129). Moreover, every picture in the documentation is influenced, more or less, by the understanding of the 17

recorder. Different exposure and different angles of photography might introduce different interpretations (Ljunge 2016: 167). For this reason, the subjective interpretation might always make the results unfittable (Berger 2008: 7–8). To obtain a more objective and comprehensive interpretation of the wheel cross, this study generally follows the intercontextual perspective as mentioned earlier (Kristiansen and Larsson 2005: 11). Different from the direct interpretation of the image on randomly selected material, the intercontextual perspective is aimed at putting the rock carving motifs in a specific time and space to make them come to life. This is also supported by Ling (2005: 453), who mentioned that the rock art motif makes power when it is connected to other elements. If one wants to find the relationship between different motifs, it is necessary to contextualize them under the cultural and historical development. Every method that ignores the connection inside the symbols will lead to biased interpretation. Besides, the motif is a medium that links the past and present, as Back Danielsson et al (2012: 1) pointed out the equal importance of the images which worked in the past and to the present society.

2.2.2. Previous understanding of the wheel cross motif The wheel cross ( ) might be the simplest but at the same time the most complex motif in rock carvings. It is simple due to its limited typological variations. It is complicated because it appears in various scenarios when relating to other motifs. As mentioned before, the previous interpretations of the wheel cross motif disperse in different publications, but the main ones are summarized below. The wheel cross appearing with a working animal ( ), with a number of two or four, means the wheel of a wagon or cart (Malmer 1981: 43). It indicates the transformation from hunter-gatherer society to agricultural society. It was suggested that before numerous ships were carved for religious activities, the cart served as the main ritual medium (Malmer 1981: 44). The cart design from Bohuslän province where Tanum is located, was influenced by Denmark at the very beginning of BA (Malmer 1981: 44). When the wheel cross appears above or near a ship ( ), with or without a stand, the ship is interpreted as the sun ship (Winter 2001: 12). The meaning is probably developed from Egypt or Eastern Mediterranean, where ships were used as vessels to transport the sun god in the local religious culture. The god sailed to the sky in the morning and entered the underworld in his evening ship, where the dead were transported to the other side. The dead then asked permission to enter the ship for the afterlife (Winter 2001: 12). However, the sun ship is also found in other circumpolar regions like Siberian, North Pacific and Canada etc. (Lahelma 2017: 163). In the context of ‘sun-cult’, the wheel cross sometimes is replaced by a cup mark. When it is found above a horse ( ), the horse is named as the sun horse that is carrying the sun chariot (Winter 2001: 23; Ling 2008: 67). The sun horse has a significant meaning not only in religion but also in cosmology since the sun is the ruler of time and seasons that are critical for an agricultural society. This interpretation is famous since the moving of the sun, when facing it at the Northern Hemisphere, is consistent with the direction of the horse, i.e., from left to right. The wheel cross also looks very similar to two foot soles together ( ). It is, however, more difficult, to distinguish the foot soles from the footprints and shoes (Skoglund et al. 2017). In the view of religious belief, foot soles indicate a relationship between the water, the earth and the sun (Kaliff 2007: 68). From the perspective of Indo-Europeans, foot soles are related to the sun, the thunder god (Kaliff 2007: 49). The double foot soles without toes are dated to the EBA (Malmer 1981: 60–62). The wheel cross also has other circular variations ( , ) that are generally dated to the EBA (Malmer 1981: 72). The wheel cross is sometimes covering a human torso ( ) and it is interpreted as a shield, usually dated to the EBA (Coles 2005: 47).

18

2.3. Previous comparative studies of Scandinavian rock carvings Compared with the numerous results from local studies of Scandinavian rock carvings (e.g. Kjellén 1976; Milstreu 1977; Burenhult 1980; Malmer 1981; Jensen 1985; Bertilsson 1987; Malmer 1989; Ullén 1994; Bradley 1997; E. M. Walderhaug 1998; Coles 1999), regional comparative studies are done to a lesser extent, at least not in a systematic way. More recently, Coles has published a detailed report on rock carvings for Enköping (Coles and Gräslund 2000) and Tanum (Coles 2005), based on several decades’ field work. Ling has also investigated them, from a seascape’s perspective and based on GIS analysis, for Tanum (Ling 2008) and Enköping (Ling 2013). However, the differences of those two regions are not discussed in so many words, and when mentioned, they are scattered among detailed contexts. The first systematic comparative study of rock carvings between Tanum and Enköping was performed by Wessman (2010). She rightly argued that the similarities were the main focus of previous studies, while the differences received much less attention to some degree. Based on a comprehension comparative study on the landscape and the motif interpretation, Wessman concluded that most rock carvings of Tanum belong to the LBA while that of Enköping to the EBA. Based on the previous work including that of Wessman (2010), Ling (2013: 98–99) summarized some general similarities and differences of rock carvings between Bohuslän and Uppland. However, considering the fact that the land of Tanum uplifted much earlier than that of Enköping (shown later in Fig.11), the conclusion drawn by Wessman (2010) is of interest to question. If the conclusion holds, then there must be some special and important reasons. Besides, the differences of the comparison trended more on the local analysis, which has been regarded to be more important than the similarities. However, as aforementioned, the regional study focusing on similarities has also to be done in order to achieve a better understanding of the transformation of culture, economics and life, as well as the communication through immigration in Scandinavian areas. There is no doubt that the local character is one aspect of studying the rock carvings since the diversities reflect the local’s values. Although the local characters contributed more to the details of overall understanding, this is not to agree that difference is the aspect that should be given more focus. As also argued by Fahlander (2008a: 69), similarity that is used to identify individual style or social status in time and space is also used for distinguishing one group from another. Comparison between the similarity and difference is a complex social process in a colonizer culture (Bhabha 2012: 122). Sometimes colony processes could have a hidden agenda (Aching 2002: 2; Varberg 2007) and make new material expressions unable to be traced to their origins (Fahlander 2007: 22). Therefore, both the similarities and the differences are equally important, no matter if the studies are either local or regional.

19

3. Historical and geographical background

This chapter is devoted to summarize the general historical and geographical background of BA Scandinavia. Such a big picture is of great importance when doing interpretations. For instance, the transition from the Neolithic to the BA is of relevance when interpreting rock carvings from the EBA, while the transition from the BA to the PRIA is pivotal when studying carvings from the LBA. Apart from the historical background, the geographical background also plays a significant role in the development of the BA society. For instance, Tanum is adjacent to the North Sea and Denmark, while Enköping is close to the Baltic Sea. The geographical locations determine what and how the surrounding culture could penetrate and influence the study areas. Finally, this chapter also includes the primary case study results, for instance the locations of rock carvings, since it is convenient to compare such data from two different areas directly in one figure. The detailed results of the interpretation of motifs are presented separately in the following chapters.

3.1. Landscape and historical remains The prehistoric landscape of Southern Scandinavia is quite different from that of the present due to the significant change of shorelines in some areas. The shoreline displacement depends on two factors, one is the LUV and the other is the slope or the average altitude of the land. The shorelines at the time of 5000 BC (when the Neolithic started) and those of 1700 BC (when the BA started), as well as the present shoreline of Sweden, are shown in Fig.11. As can be seen, the area where the shoreline changes the most is the Southeast of Sweden, where Enköping is located, although the LUV is not the highest (Fig.7). In the beginning of the Neolithic, Enköping was still under water. About half of the present land of Enköping lifted above water during the BA. Compared with Enköping, the shoreline in Tanum has a much slighter shoreline displacement during the past 7000 years. This is because Tanum is characterized by rolling hills while Enköping is characterized by flat plains, as shown in Fig.4. This geographical background should always be kept in mind, and it will be argued later in this study that geographical features are more important than previous studies have shown. Historical remains are the materials that the archaeological study is based on. Apart from rock carvings, other remains are also of importance such as settlements, graves and artefacts in the form of for instance bronze items. Unfortunately, compared with the numerous rock carvings, other historical remains that are dated to the BA are rare in Tanum (Ling 2008: 7). Despite active searching, Iron Age settlements, instead of BA settlements, were found in Tanum (Vogt 2011: 168). This might be because the development of ploughing ways from the Iron Age has made it difficult to find any trace of BA settlements (Bertilsson 1987: 17). Cairn and barrow are the traditional signs of the graves in the BA and they were located far from the shoreline, for instance, around Lake Vänern (Bertilsson 1987: 17). The limited unearthed items from settlements and graves may show that the economy of Southern Scandinavia was mainly focussed on farming and stock-breeding, as well as some fishing and hunting activities (Bertilsson 1987: 20–22). This is consistent with the figurative rock carvings that show the mixture of economic activities (Ling 2008: 8). Metal objects were used for weapons and tools, 20

and the majority belongs to the LBA (Vogt 2011: 170–172). Bronze finds in Bohuslän have revealed that its southern part was dominated with BA metal objects. On the other hand, the north has much more rock carvings, but less settlements and metal objects. Therefore, it is difficult to study the connection between rock carvings and settlements in Tanum. Fortunately, some new excavation jobs have been done recently (Petersson and Christina 2015). Compared with Tanum, Enköping has a relatively richer registration of other BA remains such as graves and heaps of fire-cracked stones (Ling 2013: 22–23). Compared with the Southwest of Sweden, the Southeast contains more instructive information about the shape of farms, burial places, methods of communication, settlement structure, which indicates that the economy was based on animal husbandry with a few examples of wheat cultivation (Coles and Gräslund 2000: 125). The settlements are small; only two farmstead sets with irregular cultivation fields and open land for pasture. Burials are near or far away from the settlement in the form of cairns, mounds and flat cemeteries. In Enköping, the fire-cracked stones and metal fragments are interpreted as settlement debris. Some hearths, fire-cracked stones and graves are dated to 2000 BC (Ullén 1994: 249). It was argued that the fire-cracked stones acted as ritual materials and played an important role in sacrificial rituals in the perspective of Indo-Europeans (Kaliff 2007). Based on the bronze objects and other metals found in Southern Scandinavia, some general remarks on the historical background can be summarized. In the LN and the EBA, Southern Scandinavia seems to have been influenced by a cultural connection to the West, i.e., the British and the Atlantic coast, whereas during the LBA, it was strongly affected by trade with the Baltic through the northern German and other Central European areas (Vogt 2011: 45).

21

Figure 11. Prehistoric shoreline of Sweden. Made by the author with the method described in Appendix 2: Reconstructing historic shoreline.

3.2. Location of rock carvings The locations of rock carvings are no doubt of great importance because they reflect where related activities have been taking place. The spatial distributions of rock carving panels of Tanum and Enköping are shown in Fig.12 and Fig.13, respectively. The BA shoreline and the water regions (sea, lake and river) are also shown. It is difficult to reconstruct prehistoric narrow rivers inland since they were probably a few meters above the prehistoric sea level. Therefore, the present rivers (width < 10 m) are shown here. Due to the land uplift, the prehistoric rivers should be a little bit wider than those of the present. However, there should be little difference in the main route. As mentioned before, the BA shoreline displacement of Tanum is not much different from now. As can be seen from Fig.12 and Fig.13, most rock carvings were located close to the BA shoreline; closer for Enköping and more extensive for Tanum, which is in agreement with previous studies (Ling 2013: 98). Compared with Enköping, much more rock carvings of Tanum were located inland, c. 10–15 km from the shoreline. Fig.12 shows that those inland rock carvings were very close to rivers. It is interesting to notice that most rivers flow from the northeast to the southwest in Tanum. Fig.13 indicates that most rock carvings of Enköping were located exactly at the BA shoreline, i.e., the carvings were made on the new rocks elevated from the sea during the BA. It is also noticeable that quite many present rivers flow from the northwest to the southeast. However, most present rivers were quite wide sea areas between two islands during the BA. The spatial distribution can also be shown by plotting the elevation of rock carving panels, as illustrated in Fig.14, where the histogram of percentage in rock carving panels of Tanum and Enköping with elevation at 5m-intervals is plotted. As shown in Table 3, the average elevation of the BA shoreline is 16–10 m and 25–16 m, respectively, for Tanum and Enköping. Fig.14 indicates that the shoreline displacement dating method is generally invalid for most rock carvings of Tanum since about 80% of the panels have an elevation higher than 20 m, an altitude not possible for a human standing at the shoreline to carve. The shoreline dating method can only be applied for individual large panels situated at lower places such as Tanum 311 (Ling 2008: 91, 237; Nimura 2015: 4), with an average altitude of 14.75 m. On the contrary, however, the shoreline dating method is an ideal method for most rock carvings of Enköping, where more than half of the panels are located below 25 m. Fig.14 also shows that more rock carvings were located much closer to the EBA shoreline but to a lesser extent to the LBA shoreline, which is consistent with previous studies, as summarized by Ling (2013: 98). However, similar conclusions cannot be made for Tanum, as most rock carvings there do not follow the BA shoreline. On the contrary, Fig.14 shows that rock carvings are almost evenly distributed from 0 m to 40 m above the BA sea level. Fig.15 illustrates the histogram of percentage in rock carving panels of Tanum and Enköping with slope at 3°-intervals. It shows that about 85% of the rock panels of Enköping are located at rather flat plains with a slope less than 12° and for half the panels less than 6°. The average slope of rock carving panels in Enköping is about 6°, even much smaller than that in the previous report, which was 10° (Coles and Gräslund 2000: 27). Compared to Enköping, rock carvings of Tanum are situated at steeper areas, with about 40% of the panels having a slope larger than 12°. This is because most rocks are located on inland hills (see Fig.4 a). It is also interesting to investigate the aspect of rock carvings. This can be easily done with the GIS software. Fig.16 demonstrates the radar plot of percentage in rock carving panels of Tanum and Enköping with azimuth at 30°-intervals.

Figure 12. Spatial distributions of rock carving panels in Tanum. Made by the author.

24

Figure 13. Spatial distributions of rock carving panels in Enköping. Made by the author.

6,6% >55 m 0,1% 6,4% Tanum Enköping 50-55 m 0,0% 8,6% 45-50 m 0,4% 10,5% 40-45 m 0,6% 9,1% 35-40 m 3,0% 11,7% 30-35 m 16,1% 17,7% Elevation 25-30 m 38,3% 15,1% 20-25 m 33,7% 11,6% 15-20 m 6,2% 2,7% 10-15 m 1,2% 0,1% 5-10 m 0,2% 0,0% <5 m 0,0%

0,0% 5,0% 10,0% 15,0% 20,0% 25,0% 30,0% 35,0% 40,0% 45,0%

Percentage

Figure 14. Histogram of percentage in rock carving panels of Tanum and Enköping with elevation at 5m- intervals. Made by the author.

5,3% >27 ° 0,6% Tanum Enköping 3,0% 24-27 ° 0,5%

3,6% 21-24 ° 1,2%

5,0% 18-21 ° 2,2%

9,9% 15-18 ° 3,8%

Slope 12,9% 12-15 ° 7,6%

18,3% 9-12 ° 14,0%

18,9% 6-9 ° 20,5%

14,8% 3-6 ° 26,5%

8,4% <3 ° 23,1%

0,0% 5,0% 10,0% 15,0% 20,0% 25,0% 30,0% Percentage

Figure 15. Histogram of percentage in rock carving panels of Tanum and Enköping with slope at 3°-intervals. Made by the author.

<30 ° 18% 330-360 ° 16% 30-60 ° 14% 12% Enköping 10% 300-330 ° 60-90 ° Tanum 8% 6% 4% 2% 270-300 ° 0% 90-120 °

240-270 ° 120-150 °

210-240 ° 150-180 °

180-210 ° Figure 16. Radar plot of percentage in rock carving panels of Tanum and Enköping with azimuth at 30°- intervals. Made by the author. As can be seen from Fig.16, rock carvings have a preference of facing E/SE for Tanum and NE/N/SW for Enköping. This is consistent with previous studies such as the reports for Tanum (Coles 2005: 93) and Enköping (Coles and Gräslund 2000: 93). However, the previous studies did not elaborate on the reasons for this. As mentioned before, most rivers in Tanum flow from NE to SW. By carving the rock on the right side of the river, the figures on the rock are perpendicular to the morning sunlight. Therefore, most rock carvings of Tanum are located on the right side of the river since they are facing SE. Similarly for Enköping, the aspect of rock carvings is determined by the direction of the BA waters; however, it seems that there is no preference about the left or right side of the waters since the numbers of rock carvings facing NE and SW are equivalent. It is interesting to notice that the water flow direction is perpendicular to the aspect of rock carving panels, forming a virtual cross that can be viewed as a lifeline. The lifeline acts as an agency to link the landscape and the outcomes of rock carving panels.

27

4. Case study of Tanum

This chapter is devoted to presenting the detailed results from the case study of Tanum. Tanum is located in the province of Bohuslän, southwest of Sweden and it is characterized by rift valleys, hills and interconnected lowland areas. Based on the FMIS database, 1489 panels of rock carvings in Tanum are selected in this study. The number of rock carving panels in different parishes of Tanum is plotted in Fig.17.

900 833 800

700

600

500

Number 400 344 300

200 161 105 100 39 1 6 0 Bottna Kville Lur Mo Naverstad Svenneby Tanum

Figure 17. Number of rock carving panels in different parishes of Tanum. Made by the author. Fig.17 shows that the Tanum parish has more than half of the rock carving panels in the selected area, i.e., the Tanum municipality. The aim of the case study is to investigate the meaning of the wheel cross and its relation with other motifs. Thus it is necessary to firstly have a review of the main motifs in the first subchapter 4.1. For sake of comparison, most motif figures are given in the second subchapter 4.2, where a detailed discussion of the wheel cross is made.

4.1. Typical rock carving motifs The classification of the motifs are sometimes ambiguous (Fredell 2004; Wahlgren et al. 2004), as it is difficult to distinguish the wheel cross figure from the various circles or foot soles together. However, such ambiguity is believed to have a small impact on the overall statistics. To handle such a larger number of rock carvings, it is necessary to use programming, see Appendix 1: Working with FMIS database (ii. Automatic statistics). The number of different motifs in rock carvings of Tanum is plotted in Fig.18, which shows that cup mark ranks first and the ship second. However, a single cup mark is not viewed as a figurative motif in general due to its simple and fixed typology. Fig.19 shows the percentage of different motifs with different number ranges in rock carvings.

28

20000

16289

15000

10000

5730 5000

2111 1143 427 327 65 0 Cup mark Ship Human Animal Circle Foot soles Bird

Figure 18. Number of different motifs in rock carvings of Tanum. Made by the author.

100%

90% Cup mark Ship Human Animal Circle Foot soles Bird 80%

70%

60%

50%

Percentage 40%

30%

20%

10%

0% None 1-3 4-6 7-9 10-12 13-15 >15 Number

Figure 19. Percentage of different motifs with different number of motifs in rock carvings of Tanum. Made by the author. As can be seen from Fig.19, most of the panels (90%) have cup marks and about half have ships. The percentage of different motifs decreases as the number of motifs increases, for all different motifs. The spatial distributions of cup mark, ship, human, animal and foot soles motifs are illustrated in Fig.20, Fig.21, Fig.22, Fig.23, Fig.24, respectively. These figures show that the entire motifs share a similar spatial pattern of distribution, for instance, wherever cup marks are present, ships, humans, animals, foot soles usually exist; and the more cup marks there are, the more other motifs there are. The general features of different motifs are summarized in subsequent subchapters.

29

Figure 20. Spatial distributions of cup mark motifs in rock carvings of Tanum. Made by the author.

30

Figure 21. Spatial distributions of ship motifs in rock carvings of Tanum. Made by the author.

31

Figure 22. Spatial distributions of human motifs in rock carvings of Tanum. Made by the author.

32

Figure 23. Spatial distributions of animal motifs in rock carvings of Tanum. Made by the author.

33

Figure 24. Spatial distributions of foot soles motifs in rock carvings of Tanum. Made by the author.

34

4.1.1. Cup mark The cup mark ( ) is a round and smooth dot, which usually presents within a group of carvings. One group of cup marks may represent a person, a family or a small community (Coles 2005: 79). In Tanum, cluster cup mark sites are located near large figurative sites, as shown in Fig.20–24. There are 16289 cup marks carved into the rock panels in Tanum and the spatial distribution is shown in Fig.20. Most cup mark motifs are randomly distributed in rock panels although some of them form a pattern. Since the cup mark is highly abstract, there are numerous ways of interpreting it. For instance, it has been suggested to be an icon, a symbol, to stand for presence, for stars, genders, blood drops, and even space aliens, etc. (Coles 2005: 79). In different ways, cup marks are related to other figurative motifs such as wheel cross (Tanum 184), wagon (Bottna 39), ship (Tanum 66, Kville 82), human (Tanum 120), and animal (Kville 162). These constellations have been suggested to indicate domestic life and the sacrifice (Thrane 1999: 126), the moon or the sun (Tanum 1, 192), the social elite (Tanum 11, 70), and the baby birth (Tanum 255), fertility (Tanum 184), and lastly a phallic male (Tanum 90) (Coles 2005: 76). It has also been argued that individual cup marks without any other motifs in their vicinity are symbols of BA settlements (Thrane 1999: 126). The cup mark has also been suggested to be a manifestation of a cult scenario in the Neolithic, that also continued to later periods (Bengtsson 2004: 167). The cup mark and circle might be the most similar motifs to the wheel cross due to their simple and circular shape. In different scenarios, it was found that the wheel cross is simply represented by the cup mark to indicate the sun or wheel, as shown in Table 4 (No. 3, 8, 9, 11). There is a famous figure in Tanum 184 panel (No. 4 in Table 4) where the wheel cross is filled with many cup marks. It is difficult to know if the cup marks were depicted at the same time as the wheel cross or at a later time.

4.1.2. Ship The ship ( ) is the most common figurative motif, with a length from 20 cm to over 2 m (Coles 2005: 22). Large ships usually carry multiple strokes, human-like figures, and small ships only have few crew strokes. There is little evidence to indicate that the ships were prepared to sail as no water-like figures were found to show that ships were on the water, but the ships are full-hulled and landlocked (Coles 2005: 22). There are 5730 ships in the rock carvings of Tanum and the spatial distribution is shown in Fig.21. The ship is also associated closely with other motifs such as human, animal, weapon, etc., for instance, with large and dominant humans carrying weapons (Tanum 325), humans raising the keel of the ship (Tanum 326), adorant humans attached to the ship (Tanum 192), a ship under a horse-like animal (Kville 162), shallowly carved cup marks above the ship stem (Bottna 43), etc. There are several different types of ships, such as stacked ships (Tanum 12A), ships with a feather-liked feature or flag of the stem (Tanum 192), ships with large double spirals over them (Tanum 62), stroke by keyhole-shaped along with central cup mark (Kville 149), round bodied human (Tanum 450, 261) besides the ship, disc drawn by horse (Tanum 261) on the ship, etc. The ship is also closely related to the wheel cross in different ways and the most common scene is the sun ship ( ) with one or two wheel crosses over the ship, with or without a stand. The concept of sun ship was suggested to be imported from Mediterranean (Kristiansen 2012; Lahelma 2017). Two examples are shown in Table 4 (No. 5, 6). The wheel cross is believed to be depicted at the same time with the ship although there are few possibilities that it was added later. In this work, the dating of a sun ship is completely based on the stylistic dating of the ship below the cross. There are two connected twin wheel crosses at the top and side of a ship in Kville 101 panel, which look like two moving ships (No. 6 in Table 4).

4.1.3. Human Like the ship, the human motif ( ) is quite special since it is neither linear nor circular, which is easy to recognize with real humans. Usually humans are carved with a cup mark head, stick arms or no arm, straight leg without feet. There are 2111 human motifs in the rock carvings of Tanum and the spatial distribution is shown in Fig.22. The styles of human bodies include line-drawn bodies, thin-bodied figures, disc-bodied figures, etc. (Coles 2005: 31). Some humans are carved with long hair, some are phallic and some have no sexual indication. Active humans over ships are seen as running or jumping (Tanum 198). Human images are usually not isolated; they are often with other humans (Coles 2005: 31). Humans are sometimes equipped with swords, spears or axes, such as some warriors (Tanum 120, 192), some axe men (Tanum 1, 12A,255). These weapons, especially the swords, are treated as special equipment, as a sign of prestige or leader (Anati 1960; Bertilsson 2017). The human motif is of special importance since it represents both human and society. The human motif is also closely associated with other motifs such as ship (Kville 157), weapon (Tanum 1), cup mark (Tanum 255) and animal (Tanum 160), as well as the wheel cross (No. 7, 8 in Table 4). No. 8 of Table 4 shows a battle scene, where humans are equipped with weapons and with the wheel cross covering the torso. In this context, the wheel cross is usually interpreted as a shield.

4.1.4. Animal The animal ( ) includes horse, snakes, fish, sheep, cow, etc. Ocean animals such as seals and dolphins seem to be absent. Domestic animals carved on the rock include cattle, bull, horse and dog. The animal was usually carved in a profile with a solid torso. A very limited number of animals are filled with lifeline ( , such as one in Tanum 311). Wild animals such as deer, elk, bear, boar, ducks, and snake are rare. It shows that animal motifs are quite different from those in Northern Scandinavia. There are 1143 animal motifs in the rock carvings of Tanum. The spatial distribution (Fig.23) shows that they are mainly present in the Tanum parish. It is difficult to recognize some animal motifs since a dog-like animal in some scenes looks like a fox or wolf, and a sheep-like animal is similar to a deer or horse. Birds carved around a disc are interpreted as an adorant scene (No. 9, 10 in Table 4), especially for those birds with triangular bodies ( ) similar to a bronze razor (Coles 2005: 60). Fish is viewed as an important supplement for the BA coastal society, however, the number is very limited. Specific animals, especially horse and cow, are usually present with wagons or discs. Horses carrying a disc, usually facing right, is called sun horse (No. 11 in Table 4), introduced to Southern Scandinavia since Period II. It shows that the sun was drawn in the sky from left to right by a horse. Interestingly, the horse driving wheel cross or wagon is usually facing left (No. 12 in Table 4).

4.1.5. Foot soles The foot soles ( ) were usually carved in life size (Bradley 1997: 317). They sometimes appeared in pairs or groups, sometimes as single foot sole, with or without toes. There are 327 foot soles motifs in the rock carvings of Tanum and the spatial distribution is shown in Fig.24. It shows that the foot soles are mainly present in Tanum and Bottna parishes. It is also interesting to notice that foot soles as well as other frequent figurative motifs discussed above are all connected with moving, i.e., foot soles, wagon, ship, horse are all important carriers for journeys and trips. Foot figures are also closely related to other motifs such as ship, wheel cross and humans. Foot soles indicated the direction of the movement (Coles 1999, 2004; Mörner 2012), like the wheel that never stops. The double foot soles without and with toes are generally dated to the 36

EBA and the LBA, respectively (Malmer 1981: 60–62). Fig.25 shows an example of foot soles related to the wheel cross, where the foot soles are located under a big wheel cross held by three humans. The size of the wheel cross and foot soles are comparable to a human motif, which indicates an important journey.

Figure 25. The foot soles related to the wheel cross in Tanum 406 panel. Credit: Dietrich Evers.

4.2. Multiple expressions of wheel cross motifs in Tanum As discussed before, there are many variations of the wheel cross which do not necessarily present themselves as a cross in a circle ( ). Table 4 summarizes some popular forms of the wheel cross in rock carvings of Tanum. The wheel cross related to other motifs such as cup mark (Bottna 39, Tanum 184), human (Tanum 90), working animal (Tanum 406), ship (Tanum 12B), weapon (Tanum 90), sun disc (Tanum 12A, 255), and sun horse (Tanum 450, 262) in different scenarios such as plough, transportation, ritual, fertility, battle, etc. The dating of different forms of the wheel cross is not straightforward but not impossible, as discussed in Subchapter 2.2.2. The wheel cross widely exists as the wheel of a cart, and it has been seen in different BA periods (Johannsen 2011: 96). As shown in Table 4, in most cases the wheel cross represents either a wheel for the transportation purpose or a symbol of the sun. In this context, the wheel cross is related to fertility and religion naturally. This is the case, since to achieve a big harvest, the cart and sunlight are of the same and significant importance as the cart could transport the seed, soil, water and the sunlight makes the crops grow. Such symbolic meaning of the wheel cross was kept but also developed in the LBA, with more expressions, for instance, the wheel cross was replaced by other circular motifs (disc, circle) and was decorated by cup marks.

37

Table 4. Different variants of the wheel cross in rock carvings of Tanum. Interpretations marked with (*) are inspired by Coles (2005). All images are adapted from SHFA.

Image Panel Forms No. Features RAÄ number Images Dating credit elevation

The wheel cross was drawn by the Bertil 1 Tanum 406 25 m LBA12 animal Almgren

Horse wagon, The horse or oxen like animal dragged the wagon. The wheel disc or circle Bertil 2 used the disc to represent. The Tanum 120 19 m LBA12 connected to a wagon Almgren or draught animals animals were positioned back to back*

The wheel was represented by a Torsten 3 cup mark in a circle, no spoke Bottna 39 48 m LBA13 Högberg wheel

12 This form of wagon has been found in different BA periods in Scandinavia (Johannsen 2011: 96). It is dated to LBA here by referring to other motifs nearby. 13 A typical cart from LBA (Johannsen 2011: 96).

Image Panel Forms No. Features RAÄ number Images Dating credit elevation

The cup marks inside the wheel The wheel cross with cross indicated the wish of cup marks as 4 Tanum 184 Ellen Meijer 45 m LBA?14 fertility and the social decorations organization*

The ship transported a big wheel cross to its beginning at night*; a EBA15 5 Tanum 12B SHFA 23 m wagon with four smaller wheels Period II was under a ship

The wheel cross over or below ship

The twin wheel cross were 6 connected at the top and side, Kville 101 Åke Fredsjö 33 m LBA15 which made it look like a ship

14 Not possible to date directly as the cup mark could be made at a later time after the wheel cross was made. 15 Dated by referring to the ship motifs nearby.

39

Image Panel Forms No. Features RAÄ number Images Dating credit elevation

Either the concentric disc or the wheel cross covered the human Gustav 7 torso as a shield; The humans Tanum 90 24 m EBA16 Monberg were carrying weapons and facing different directions* The wheel cross covering the human torso

A human was standing on a circle Carl Georg EBA16 8 fighting against another; Above Tanum 12A 23 m Brunius Period II the humans was a sun disc

The sun disc was pushed by two women, towards the same Catarina Sun disc 9 direction with the bottom hunting Tanum 12A 23 m EBA17 Bertilsson scene; One was standing on a cup mark (wheel)*

16 Battle axes held by humans are generally dated to EBA (Coles 2005: 79–80). 17 The sun disc surrounded by birds is an EBA symbol, see the design on Wismar trumpet (Fig.38). The sun disc here, however, could also be made in LBA.

40

Image Panel Forms No. Features RAÄ number Images Dating credit elevation

Two sun discs, one finished and 10 Tanum 255 Åsa Fredell 46 m EBA18 the other unfinished

A cup mark was over the horse EBA19 11 and the horse was facing the right Tanum 262 Ellen Meijer 45 m Period II direction Sun horse

Two animals dragged the wheel cross to the left and another Bertil 12 Tanum 450 18 m LBA20 animal dragged a sun disc to the Almgren right*

18 Suggested by nearby foot soles and ships. 19 This is the classic early form of sun horse. 20 Suggested by shoreline dating.

41

5. Case study of Enköping

Similar to Chapter 4, this chapter presents the case study result of Enköping. Enköping is located in the southwest of Uppland province, with a fairly flat landform surrounded by forest covered hills. The rock outcrops are quite different from those in Tanum, with smaller sizes, more irregular shapes and rougher surfaces. There are 1847 panels of rock carvings registered in FMIS database. The number of rock carving panels in different parishes is shown in Fig.26.

500 450 425 400 350 300 282 250 205 200 Number 144 155 150 102 93 100 80 69 50 34 19 22 21 24 29 32 19 7 5 3 4 8 1 2 13 7 3 2 14 3 7 7 1 5

0

Löt

Bred

Teda

Fittja

Gryta

Torsvi

Vallby

Altuna

Hjälsta

Tillinge

Giresta

Litslena

Nysätra

Härnevi

Boglösa

Hacksta

Lillkyrka

Simtuna

Villberga

Långtora

Torstuna

Frösthult

Enköping

Fröslunda

Veckholm

Sparrsätra

Österunda

Svinnegarn

Härkeberga

Vårfrukyrka

Biskopskulla

Kungs-Husby Husby-Sjutolft Enköpings-Näs Figure 26. Number of rock carving panels in different parishes of Enköping. Made by the author. Fig.26 shows that the rock carvings are also unevenly distributed in Enköping, with Boglösa, Litslena and Vårfrukyrka parishes in top three. Similar to Chapter 4, typical rock carving motifs are introduced in Subchapter 5.1, followed by Subchapter 5.2 describing the wheel cross.

5.1. Typical rock carving motifs The number of different motifs in rock carvings of Enköping is plotted in Fig.27, which clearly shows that the cup mark is the dominant motif. Fig.28 shows the percentage of different motifs with different number ranges in rock carvings, which has a similar pattern to that of Tanum. All figurative motifs have a very limited number (< 3) in most panels.

25000

20996 20000

15000

10000

5000

1486 460 277 212 169 143 0 Cup mark Ship Oblong Foot soles Animal Human Circle

Figure 27. Number of different motifs in rock carvings of Enköping. Made by the author.

100%

90% Cup mark Ship Oblong Foot soles Animal Human Circle

80%

70%

60%

50%

Percentage 40%

30%

20%

10%

0% None 1-3 4-6 7-9 10-12 13-15 >15 Number

Figure 28. Percentage of different motifs with different number of motifs in rock carvings of Enköping. Made by the author. The spatial distributions of cup mark, ship, foot soles, animal and human motifs are illustrated in Fig.29, Fig.30, Fig.31, Fig.32, Fig.33, respectively. Different from Tanum, these figures show that only the cup mark spreads out in the area of Enköping, however, most other motifs are concentrated in one place, i.e. the Boglösa parish. This suggests that Boglösa is the economic, political, and cultural centre of BA Enköping. On the other hand, it also suggests that the cup mark is quite different from other figurative motifs in Enköping; the latter are perhaps given other or special meanings compared to the former. There are some ship motifs in other panels and other figurative motifs such as human and animal, but they are almost absent outside Boglösa.

43

Figure 29. Spatial distributions of cup mark motifs in rock carvings of Enköping. Made by the author.

Figure 30. Spatial distributions of ship motifs in rock carvings of Enköping. Made by the author.

45

Figure 31. Spatial distributions of foot soles motifs in rock carvings of Enköping. Made by the author.

46

Figure 32. Spatial distributions of animal motifs in rock carvings of Enköping. Made by the author.

Figure 33. Spatial distributions of human motifs in rock carvings of Enköping. Made by the author.

5.1.1. Cup mark The cup mark ( ) in Enköping generally has a diameter of 3–5 cm and 1–3 cm in depth, it is abundant and simply carved (Coles and Gräslund 2000: 31). Cup marks tend to separate themselves in a variety of ways from other figures or motifs. Some sites have scattered cup marks, some have rare or none, some have groups of cup marks alongside other figures, and some only have many cup marks (as also indicated by Fig.29–33). Individual cup marks are associated with two features types. One association is with a groove, where the cup mark position is suggestive of a dumbbell, or even a penis, another with an elongated shape to make a snake like image (Coles and Gräslund 2000: 31). When the cup marks are present with other motifs, they are usually aligned in a single row (Litslena 138, Vårfrukyrka 192), and/or in parallel lines (Långtora 15, Härnevi 12). The cup mark is also related to other motifs such as ship (Litslena 221, 301, Biskopskulla 69, 106, Vårfrukyrka 44), weapon (Villberga 72), and foot soles (Boglösa 158, Vårfrukyrka 44). The cup mark inside a circle or disc could represent a wheel cross, as shown in No. 3, 5 and 10 in Table 5. In this case, the cup mark means the hub of a wheel while the spokes (rods) are omitted.

5.1.2. Ship The styles of ship ( ) vary significantly in Enköping, such as ships represented by carved lines with unturned ends (Boglösa 61), standard styles with a keel line and gunwale (Boglösa 61), keel lines with prows (Litslena 221) and stems without gunwale (Boglösa 53,141), hammered hulls (Boglösa 109, 305), and ships with animal heads (Boglösa 111, Villberga 51) (Coles and Gräslund 2000: 57). There are 1486 ships in the rock carvings of Enköping and the spatial distribution is shown in Fig.30. There are only a few ships outside Boglösa parish. A ship motif sometimes reused the nature groove as its hull, e.g., Boglösa 416. This reuse action was explained as a narrative tool (Bradley 2005; Goldhahn 2008; Tilley and Bennett 2008). Large ships occupying the whole rock surface might indicate a competition scenario, that is social and ritual aspects of competition (Sognnes 2001; Fahlander 2012). The sun ship is also found in Enköping, as shown in No. 4 and 5 in Table 5.

5.1.3. Foot soles The foot soles ( ) are the second richest figurative motifs in Enköping, with a total number of 277, just following the ship. This number is comparable to that of Tanum (which has 327 foot soles in total). As shown in Fig.31, Boglösa parish owned the greatest number of foot soles in Enköping. There are also many variations of foot soles, such as naked feet with toes (Boglösa 131), fully groundout foot soles like ellipses (Vårfrukyrka 44), small foot soles used as rays of the disc (Boglösa 160), foot soles with narrow insteps (Boglösa 158), etc. The foot soles have been argued to be a symbol for connections with the outside world, or the world of death (Coles 2004). It has also been suggested as being the footprint of a god, but also as having other cosmological meanings (Bertilsson 2013: 162). The foot soles together ( ) are very similar to the shape of the wheel cross ( ) in Enköping, for instance, in the famous Boglösa 138 panel (No. 7 in Table 5).

5.1.4. Animal The animals ( ) in Enköping include horse, deer, cattle, pig, bear, dog, long bodied animal (e.g. Vårfrukyrka 275). The spatial distribution of animal motifs (Fig.32) shows that they are only present in Boglösa and not anywhere else in Enköping. The connection between the animal and wheel cross in Enköping is similar to that in Tanum. For example, a horse-like animal drives a car (No. 3 in Table 5) and a horse drags a sun disc (No. 8 in Table 5).

5.1.5. Human Compared to Tanum, the human motifs are not so common in Enköping and they are only present in Boglösa parish, as shown in the spatial distribution (Fig.33). Human motifs are sparsely carved by small sizes, most basically consisting of a body with legs and heads with few details (Coles and Gräslund 2000: 48). There is a famous scene where humans decorated with clothes are paddling (Boglösa 109, see Table 2). The wheel cross relating to humans could indicate a portal function having religious or cosmological meanings (Kristiansen and Larsson 2005; Kaliff 2007). For instance, as shown in Fig.34, a wheel cross above might indicate the sky; the circle below the feet of the human might indicate the earth. The human is carrying a weapon, a sword, and seemingly stands in a portal, perhaps in a betwixt and between state between ordinary life and going to war, and/or in the portal between the land of the living and the land of the dead. The wheel cross above the portal might act as a protective symbol of life.

Figure 34. The wheel cross related to the human and cup mark in Boglösa 126 panel. Credit: Einar Kjellén.

5.2. Multiple expressions of wheel cross motifs in Enköping Similar to that of Tanum, the wheel cross in rock carvings of Enköping varies in many forms and relates to other motifs such as ship, disc, cup mark, foot soles, human, etc. For instance, some are set beside cup marks (Boglösa 162), some are hovered over a ship (Boglösa 138), some are close to a group of discs (Vårfrukyrka 273), some to spoke wheels (Boglösa 58) or present in wagons, etc. The disc motif sometimes is another form of the wheel cross. Most disc motifs are found in the panels of Boglösa, where some are designed with a cup mark at the centre (Boglösa 111, 138), some are carved with internal lines (Långtora 54B, Litslena 230, 240), some are hovered over a ship (Härkeberga 41), some are attached to lines which looks like a human with legs but without arms and head (Biskopskulla 71). Other variants of 50

the wheel cross are plain circle, concentric lined disc, triple-lined disc, internal line and small disc (Boglösa 131), etc. Different variants of wheel cross motifs in rock carvings of Enköping are listed in Table 5, where the motif image, panel elevation as well as the description and dating are given. It is interesting to notice that the Boglösa 138 panel has several different forms of wheel cross expressions from different periods, making it quite special. As shown in Table 5, when the wheel cross is present in spoke wheels or wagons, it is usually dated to the beginning of EBA, i.e. Period I–II. Later in Period II–III, the wheel cross is found over a ship and the wheel cross could be replaced by a cup mark during Period V. It is also noticeable that the human torso covering the wheel cross (No. 6 in Table 5) is dated to the EBA, which is similar to that in Tanum. Boglösa 138 panel contains a section showing many regular and irregular foot soles together that are very close to the wheel cross (No. 7 in Table 5) in appearance, dated to Period V.

51

Table 5. Different variants of the wheel cross in rock carvings of Enköping. Interpretations marked with (*) are inspired by Coles (2000). All images are adapted from SHFA.

Forms No. Features RAÄ number Images Image credit Panel elevation Dating

EBA21 1 Spoke wheel cross Boglösa 131 Eva Kjellén 26 m Period I–II

Spoke wheel

Many spokes were EBA22 decorated on the outer side 2 Boglösa 58 Einar Kjellén 24 m Period II– of the wheel III

Two horses dragged a EBA23 Wagon 3 wagon with a human Boglösa 131 Dietrich Evers 26 m Period II standing on it

21 Dated by referring to the ship motifs nearby and the panel elevation. 22 Suggested by shoreline dating since there are no other motifs nearby. 23 Very similar to the cart found in the famous Kivik grave (Malmer 1981: 17; Johannsen 2011: 98-99).

Forms No. Features RAÄ number Images Image credit Panel elevation Dating

Two wheel cross motifs EBA21 4 were above the line crew Boglösa 138 Eva Kjellén 24 m Period II– of a ship III

Sun ship

A disc was above a ship; A LBA21 5 wheel cross was below the Boglösa 138 Bertil Almgren 23 m Period V ship

The wheel cross EBA21 Wheel cross covering related to 6 Boglösa 131 Eva Kjellén 26 m Period II– human torso humans III

Regular or Some foot soles together LBA21 irregular foot 7 looked very similar to the Boglösa 138 Bertil Almgren 23 m Period V soles wheel cross

53

Forms No. Features RAÄ number Images Image credit Panel elevation Dating

A big cup mark worked as the sun with foot rays; A horse dragged the sun in Boglösa 160 Sun disc 8 Einar Kjellén 25 m EBA24 the right corner of the image, similar to sun horse*

The wheel cross and the 9 disc indicated they were Vårfrukyrka 273 Einar Kjellén 33 m EBA22 moving

Concentric disc, cup mark A concentric disc with a Härkeberga 41 EBA21 10 curve or handle was Einar Kjellén 26 m Period I–II located above a ship*

24 Dated by referring to the sun horse motif nearby and the panel elevation.

54

Forms No. Features RAÄ number Images Image credit Panel elevation Dating

The disc worked as the EBA21 11 Litslena 230 Einar Kjellén 24 m wheel of the ship Period I–II

A life sized disc or shield EBA21 12 Boglösa 141 Einar Kjellén 25.6 m was crossed by a ship* Period I–II

55

6. Discussion

6.1. Wheel cross motifs in a cultural-historical-geographical context In Chapter 4 and 5, different variants of the wheel cross in South Scandinavian rock carvings have been summarized. It is found that the wheel cross ( ) has many different meanings that were expressed through its own variations such as wheel ( ), circle ( ), disc ( ), cup mark ( ), as well as by relating to other motifs such as ship ( , ), human ( , ), animal ( , ), foot soles ( ), cup mark ( ), etc. These different scenarios from rock carvings of Tanum and Enköping suggest that the wheel cross might not only have a practical meaning as the wheel of wagon or as the shield covering human torso, a cosmological meaning as the sun, but also have symbolic and religious meanings. However, one natural question is: did the wheel cross really have so many different meanings? To answer this question, it is necessary to perform the ideological analysis, which was proposed by Malmer (1981: 109) as contemporary cultural context analysis, later developed by Hultkrantz (1986: 42) as the comparative study and holistic interpretation. Recently, the method was extended to the intercontextual strategy of interpretation (Kristiansen and Larsson 2005: 12). The intercontextual analysis suggests probing archaeology with a holistic cultural historical framework (Kristiansen and Larsson 2005: 372). Based on this framework, the theoretical and interpretative studies are done from the perspective of comparative, contextualised culture history. In short, the interpretation of rock carvings should not be restricted by archaeological material data in selected areas, but needs to include other materials and material culture as well as the materiality study as much as possible. The more culture history is involved and analysed, the closer it is to the original meanings of interpretation. To reach this point, it is necessary to at least include three examples of material culture, related to the wheel cross. The first example is the famous bronze sun chariot from Trundholm, Denmark, as shown in Fig.35. It was dated to about 1400 BC, i.e., the BA Period II. The wheel cross is so tightly connected with the sun that in many cases it is also named as sun cross, an important symbol in BA society (Nilsson 1971: 421).

Figure 35. Trundholm sun chariot. Height: 35 cm, width: 54 cm. Credit: National Museum of Denmark.

The second example is the stone slabs found in the Kivik grave, as shown in Fig.36. This grave is located in Bredarör, Scania of Sweden. Like the Trundholm sun chariot, it is also dated to between 1400 BC and 1300 BC. As can be seen from Fig.36, a large pair of wheel crosses was depicted on slab 4 and 6 and a cart was depicted on slab 7. Large curved axes, typical symbols from the EBA, used for ritual or ceremony, are also found on slab 1 and 6. It indicates the cart was used in the upper class of the BA society since the EBA. The cart in slab 7 was probably imported, originally, from Egypt (Johannsen 2011: 103). Fig.37 shows a well-preserved Egyptian cart, dated to about 1350 BC. Compared with the ship, which is no doubt the most important transportation in BA Scandinavian society, the wagon and cart should have a more pure religious meaning than their practical meaning (Johannsen 2011: 106). In this context, the wheel cross acts more as a virtual wheel, for instance, the wheel of life for prehistoric elites (Kaplan 1975; Sopa 1984). The stone slab 7 in Fig.36 shows a row of carved humans who have very different postures at the bottom and four humans with similar postures in the up left corner. This slab probably describes a traditional cart ceremony. The slab 6 has twin axes carved above twin wheel crosses and this indicates that they have very close meanings or functions in BA society. Besides, the slab 2 has a ship-like carving. All the carvings on the slabs were very similar to the rock carvings found in Tanum and Enköping. However, the carvings on stone slabs are different from the rock carvings in the sense that the former already separates the decorations into different groups more distinctly than the later (Malmer 1981: 33). The stone slab carvings are more neat and symmetrical and more aesthetically significant.

Figure 36. The eight stone slabs decorated with carvings from the Kivik grave. Adapted from Goldhahn (2009: 362).

57

Figure 37. The photo of Florence chariot. Credit: The Egyptian Museum of Florence. The third example is the bronze horn or trumpet, found in Wismar, Mecklenburg, Germany, as shown in Fig.38. This trumpet was dated to EBA Period III (Malmer 1981: 33). It is of great importance in the chronological study of rock carvings since there are different kinds of symbols or pictures depicted on the trumpet such as ship, concentric rib, warrior as well as the wheel cross and sun disc.

Figure 38. Wismar trumpet (up) and the decorations on it (down). Credit: Danish art. As can be seen from Fig.38, the Wismar trumpet was decorated with four pairs of wheel cross motifs, one sun disk surrounded with birds, two warriors with concentric disc torso and spear, several ships and many connected concentric ribbons, all are similar to the rock carvings discussed before. It is interesting to notice that the above three examples, Trundholm sun chariot, Kivik 58

stone slabs and Wismar trumpet are located in the Southmost part of Scandinavia, a location that not only connects Scandinavia and Continental Europe, but also connects the North Sea and the Baltic Sea, as shown in Fig.39. Kivik of Sweden, Trundholm of Denmark and Wismar of Germany form a triangle that connects South Scandinavia, North Sea, Baltic Sea and Continental Europe. This part of the region should have played a significant role not only in prehistory but also nowadays, in various aspects such as economy, culture and religion etc.

Figure 39. Schematic diagram showing how Southern Scandinavia is connected with the outside world. The positions of Kivik, Trundholm and Wismar are also shown. The background map is retrieved from Google Map. The wheel cross motif, as an icon or symbol, reflects the cultural identity (Werbart 2001: 1). It indicates the cultural interaction between Southern Scandinavia and Continental Europe as well as the eastern Mediterranean during the LN and BA. The evidence such as goods and symbols have shown these interactions are reflected in various aspects such as religion, economy and politics (Kristiansen and Larsson 2005; Uhnér 2005; Vogt 2011; Uhnér 2012; Rowlands and Ling 2013). The social transformation and communication between Southern Scandinavia and the eastern Mediterranean have an impact on rock carvings, and the wheel cross is not an exception. Therefore, a big picture like Fig.39 puts the wheel cross expression in a cultural-historical-geographical context. This context or framework enables us to obtain a comprehensive and holistic interpretation of South Scandinavian rock carvings. Fig.39 shows that the cultural diffusion forms a virtual cross that connects the West and East, the North and South, which can also be viewed as a lifeline. The lifeline acts as an agency to link Southern Scandinavia to the outside world. It enriches and shapes the Scandinavian society. 59

6.2. The wheel cross beyond expressions and representation During the transition from LN to BA and in Period I, the wheel cross is only present as a wheel ( ), found in a very limited number in Enköping (Table 5) and even less in Tanum. A few more were found in Scania and West Denmark during the EBA. Malmer (1981: 44) suggests that the wagon was used for specific cult ceremonies in LN but it was then replaced by rock carvings of other motifs, mainly ships. During Period I, there were some motifs that were showing the concentric disc ( ) above the ship ( ). At this time, it seems that there is no connection between the wheel cross and the disc. During BA Period II–III, the culture of sun horse and sun ship diffused to both Tanum and Enköping. For the sun horse ( ), the sun was represented by a cup mark or a circle, without a cross inside. And for the sun ship ( ), the sun was represented by a wheel cross (see Table 4 and Table 5). It is generally accepted that the idea of sun cult and sun god related to the wheel cross originally came from the Near East, Egypt and Greece (Larsson and Orrling 1995; Winter 2001; Kristiansen 2005, 2012). During the transition from Period III to IV, the connection between the wheel cross and sun has been emphasized in such a way that more variations of the wheel cross were developed during the LBA. Therefore, it is the sun ship and sun horse that add cosmological and religious meanings (sun and sun god) to the wheel cross. With the importing of bronze objects and technology from Continental Europe and Mediterranean, Southern Scandinavia also introduced BA culture and religion, however, with adaptation and development later. The overview of the introduction and duration of different variations of the wheel cross in South Scandinavia rock carvings during different BA periods is illustrated in Fig.40.

1700 1500 1300 1100 900 700 500, BC I II III IV V VI Period

Figure 40. Overview of the introduction and duration of different variations of the wheel cross in South Scandinavian rock carvings during different periods of Bronze Age. Made by the author. The dark grey region shows the periods where the corresponding motifs are found more than once. The light grey colour marked the periods where the motifs are only found once or confirmed dating is not available. 60

Fig.40 clearly shows how the wheel cross was introduced and developed during the BA and it also reflects the relation between the wheel cross and other motifs such as ship, animal, human, etc. Such a dynamic picture is the epitome of the development of the BA rock carvings as well as the society of Southern Scandinavia. Based on the above analysis, the case of the wheel cross covering the human torso has more religious meaning than practice, i.e., it probably means an amulet instead of a shield. Therefore, the human whose torso was covered by a wheel cross probably wanted the sun god or other religious entity to protect them and send good luck to them. In this context, the wheel cross decorated by many cup marks inside probably indicates a prayer ceremony for fertility, or asking for help to fight the fear of death (Bradley 2005: 3–40). The wheel cross relating to the cup mark can also be used as important elements for fire sacrifice rituals in South Scandinavia (Kaliff 2007: 121–134). As a summary, the wheel cross shown in Scandinavian rock carvings has a meaning beyond its practical meaning and visual expressions as the wheel of wagon or as the shield covering human torso. With a cosmological meaning as the sun, and religious meanings as god, the wheel cross becomes the symbols of moving, fertility, power and life, which has been integrated into not only the rituals but also the domestic life of the BA society of Southern Scandinavia. Inspired by Goldhahn (2002: 29) who has tried to interpret rock carvings in North Sweden with an audio-visual perspective (Scarre 1989; Waller 1993; Lawson et al. 1998; Devereux 2001; Watson 2001), it is interesting here to notice that the wheel cross also have special audio-visual experience. One could imagine a picture where a human standing close to the BA shoreline and heard the sound of horse hoofs and rotating wheels coming from afar. The wheel cross, as a special motif that connects other different motifs and aspects, provides a possibility to study the rock carvings from the background of landscape with a visual scene and hearing sense. The visual and audio experiences of the rock carvings are evaluated in two different dimensions, which also form a virtual cross in an orthogonal coordinate system. From this perspective, this form of lifeline enables a multidimensional expression and representation of the wheel cross as well as other motifs in Scandinavian rock carvings. As can be seen from Fig.40, the sun horse and sun ship motifs lasted until the LBA. During this period, the expression of the wheel cross became much more diverse and flexible, e.g., the wheel cross sometimes was replaced by the disc, circle or even cup mark. Such a mixed usage of the wheel cross indicates that the wheel cross acts as a lifeline that connects other motifs naturally, not only from the typological point of view, but also from the perspective of material culture and materiality. With regards to this, the wheel cross is so special that none of the other motifs can replace it. Like a wheel without the spoke could not physically support a BA cart, a society without the wheel cross could not be mentally supported. Therefore, the cross in the circle is more than the lifeline of the wheel cross, but rather, it is the lifeline of the BA society. From the perspective of materiality, the wheel cross is not only the agency to assemble and link other motifs together, but also the agency to connect the culture, the religion, the human, the society and other materialities.

6.3. Comparative analysis of rock carvings of Tanum and Enköping When it comes to comparisons between the rock carvings of Tanum and Enköping, Table 6 lists their similarities and differences from different perspectives, such as landscape, locations of rock carvings, motifs, economic basis, social structure, etc. It is clear to see that although Tanum and Enköping share many similarities, they are also different in many ways. Note that the aim of Table 6 is to illustrate a general overview of the similarities and differences between Tanum and Enköping. There might be some cases that are exceptions, making this table not complete and universal. However, it is believed that the contributions of Table 6 in achieving a big picture are more important than the incomplete information it potentially contains.

61

It has to be mentioned that all different aspects that are compared in Table 6 are connected with each other. Each line of comparison shall not be understood alone. For example, rock carvings are generally more abstract in Enköping and more elaborative in Tanum. This might be misunderstood since to some degree this justification is incomplete. Note that most motifs from Enköping are dated to the EBA while Tanum to the LBA. To this end, the differences of motif expression between Enköping and Tanum are also the differences of motif between the EBA and the LBA. Thus, it is argued that comparison in one specific aspect for two regions is unequal. For instance, in LBA Enköping, there are also well- polished motifs with many decorations such as the paddler carved on the ship in Boglösa 109 panel (see the corresponding image in Table 2). It shows that the crew on the ship are decorated with clothes and the nearby animals are more lively and particular. Table 6. Comparison of Tanum and Enköping in different aspects. Made by the author. Most conclusions are based on this study with some supported from other literature.

Aspect Tanum, Bohuslän Enköping, Uppland

Simi. Seascape, close to the sea and easy access to water

Landscape Rolling hills, rock with smooth surface Flat plain, rock outcrops with glacial Diff. (Coles and Gräslund 2000: 11) striation on the surface

Simi. BA Shoreline have elevated more than 10 m Average uplift velocity 3.75 mm/year Average uplift velocity 5.75 mm/year Land uplift Diff. BA shoreline is 16–10 m.a.s.l. and BA shoreline is 25–16 m.a.s.l. and slightly different from present significantly different from present Simi. Close to shoreline or water Locations of rock Elevation 14–55 m with many located Elevation 15–35 m with most close carvings Diff. inland; most panels are facing east and to shoreline; most panels are facing south east north and north east

Dating of rock Simi. Most rock carvings dated to the BA carvings Diff. More belonging to the LBA More belonging to the EBA Simi. Cup mark motifs dominate and spread out, with an equivalent number Cup mark Usually appears independently, quite Diff. Usually appears with other motifs different from other motifs Simi. All have different motifs such as ship, human, animal, etc. Figurative motifs 5–10 times more motifs in number, Most motifs are abstract and generic Diff. more elaborative (Ling 2013: 99)

Wheel cross Simi. Multiple expressions relating to ship, wagon, cup mark and foot soles expressions Diff. Most dated to the LBA Most dated to the EBA

Impact of foreign Simi. Central Europe and Mediterranean culture Diff. Also from West Denmark (Jutland) Also from Baltic regions and Russia Simi. Agriculture and husbandry Economic basis More diversity, with additional hunting Diff. More uniform and fishing Simi. Public activity with ritual and religious ceremony Social structure Diff. Several centres develop simultaneously Centralisation at Boglösa

62

It should be emphasized that most differences between Tanum and Enköping are determined or affected by their geographical features, the landscape, landform, rock properties, etc. For instance, as discussed in Subchapter 3.2, the aspects of rock carving panels are affected by the direction of river or water. Most rock panels are facing specific directions inherently with their locations. In other words, it might not be the leader or craftsman who decided their aspects. Therefore, on one hand, rock carvings in Tanum and Enköping are different, but on the other hand, they are the same, both reflecting the geographical impact. The second example is the difference of economic basis. This could be understood by the fact that most land of Enköping elevated during BA and it is a kind of new land (Fig.11), making Enköping influenced much easier than other places with much smaller shoreline displacement such as Tanum. The previous studies have shown that most EBA rock carvings in both Tanum and Enköping had a trend of pursuing the BA shoreline, whereas the LBA rock carvings deviated from the shoreline (Ling 2013: 98). Let us take the Boglösa 138 panel as a first example. As shown in Fig.10, the ship motif in the upper right corner was carved at an altitude of 25.2 m. This position was only a few meters from the shoreline of Period I, which still covered part of the panel. Fig.10 indicates that the Boglösa 138 panel has been elevated out of the sea level since about 1300 BC. In order to illustrate the later shorelines, another landscape overview is shown in Fig.41. Apart from the Boglösa 138, other rock carving panels are also plotted with purple dots. The richer panels of Enköping, defined as the number of ship motifs larger than 15, are also marked with white dots. It is interesting to find that most of the richer panels are located at the shoreline of 1700 BC. None of the richer panels are located at the shoreline later than 900 BC. Getting back to Fig.10, the right middle ship, which is dated to Period V (shoreline altitude is about 18 m according to Table 3), was carved at an altitude of 23.2 m. This was about 300 meters from the shoreline of the start of Period V (900 BC) according to Fig.41. It indicates that the LBA rock carvings of Enköping did not pursue the shoreline closely anymore, although there were many new rocks elevated from the sea, as shown in Fig.41. Similar to the Boglösa 138 panel, let us take another example, the Tanum 311 panel. The BA landscape and the rock carvings around the panel are shown in Fig.42. The richer panels of Tanum, defined as the number of ship motifs larger than 30, are marked with white dots. Different from that of Enköping, very few rock panels are located at the shoreline later than 1700 BC. A photo of Tanum 311 with the surrounding environment taken in 2006 is shown in Fig.43. As can be seen from Fig.42, there was new exposed land during the EBA around Tanum 311, although it was much smaller than that of Enköping. There must be important reasons behind the fact that most LBA rock carvings of Tanum were not pursuing the shoreline. As indicated by Fig.43 and presumably, the new land was just a shallow bank of silt, without new rock that was suitable for carving. The above analysis shows that the different geographical and rock features of Tanum and Enköping could significantly affect the outcomes of the carvings. In other words, the rock and the landscape had the agency to influence the carving motifs and their placing, i.e., their agency created certain relations with humans that made them carve in certain ways. To this end, all the properties of materials have to be identified, categorized, and linked to different materialities such as artefacts, landscape, settlements, vegetation, etc., to find their processual and relational aspects. Therefore, it is of importance to apply the theory of materiality – the multidimensional interpretation framework – to obtain a comprehensive and unbiased understanding.

63

Figure 41. BA landscape around the Boglösa 138 panel. Made by the author.

Figure 42. BA landscape around the Tanum 311 panel. Made by the author.

65

Figure 43. Photo of Tanum 311 panel taken in 2006. Credit: Åsa Fredell.

7. Conclusion

In this thesis work, case studies of Tanum and Enköping have been performed to investigate the meaning of the wheel cross motif in South Scandinavian rock carvings. The main conclusions of this work are summarized as follows. To understand the shifting meaning of the wheel cross in time, it is necessary and important to date it as correctly as possible. In this work, both the stylistic dating and shoreline dating methods are employed. It is found that the shoreline dating method can only be used to analyse individual rock carvings situated close to the BA shoreline. It can only suggest an upper bound age of the rock carving but cannot provide an overall solution to the dating. The shoreline dating method acts as an effective auxiliary method to the traditional stylistic dating method. Moreover, with the reconstruction of the BA shoreline and landscape, it is shown that most EBA rock carvings have a much closer attachment to BA shorelines than those of the LBA in Southern Scandinavia. A detailed discussion on the meaning of the wheel cross motif is made in a cultural- historical-geographical framework. During Period I, the wheel cross had very limited expressions such as the wheel of a wagon. However, as the culture of ‘sun ship’ and ‘sun horse’ diffused to South Scandinavia during Period II and III, the wheel cross was given more complex meanings in different aspects such as cosmology, life, ritual, and religion. The meaning of the wheel cross motif was then developed by the local elites during the LBA, resulting in multiple expressions of the wheel cross in different scenarios. The mixed usage of the wheel cross with other motifs such as cup mark and sun disc indicates that the wheel cross could be used as a link that connected not only other symbolic motifs, but also the landscape, culture, religion and other materialities. It is argued that the wheel cross is so special in terms of typology and religion that none of the other motifs can replace it. The shifting meaning of the wheel cross is the epitome of the development of the BA rock carvings as well as the society of Southern Scandinavia. Comparative analysis is conducted to process the material data from the selected case study areas. There are many differences between the rock carvings of Tanum and Enköping in many aspects such as locations, dating, styles and expressions. These differences indicate their dissimilarities in economic bases and social structure. Enköping showed a much simpler economic form and a centralisation form of social structure; whereas Tanum had more mixed economic components and clustered social units that developed simultaneously. It is also found that rock carvings from Period I of Tanum and Enköping generally pursue the shoreline. The locations of rock carvings deviate from the shoreline from Period II and Period IV for Tanum and Enköping, respectively. However, all the differences between Tanum and Enköping are tightly connected with and affected by their own geographical features, i.e., the landscape had agency to affect the outcomes of the carvings. It is also argued that any single comparison between two areas might be misleading and should be avoided. The theories of materiality and lifeline have been applied to analyse the materials and interpret the shifting meanings of the wheel cross in space and time. It is argued that the lifeline also has multiple expressions and it plays an important role not only in the aspect of rock carving panels but also in the connection between Southern Scandinavia and the outside world.

67

8. Summary

The thesis is aimed at exploring the multiple expressions and shifting meanings of the wheel cross motif throughout space and time in South Scandinavian rock carvings. The aim can be specifically formulated into three research questions: 1). Did different expressions of the wheel cross in South Scandinavian rock carvings share the same meaning? If Yes/No, what might their meanings have been? 2). Did the meanings of the wheel cross change in different BA periods, and how? 3). Did different regions share the same answers to the above questions? In other words, what are the differences and similarities about the expressions of the wheel cross in different study areas? To solve the above questions, two typical regions with extremely rich rock carvings were selected as case study areas, i.e., Tanum and Enköping municipalities. The material data of rock carvings, including geographic coordinates, motif types and quantities, was downloaded from FMIS. It included 1489 and 1847 rock carving panels from Tanum and Enköping, respectively. The rock carving materials with the wheel cross motif were processed by chorological and chronological studies. In addition to the traditional stylistic dating method, the shoreline dating method was also used to perform chronological analysis of the wheel cross. In order to overcome the difficulty encountered in reconstructing prehistoric shoreline, a simpler and more general GIS method was proposed and it was only dependent on the local land uplift velocity. The results computed with the new method are in good agreement with previous studies. To handle the FMIS database more efficiently, the thesis also provided the scientific community of rock carvings study with a programming code to automatically count the number of different rock carving motifs. The chronological study provided an overview of the introduction and duration about different variations of the wheel cross in South Scandinavian rock carvings during different periods of the Bronze Age. The results of the chronological study were then analysed in a cultural-historical-geographical context using an intercontextual interpretation and comparative analysis method. The intercontextual interpretation suggested that the various visual expressions of the wheel cross such as the wheel of wagon or as the shield covering a human torso, had unique symbolic meanings beyond their practical and physical counterparts. With a cosmological meaning as the sun, and a religious meaning as god, the wheel cross becomes the symbol of moving, fertility, power and life, which has been integrated into not only the rituals but also the domestic life of South Scandinavian society. Although the wheel cross motif is limited in number, its various forms relating with other motifs are able to reflect how the Bronze Age society of Scandinavia imported and developed ideas, artefacts, stories, etc. from foreign cultures in Continental Europe and the Mediterranean area. Like the cross spoke physically supporting the cart, the wheel cross acts as the lifeline and agency that closely linked landscape, ritual, artefact, human, and other materialities to mentally support South Scandinavian society. The results of comparative analysis of rock carvings from Tanum and Enköping are summarized with a detailed table that lists both the similarities and differences from different perspectives, such as landscape, locations of rock carvings, motifs, economic basis, social structure, etc. It is shown that all the differences are tightly connected with and affected by their own geographical features, i.e., the landscape had the agency to affect the outcomes of the carvings. Therefore, it is important to apply the theory of materiality to obtain a comprehensive and unbiased interpretation of rock carvings.

68

Bibliography

Aching, G. 2002, Masking and power: Carnival and popular culture in the Caribbean (U of Minnesota Press). Anati, E. 1960, 'Bronze Age Chariots from Europe', Proceedings of the Prehistoric Society (26: Cambridge University Press), 50-63. Andrén, A. 2014, Tracing Old Norse cosmology: the world tree, middle earth, and the sun from archaeological perspectives (16: Nordic Academic Press). Back Danielsson, I.-M., Fahlander, F., and Sjöstrand, Y. 2012, 'Imagery beyond representation', in I.-M. Back Danielsson, F. Fahlander, and Y. Sjöstrand (eds.), Encountering imagery: materialities, perceptions, relations (Stockholm: Stockholm University), 1-12. Bengtsson, L. 2004, 'Cup marks of the common people, images of the elite', in G. Milstreu and H. Prøhl (eds.), Prehistoric pictures as archaeological source (Tanumshede: Tanums Hällristningsmuseum Underslös), 167-177. Berger, J. 2008, Ways of seeing (Penguin Books UK). Berntsson, A. 2006, 'The isolation of Lake Raftötångstjärnet and its implications of the dating of rock art in Tanum, Bohuslän', (Stockholm: Institutionen för naturgeografi och kvartärgeologi, Stockholms university). Bertilsson, U. 1987, 'The rock carvings of northern Bohuslän : spatial structures and social symbols', Doctoral thesis, monograph (Stockholm University). --- 2013, 'Divine footprints. Traces of cosmological archetypes and prehistoric religion on the rock faces', Art as source of History, 163-172. --- 2017, 'The Spear-Digital documentation sheds new light on Early Bronze Age spear carvings from Sweden–an analysis with some comparative examples from Valcamonica, Italy', Adoranten, 64-83. Bhabha, H. K. 2012, The location of culture (Routledge). Bradley, R. 1997, 'Death by water: boats and footprints in the rock art of western Sweden', Oxford Journal of Archaeology, 16 (3), 315-324. --- 2005, Ritual and domestic life in prehistoric Europe (Psychology Press). Burenhult, G. 1980, 'Götalands hällristningar', (Institute of Archaeology, Stockholm University). Clemmensen, L. B., Richardt, N., and Andersen, C. 2001, 'Holocene sea-level variation and spit development: data from Skagen Odde, Denmark', The Holocene, 11 (3), 323-331. Coles, J. 1999, 'The dancer on the rock: record and analysis at Järrestad, Sweden', Proceedings of the Prehistoric Society (65: Cambridge University Press), 167-187. --- 2004, 'Bridge to the outer world: Rock carvings at Bro Utmark, Bohuslän, Sweden', Proceedings of the prehistoric Society (70: Cambridge University Press), 173-205. --- 2005, Shadows of a northern past: rock carvings in Bohuslän and Ostfold (Oxbow Books). Coles, J. and Gräslund, B. 2000, Pattern's in a Rocky Land: Rock Carvings in South-West Uppland, Sweden (Department of Archaeology and Ancient History). Cornell, P. and Fahlander, F. 2009, Encounters| Materialities| Confrontations: Archaeologies of social space and interaction (Cambridge Scholars Publishing). Demarrais, E., Gosden, C., and Renfrew, C. 2004, Rethinking materiality: The engagement of mind with the material world (McDonald Institute for Archaeological Research). Derrida, J. 1978, Writing and difference (University of Chicago press). Devereux, P. 2001, Stone Age soundtracks: the acoustic archaeology of ancient sites (Sterling Publishing Company, Inc.). Eronen, M., et al. 2001, 'Rates of Holocene isostatic uplift and relative sea-level lowering of the Baltic in SW Finland based on studies of isolation contacts', Boreas, 30 (1), 17-30. Fahlander, F. 2004, 'Archaeology and anthropology–brothers in arms?', Material culture and 69

other things, 157. --- 2007, 'Third Space Encounters: Hybridity, Mimicry and Interstitial Practice', in P. Cornell and F. Fahlander (eds.), Encounters, Materialities, Confrontations: Archaeologies of social space and interaction (Cambridge Scholars Publishing), 15-41. --- 2008a, 'Same, same, but different? Making sense of the seemingly similar', Facets of archaeology. Essays in honour of Lotte Hedeager on her 60th birthday. Oslo, 67-74. --- 2008b, 'Differences that matter: Materialities, material culture and social practice', in G. Håkon and H. Lotte (eds.), Six essays on the materiality of society and culture (Bricoleur Press), 127-154. --- 2012, 'Articulating stone. The material practice of petroglyphing', in I.-M. Back Danielsson, F. Fahlander, and Y. Sjöstrand (eds.), Encountering imagery: materialities, perceptions, relations (Stockholm: Stockholm University), 97-116. Fredell, Å. 2004, 'To let the picture talk', in G. Milstreu and H. Prøhl (eds.), Prehistoric pictures as archaeological source (Tanumshede: Tanums Hällristningsmuseum Underslös), 137-147. Gell, A. 1998, Art and agency: an anthropological theory (Clarendon Press). Gibson, W. 2006, 'Material culture and embodied action: sociological notes on the examination of musical instruments in jazz improvisation', The Sociological Review, 54 (1), 171-187. Goldhahn, J. 2002, 'Roaring rocks: an audio-visual perspective on hunter-gatherer engravings in northern Sweden and Scandinavia', Norwegian Archaeological Review, 35 (1), 29- 61. --- 2008, 'Rock art studies in northernmost Europe, 2000–2004', in P. Bahn (ed.), Rock Art Studies News of the World (3; Oxford: Oxbow Books), 16-36. --- 2019, Birds in the Bronze Age - a North European Perspective (Cambridge University Press). Goldhahn, J., Fuglestvedt, I., and Jones, A. 2010, Changing Pictures: rock art traditions and visions in northern Europe (Oxbow Books). Hallström, G. 1938, Monumental Art of Northern Europe from the Stone Age: I. The Norwegian Localities (Thule). Helskog, K. 2014, 'Petroglyphs of Boats as Evidence of Contact between the Caspian Sea and Scandinavia', in V. Roggen (ed.), Thor Heyerdahl’s Search for Odin: Ancient Links between Azerbaijan and Scandinavia? (Novus Press), 202-219. Hjarthner-Holdar, E. and Risberg, C. 2001, 'The innovation of iron: From Bronze Age to Iron Age societies in Sweden and Greece', BAR International Series, 985, 29-42. Hodder, I. 2006, The leopard's tale: revealing the mysteries of Çatalhöyük (Thames & Hudson). Hultkrantz, Å. 1986, 'Rock drawings as evidence of religion: some principal points of view', in G. Steinsland (ed.), Words and Objects: Towards a Dialogue Between Archaeology and History of Religion (70: Oxford University Press), 42-66. Hünicke, B., et al. 2015, 'Recent Change—Sea Level and Wind Waves', in Bacc I. I. Author Team The (ed.), Second Assessment of Climate Change for the Baltic Sea Basin (Cham: Springer International Publishing), 155-185. Ingold, T. 2006, 'Rethinking the animate, re-animating thought', Ethnos, 71 (1), 9-20. --- 2007, 'Materials against materiality', Archaeological dialogues, 14 (1), 1-16. Jensen, R. 1985, 'Bronze Age settlement patterns in the Mälaren basin–exological and social relationships. I: H-Å Nordström & A', Knape (red.) Bronze Age studies: transactions of the British-Scandinavian Colloquium in Stockholm, 133-150. Johannsen, J. W. 2011, 'Carts and wagons: on Scandinavian rock carving sites', Adoranten, 95. Jones, A., Díaz-Guardamino, M., and Crellin, R. J. 2016, 'From artefact biographies to ‘multiple objects’: a new analysis of the decorated plaques of the Irish Sea region', Norwegian Archaeological Review, 49 (2), 113-133. Kaliff, A. 2007, Fire, water, heaven and earth: ritual practice and cosmology in ancient Scandinavia: an Indo-European perspective (Riksantikvarieämbetet). Kaplan, R. W. 1975, 'The sacred mushroom in Scandinavia', Man, 10, 72-79. Kaul, F. 1998, Ships on bronzes: a study in bronze age religion and iconography: catalogue of Danish finds. II (National Museum of Denmark, Department of Danish Collections). 70

Kjellén, E. 1976, Upplands hällristningar (The rock carvings of Uppland, Sweden), trans. Åke Hyenstrand (Lund: Vitt.- Hist.- O. Ant.- Akad). Kristiansen, K. 1984, 'Ideology and material culture: an archaeological perspective', Marxist perspectives in archaeology, 72-100. --- 2005, 'Cosmology and consumption in the Bronze Age', in J. Goldhahn (ed.), Mellan sten och järn (Göteborg: Gotarc Serie C. Arkeologiska Skrifter), 135–149. --- 2012, 'Rock art and religion: the sun journey in Indo-European mythology and Bronze Age rock art', Adoranten, 69. Kristiansen, K. and Larsson, T. B. 2005, The rise of Bronze Age society: travels, transmissions and transformations (Cambridge University Press). Kveiborg, J. 2019, 'Traversing Sky and Earth. The Nordic Bronze Age horse in a long-term perspective', Praehistorische Zeitschrift, 93 (2), 225-264. Lahelma, A. 2017, 'The circumpolar context of the ‘sun ship’–motif in South Scandinavian rock art', in P. Skoglund, J. Ling, and U. Bertilsson (eds.), North Meets South: Theoretical Aspects on the Northern and Southern Rock Art Traditions in Scandinavia (Oxford: Oxbow Books), 144-171. Larsson, T. B. and Orrling, C. 1995, 'Symbols in a European Bronze Age cosmology', Communication in Bronze Age Europe. Transactions of the Bronze Age Symposium in Tanumstrand, Bohusländ, Sweden, 7-5. Latour, B. 1992, 'Where are the missing masses? The sociology of a few mundane artifacts', in W. E. Bijker and J. Law (eds.), Shaping technology/building society–studies in sociotechnical change (Cambridge: MIT Press), 225-258. Lawson, G., et al. 1998, 'Mounds, megaliths, music and mind: some thoughts on the acoustical properties and purposes of archaeological spaces', Archaeological Review from Cambridge, 15 (1), 111-134. Ling, J. 2005, 'The fluidity of rock art', in J. Goldhahn (ed.), Mellan sten och järn (Göteborg: Gotarc Serie C. Arkeologiska Skrifter), 437-460. --- 2006, 'Elevated rock art. Maritime images and situations', Adoranten, 5-32. --- 2008, Elevated Rock Art: towards a maritime understanding of rock art in northern Bohuslän, Sweden (GOTARC Serie B, Archaeology Thesis, 49; Stockholm: Intellecta Solna). --- 2013, Rock art and seascapes in Uppland (Oxbow Books). Ljunge, M. 2013, 'Beyond ‘The Phenomenological Walk’: Perspectives on the experience of images', Norwegian Archaeological Review, 46 (2), 139-158. --- 2016, 'Picturing the meaning of scandinavian rock art: graphic representations, archaeological interpretations and material alterations', Current Swedish Archaeology, 24, 163-190. Malmer, M. 1981, A chorological study of north European rock art (Almqvist & Wiksell International). --- 1989, 'Principles of a non-mythological explanation of North-European Bronze Age rock art', Bronze Age Studies, 6, 91-99. Meskell, L. 2008, Archaeologies of materiality (John Wiley & Sons). Mikkelsen, E. 1983, 'Mesolithic hunthers’ rock carvings at Geithus, Buskerud, Norway', Ars Praehistorica, 1, 61-86. --- 1986, 'Religion and ecology: motifs and location of hunters’ rock carvings in eastern Norway', in G. Steinsland (ed.), Words and Objects: Towards a Dialogue Between Archaeology and History of Religion (70: Oxford University Press), 127-141. Miller, D. 2005, Materiality (Duke University Press). Milne, G. and Shennan, I. 2013, 'SEA LEVEL STUDIES | Isostasy: Glaciation-Induced Sea- Level Change', in Scott A. Elias and Cary J. Mock (eds.), Encyclopedia of Quaternary Science (Second Edition) (Amsterdam: Elsevier), 452-459. Milstreu, G. 1977, Rock carvings in Tanum (Tanum Turistsentrum). --- 2017, 'Re-cut rock art images (with a special emphasis on ship carvings)', New Perspectives on the Bronze Age Proceedings of the 13th Nordic Bronze Age Symposium (Gothenburg). Mörner, N.-A. 2012, 'Strict solar alignment of Bronze Age rock carvings in SE Sweden', Journal of Archaeological Science, 39 (10), 3301-3305. Nilsson, M. P. 1971, The Minoan-Mycenaean religion and its survival in Greek religion (9: Biblo & Tannen Publishers). 71

Nimura, C. 2012, 'Rock art and coastal change in Bronze Age Scandinavia', in I.-M. Back Danielsson, F. Fahlander, and Y. Sjöstrand (eds.), Encountering imagery: materialities, perceptions, relations (Stockholm: Stockholm University), 117-132. --- 2015, Prehistoric rock art in Scandinavia: agency and environmental change (Oxbow Books). Olsen, B. 2003, 'Material culture after text: re‐membering things', Norwegian Archeological Review, 36 (2), 87-104. Petersson, H. and Christina, T. 2015, Domestic life in the Tanum rock carving area : settlements, graves and rituals in the vicinity of the world heritage rock carvings of Tanum (: Bohusläns museums förlag). Plikk, A. 2010, 'Shore displacement in Fjärdhundraland, SW Uppland, and the northern coastal areas of Lake Mälaren since c. 1000 BC', (Stockholms University). Påsse, T. 2001, 'An empirical model of glacio-isostatic movements and shore-level displacement in Fennoscandia', (Swedish Nuclear Fuel and Waste Management Co.). Ramqvist, P. H., Forsberg, L., and Backe, M. 1985, 'and here was an elk too... A preliminary report of new petroglyphs at Stornorrfors, Ume River', Honorem Evert Baudou, 313- 337. Rédei, A. C., Skoglund, P., and Persson, T. 2019, 'Applying cartosemiotics to rock art: an example from Aspeberget, Sweden', Social Semiotics, 29 (4), 543-556. Rowlands, M. and Ling, J. 2013, 'Boundaries, flows and connectivities: mobility and stasis in the Bronze Age', Counterpoint: essays in archaeology and heritage studies in honour of Professor Kristian Kristiansen, 517-529. Sanz, I. D. and Fiore, D. 2014, 'Style: Its Role in the Archaeology of Art', in Claire Smith (ed.), Encyclopedia of Global Archaeology (New York, NY: Springer New York), 7104-7111. Scarre, C. 1989, 'Painting by resonance', Nature, 338 (6214), 382-382. Skoglund, P. 2016, Rock art through time: Scanian rock carvings in the Bronze Age and earliest Iron Age (Oxbow Books). Skoglund, P., Nimura, C., and Bradley, R. 2017, 'Interpretations of footprints in the Bronze Age rock art of South Scandinavia', Proceedings of the Prehistoric Society (83: Cambridge University Press), 289-303. Sofaer, J. 2008, Material identities (John Wiley & Sons). Sognnes, K. 1995, The social context of rock art in Trøndelag, Norway: rock art at a frontier (na). --- 2001, Prehistoric imagery and landscapes: rock art in Stjørdal, Trøndelag, Norway (998: British Archaeological Reports Ltd). --- 2003, 'On shoreline dating of rock art', Acta Archaeologica, 74 (1), 189-209. Sopa, G. L. 1984, 'The Tibetan" Wheel of Life": Iconography and doxography', Journal of the International Association of Buddhist Studies, 125-146. Sund, C. 2010, 'Paleogeografiska förändringar i östra Svealand de senaste 7000 åren', (Stockholms University). Thrane, H. 1999, 'Bronze Age settlement in south Scandinavia: Territoriality and organisation', Experiment and Design, 123-132. Tilley, C. and Bennett, W. 2004, The materiality of stone: explorations in landscape phenomenology (Berg publishers). --- 2008, Body and image: explorations in landscape phenomenology (Left Coast Press). Tilley, C., et al. 2006, Handbook of material culture (Sage). Uhnér, C. 2005, 'Tells and the tell-building tradition in the Carpathian Basin during the Middle Bronze Age', in J. Goldhahn (ed.), Mellan sten och järn (Göteborg: Gotarc Serie C. Arkeologiska Skrifter), 745–753. --- 2012, 'Society and power: political economy in Bronze Age tell-building communities', Enclosed space–open society. Contact and exchange in the context of Bronze Age defensive settlements in Central Europe, 347-364. Ullén, I. 1994, 'The power of case studies. Interpretation of a late-Bronze-Age settlement in central Sweden', Journal of European Archaeology, 2 (2), 249-262. Wahlgren, K., Milstreu, G., and Prøhl, H. 2004, 'Switching images on and off: Rock carving practices and meaning in Bronze Age life-world', in G. Milstreu and H. Prøhl (eds.), Prehistoric pictures as archaeological source (Tanumshede: Tanums Hällristningsmuseum Underslös), 149-165. 72

Walderhaug, E. M. 1998, 'Changing art in a changing society: the hunters' rock-art of western Norway', in C. Chippindale and P. Taçon (eds.), The archaeology of rock-art (Cambridge University Press), 285-301. Walderhaug, O. and Walderhaug, E. M. 1998, 'Weathering of Norwegian rock art - a critical review', Norwegian Archaeological Review, 31 (2), 119-139. Valdez-Tullett, J. 2019, Design and connectivity: the case of Atlantic rock art, eds A. Jones and I.-M. Back Danielsson (Oxford: BAR Publishing). Waller, S. J. 1993, 'Sound reflection as an explanation for the content and context of rock art', Rock Art Research: The Journal of the Australian Rock Art Research Association (AURA), 10 (2), 91-101. Varberg, J. 2007, 'Dawn of a new age: the Late Neolithic as third space', in P. Cornell and F. Fahlander (eds.), Encounters, Materialities, Confrontations: Archaeologies of social space and interaction (Cambridge Scholars Publishing), 58-82. Vastokas, J. M. and Vastokas, R. K. 1973, Sacred art of the Algonkians: A study of the Peterborough petroglyphs (Mansard Press). Watson, A. 2001, 'The sounds of transformation: acoustics, monuments and ritual in the British Neolithic', The archaeology of shamanism, 178-192. Werbart, B. 2001, 'Introduction: Cultural interactions in Europe and the eastern Mediterranean during the Bronze Age (3000-500 BC)', BAR International Series, 985, 1-6. Wessman, A. 2010, 'Hällbilder, landskap och sociala logiker. En komparativ och explorativ studie kring hällristningar och bronsålder i Enköpingsområdet', (Göteborgs universitet). Vestøl, O., et al. 2016, 'NKG2016LU, an improved postglacial land uplift model over the Nordic-Baltic region', (NKG working group of geoid and height systems). Whitehouse, P. L. 2018, 'Glacial isostatic adjustment modelling: historical perspectives, recent advances, and future directions', Earth Surf. Dynam., 6 (2), 401-429. Whitley, D. S. 2001, Handbook of rock art research (Rowman & Littlefield). Winter, L. 2001, 'Cultural Encounters. Symbols from the Mediterranean World in the South Scandinavian Rock Carving Tradition during the Bronze Age', BAR International Series, 985, 9-28. Vogt, D. 2011, Rock carvings in Østfold and Bohuslän, South Scandinavia: an interpretation of political and economical landscapes (Instituttet for Sammenlignende Kulturforskning).

73

List of figures

Figure 1. (a) Photo of an original rock carving panel (Tanum 311); (b) Photo of a renewed rock carving panel (Bottna 43)...... 1 Figure 2. Examples of wheel cross motifs: (a) in Tanum 465; (b) in Tanum 12; (c) in Boglösa 131; and (d) in Tanum 470...... 2 Figure 3. Overview of the location of study areas, Enköping and Tanum municipalities, in the map of Scandinavia...... 3 Figure 4. Elevation and coastal view of (a,c) Tanum and (b,d) Enköping...... 4 Figure 5. Schematic diagram showing the connection among material, material culture and materiality...... 5 Figure 6. Structure and components of a BA ship...... 9 Figure 7. Land uplift velocity of Sweden with the study areas marked...... 13 Figure 8. Schematic diagram showing sea level variations during land uplift...... 14 Figure 9. The historic shoreline of Enköping and Tanum...... 15 Figure 10. The picture of rock carving panel Boglösa 138 in Enköping. The estimated shoreline is denoted by dash lines...... 16 Figure 11. Prehistoric shoreline of Sweden...... 22 Figure 12. Spatial distributions of rock carving panels in Tanum...... 24 Figure 13. Spatial distributions of rock carving panels in Enköping...... 25 Figure 14. Histogram of percentage in rock carving panels of Tanum and Enköping with elevation at 5m-intervals...... 26 Figure 15. Histogram of percentage in rock carving panels of Tanum and Enköping with slope at 3°-intervals...... 26 Figure 16. Radar plot of percentage in rock carving panels of Tanum and Enköping with azimuth at 30°-intervals...... 27 Figure 17. Number of rock carving panels in different parishes of Tanum...... 28 Figure 18. Number of different motifs in rock carvings of Tanum...... 29 Figure 19. Percentage of different motifs with different number of motifs in rock carvings of Tanum...... 29 Figure 20. Spatial distributions of cup mark motifs in rock carvings of Tanum...... 30 Figure 21. Spatial distributions of ship motifs in rock carvings of Tanum...... 31 Figure 22. Spatial distributions of human motifs in rock carvings of Tanum...... 32 Figure 23. Spatial distributions of animal motifs in rock carvings of Tanum...... 33 Figure 24. Spatial distributions of foot soles motifs in rock carvings of Tanum...... 34 Figure 25. The foot soles related to the wheel cross in Tanum 406 panel...... 37 Figure 26. Number of rock carving panels in different parishes of Enköping...... 42 Figure 27. Number of different motifs in rock carvings of Enköping...... 43 Figure 28. Percentage of different motifs with different number of motifs in rock carvings of Enköping...... 43 Figure 29. Spatial distributions of cup mark motifs in rock carvings of Enköping...... 44 Figure 30. Spatial distributions of ship motifs in rock carvings of Enköping...... 45 Figure 31. Spatial distributions of foot soles motifs in rock carvings of Enköping...... 46 Figure 32. Spatial distributions of animal motifs in rock carvings of Enköping...... 47 Figure 33. Spatial distributions of human motifs in rock carvings of Enköping...... 48 Figure 34. The wheel cross related to the human and cup mark in Boglösa 126 panel...... 50 74

Figure 35. Trundholm sun chariot...... 56 Figure 36. The eight stone slabs decorated with carvings from the Kivik grave...... 57 Figure 37. The photo of Florence chariot...... 58 Figure 38. Wismar trumpet (up) and the decorations on it (down)...... 58 Figure 39. Schematic diagram showing how Southern Scandinavia is connected with the outside world...... 59 Figure 40. Overview of the introduction and duration of different variations of the wheel cross in South Scandinavian rock carvings during different periods of Bronze Age. . 60 Figure 41. BA landscape around the Boglösa 138 panel...... 64 Figure 42. BA landscape around the Tanum 311 panel...... 65 Figure 43. Photo of Tanum 311 panel taken in 2006 ...... 66

75

List of tables

Table 1. Example of ship motifs in different periods from rock carvings of Tanum...... 10 Table 2. Example of ship motifs in different periods from rock carvings of Enköping...... 11 Table 3. The computed average shoreline altitude of Enköping and Tanum during different periods...... 15 Table 4. Different variants of the wheel cross in rock carvings of Tanum...... 38 Table 5. Different variants of the wheel cross in rock carvings of Enköping...... 52 Table 6. Comparison of Tanum and Enköping in different aspects...... 62 Table 7. Number of rock carvings record of different layers in FMIS...... 77

76

Appendix 1: Working with FMIS database

This appendix introduces the most important information that can be helpful when working with the FMIS database. i. Introduction The general information with the geographic coordinates of the rock carving panels, e.g. Tanum 121:1, is stored in the point layer, identified with its object_id or raa_number. The general information of rock carving units, e.g. Tanum 121:1(1), is stored in the nil layer. The nil layer contains the type (e.g. motif type) and quantity information of rock carvings but does not have coordinate information. Moreover, the rock carvings are categorized as Hällbild in the nil layer, instead of Hällristning. If there are more than one unit in one panel, each unit will be assigned an attribute referred to as samsatt_id, which is the object_id of the corresponding panel and can be used to link to the point layer. In the FMIS database accessed at 2019-3-11, the item number of different layers is listed in the table below. The area layer can be transformed into a point layer using the function Polygon Centroids. Table 7. Number of rock carvings record of different layers in FMIS. Municipality Point layer Area layer Line layer Nil layer Tanum 1478 panels25 11 panels 0 3641 units26 (1489 panels) Enköping 1822 panels27 25 panels 0 2588 units (1847 panels) ii. Automatic statistics There are many different ways to count the total number of different motifs in each panel. However, due to the deficiency in the original data, it is more convenient to use a programming language. Although programming itself also takes time, it is more accurate and much less errors would be introduced, as also recommended by Malmer (1981: 9). In this study, the VBA (Visual Basic for Applications) with Microsoft Excel is used. The code and result are available for download as an attachment of the PDF version of the thesis at http://www.diva-portal.org/. The main part of the code is printed below.

25 Excluding error logs: Lur 206:1 and Lur 427:1. 26 Excluding error logs: Tanum 457:1,855:1,876:1,1142:1,2188 and Kville 832:1. 27 Excluding error logs: Bred 24:2; Härkeberga 158:1; Litslena 319:1; Nysätra 59:2,59:4; Tillinge 113:3,34:1; Vårfrukyrka 551:1. 77

iii. VBA code

1. Public Sub DataToResult(data As String, results As String) 2. 'Subfunction to compute 3. Worksheets(results).Cells.Clear 'clear all cells 4. Dim N As Integer 'total rows 5. N = Worksheets(data).Range("A1").CurrentRegion.Rows.Count 6. Dim i As Integer 'row in data sheet 7. Dim i2 As Integer 'row in results sheet 8. i2 = 2 9. Dim j2 As Integer 'column in results sheet 10. Dim j2N As Integer 11. j2N = 2 12. Dim IsNum As Boolean 13. Dim Site As String 14. Dim Motif As String 15. Dim MotNum As Integer 16. Dim isNewPanel As Boolean 17. Dim duplicateMotif As Boolean 18. Dim isNewMotif As Boolean 19. 'loop in data sheet 20. For i = 2 To N 21. duplicateMotif = False 22. isNewMotif = False 23. Site = Worksheets(data).Cells(i, 1).Value 24. Site = Split(Site, "(")(0) 25. Motif = Worksheets(data).Cells(i, 3).Value 26. Motif = Right(Motif, Len(Motif) - 10) 27. IsNum = IsNumeric(Worksheets(data).Cells(i, 4)) 28. If IsNum Then 29. MotNum = Worksheets(data).Cells(i, 4).Value 30. Else 31. MotNum = 0 32. End If 33. If i = 2 Then 34. Worksheets(results).Cells(1, 1).Value = "raa_nummer" 35. Worksheets(results).Cells(2, 1).Value = Site 36. Worksheets(results).Cells(1, 2).Value = Motif 37. Worksheets(results).Cells(2, 2).Value = MotNum 38. Else 39. isNewPanel = StrComp(Site, Worksheets(results).Cells(i2, 1).Value) <> 0 40. If isNewPanel Then 41. i2 = i2 + 1 42. Worksheets(results).Cells(i2, 1).Value = Site 43. Else 44. 'the same panel and motif 45. duplicateMotif = StrComp(Worksheets(data).Cells(i, 3).Value, Worksheets(data).Cells (i - 1, 3).Value) = 0 46. End If 47. For j2 = 2 To j2N 48. If StrComp(Motif, Worksheets(results).Cells(1, j2).Value) = 0 Then 49. If duplicateMotif Then 50. 'duplicate motif, compare values and exchange 51. If StrComp(Motif & " 2, smaller value", Worksheets(results).Cells(1, j2 + 1).Val ue) <> 0 Then 52. Worksheets(results).Cells(1, j2 + 1).EntireColumn.Insert 53. Worksheets(results).Cells(1, j2 + 1).Value = Motif & " 2, smaller value" 54. End If 55. If MotNum > Worksheets(results).Cells(i2, j2).Value Then 56. Worksheets(results).Cells(i2, j2 + 1).Value = Worksheets(results).Cells(i2, j2 ).Value 57. Worksheets(results).Cells(i2, j2).Value = MotNum 58. Else 59. Worksheets(results).Cells(i2, j2 + 1).Value = MotNum 60. End If 61. 'mark duplicate motif blue 62. Worksheets(results).Cells(i2, j2 + 1).Interior.Color = RGB(153, 204, 255) 78

63. isNewMotif = False 'this is important 64. Exit For 65. End If 66. Worksheets(results).Cells(i2, j2).Value = MotNum 67. If Not IsNum Then 68. 'mark non-numeric red 69. Worksheets(results).Cells(i2, j2).Interior.Color = RGB(255, 153, 153) 70. End If 71. isNewMotif = False 72. Exit For 73. Else 74. isNewMotif = True 75. End If 76. Next j2 77. If isNewMotif Then 78. j2N = j2N + 1 79. Worksheets(results).Cells(1, j2N).Value = Motif 80. Worksheets(results).Cells(i2, j2N).Value = MotNum 81. If Not IsNum Then 82. 'mark non-numeric red 83. Worksheets(results).Cells(i2, j2N).Interior.Color = RGB(255, 153, 153) 84. End If 85. End If 86. End If 87. Next i 88. 'Summary 89. Worksheets(results).Cells(i2 + 1, 1).Value = "Sum" 90. Worksheets(results).Cells(i2 + 1, 1).Interior.Color = RGB(255, 0, 0) 91. Dim k As Integer 92. For k = 2 To j2N 93. Dim cell1 As String 94. cell1 = Worksheets(results).Cells(2, k).Address(0, 0) 95. Dim cell2 As String 96. cell2 = Worksheets(results).Cells(i2, k).Address(0, 0) 97. Worksheets(results).Cells(i2 + 1, k).Value = "=SUM(" & cell1 & ":" & cell2 & ")" 98. Next k 99. End Sub

79

Appendix 2: Reconstructing historic shoreline

This appendix illustrates how to reconstruct the shoreline of Scandinavia in 1700 BC using the software package QGIS, following the Subchapter 2.1.2. Similar method can be used to reconstruct the shoreline of any place and time. i. Required data  Raster layer of present elevation data, for instance named as RE2000.tif (unit: m)  Raster layer of levelled land uplift velocity28, for instance LUV.tif (unit: mm/year). ii. Steps  Use the Raster Calculator, with expression "RE2000.tif"-("LUV.tif"*3700*3700/24000 +3700*"LUV.tif")/1000, to calculate the elevation in 1700 BC, resulting in a raster layer named RE1700BC.tif.  Use the Raster Calculator, with expression "RE1700BC.tif">0, to generate a mask of the elevation map where the land and sea areas are assigned values of 1 and 0, respectively.  Use the Translate tool to assign Nodata to the region with 0 values.  Use the Polygonize tool to generate the vector layer of the boundary, i.e., the shoreline.

28 The original .dat file provided by NKG working group (Vestøl, O., et al. 2016) uses the WGS 84 CRS, which needs to be converted to SWEREF99 CRS which FMIS data are projected to. The .dat file can be converted to .tif raster layer with QGIS by processing the following steps: (1) add .data file as Delimited Text layer; (2) resave to SWEREF99 CRS; (3) use interpolation tool to generate a raster layer. 80