L. Holmquist, S. Kalmring & C. Hedenstierna-Jonson (eds.),

New Aspects on -age Urbanism, c. 750-1100 AD.

Proceedings of the International Symposium at the Swedish , April 17-20th 2013.

Theses and Papers in B THESES AND PAPERS IN ARCHAEOLOGY B

New Aspects on Viking-age Urbanism, c. 750-1100 AD.

Proceedings of the International Symposium at the , April 17–20th 2013

Lena Holmquist, Sven Kalmring & Charlotte Hedenstierna-Jonson (eds.)

Contents

Introduction : royal site and Christian and the Lena Holmquist, Sven Kalmring & regional perspective, c. 980-1100 Charlotte Hedenstierna-Jonson...... 4 Sten Tesch...... 107 Sigtuna and excavations at the Urmakaren Early northern as special economic and Trädgårdsmästaren sites zones Jonas Ros...... 133 Sven Kalmring...... 7 No Kingdom without a town. Olofs- Spaces and places of the urban settlement of son’s policy for national independence and its materiality Charlotte Hedenstierna-Jonson...... 16 Rune Edberg...... 145

Birka’s defence works and harbour - linking The Schleswig waterfront - a place of major one recently ended and one newly begun significance for the emergence of the town? research project Felix Rösch...... 153 Lena Holmquist...... 28 Starup Østertoft – a landing place and set- Birka in tradition and research tlement with a church from the 11th–12th Björn Ambrosiani...... 40 Centuries Anders Hartvig...... 168 Elites, networks and the Finnish connection in Birka and in Carolingian and Ot- Ingrid Gustin...... 43 tonian times and the Viking raids Sebastian Ristow...... 177 ’s demise in the late and the shift to Schlesvig () – a ninth century Viking Volker Hilberg...... 55 town? From Early Medieval trading place to Ottonian power centre Dries Tys & Barbora Wouters...... 188 Geomagnetic survey at Hedeby Hochburg Andreas Viberg & Sven Kalmring...... 74 What Scandinavians may have seen and experi- enced in Miklagarðr: the urban fabric of Constan- Kosel-ost: a Viking-age settlement in the hin- tinople during the terland of Hedeby Jesper Blid...... 196 Tobias Schade...... 82 and Samarra. Imperial capitals of the Husby and the equestrian graves in Angeln Abbasid and – different chronological stages in the development of a royal adminis- Andrew D. Petersen...... 203 tration? Thorsten Lemm...... 89 New Aspects on Viking-age Urbanism, c. 750-1100

Introduction

Fig. 1. Well attended lecture room at The Swedish History Museum. Conference participants engaged in commited discussion (Photo: Charlotte Hedenstierna-Jonson).

The idea to hold a joint conference comparing the Vi- central to the conference, were addressed. age sites of Birka and Hedeby in a larger context An attempt to relate and compare these two sites arose when a guest researcher was sent from the Centre solely with each other is not at all: as early as in of Baltic and Scandinavian Archaeology in Schleswig to 1926, Sune Lindqvist pointed out the close connec- the Arkeologiska forskningslaboratoriet at tions between the two maritime trading centres in his University, an initiative sponsored by the Alexander article “Hedeby och Birka”, published in the periodical von Humboldt foundation. This act led to a vivid de- Fornvännen. In 1941 Herbert Jankuhn penned an ar- bate about the two iconic early urban sites among the ticle on “Birka und Haithabu” in the Ahnenerbe peri- organisers. An inspiring workshop hosted by Laila Kit- odical Germanien on the sites’ different nature in . zler-Åhfeldt and Riksantikvarieämbetet, entitled “Där Today, scholars perceive Birka and Hedeby as two of förändringen börjar”, was held in February 2013 in the very few proto-urban centres in rural Northern Eu- , where many related topics, which would become rope, both of them nodal points in the long-distance

4 Introduction

Fig. 2. Post-conference exkursion to Gamla . Conference participants climbing the Royal Mounds (Photo: Jana Marikova). trading network of their time. In fact, both sites not The specific aim of the conference was the discus- only have similarities in terms of their extraordinary sion of structural aspects of Viking age urbanism at two size and differing structural logistic details, but also specific in lake Mälaren in Eastern and feature direct links, such as the attempts by at the at the margin between North and the church of - and Anskar († Continental . There is still no simple explanation 865) or certain comes or prefectus vici safeguarding the for Viking age urbanism and rather old models, such as interests on-site. However, the traditional roy- German geographer Walter Christaller’s (1893 –1969) al-administrational and cultic centres of the regions, as central place theory, are still in use. In recent years, im- e.g. for , remained seemingly portant progress has been made regarding questions of unaffected by the establishment of the new and quite structural details, creating a new basis for the inspired alien economic centres. It was only with the Christian discussion: Birka research has focused on the fortifica- high medieval towns of Sigtuna and Schleswig that all tions, the still submerged harbour areas and the general of the functions merged at one place. layout by means of large scale geophysical surveys, while

5 New Aspects on Viking-age Urbanism, c. 750-1100 there are ongoing excavations in Gamla Uppsala due to outs and Central-Site Functions of Contemporaneous the development of the Ostkustbanan. For Sigtuna, the Superregional Centres” on day two introduced the the- insightful Trädgårdsmästaren excavations have recently matic extension. Where possible, pairs of sites were cho- been reported. In Hedeby, large-scale geophysical and sen to test the question of the very location of adminis- metal detector surveys have been carried out and im- tration, cult/religion and trade. In the portant aspects of the large settlement excavations, its session, /Hamwic (Martin Biddle) and Co- harbour and its burial grounds have finally been pub- logne/Aachen (Sebastian Ristow) as well as in the Cen- lished. With Füsing and Husby in Angeln, potential roy- tral and session Rjurikowo Gorodischt- al sites have been located by means of metal detection, sche/Nowgorod (Evgenij N. Nosov) and the Hradschin while Schleswig is currently the focus of a large Volkswa- and Suburbium of (Jana Marikova-Kubkova) gen research project. Finally, with the Swedish-German were presented. After a poster session, the section was excavation at the hill fort of Hochburg next to Hedeby, concluded by the Southern Europe and Middle East a new level of cooperation has been reached. sessions on early medieval (Jesper Blid In order to assure the thematic frame of the confer- Kullberg) and Abbasid Baghdad/Samarra (Andrew Pe- ence colleagues were asked to give papers on predefined tersen) followed by a synthesis of the section (Barba- topics related to the problem. The symposium itself ra Crostini). In the evening, the participants enjoyed was divided into two sections: the first part was a conference dinner at the Swedish History Museum. entirely focused on Birka and Hedeby, their broader The next day a post-conference tour was led in Gamla hinterland and declared successors, and the second part Uppsala (fig. 2), where the current excavations (Lena emphasized developments in the presented regions, Beronius) and the site itself (Ingmar Jansson) were pre- where participants were confronted with examples sented. from the most important centres of the world at that As a cooperation between the Swedish History Mu- time in order to put the results into perspective (fig. 1). seum, Stockholm University and the Centre for Baltic Here, speakers on Anglo-Saxon , the Carolin- and Scandinavian Archaeology, the international sym- gian Realm, Bohemia, the Rus’, the posium was well received and attended by more than and even the Muslim were enlisted. With 100 auditors. We are much obliged to the Swedish His- the examples of Constantinople and Baghdad/Samarra tory Museum for hosting the symposium which was especially, the latter two as major capitals of the then kindly funded by the Letterstedtska föreningen. The known-world were recognised as the utmost important subsequent publication of the conference proceedings thematic opening. was made possible by the support of the Centre of Bal- The symposium started off with a public evening lec- tic and Scandinavian Archaeology. ture on “Urban in Birka and Hedeby” (Björn Ambrosiani & Sven Kalmring), which summed up The editors the history of research of both sites in order to prepare Lena Holmquist, Sven Kalmring & ground for the academic debate to follow. After a warm Charlotte Hedenstierna-Jonson welcome by the Swedish History Museum and an intro- duction to the conference aim (Lena Holmquist), the first day was dedicated to the section “Viking-age Ur- banism at Lake Mälaren and the Schlei Fjord”. It was grouped in thematic pairs and comprised sessions on Birka (Charlotte Hedenstierna-Jonson; Ingrid Gustin) and Hedeby ( Schulze; Volker Hilberg), Gamla Uppsala (Lena Beronius; Alexandra Sanmark) and Füs- ing/Husby (Andres Dobat; Thorsten Lemm) as well as Sigtuna (Sten Tesch; Jonas Ros) and Schleswig (Felix Rösch; Michaela Schimmer). After the synthesis (Fredrik Svanberg), a reception followed in the Guldrummet of the museum at the end of the first conference day. The section “Early Medieval Urbansim. Town Lay-

6 Early northern towns as special economic zones

Early northern towns as special economic zones

Sven Kalmring

One of the major proto-urban centres of the early me- dieval world in was, without a doubt, Hedeby. Hedeby was situated on the land bordering and Continental Europe, connecting the with the Baltic by a portage. Its success story as a trading hub is inseparably connected to the destruc- tion of the emporium (cf. Tummuscheit 2003) by the Danish king Göttrik in AD 808. In order to control and the on-going trade, merchants were subsequently relocated to Hedeby. But whilst an emporium in the area of influence of the West Slavic tribe of the was father to the site in the contemporary historical sourc- es, Hedeby was rather referred to as a portus (Kalmring 2010:42–47). Its continental denomination as Sliaswīk, however, includes the element “wic” deriving from vicus (Laur 1992:575). Only in the report about its de- struction in AD 1066 ( lib. III, cap. 50, schol. 81), and in younger sources (cf. of , c. AD 1167), was it denoted a civitas. Hedeby’s counter- part, Birka, was situated on Björkö in Lake Mälaren, a small island located not only on the borderland between Uppland and Södermanland, but also on the crotch be- tween the Uppland of , Tiunda- land and Fjärdhundraland. Its position was traffic-geo- graphically favourable, situated along the waterway of the Fyrisleden, leading from the Baltic and Södertälje via Fig. 1. Word cloud on technical terms within the urbanism debate (the author; based on wordle.net). Birka to Old Uppsala and Vendel. Birka is referred to as a portus and vicus in the contemporary written sources. Apart from that, it was also denoted as civitas, urbs and was frequented by Friesians ( cap. 20), , even (Mohr 2003:98–101; discussion cf. Kal- , , Sambians and “” (Adam of mring in print a). The passed-down descriptions of both Bremen lib. I, cap. 60) and according to Rimbert (cap. of the sites characterised by these terms are indeed quite 19) “contained many rich merchants and a large amount alike: Hedeby has been described as a Danish port, where of and money”. “merchants from all parts [of the world] congregated” and the attendance of and from Dor- Lost in Terminology – The Urbanism Debate estad, Hamburg and Bremen led to “an abundance of goods [that] converged there” (Rimbert cap. 24) . Birka From this brief contrasting juxtaposition it becomes

7 New Aspects on Viking-age Urbanism, c. 750-1100 immediately emanate that both Hedeby and Birka had 2005). Following the ideas of K. Polanyi (1963), these many features in common. They possessed a specifically kinds of sites are also subsumed by the slightly out of favourable traffic-geographic situation and were situated fashion term ports of trade, which stresses their maritime in borderlands and could attract many merchants from component. Ports of trade denote sites with some sort of sometimes far-distant places, resulting in considerable administrational organisation of trade by an authority economic prosperity. Even their denotations in the con- in a society yet without markets. Moreover, these sites temporary written sources partly overlap. But are the tend to be located on the border of different political latter able to give further insights into the closer nature and economic entities without directly belonging to a of these sites? The way in which the Continental sourc- certain domain. As ports in a borderline situation, they es label these kinds of trading places has long been the form gateways in different political or economic forms matter of academic discussion and is immediately con- of organisation and are hubs for the exchange of goods nected to the vast field of “early stages of the European between their hinterlands and abroad (cf. Steuer 2003). town” and urbanisation in Central and Northern Eu- In 1972, the international conference “Vor- und rope in general. Apart from the process of the Christian- Frühformen der europäischen Stadt im Mittelalter” isation of the North, there is hardly any other topic that in Göttingen aimed, amongst other things, to clarify has been debated as intensively and controversy within conceptual problems, the definition of the object of re- the field of Early and Viking age search “town” and the terminological possibilities where studies. Hence, it is not a coincidence that there still the term did not yet seem appropriate (Jankuhn et al. seems to be no generally accepted, straightforward in- 1973:8–9). In this context, G. Dilcher (1973) elabo- terpretation (fig. 1). Current publications try instead to rates on the legal-historical aspects of the town-notion avoid contemporary Latin terms as they “are not used of the high medieval town, particularly characterised by consistently by present scholars. Furthermore, their cur- the four main elements municipal peace, municipal free- rent use tends to »create« specific categories of towns; dom, town law and civic constitution. Based on these ju- these then become the focus of attempts to characterise ridical indicators he concludes that there is no clear con- emporia, civitates, portus or wiks as specific and different stitutional exclusion of the town and its inhabitants from types of urban settlement. However, this modern typol- the rural countryside when it comes to the early form of ogy of town-types was not necessarily the same as early town of the 9th–11th centuries (Dilcher 1973:24–27). medieval perceptions of towns, and the differences be- In the same volume W. Schlesinger (1973: 262) intro- tween the many types of settlements with these urban duces the terms “Vorform” (proto-town) and “Frühform” features or functions are not clear-cut in the contem- (early town) of towns. At the same time he clarifies that porary evidence” (Verhaeghe 2005:260–261). However, these perceptions, however, should not be understood in despite this appraisal, the sometimes elaborate scholarly an evolutionary sense; all towns did not necessarily have attempts to work with this body of historically docu- to run through this process before becoming one of the mented terms continues in the present (cf. Kleingärtner “fully developed” of the high and , and 2013: 177–191 tab. 10 fig 30). even if abandoned earlier, they could have already been proper towns. Despite a pluralistic theoretical approach Seehandelsplätze, Proto-Towns, Emporia and within the Swedish medeltidsstaden-project (the medieval Viking Age Towns town-project), the universal burden of the historical defi- nition is obvious: According to H. Andersson (1979), a The notionSeehandelsplätze was developed by H. site’s nature and position could either be defined by its Jankuhn (1958), who recognised the maritime trading functions within a spatial network, the topography of the places along the coasts of the North Sea and the Baltic as settlement’s internal layout, or else by its inner legal-ad- a group of their own, offset from the early medieval vici ministrative structure i.e. town privileges, municipal law, occurring inland. To him, the former were particularly town council, mayor and own coat of arms. Depend- characterised by their connection to maritime trade and ing on the particular emphasis, the respective approach involvement in long distance trading networks as well would allow to test if the site was about a “central site”, as their appearance in contemporary written sources, a “population centre” or just a town in a formal meaning in addition to their abandonment in the course of the (Andersson 1979:6–7; emphasis added). and replacement by new towns (cf. Steuer Most frequent is the appliance of the term empori-

8 Early northern towns as special economic zones um as defined by R. Hodges (1982:50–52; cf. Hodges 2000:76–92; 2006:63–71), who dis- tinguishes three types of gateway communities and therewith was the first to classify distinct phases of a process. A common feature of these gateway communities is the placement on or near territorial boundaries in addition to the location near or on coastlines. Type A emporia constitute the earliest category and are about fairs visited for short periods re- spectively, seasonally only. The “classic” type B emporia may be rooted in type A emporia and emerge from AD 725 onwards. They are heavily involved in long-distance trade, fea- ture centralised planning in terms of streets and dwellings in addition to a mercantile community of alien traders and native crafts- men. If not abandoned due to a decline in trade type B emporia could further develop into type C emporia with a function based rather in the regional economy. According to Hodges they are characterised by an adminis- trative function, and a partially commercialised production level. Fig. 2a. Language areas of medieval Schleswig- reflecting the In 1994 J. Callmer claimed that the discussion of the four main ethnicities: Frisian (1), Danish (2), Saxon/German (3), Sla- vonic (4). Note Hedeby’s strategic borderland position at the intersec- field of early urbanism in Northern Europe was left too tions of these areas (after Müller-Wille 2002:fig. 646, with additions). much to and – against a background of a con- stantly growing body of archaeological source material – one should rather try to formulate their arguments on an archaeological basis. In a circum-Baltic comparison, the Baltic there was a “complete breakdown in special- he is able to establish an elaborate phasing for trading ised production” (950-1000 AD). At the very end of places and early urban sites. Callmer starts off with the the 10th century it is replaced by new towns as royal appearance of new non-permanent trading places due foundations after West European models. His consid- to a North Sea impetus (AD 700-750) after an indis- eration ends with a continuous reorganisation and a tinct break with late Roman Iron Age/ detached development on the Southern Baltic coast as conditions. It is followed by a phase of consolidation “a different branch of European urbanisation” (Callmer and a change to permanent or semi-permanent sites; a 1994:80) (AD 1000-1050) and a change in character growth in size and an increasing number of trading plac- towards a denser population and brick buildings (AD es (AD 750-800); the subsequent heyday of early ur- 1050-1100). banisation with a developed hierarchy among the sites, U. Näsmans’ (2000) contribution to the discussion and a closer control and standardisation in craft pro- is not clear terminologically. He applies “central places duction (AD 800-850). He then surmises an impelling of rural character” as opposed to “central places/trad- collapse, where “second-level trading places” lose their ing sites” and “proto-towns or early towns” (Näsman importance or become abandoned due to “political in- 2000:42) as well as “new economic centres, the emporia” stability” (AD 850-900), resulting in the engagement of (Näsman 2000:62) – thus it is a discussion of the prem- local elites on the sites and a reorganisation of trade in ises of the initial urbanisation. The starting point of his terms of strong regulation and control (AD 900-950). reflections is end of the Saxon Wars, when, for the first Even so, the system of the old system collapses and, ac- time, the Carolingian and Danish realms shared a com- cording to Callmer (1994: 72), in considerable parts of mon border resulting in the “rapidly growing political

9 New Aspects on Viking-age Urbanism, c. 750-1100

these towns as alternative royal power bases in order to finance military and political ex- penses. It was only in the 10th century that, against the background of conversation and direct royal rule, a new phase of urbanisation took place. With reference to the work of S. Reyn- olds (1977; 1987; 1992), current research tends to denote early medieval urban settle- ments simply as “towns of their time” or else as “Viking-age towns” (Clarke & Ambrosiani 1991:3–4; Verhulst 2002:91; Verhaeghe 2005:261; Skre 2007:45–46). For our mat- ter, however, it is most interesting that the latter is quite restrictive in its application of the term. D. Skre (2007: 453–455 fig. 1.4) writes: “(...) there are only few sites in Scan- dinavia – Hedeby, Birka, and – that show a significant connection with long-distance trade systems before the . This connection is represented by their large quantities of goods imported from outside Scandinavia. Furthermore, the quan- tity of balances, weights and coins, and ev- Fig. 2b. Medieval administrative provinces of Uppland (inclusive coast- idence of the use of as a currency, is al Roden). Note their pie slice-shaped layout with Birka at the southern much higher at these four sites than at smaller sites. summit of the three regions (after Andersson 2004:fig. 1, with addi- Similarly, these four sites have a broad range of craft- tions). work that made use of imported raw materials, mainly metal casting and glass-bead production.” Thus he con- and economic impact” of the Carolingians and Ottonians cludes: “With what is known now about the specialised in Southern Scandinavia (Näsman 2000:37). While he sites for trade and craft production in Scandinavia at alludes to an increasing number of centres and landing the beginning of the Viking Period therefore, only Ribe, places from the 5th century on as “non-urban centres” Birka, Hedeby and Kaupang can be classified as towns (Näsman 2000: 53), the first proto-towns established in (…).” the were modelled on Western European proto-types (e.g. , or Hamwic). In addition, he differentiates between proto-towns devel- Central Place and Network Theories oping into early towns (Ribe, Hedeby) on the one hand, Another approach to the problem is borrowed from hu- and shore-markets attached to a manor/central-place man geography and deals with the geographical town with roots in the Migration/Merovingian period (Åhus concept and spatial/functional method of approach. with Vä) on the other. To Näsman (2000:64–68), early The application of the term central places is still root- towns were developed both as royal strongholds and as ed in German geographer W. Christaller’s old model gateways to the advanced monetary economies of West- of central place theory (Christaller 1933; cf. Denecke ern Europe, while the lingering central-place system 1973; Schenk 2010), known by its characteristic Thies- was still deeply rooted in non-monetary barter-markets. sen polygons. Less sophisticated attempts to reconstruct With Ribe and Hedeby, an urbanised economy was first the delimitation of major Viking age towns in terms of implemented in south-western , subsequently their catchment areas have been undertaken with the turning it into the main of Viking Denmark. help of site catchment analysis (Randsborg 1980:77 According to his reading, the kings actively supported fig. 20). However central places or central place areas are

10 Early northern towns as special economic zones settlement agglomerations with a bundle of functions places in Northern Europe, East and Southeast – power, protection, resources and craft, trade and cult Asia in order to gain new input for a bewildering and – situated in a hierarchical structured space. As a term, paradigmatic discussion (Mogren 2013). Amongst other central place is mostly applied to sites from the Roman things, M. Mogren (2013) demands the deconstruction Iron Age to the Migration Period (e.g. Fabech 1999; of the “central place” concept, highlighting the function, Larsson & Hårdh 2002; Ludowici et al. 2010), which agency and communication instead of morphology and are often characterised by remarkable chronological structure as essential parameters for “urban” and empha- continuity. Central places have an impact on the sur- sises the aspects of initial self-organisation of traders- and rounding settlements, which can be divided into places craftsmen-communities. Most importantly, however, he with fewer functions and thus only with super-regional, takes up the model of orthogenetic and heterogenetic sites local or regional significance. However, when all func- as a conceptual pair (cf. Miksic 2000). Here orthogenetic tional criteria are fulfilled, H. Steuer (2007:882) likes sites are associated with stability and heterogenetic sites to recognise central places as in fact correlating to “early are related to change and entrepreneurship. While the towns”. first are characteristically “(…) placed in areas of sur- Due to the constantly increasing number of known plus agricultural production, are always in some respect trading sites around the Baltic, S. Sindbæk tries to dis- central, are populated by a civil, religious and military cern between local markets and trading places of urban bureaucracy, might have an aristocratic/royal presence, character engaged in long-distance trade by applying and are distinguished by some form of monumentality complex network theory. Relating to P.M. Hohenberg (…)” the latter are often – as at Hedeby and Birka, too & L.H. Lee’s network urbanism (Hohenberg & Lee – “liminally placed, between ecological zones, ethnic or 1996:165) as a point of origin, he recognises trading political regions, or different transport zones (…) and places as traffic junctions in a network. By means of a often lack monumentality (…)”, but have “production quantitative analysis of imported goods and crafts in re- and trade of commodities (…) [as] most distinguishing lation to the size of miscellaneous excavations, he con- feature” (Mogren 2013:80). cludes that with Ribe, Kaupang, Birka, Åhus, , Re- ric and Hedeby there was only a small distinct group of nodal points heavily involved in long-distant trade (Sind- The Viking World and the Otherness of the bæk 2007a). Such a small-world trading network with Town scale-free architecture – a few nodal points (or hubs) with Via the example of Birka, M. Bäck (1997:151) quite many external links along the major sailing routes and an accurately observed: “One of the most fundamental abundance of depended local markets closely associated preconditions for people from a »civilised« Europe ven- with the core and a few links only – was a very fragile turing into the heart of the barbarous north is that there one in the face of crisis and/or threat against one of its must have been a place where they could safely transact indispensable hubs (Sindbæk 2007b). their business, a kind of free zone. (…) The face that The idea was developed further towards “maritime Birka being situated on an island also helped to create a network urbanism” and lead to the entrepôt-project, neutral space, segregated from the mainland (…).” This which analysed global maritime exchange patterns in appraisal can be understood as an adjusted re-reading the by comparing contemporary de- of Polanyi (1963) and his historico-cultural concept of velopments in Northern Europe, the Mediterranean, “ports of trade“ as a neutrality device offering security Eastern Africa, and South-. In the con- requirements for oversea trade at the dawn of a fully text of urbanism, however, entrepôt is a rather vague developed system of market economy. Yet in order to term generally describing “a port, , or other centre be able to describe their generally presupposed distinct to which goods are brought for import and export and otherness, and assumed exceptional position within the for collection and distribution” (OED 2010:586). Con- surrounding society, nothing less than the attempt to sequently, in the catalogue to the exhibition present- a brief outline of the keystones of the Viking world is ing the project’s output, it is the term emporium that is indispensable. applied by the editors (Sindbæk & Trakadas 2014). A In order to make the argument presented below, the comparable project in is also studying the caus- most important facts shall be briefly summarised (Kal- es of early urbanisation by focusing on contemporary mring forthcoming). The Viking age was still very much

11 New Aspects on Viking-age Urbanism, c. 750-1100 rooted in Iron Age societal structures: It was about a kings granted market peace within the early northern rural agrarian society with communities living in single towns and sometimes even personal royal protection on farmsteads of hamlets subdivided into distinct regional sea routes. Anyone found outside their native legal sphere groups. Traditional centres of power and cult with roots would lack his/her legal security. Yet while long-distance in the younger Roman Iron Age and Migration periods merchants were much desired at these exterritorial open were to legitimise royal power. The gathering of a crowd spaces for trade, their presence was also rigidly limited at religious festivities was utilised for holding juridical as- to the margins of the early northern towns themselves semblies and local fairs. Fairs and cult practice were like- (cf. Kalmring in print b: fig. 3); as aliens, they were lord- wise performed at periodically occupied royal manors. less men constituting a social threat for the surrounding Against this background, then, early northern towns, traditional society and thus were hardly allowed to roam with vast and multi-ethnic populations and year-round the . And even if the first urban communities engagement in long-distance market-trade, constituted relied heavily on their hinterlands in terms of supply an utterly alien element only able to exist at the fringes of with consumables, their archaeological footprint in the the traditional society; royal manors have not been found material culture of the towns’ environs is rather limited. within the boundaries of the towns, only in the immedi- However, there seem to be the first faint indications for ate vicinity to the urban communities. And though un- the diffusion of innovations. Certainly not by chance, doubtfully traffic-topographically in a favourable posi- this goes well along with insights from modern cultural tion, their geographic situation in the borderland was, at geography that innovations initially tend to spread via the same time, administratively detached from the power super-regional centres and only then diffuse hierarchi- structures of the surrounding countryside (fig. 2a–b); the cally from there (Kulke 2009:258). latter is not least emphasised by the indications for an own -assembly. In their initial phases early northern towns also started out as quite ordinary beach markets Special Economic Zones of Underdeveloped and as such, as everyday institutions of the ordinary soci- Systems etal structure. Yet due to their geographic situation and/ At first glance, the evidence from Northern Europe or success in trade and commerce, they attracted the at- seems to accord with Verhaeges’ assessment of the Car- tention of local potentates who started to intervene and olinginan attitude towards their towns: “(…) we still deliberately promote a few select sites, most probably by know too little about royal perceptions of, and attitude the concession of a divergent and unique legal status. towards, towns. Indirect intervention seems clear – no- Archaeologically this special jurisdiction is both in- tably by influencing the economy with measures related dicated by symbolic boundary ditches around trading to coinage, tolls and markets – but that is about as far as sites and – most probably and at a later stage – harbour we can go at present. (…) In fact, it is also quite possible chains, too. Beyond that, from both the contemporary that the Carolinians had only a limited or perhaps even Frankish World and from Anglo-Saxon England we are parasitic interest in the urban phenomenon, given their informed that an extraction of royal naval duties and predilection for the rural world that was the basis of trading tariffs, as well as of the existence of distinctive their power and income, and where they chose to build harbour administration, was maintained by a row of their palaces.” (Verhaeghe 2005: 284; emphasis added). officials. In the North, analogous legal regulations can If this is true then one has to ask what exactly the “par- only be deduced from younger law codes and , but asitic interest” might have consisted of. For Viking age with the mention of the comes respectively prefectus vici, Scandinavia, one obvious and very true answer is that the accordant royal officials were handed down, too. the kings were perpetually obliged to be lavish and gen- As representatives of the king, they were likely to have erous in order to pledge their retainers to themselves and been responsible for jurisdiction, defense and, most no- to establish and up political alliances. From Skaldic tably, for the collection of corresponding custom duties. and sagas we learn that gift-giving was one of the Scandinavian potentates did not only adapt to custom major social principles in societies. Gift-giv- duties themselves, but also tried to arrange commercial ing accumulated social prestige and created bonds be- agreements on their own and sometimes even actively tween people, but, if violated, it could just as well lead to campaigned against Viking pirates threatening the trade a negation of social communication, feud and even war routes. As a perquisite to attract traders from abroad, (Vestergaard 1991). In his bright article on wealth and

12 Early northern towns as special economic zones gift-bestowal A. Gurevich (1968) emphasised that the the offer both financial incentives, such as lower acceptance of any gift would have required the recipient and tariffs, and subsidies as well as the substantial dereg- to bring an action for the debt towards the donor – as ulation of the legal and administrative framework or the e.g. to put oneself under the king’s power – if not com- provision of legal privileges” (Ahrens & Meyer-Baudeck pensated by a return gift. In such a way, retainers were 1995:88; emphasis added). obliged to allegiance and loyalty towards the wealthy Moreover “(…) in addition to the direct economic and generous lord. Moreover, a successful and gener- benefits that can be derived from boosting trade (…) ous king was equated with a guarantee for peace, good SEZs can also carry indirect economic benefits, which harvest and abundance in his domain. Since objects in turn can drive local innovation. Indeed, while at- were believed to absorb some of the owner’s personal tracting »content-rich« FDI [foreign direct investment] qualities, the perpetual accumulation of wealth was not and stimulating trade, SEZs tend to favor the acqui- intended for enrichment, but rather a means of both sition of international knowledge, which is crucial to preserving and increasing the own success and good the development of innovation capabilities” (White fortune. Thus even gifts had this magical connotation 2011:184). Herewith the “new” innovations are defined and carried a part of the donor’s success and good for- as the “implementation of new or improved products, tune, which was desirable to share in (Gurevich 1968). processes, marketing, or organizational methods in Consequently, the Scandinavian kings in their role as business practices and workplace organization”, which, lavish gift-givers were in permanent need of objects of as radical innovations, can be about “the creation of value, which could be either obtained by warfare and »new-to-the-world« innovations”, but as incremental looting, or else by controlling trade demanding dues for innovations can also be merely about “applying existing the warrant of market peace, collecting trade tariffs by global that are new to the local context or their reeves and – in the case of Hedeby – even by mint bringing small improvements to existing technologies” and seigniorage. (White 2011:183). Via the diffusion of innovations, the But is there more to the matter? In light of the gathered development of shielded special economic zones likely evidence it becomes apparent that early northern towns helped upgrade the local conditions in the respective and their interdependence on the surrounding conser- hinterland and thus, as a catalyst for the promotion vative social order at that time had many characteristics of innovation, were located particularly in developing in common with what in modern economics has termed (White 2010:305, 329). Interestingly enough a “special economic zone”. Special economic zones are vis-a-vis Viking age urban settlements in the , universally defined as “(…) demarked geographic areas Hodges says: “Indeed, the Vikings, when as conquerors contained within a country’s national boundaries where and kings rather than raiders they took command of the rules of business are different from those that prevail regions, conspicuously adopted the portus model, cre- in the national . The differential rules principal- ating solar central-places [= class C sites, after AD 850] ly deal with investment conditions, such as Ipswich, , Stamford, Lincoln and . and , taxation, and the regulatory environment; Their leadership led to an infusion of economic activity whereby the zone is given a business environment that in previously underdeveloped regions like East and is intended to be more liberal from a policy perspective ” (Hodges 2006: 70; emphasis added). and more effective from an administrative perspective than that of the national territory” (Farole 2011:23). In their article “Special Economic Zones: Shortcut or Conclusion Roundabout Way towards ?” J. Ahrens and The author does not want to introduce yet another term A. Meyer-Baudeck’s definition is in many aspects quite to an already bewildering academic discussion. It seems similar, yet they name some important additional fac- unlikely that one more suggestion, one more denota- ets, too: “SEZs are geographically or functionally limited tion for the first sparks of the urban phenomenon in the parts of an economy in which rules and other institutions North, manifested by four recognised sites only (sic!), concerning the production and the distribution of goods will bring us any further in the future debate of this and services differ from those in the rest of the economy. major field of research in Scandinavian Early Medieval These special institutions are realized in order to promote archaeology. It may even be raised the heretical question and favour economic activities in a specific area. Generally, if our main concern is still a question of terminology.

13 New Aspects on Viking-age Urbanism, c. 750-1100

Regardless of their academic denotation, the issue of what these places were about remains as primary as the problem. But what apart from the mere “trade and pro- duction” as commonplace is at the very core of that dis- continuous phenomenon of the dawn of high medieval towns that these very special sites constitute? It might be about time to shift the emphasis from a mere what? to a more elaborate why? Obviously that answer can- not be achieved by focusing on the early urban centres only. It requires nothing less than a sincere holistic ap- proach: A thorough examination of the societal frame- work of the circumjacent ordinary world, mirrored in a super-regional comparison with Anglo-Saxon England, Continental Europe as well as Byzantium and lastly – since we are dealing with protohistoric archaeology! – an inclusion of the sparse but not less important infor- mation from chronicles and legal texts. What sounds like a vast investment in research is indeed quite achiev- able since most of the jigsaw pieces have partly been known for quite some time within the scientific debate. It is on us to place them out into one coherent picture. Such an attempt has been coarsely sketched above and will be elaborated on in detail elsewhere. In light of the outlined evidence, the implementation of the concept of “special economic zones” seeks to contribute to help- ing understand the specific nature and function of these very distinct sites for the traditional Viking age society at the earliest stage of urbanism in Northern Europe.

Corresponding Author Sven Kalmring Centre for Baltic and Scandinavian Archaeology (ZBSA), Schloss Gottorf Schleswig, [email protected]

14 Early northern towns as special economic zones

in Anwendung auf historische Quellen. In: H. References Jankuhn W. Schlesinger & H. Steuer (eds.) Vor- und Frühformen der europäischen Stadt im Mit telalter I. Abhandlungen der Akademie der Wissen Primary Sources schaften in Göttingen. Philolgisch-historische Adam of Bremen. History of the of Ham Klasse. Dritte Folge. No. 83. Göttingen. 33 – 55. burg-Bremen. Transl. F.J. Tschan 2002. New York. Dilcher, G. 1973. Rechtshistorische Aspekte des Stadt Helmold of Bosau. The chronicle of the Slavs. Transl. begriffs. In: H. Jankuhn W. Schlesinger & H. F.J. Tschan 1966. New York. Steuer (eds.) Vor- und Frühformen der europäischen Rimbert. Vita Ansgarii. Anskar. The Apostle of the Stadt im Mittelalter I. Abhandlungen der Aka North, 801–865. Transl. Ch.H. Robinson 1921. demie der Wissenschaften in Göttingen. Philolo gisch-historische Klasse. Dritte Folge. No. 83. Göt tingen. 12– 32. Fabech, Ch. 1999. Centrality in Sites and Landscapes. Secondary Sources In: Ch. Fabech (ed.) Settlement and Landscape. Ahrens, J. & Meyer-Baudeck, A. 1995. Special Econom ic Zones: Shortcut or Roundabout Way Towards Capitalism? Interecomomics. Review of Europen Economic Policy 1995 (30):2. 87 – 95. Andersson, H. 1979. Urbaniseringsprocessen i det medeltida Sverige. En forskningsöversikt. Medeltids staden. Den tidiga urbaniseringsprocessens kon sekvenser för nutida planering 7. Göteborg. Andersson, Th. 2004: Svethiudh, det svenska rikets kär na. Namn och bygd 2004 (92). 5 – 18. Bäck, M. 1997. No Island is a Society. Regional and Interregional Interaction in Central Sweden during the Viking Age. In: H. Andersson, P. Carelli & L. Ersgård (eds.) Visions of the Past. Trends and Tradi tions in Swedish Medieval Archaeology. Lund Stud ies in Medieval Archaeology 19 = Riksantik varieämbetet arkeologiska undersökningar skrifter 24. Stockholm. 129 – 161. Callmer, J. 1994. Urbanisation in Scandinavia and the Baltic c. AD 700 – 1100: Trading Places, Centres and Early Urban Sites. In: B. Ambrosiani & H. Clarke (eds.) Developments around the Baltic and North Sea in the Viking Age. Birka Studies 3 = The Twelfth Viking Congress. Stockholm. 50 – 90. Christaller, W. 1933. Die zentralen Orte in Süddeutsch land eine ökonomisch-geographische Un tersuchung über die Gesetzmäßigkeit der Verbreitung und En twicklung der Siedlungen mit städtischen Funk tionen. . Clarke, H. & Ambrosiani, B. 1991. Towns in the Vi king Age. & London. Denecke, D. 1973. Der geographische Stadtbegriff und die räumlich-funktionale Betrachtungs weise bei Siedlungstypen mit zentraler Bedeutung

15 New Aspects on Viking-age Urbanism, c. 750-1100

Spaces and places of the urban settlement of Birka

Charlotte Hedenstierna-Jonson

Birka holds a special place in research on Viking Age A Place for Remembrance and Valitdation – society. Over the centuries, the archaeological struc- Graves and Burial Grounds tures have been studied, excavated and published by a number of different researchers and institutions. The The burial grounds are perhaps the most striking feature diverse picture that this multitude of research has pro- of Birka. Vast areas of grave mounds encompass the set- vided adds both breadth and complexity to the discus- tlement area and the . Today these grounds are sion, but also makes a comprehensive outline of the site divided by arable fields, but they may well have consti- rather difficult. The monumentality of the place and tuted a continuous unit surrounding most of the Birka the wide-ranging research history tend to influence the structure. Out of the 7 defined burial grounds, the so- theories we bring forth and effect the questions that we called Hemlanden is by far the most extensive, cover- pose. Birka, and the island of Björkö, has great sym- ing over 17 hectares of land and containing more than bolic value and its extensive archaeological material is 1600 presumed graves (Gräslund 1980:4–6). Hjalmar often used as a key to Viking Age material culture. But Stolpe’s comprehensive excavations of over 1100 of the in most respects, Birka was and is unique. Using it as a identified ca 3000 graves in the late 19th century formed template for the development of the Viking Age in gen- the starting point for modern Birka research. The results eral generates a discrepancy between the interpretations were published in the 1940s by Holger Arbman and made and the actual archaeological evidence. the publications remain the most influential source for To avoid deceptive generalisations, a possible ap- further research as well as provide a catalogue of Viking proach is to return to the archaeological evidence and Age material culture in general (Arbman 1940–43). what it actually conveys, to work from the restricted to Since Stolpe’s excavations, a minor number of graves the complex rather than establishing a complex inter- have been examined, mainly as a of the studies pretation and then placing the factual context into the of other structures such as the town and fortification model. In such an approach the material culture and ramparts and various buildings in the settlement and archaeological contexts could be considered as parallel garrison contexts (e.g. Holmquist 1990; Fennö Muyin- and interactive components that together create and de- go 2000; Hedenstierna-Jonson 2000; Bergström 2013). velop a complex site and reflect its changes over time in Other inquests have dealt directly with graves (Arrhe- both purpose and character. In a previous study I have nius 1990; Holmquist Olausson 2004; Andersson et al tried to interpret the development of Birka using four 2000; Andersson et al 2015). These additional graves are different burials from various time periods as a starting few in numbers but in several cases display interesting point (Hedenstierna-Jonson 2016 in press). The aim of and slightly deviant practices. The development of new this paper is to address the spaces and places that to- archaeological techniques and analyses also contribute gether form the Birka structure, how they changed over to new knowledge but have also conferred the high qual- time and how they together provide the prerequisites ity of Stolpe’s old excavations and documentation, and for the first urban settlement in the region. strengthened the overall interpretation of a complex set of burial traditions that included both cremations and

16 Spaces and places of the urban settlement of Birka

Fig. 1. Map over Björkö and Birka by 17th century cartographer Carl Gripenhielm. Drawn in the 1680’s (ATA). various forms of inhumations such as coffin burials be deceiving. A brief run-through of the grave gifts and chamber-graves. from the cremations indicates a wealth and variety in and quality of objects that reflect the situation in the inhumations quite well. The fact that the cremations Grave, Practice and Individual are situated in grave fields not visibly connected to the A more detailed study of burial practice provides an family’s in itself constitutes a break in tradition. important source for gaining insight into the religious Over time, the burial-practice will become even more beliefs of the people it represents (cf. Gräslund 2010). distant from the regional customs. Although present Out of the excavated graves, a little more than half during the first phase, 10th century Birka sees an in- were cremations. At first glance they correspond to crease in various forms of inhumations. The increase is the predominant burial practice in the surrounding noticeable among inhumations with and without cof- hinterland and may be thought to reflect the region- fins and has been discussed in terms of possible Chris- al traditions. The cremations in Birka are, however, tian influences (cf. Gräslund 1980:83ff and there cited the graves least investigated, and first glances may references). Another type of inhumations is the remark-

17 New Aspects on Viking-age Urbanism, c. 750-1100 able chamber-graves. Even if there are a limited number from the 9th and possibly even the 8th century, the majority of the chamber-graves date from the 10th century. Representing a particular type of site rather than a geographical region for the chamber-graves signify a group of people di- rectly linked to the Viking Age urban settlements (Hedenstierna-Jonson 2015b:144). How are we to understand the many variet- ies of burial practice present in the Birka burial grounds? “Any culture’s responses to mortality are intimately bound up with their world-views and perspectives on life” (Price 2012:82). The inhabitants of Birka were in the middle of a reli- gious transition that also involved great changes in society. A gradual change in burial practice with an increase in inhumations may signify re- ligious stances. The well-known mission of Ans- gar in the first half of the coincides roughly in time with the appearance of inhuma- tions and the first Christian objects in burials (cf. Gräslund 1980; Trotzig 1996; 2001; 2004). This should not, however, be regarded as evidence that the burials in question were Christian, but rather that new influences were having an im- pact on older traditions. The Christian mission is known to us by the texts of ’s successor Rimbert, the archaeological material clearly shows a strong pres- Fig. 2. ’s note book from 1874 with drawings and notes of grave Bj 199 (ATA). ence of pagan beliefs. A relatively high number of graves contain pagan amulets, often in the shape of Thorsham- mer rings. They are mainly connected to the cremation on diet in the Baltic area using stable isotopes (Lidén burials where they have been placed in the grave, not & Nelson 1994). Additional isotopes studies have been as a grave gift but rather as a part of the burial rite (cf. conducted through the years, compiling information on Ström 1970; Andersson 2005:71ff). the mobility pattern and diet of a number of individuals The skeletal material is really the key to the inhab- (Linderholm et al. 2008; Kjellström 2012; Hedenstier- itants of Birka. Most efforts to determine gender, age na-Jonson et al. 2013). The results show that there were and possible place of origin have been based on archae- variations in dietary patterns and possible place of ori- ological context, i.e. grave gifts and burial practice. Even gin between several of the individuals and that they all though parts of the skeletal material have undergone differ from the animals judged to be local to the island. osteological assessment, the results have rarely been re- Adding the results of the isotope analysis to the mate- lated to the archaeological context of the grave (for an rial culture and social practices that can be interpreted overview of earlier studies see Kjellström 2012:71 and from the archaeological remains, it is obvious that the ref. cited therein). In a few particular cases there have population in Birka was heterogeneous and came from been osteological analyses of the skeleton directly linked a widespread geographical region. to the excavation of a grave (e.g. Boije 2000; During In a study targeting health, age and mobility in the 2000; 2004). But new research involving osteology and late Viking Age and early Middle Ages, Anna Kjell- biochemistry has given insights into formerly unex- ström has compared individuals from Birka with indi- plored areas such as health, diet, mobility and relations. viduals from the rural hinterland of the Lake Mälaren Individuals from Birka were included in a pioneer study

18 Spaces and places of the urban settlement of Birka region (Kjellström 2005; 2012; Ahlin Sundman & suggested that this space contained the graves of people Kjellström 2013). Based on osteological data and from a farmstead or small outside urban Birka supported by isotope analyses, Kjellström shows (Hallström 1913:10). that there were gender-related differences regard- Burial grounds are and were spaces organised by the ing age and health where women showed a higher living, fulfilling the need for both remembrance and level of physical stress both in comparison to men continuity. Links to earlier generations are manifested in Birka and in comparison to women in rural con- here and thus provide the descendants with a sense of texts. There were also differences in age distribution belonging and validation. In a sense, the inhabitants of between Birka and the hinterland. The results show Birka had lost their original link to their families’ farms that there were variations between urban and rural and history and the novel contexts within which these life and that Birka in terms of health issues already people had lived created a need for new practices linked shows the problems of the medieval town. to death and burial. The diversity of Birka’s graves and burial grounds could therefore be a reflection of the dis- continuity and change the inhabitants experienced in Spatial and Chronological Distribution this new setting. Apart from studying the individual burial and the interred, the spatial distribution of the graves with- in and between the burial grounds may provide A Space for Production, Exchange and new perspectives on how the Birka structure devel- Contacts – The Settlement and Harbour oped over time and how we are to interpret these The sheer size of the settlement and the density of its changes. Studies, first and foremost by Anne-Sofie population are two of the features that make Birka Gräslund (1980), have shown differences between unique in its region. This was the first time that this par- the burial grounds regarding burial-practices. The ticular area was exposed to the riches and difficulties of outer regions of the largest grave field, Hemlanden, a specialised centre for trade and crafts. One of the chal- and the burial grounds, South of Borg, only con- lenges was to organise the town structure in itself, in- tain cremations. The inhumations, including the cluding the provision of food for a growing population. chamber-graves, were situated in close proximity to Another task was to provide the infrastructure needed the settlement-area, inside or just outside the town for the influx of goods and people. and in the topographically distinguished Birka’s harbour has generally been described as part burial ground North of Borg. These spaces are spa- of the town rather than at the heart of its undertakings. tially closely linked to the urban settlement and also But with the location of the settlement, on a small island contribute with the majority of the burials diver- in the middle of the waterways, and the character of the gent in practice from the regional tradition. They activities with an emphasis on exchange and contacts, can be considered to represent the urban population the presence of an extensive seaport was a prerequisite. and its changes over time. There are chronological As a construction the harbour had to enable ships to differences in the spatial distribution, even though land, and loading and unloading to take place. It should these are less prominent. Many of the graves are yet also offer a safe haven for visiting ships, protecting them to be dated, especially the cremations. The datable from bad weather and hostile attacks. The town was cen- early inhumations were concentrated to two main tred on the harbour serving its needs of repairs, man- areas, the part of Hemlanden denominated 1B power and supplies. in Gräslund’s study and North of Borg, 2A & B (Gräslund 1980). However, these areas were in con- tinuous use and also contain some of the youngest The harbour-zone burials (e.g. Hedenstierna-Jonson 2015a). It is im- Though previously touched upon by several scholars, it is portant to bear in that the burial grounds on only during recent years that a fuller image of the struc- the island of Björkö need not exclusively represent ture and activities of the harbour-zone has begun to evolve the population of the urban settlement. The burial based on recent marine archaeological excavations and ground of Kärrbacka (6) encompasses some of the surveys (Lindström et al. 2012) and the now published youngest inhumations on Björkö and already in the stratigraphy of the settlement (Ambrosiani 2013a). Un- beginning of the 20th century Gustaf Hallström der water archaeological mapping and excavations have 19 New Aspects on Viking-age Urbanism, c. 750-1100

resulted in a greater understanding of the function and shoreline, foundations to jetties and the structure of a magnitude of the harbour construction (Lindström et breakwater. The oldest excavated jetty has been dated al. 2012). The numerous remains of construction tim- to the mid 8th century, making it contemporary to ber and stone cists along the waterfront that have been the oldest town plots. The breakwater was construct- identified provide an important new insight into the ed in the late 8th century enabling the town-plots to activities of the harbour. The results so far reveal an en- spread in the direction of the waterfront. Excavations hancement and enlargement of the harbour structures also showed that the settlement in the harbour-zone during the first part of the 10th century. The harbour was subject to more or less constant construction work takes on a new scale and can no longer be defined as a during the first part of the 9th century (Ambrosiani et number of jetties. The maritime excavations have also al. 1973; Ambrosiani 2013a: 9). provided research with a completely new set of finds Additional excavations on land within the har- that are absent from the dry contexts on land – that of bour-zone were initiated during 2015 (Holmquist objects in organic materials such as wood and textiles. & Kalmring 2015). By connecting the pediments of By excavating maritime remains on land, Björn the jetties resting on land with the waterfront, the Ambrosiani has been able to identify the Viking Age aim is to get a further understanding of the develop-

Fig. 3. Areal photo of Birka’s hillfort (Photo: Lars Bergström KVHAA 1964).

20 Spaces and places of the urban settlement of Birka ment and activities of maritime Birka. As with all other were other varieties of buildings in other contemporary spaces of the town, the constructions go through several spaces of Birka. As Lena Holmquist has shown with phases, changing in scale and possibly even in function her excavations of and close to the town rampart, this over time. This “wet zone” of the settlement was also area held more traditional long houses at an early stage continuously affected by land elevation, intentional ex- (Holmquist Olausson 1993). Interpreted as the living pansion of the harbour and garbage disposal setting its space of the higher levels of society, these houses were mark on both stratigraphy and archaeological material. erected, elevated above the smells and bustle of the town, on terraces following the stretch of the town rampart. With time this particular area transformed into a place The settlement for metal-working with a sunken-floor , hearths While the harbour provided a space for exchange and and pit forge (Holmquist 1993; Holmquist Olausson contacts, the settlement constituted the main area for & Fennö Muyingo 1995; Söderberg 1996). Although production. Generally known as Svarta jorden (the excavations so far only reveal a minor part of the entire Black Earth), the settlement covers approximately 6 area, the results show that there was a differentiation hectares of land out of which less than 1% has been ex- within the structure with varying types of houses and cavated. Even less has been documented and published. activities concentrated to specific places. The dynamics During later years, archaeological prospection using of the settlement also included continuous changes over radars and magnetometers have been implemented on time with periods of expansion and condensation. the entire settlement area, but the results are being pro- cessed and have not yet been made accessible (Trinks et al. 2014). The overall structure of the settlement there- Structures of Restriction and Control – fore remains more or less unknown. Hjalmar Stolpe The conducted excavations in the settlement area during the Even in the first stages of planning and organising the late 19th century. Though the finds are preserved very urban layout of Birka, the rudimentary structures of little is known about the contexts. The results were nev- demarcation and fortification were included. The dif- er published and the documentation was either lost or ferent parts of the fortifications, recognised early by is non-existent (cf. Ambrosiani 2013b:10). Hallström (1925, unpublished report), have been exca- More than a century later, Björn Ambrosiani initiat- vated, problematized and published by Lena Holmquist ed new excavations in the Black Earth aiming “to study (Holmquist Olausson 2001; 2002a; 2002b; - an entire block with one or more plots in the original stierna-Jonson et al. 2013). Holmquist has shown that town plan” (Ambrosiani 2013a:15). The results show beginning in the mid 8th century, the settlement was that within this particular space, the plots of the town surrounded by a town rampart. Initially it was of minor changed over time and that the original plots from the dimensions but with a wooden super structure – a pali- foundation of the town in the mid 8th century were laid sade or wooden wall. Even some of the more martial el- out in a fan-shaped pattern with the boundary ditches ements of the defence system were initiated in this first at right to the shore (Ambrosiani 2013a:210f). phase. The town rampart enclosed the settlement area The block was situated close to the harbour-zone and while the hillfort presided in solitude on the hilltop of was affected by land uplift and shore displacement. Of Borg. Remains in the waters outside indicate structures, particular interest is the building-technique, as all of possibly pile-barricades at this stage (cf. Lindström et al. the houses had roof-bearing wall structures indicating 2012:31 and ref. cited therein). a new era in construction and setting the town-houses apart from the other excavated houses in Birka. Ambro- siani emphasizes that this makes Birka – and Hedeby – On the Inside different from the contemporary structures of Kaupang The of the settlement can be interpreted in and . This new type of house can also be found different ways; the first being to protect and control in the hinterland, but possibly the knowledge developed the wealth that the activities in the town generated. in Birka and was then spread (Ambrosiani 2013a:217). This martial aspect was the focus of the research proj- In contrast to the planned plots and “new-type” ar- ect Strongholds and Fortifications within which the chitecture of the block close to the harbour-zone, there different parts of the defence system were excavated.

21 New Aspects on Viking-age Urbanism, c. 750-1100

Aiming at resolving the constructions and dating of ring introduces the term special economic zones, used the fortification structures, the excavations provided in modern economics, to the archaeological discourse, new knowledge of the advanced military structure that in an attempt to understand the specific nature and presided on Birka. Even though there was a martial ele- function of these developing urban contexts (Kalmring ment from the start, the first phases of the fortifications in this volume; cf. Farole 2011:23). Within the special were weak in comparison to what was to come. During economic zones, the rules for trade were different from the late 9th – early 10th century the town rampart was those outside and the expected archaeological evidence enlarged and strengthened and stretched out to reach would then be demarcations in the form of town ram- and connect to the hillfort (Arbman 1933; Holmquist parts and structures of control such as a hillfort. Olausson 1993:chapter 6; Wåhlander 1998; Trinks et In this context mention must be made of the so al 2014). The hillfort was erected in the second half of called Bjärköarätt, the first known legislation regulating the 8th century and crowned with a or wood- life and trade in the early medieval towns in Scandina- en construction. Built according to traditional hillfort via. A possible link between Birka and the Bjärköarätt technique of earlier periods, it was a solitaire in its re- has been long debated. From an archaeological point gion (Fennö Muyingo 1998; Hedenstierna-Jonson et of view, Birka is a possible or even likely candidate al. 2013:294ff). Activities within the hillfort have been (Skre 2008:88; cf. Wessén 1923), a conclusion that has difficult to prove. There are a number of burials dating found support among legal historians (cf. Ambrosiani from mainly the 9th century, but no traces of cultur- 2013b:28). The question of trade regulations is at the al layers or house constructions. This first phase of the heart of the urbanisation process, distinguishing sites of hillfort was burnt down in the late 8th – early 9th cen- special character and enabling activities outside the or- tury only to be rebuilt in a larger fashion. Just outside dinary societal structure. Again, the importance of the the hillfort and overlooking the waterfront, the war- harbour-zone can be stressed, as a space for interaction rior’s garrison was established in the mid 10th century. and agreements that were not covered by ordinary prac- The site included a great assembly building much like a tices. Similar to the particular regulations connected to hall, a smithy, for water and a number of smaller the thing, the harbour-zone including ships, jetties and buildings of different functions. The find material is vast market-place constituted a space where special peace and contains weaponry, personal belongings and lux- (frid) was required (Larsson 2012:163; cf. Brink 2008 ury items. The fortification structures of the garrison, and there cited references; cf. Skre 2008:85f). and possibly the hillfort, came to a dramatic end with a devastating battle and the burning down of the hall. This final battle seems to have taken place in the late On the Outside – the Hinterland 10th century. The different phases of the fortifications The relation between Birka and its surrounding region correlate to the power-politics and the perceived level of constitutes an important key to the instigation, devel- threat in the region. The strengthening of the fortifica- opment and end of the urban settlement. It is gener- tion structures that took place in the beginning of the ally regarded as a reciprocal relation where the town’s 10th century and the establishment of an elite troop of need for supplies from the hinterland was equalled by based in the garrison reflect an increased level the opportunities to acquire sought after goods that the of tension in the region, rather than a distant threat. town provided. Birka has been considered a feature or We know from the contexts of the garrison that it was function firmly established in the region (Ambrosiani attacked and burned – but the identity of the attacker 1998; 2008:98). At the same time Birka has always remains a question. been regarded as a solitaire in its region, only paralleled Apart from the martial features there are also the by similar structures in other regions, such as Hedeby, aspects of controlling the economy and the movement Staraja Ladoga, Truso etc. of people and valuable goods. The demarcations of the The hinterland began on the very island of Björkö town could thus be an indicator of administered trade, with a small village or farmstead situated outside the taxes and possible judicial differences between the in- enclosed settlement and in the vicinity of the village of side and the outside of the town limits. This model fits Björkö by, dating back at least to the well into the discussion on the early Northern towns (Bäck 2009; 2010; 2012). The farmstead, initially con- as presented by Sven Kalmring in this volume. Kalm- sisting of at least one timber-frame house and a sunk-

22 Spaces and places of the urban settlement of Birka en-floor hut was in use from the 8th – 10th century, i.e. the urban settlement and the predominant rural society contemporaneous with the town structures. Fragments was complex and multi-facetted and in order to under- of 11th century rune stones from Björkö by and early stand it properly, the perspective must be extended to Christian graves in the surrounding burial grounds indi- include a much greater area of contact. cate that life in the farmstead continued well beyond the end date of Birka (Holmquist Olausson 1993:65; Bäck 2012:52f; Källström 2012). The Components of an Urban Settlement We can assume that Birka depended on its hinterland Birka has been called many things, covering most struc- for various resources and that this required an organised tures from trading post to Viking Age town. The ter- and centralised form of power. Ambrosiani has defined minology is continuously discussed and often the site three “circles” with which Birka mainly interacted: the is characterised as a structure slightly out of the box: immediate hinterland, the regional hinterland and “in- a centre for crafts and trade rather than a trading post ternational” trade (Ambrosiani 1985:103ff). The level or a proto-town rather than a full scale medieval town. of interaction and interchange between Birka and its The need for a new terminology is highlighted by Sven surroundings was not, however, as obvious as could be Kalmring in this volume. A complementary approach expected. Imported luxury goods particular to Birka to defining the structure itself in terms of town, trad- are rarely found outside for example, even in the im- ing post, emporia etc has been to define the character- mediate hinterland. This is true for the manifest struc- istics of the way of life particular to these special sites. tures of Hovgården on the island of Adelsö, with great This approach has in many ways been dominated by mounds and a royal manor (Rydh 1936). Having long the concept of urbanism and urbanisation (e.g. Callmer been considered a part of the Birka structure represent- 1994:50). In a comprehensive work on urban geography, ing administration and royal power, the material culture Pacione has listed four principal methods used of the manor and the grave goods in the various burials to identify urban places (2009). Though modelled after lacks most of the objects typical for Birka’s extensive and modern contexts and therefore not fully applicable to long-distance network of contacts. The burial practice the sites of the Viking Age, they still conclude the main displayed in the great mound of Skopintull (the only features separating an urban site from a rural; features great mound excavated on the site) conforms to regional such as population density, an economy dominated by traditions rather than the contemporaneous inhuma- non-agricultural work, legal and administrative criteria tions in Birka. Skopintull, however, contains grave goods and functionality (Pacione 2009:19ff; cf. Clarke & Am- similar to that of the wealthiest chamber graves. The brosiani 1993:138). Adding an archaeological perspec- structure at Hovgården proceeds into medieval times tive on urban sites, the division of the settlement into and regains its political importance in the fixed plots and the volume of standardised production, with the construction of the palace Alsnö hus. By then reflected through waste and specialised structures, form Birka is long gone and without any visible remains of the other relevant features (cf. Callmer 1994:51 and there urban settlement. cited references). When the town plots were laid out The discrepancies in material culture and practices and the first phase of the town rampart was constructed in Birka compared to both the immediate and regional in the mid 8th century the conceptual and organisa- surroundings call for a more nuanced approach to the tional transition from a rural community to something question of hinterland. The area of resources and sup- else had already begun. And even though Birka in its plies was perhaps more dispersed than limited to the region was without forerunners, there must have been vicinity of the island. This area of research needs further paragons after which Birka was modelled. The layout study, but a possible approach is to identify sites where of the town was, at least partly, in place already from objects typical for and otherwise restricted to the urban the beginning, and so were the different functions that context of Birka can be found. Examples of such plac- contributed to making Birka work as a structure. es are e.g. the Åland islands (Ahola, Frog & Lucenius Apart from the structural differences between urban 2014) and the Storsjö region in Jämtland, in particular and rural sites, the concept of urban is also a quality re- Frösön and the area round Östersund (Kjellmark 1906; lated to the meaning and activities of urban places and Welinder 2003; Zachrisson 2006; Holm 2015; Lind- to the effect they have on people’s lifestyles (Pacione holm & Ljungkvist 2015). The interrelation between 2009:21). The activities and life style of Birka was funda-

23 New Aspects on Viking-age Urbanism, c. 750-1100 mentally different from that of the surrounding region. 2015c:84f). The impact that this had on trade relations As a general overview, the contexts could be said to rep- as well as political connections is not fully recognised, resent different aspects of these meanings, activities and but could well be a contributing factor behind the de- people of the Viking Age settlement. During their time cline of Birka. of use the burial grounds created a visual and temporal bond to between the town and the new generations of in- habitants, but today it also mediates an image of how the Corresponding Author people of Birka perceived life and their array of religious Charlotte Hedenstierna-Jonson beliefs. Within the town rampart the settlement was a Department of Archaeology and , regulated area of production and exchange but also of the activities linked to everyday life. The harbour constitut- Uppsala, Sweden ed a space for contacts and interaction with the outside [email protected] world while the fortifications give a perception of how the outside world was seen in terms of estimated threat. Although there would have been an incessant need for “security” when transporting and gathering valuable goods in “unpoliced” areas (cf. Polanyi 1963:30), there were also other purposes behind the martial presence on site.

It is important to recognise that Birka changed over time. What can be seen in and deducted from the ar- chaeological contexts of the early phases cannot provide a prolonged interpretation of the later phases. The pre- requisites changed, as did the activities and the people. Birka was in many ways the beginning of a new era with specialisation, high level of organisation and pro- fessionalization in trade, crafts and warfare. It was also the end as it ceased to exist when the societal structure of the Early Middle Ages was established and consoli- dated with landowning aristocracy and as two important new elements of power. Despite various contacts with the Christian world, Birka essentially re- mained pagan. In the meantime Viking Age society was on the verge of Christianity. The complexity and the pre-planning of the struc- ture clearly show that there was an organised initiative and certain level of power behind the development of Birka. Most likely kings and petty kings were instru- mental in the initial phase and Birka was a product of the needs and demands of a certain level of people in the Lake Mälar region (cf. Ambrosiani 1985:105; Skre 2008:87). But the development of the urban settle- ment introduced a societal structure for which the Lake Mälar region was perhaps not fully prepared. The rein- forcement of the fortifications and the establishment of a permanent garrison in the 10th century could be in- terpreted as an increased level of tension between Birka and the surrounding region (cf. Hedenstierna-Jonson

24 Spaces and places of the urban settlement of Birka

Arrhenius, B. 1990. Utgrävning av den östligaste sto References rhögen på gravfältet Ormknös. Raä 111, Björkö, Ahlin Sundman, E. & Kjellström, A. 2013. Signs of Adelsö sn, Uppland. Laborativ Arkeologi 1990 (4). sinusitis in times of in Viking Age-Ear Stockholm. 65 – 80. ly Medieval Sweden. Journal of Archaeological Sci Bergström, L. 2013. Hus och hantverk. Arkeologisk un ence 2013 (40). 4457 – 4465. dersökning av de övre terrasserna i Birkas garnison. Ahola, J., Frog & Luncenius, J. (eds.) 2014. The Viking RAÄ 173, Björkö, Adelsö sn, Uppland, 2001–2004. Age in Åland. Insights into Identity and Remnants of Unpublished report. universitet. Culture. . Boije, M. 2000. Osteologisk analys. In: K. Andersson, Ambrosiani, B. 1985. Specialization and urbanization G. Werthwein & M. Boije, En dubbelbegravning på in the Mälaren valley – a question of maturity. In: Birka. Arkeologisk undersökning och restaurering av S-O Lindquist (ed.) Society and trade in the Baltic en skadad stensättning på Hemlanden, Raä 118, during the Viking Age. Visby. 103 – 112. Björkö 5:1, Adelsö , Ekerö kommun, Upp – 1998. Birka och omlandet. In: P. Bratt (ed.) Forntid i land. Rapport 2000:21. Stockholms länsmuseum. ny dager – arkeologi i Stockholmstrakten. Stockholm. Bilaga 2. 199 – 207. Brink, S. 2008. Law and Society. Polities and legal – 2008. Birka. In: S. Brink & N. Price (eds.) The customs in Viking Scandinavia. In: S. Brink & N. Viking World. London & New York. 94 – 100. Price (eds.) The Viking World. London & New York. – 2013a. Stratigraphy Vol. 1. Excavations in the Black 23 – 31. Earth 1990–1995. Birka studies 9. Stockholm. Bäck, M. 2009. Vikingatida bebyggelse funnen utan – 2013b. Birka 1680–2013. Ett långvarigt forskning för Birkas stadsvall. Fornvännen 2009 (104):4. 262 sprojekt. Stockholm. – 273. Ambrosiani, B., Arrhenius, B., Danielsson, K., Kyhl – 2010. Mellan byn och staden - Birkas och Björkös be berg, O. & Werner, G. 1973. Birka. Svarta jordens byggelse ur nya synvinklar. Amatörarkeologiska hamnområde. Arkeologisk undersökning 1970–1971. undersökningar år 2005-2007. Uppland, Adelsö Riksantikvarieämbetet rapport C1 1973. Stock socken, Björkö by och Grindsbacka, RAÄ 171. holm. Arkeologisk undersökning. Rapport UV Mitt Andersson, G. 2005. With on our Side. The Sym 2010:40. Riksantikvarieämbetet. bolism of the Thor Hammer-ring in Viking Age – 2012. På andra sidan Birka. In: C. Hedenstier Burial Ritual. In: T. Artelius & F. Svanberg (eds.) na-Jonson (ed.) Birka nu. Pågående forskning om Dealing With the Dead. Archaeological Perspectives on världsarvet Birka och Hovgården. The National Prehistoric Scandinavian Burial Ritual. Stockholm. Historical Museum, Stockholm. Studies 22. Stock 54 – 62. holm. 45 – 68. Andersson, G., Näversköld, K.O. & Vedin, E. 2015. Callmer, J. 1994. Urbanization in Scandinavia and the Arkeologisk efterundersökning av grav Bj 750 och c. AD 700–1100: Trading Places, intilliggande ytor, RAÄ 118, Adelsö socken, Uppland. Centres and Early Urban Sites. In: B. Ambrosiani Statens historiska museer. FoU rapport 12. & H. Clarke (eds.) Developments Around the Baltic Andersson, K., Werthwein, G. & Boije, M. 2000. and the North Sea in the Viking Age. Birka Studies 3. En dubbelbegravning på Birka. Arkeologisk un Stockholm. 50 – 90. dersökning och restaurering av en skadad Clarke, H. & Ambrosiani, B. 1993: Towns in the Vi stensättning på Hemlanden, Raä 118, Björkö 5:1, king Age. London. Adelsö socken, Ekerö komun, Uppland. Rapport During, E. 2000. Analys av ett obränt skelettmaterial 2000:21. Stockholms länsmuseum. från utgrävningarna av Birkas borgvall. In: H. Fen Arbman, H. 1933. Rapport rörande undersökningarna nö Muyingo, Borgvallen II. Utvidgad undersökning på Björkö den 12–29 juli 1932. Unpublished exca av Borgvallen och underliggande grav 1997. Arkeolo vation report. Antikvarisk-topografiska arkivet. gisk undersökning Dnr. 20300. 1997. Uppland, Adelsö socken, RAÄ 34. Arkeologiska – 1940–43. Birka I. Die Gräber. Tafeln & Text. Stock forskningslaboratoriet, Stockholms universitet. holm. Bilaga 6.

25 New Aspects on Viking-age Urbanism, c. 750-1100

– 2004. Bestämning av benmaterial från fornlämning archaeology 62. Stockholm. 73 – 91. 116, Björkö, Adelsö sn, Up. In: L. Holmquist – 2016 in press. Med utgångspunkt från människorna. Olausson & C. Augustsson Edstam, Delundersökning Sociala roller och samhälleliga funktioner i Birka av gravfält, fornlämning 116, Björkö, Adelsö sn, med utblick mot Gamla Uppsala. Uppland. Rapport. Arkeologiska forskningslaborato Hedenstierna-Jonson, C., Holmquist, L. & Olausson, riet, Stockholms universitet. M. 2013. The Viking Age Paradox: continuity and Farole, T. 2011. Special Economic Zones in Africa: Com discontinuity of fortifications and defence works in paring Performance and Learning from Global Experi Eastern Scandinavia. In: J. Baker, S. Brookes & A. ences. Washington D.C.: . Reynolds (eds.) Landscapes of Defence in Early Fennö Muyingo, H. 1998. Borgvallen. Arkeologisk Medieval Europe. Turnhout. 285 – 301. undersökning 1996. Uppland, Adelsö socken, Hedenstierna-Jonson, C., Kitzler, L. & Stjerna, N. Björkö, RAÄ 34. Arkeologiska forskningslaboratori 1998. Garnisonen II. Arkeologisk undersökning et, Stockholms universitet. 1998. Rapport. Stockholms universitet. – 2000. Borgvallen II. Utvidgad undersökning av Holm, O. 2015: Trading in Viking-period Scandinavia Borgvallen och underliggande grav 1997. Arkeologisk – A Business Only for a Few? The Jämtland Case. undersökning 1997. Uppland, Adelsö socken, RAÄ Viking and Medieval Scandinavia 2015 (11). 79 – 34. Arkeologiska forskningslaboratoriet, Stockholms 126. universitet. Holmquist, L. 1990. ”Älgmannen” från Birka: en pre Gräslund, A.-S. 1980. The Burial Customs. A study of sentation av en nyligen undersökt krigargrav med the graves on Björkö. Birka. Untersuchungen und människooffer. Fornvännen 1990 (85). 175 – 182. Studien 4. Stockholm. Holmquist, L. & Kalmring, S. 2015. Drakhuvudet och – 2010. Religionsskiftet speglat i gravskicket. In: B. birkas hamn som gick på land. Populär arkeologi Nilsson (ed.) Från hedniskt till kristet. Stockholm. 2015 (4). 20 – 23. 131 – 164. Holmquist Olausson, L. 1993. Aspects on Birka. Investi Hallström, G. 1913. Birka I. Hjalmar Stolpes gra gations and surveys 1976-1989. Diss. Stockholm. fundersökningar. Stockholm. – 2001. Birkas befästningsverk – resultatet från de – 1925. Till riksantikvarien med iakttagelser rörande senaste årens utgrävningar. In: M. Olausson (ed.) befästningar och huslämningar på Björkö samt med Birkas krigare. Stockholm. 9 – 15. förslag till kommande fältarbeten därstädes. Un – 2002a. The fortification of Birka. In: A. Nørgard Jør published report.Antikvarisk-topografiska arkivet. gensen, J. Pind, L. Jørgensen & B. Clausen (eds.) Dnr. 2978/25. Maritime warfare in Northern Europe. , Hedenstierna-Jonson, C. 2000. Borg. Arkeologisk un organisation, logistics and administration 500 BC- dersökning 1998/1999. Uppland, Adelsö socken, 1500 AD. Publications from The National Mu Björkö, Raä 34. Arkeologiska forskningslaboratoriet, seum. Studies in Archaeology & History 6. Copen Stockholms universitet. . 159 – 167. – 2015a. She came from another place. In: M. Hem – 2002b. Patterns of settlement and defence at the Pro Eriksen, U. Pedersen, B. Rundberget, I. Axelsen & to-town of Birka, Lake Mälar, Eastern Sweden. In: J. H. Lundberg (eds.) Viking Worlds. Things, Spaces Jesch (ed.) The Scandinavians from the and Movement. Oxford. 90 – 101. to the tenth century. An ethnographic Perspective. – 2015b. Close Encounters with the Byzantine Border Woodbridge. 153 – 167. -Zones - on the Eastern Connections of the Birka Holmquist Olausson, L. & Fennö Muyingo, H. 1995. . In: O. Minaeva & L. Holmquist (eds.) Rapport från en arkeologisk undersökning vid Birkas Scandinavia and the : Cultural Interactions stadsvall 1995. Arkeologiska Forskningslaboratoriet, with Byzantium and Eastern Europe in the First Stockholms universitet. Millennium AD. Cambridge. 139 – 152. Kjellmark, K., 1906. Ett graffält från den yngre järnål – 2015c. To own and be owned: the warriors of Birka’s dern i Ås i Jämtland. Ymer 1906. Stockholm. garrison. In: A. Klevnäs & C. Hedenstierna-Jonson Kjellström, A. 2005. The Urban Farmer. Osteological (eds.) Own and be owned. Archaeological approaches Analysis of Skeletons From Medieval Sigtuna Interpret to the concept of possession. Stockholm studies in edin a Socioeconomic Perspective. Stockholm.

26 Spaces and places of the urban settlement of Birka

– 2012. Projektet Människor i brytningstid. Skel Stockholms Universitet. Stockholm. ettgravar i Birka och dess nära omland. In: C. Trinks, I., Neubauer, W. & Hinterleitner, A. 2014. Hedenstierna-Jonson (ed.) Birka nu. Pågående forsk First High-resolution GPR and Magnetic Archaeo ning kring världsarvet Birka och Hovgården. Historis logical Prospection at the Viking Age Settlement of ka museet Studies 22. Stockholm. 69 – 80. Birka in Sweden. Archaeological Prospection 2014 Källström, M. 2012. Runstensfynd i Björkö by, Adelsö (21). 185 – 199. sn, Uppland. Rapport. Riksantikvarieämbetet, dnr. Trotzig, G. 1996. Paradisspännen. Ikonografiska 322-04083-2012. reflexioner kring några vikingatida smycken. Iconog Larsson, G. 2012. Birka och båtsymbolen. In: C. raphisk post 1996 (2). Riksantikvarieämbetet. Stock Hedenstierna-Jonson (ed.) Birka nu. Pågående forsk holm. 1 – 13. ning kring världsarvet Birka och Hovgården. His – 2001. Symbolernas kamp i religionsskiftets Birka. In: toriska museet Studies 22. Stockholm. 151 – 172. B. Magnus, C. Orrling, M. Rasch & G. Tegnér Lidén, K. & Nelson, E.D. 1994. Stable carbon (eds.) Vi får tacka Lamm. The Museum of National isotopes as dietary indicator, in the Baltic area. Antiquities, Studies 10. Stockholm. 213 – 218. Fornvännen 1994 (89):1. 13 – 21. – 2004. Trons försvarare i Birka. Fornvännen 2004 Linderholm, A., Hedenstierna-Jonson, C., Svensk, (99):3. 196 – 208. O. & Lidén, K. 2008. Diet and status in Birka: Welinder, S. 2003. Christianity, politics and ethnicity stable isotopes and grave goods compared. Antiquity in Early Medieval Jämtland, Mid Sweden. In: M. 2008 (82). 446 – 461. Carver (ed.) The cross goes north. Processes of con Lindholm, K.-J. & Ljungkvist, J. 2015. The Bear in version in Northern Europe, AD 300–1300. Wood the Grave: Exploitation of Top Predator and bridge. 509 – 530. Herbivore Resources in First Millennium Sweden— Wåhlander, L. 1998. Birkas stadsvall och dess förläng First Trends from a Long-Term Research Project. ning till Borg. CD-uppsatser i laborativ arkeolo European Journal of Archaeology 2016 (19):1. 3 – gi 1997/98:1. Stockholm. 27. Zachrisson, I. 2006. Väskan från Röstahammaren Lindström, J., Olsson, A. & Rönnby, J. 2012. Mariti i Ås och gravfältets etniska tillhörighet. Fornvännen ma Birka. Undersökningar av vikingastadens ham 2006 (101):1. 19 – 28. nområden. In: C. Hedenstierna-Jonson (ed.) Birka nu. Pågående forskning kring världsarvet Birka och Hovgården. Historiska museet Studies 22. Stock holm. 29 – 43. Pacione, M. 2009. Urban geography. A global perspec tive. New York. Polanyi, K. 1963. Ports of Trade in Early Societies. Journal of Economic History 1963 (23). 30 – 45. Price, N. 2012. Wooden Worlds. Individual and col lective in the chamber graves of Birka. In: C. Hedenstierna-Jonson (ed.) Birka nu. Pågående forsk ning kring världsarvet Birka och Hovgården. His toriska museet Studies 22. Stockholm. 81 – 93. Rydh, H. 1936. Förhistoriska undersökningar på Adelsö. Stockholm. Skre, D. 2008. The Development of urbanism in Scan dinavia. In: S. Brink & N. Price (eds.) The Viking World. London: Routledge. 83 – 93. Söderberg, A. 1996. Schmelzkugeln - identifikation av en hantverksprocess. Fyndmaterial från Birka och Sigtuna. CD-uppsatser i Laborativ Arkeologi 95/96:2. Arkeologiska Forskningslaboratoriet,

27 New Aspects on Viking-age Urbanism, c. 750-1100

Birka’s defence works and harbour - linking one recently ended and one newly begun research project

Lena Holmquist

Introduction Throughout Birka’s long research history, covering sev- eral centuries, archaeologists and others have tried to comprehend the reason for the town’s existence, how it was constructed and why it was finally abandoned. It is a curious bird in its surroundings and differs greatly from the neighbouring settlements in the countryside. The town, focussing on trade and craftwork, was set up during the second half of the 8th century, at the same time as similar sites that lie spread out along the coast of the , like nodal points in a transport network between the interior and the coast (Callmer 1994; Sind- bæk 2007). These included and Truso in pres- ent-day , Ribe in Denmark, Staraja Ladoga and Wiskiauten in as well as Hedeby in Germany. These were founded around the same time and seem to have arisen from similar motivations, achieving a life- span of about a couple of centuries. These places gave the elite an opportunity for trade, with the manufacture and production of special products to fund their ambi- tions for power. This was the purpose of their founding, before the time was ripe for the establishment of towns in a medieval sense. Terms such as “proto-towns” or Fig. 1. Map of Birka. (Drawing of F. Sieurin-Lönnqvist). “emporia” have been used to distinguish them from the medieval town (Kalmring, this volume). The components essential for the towns to function mediary link between the towns of were assembled within a well-defined area, surround- and those of the medieval period, wherein ground plans ed by a town rampart in several cases, as in Birka and go back to Roman schemes. In my opinion, they are not Hedeby, for both external and internal control. It was linked to either, but their model, with regard to layout important to be able to distinguish inhabitants from and structure, should be sought in another area, such as visitors. The rampart may also have marked an admin- the great contemporary trade capitals. istrative area where special laws were in force. Trade was Birka has also been considered as a link in a chrono- totally dependant on a functioning commercial system. logical chain surrounding Lake Mälaren starting It has been claimed that proto-towns were the inter- with Helgö, and followed by Sigtuna (cf. Ambrosiani

28 Birka’s defence works and harbour

Fig. 2. Arial view of Birka. The largest cemetery Hemlanden and the town wall in the foreground. The open field in the background contains the site of the town (Photo: Lars Bergström).

1975:63). They all show urban traits but are very differ- of an outcrop of bedrock. Defensive pilings and other ent in character. They overlap one another chronologi- structures including jetties have been found in the water cally while co-existing for some periods. of the harbour below the town (fig. 1). In other words In attempting to understand the process of urbaniza- Birka was totally surrounded by fortifications of vary- tion, researchers have sought after explanatory models ing importance and purpose. The town did not remain but these can lead to an oversimplification. Old Uppsa- static throughout its span of life; instead changes took la (Gamla Uppsala) (cf. Sundqvist & Vikstrand 2013), place over time, both in function and in the extent of for example, should be included in the discussion. the settlement and buildings. Cemeteries of varying sizes, comprising two thou- sand visible graves, lay just outside the town. Beyond Birka the ramparts there existed buildings of a more agrari- Birka was in existence between the years 750 and 975 an character (Holmquist Olausson 1993, Bäck 2010). AD. Its built area consisted mainly of a town settle- Their extent and dating are however not yet determined ment – the Black Earth – surrounded by a town ram- as research into these important areas has only just be- part that is today c. 450 metres long and encompasses gun. almost 12 hectares, but this includes additional land Archaeological excavations in Birka have mainly tak- caused by postglacial rebound. The town rampart was en place in the town area (an estimated 5% has been connected to a hillfort which consisted of a semicircular excavated; cf. Ambrosiani 2013) as well as the cemeter- rampart about 350 metres long surrounding the slopes ies (with c.1100 excavated graves; Gräslund 1980; cf.

29 New Aspects on Viking-age Urbanism, c. 750-1100

Kalmring 2012). On the other hand the defence works the area here was subjected to intense tillage, and any have not been studied to the same extent. To fill that rampart would have been ploughed out at a very early gap in our knowledge a research project was begun at stage. With the aim of finding the continuation of the the Archaeological Research Laboratory, entitled Birka’s rampart at this spot, geophysical and geochemical in- Defence Works and followed-up by a project financed by vestigations were conducted in the open field between the Bank of Sweden Tertian Fund; and defence the rampart and the hillfort (Wåhlander 1998). The works in Central Sweden 400-1100 AD. Excavations most significant result was a sharp drop in extremely continued in various stages from the end of the 1980s to high phosphate values that revealed the rampart’s orig- 2004 (cf. Holmquist Olausson 2001; 2002a; 2002b). inal delineation. An excavation at the end of the existing rampart in the 1930s documented row after row of clinch nails The Town rampart (Arbman 1939; cf. Larsson 2007: 359). The same The town rampart measures 450 m today and consists types of clinch nails were also found during later ex- of seven segments with openings between them, some cavations of the hillfort rampart in the 1980s, which of which may have functioned as entrances during Bir- indicates that the rampart was of a unified construction ka’s main period (fig.2). The northern section of the throughout all of its length. This was originally 1000 rampart leads into a harbour bay, known as Kuggham metres. The results of the excavation of the town ram- (Ingelman-Sundberg 1972; Olsson 2004). The south- part (Holmquist Olausson 1988; 1993:77ff) showed ern section stops abruptly. After Birka’s abandonment that this was a well-planned structure consisting of a

Fig. 3. The hillfort at Birka. (Photo: Jan Norrman).

30 Birka’s defence works and harbour

Fig. 4. Map Map of the Garrison. (Map after Holmquist Olausson & Kitzler Åhfeldt 2002).

double wall of stone with between them. The ram- nö Muyingo & Holmquist Olausson 1995; Söderberg part was built in several phases, and beneath it lay an & Holmquist Olausson 1996). These were high-status older occupation layer which proved that here had been buildings. The oldest finds indicate that the inhabitants a settlement that was older than the town. 14C-datings already had long-range trading contacts before the town show that the earliest rampart was built when the town was established. It must however be added that only one was founded. In other words, during its 200 years’ ex- longhouse on one of the fifteen or so terraces that are istence, Birka had gradually developed its focus, with preserved close to the town rampart, has been excavat- different types of settlement and activities. The answer ed. to the origin of the town is to be sought up against the The town rampart should not be considered to be town rampart where longhouses stood on built-up ter- purely for defence. It most likely served an administra- races, of the same type as those in the contemporary ru- tive function (Holmquist Olausson 2002a; 2002b; now ral surroundings (Holmquist Olausson 1993:90ff.; Fen- latest also Kyhlberg 2012). The rampart created a clearly

31 New Aspects on Viking-age Urbanism, c. 750-1100 marked delineation between the town area and the surrounding countryside. “A city rampart is almost of necessity a symbol of sovereign power, because no government of more than nominal authority will permit the massive mobilization of labour and that rampart-building re- quires, to proceed without its approval” (Tracy 2000:5). The rampart was vital not only to monitor visitors to the town and to maintain peace but also to guarantee safe trade. “The most likely to build ramparts were those that had a strong merchant community, or served a territo- rial ruler´s need to secure his borders” (Tra- cy 2000:4; 71–87). Bjärköarätten, the oldest known municipal law in Sweden, has been con- sidered by some researchers to have originated in Birka (Hagland & Sandnes 1997). For these trading places to function it was also necessary to be open to and tolerant of for- eign cultures and their people, in a way that was unfamiliar to the fiercely ordered society of the Later Iron Age. Studies of later trading towns in the Mediterranean such as and Smyrna, which housed deeply mixed populations, prove Fig 5. GIS-map of the town site (The black Earth). Investigated areas that the free practice of religion was highly beneficial 1871-2012 are marked as yellow trenches. (Kalmring 2012). for trade and exchange and strengthening of contacts (Mansell 2011). put forward. One of these placed it within the Migra- Birka’s hillfort, Borg tion-period hillfort tradition in the Mälar Valley, thus dating it to the 6th century AD. However Birka’s fort The dating of the town rampart and its extension has differs from these in its direct proximity to the town been a matter of much discussion but our excavations proper. The Migration-period hillforts lie separately and clearly show at what point and in what manner the town at a long distance from any settlement. rampart links up with the hillfort. Another theory was that the hillfort was added later, More recently, georadar investigations have shown though not before the 10th century. This idea was based that there might have been an earlier shorter rampart on the graves found partly underlying the rampart. inside that visible today (Ambrosiani 1998; Trinks et. al Thus, the most important issues within our project were 2014). This suggests that the town settlement at some the date of the hillfort, its method of construction and time expanded and that the town rampart was altered whether later expansion phases could be substantiated. and lengthened appropriately. This raises a number of Our excavations showed that the rampart was built in a issues concerning the late 8th century dating of the town well-planned manner resulting in a sturdy construction rampart (Holmquist forthcoming). of a double wall of stone with soil infill (Sw. skalteknik) Birka’s hillfort is one of the few monumental struc- (Holmquist Olausson 1997; Fennö Muyingo 1998; tures from the Viking Age to survive in Sweden. It con- 2000). Up to 20 layers could be documented in its sists of a 350-meter long semicircular rampart delimited make-up, comprising both natural and man-made de- to the south by a sharp cliff which falls directly towards posits. These layers could be grouped into two distinct Lake Mälaren (fig.3). Before the hillfort was excavat- main phases comprising an earlier and a later rampart. ed several different hypotheses concerning its date were

32 Birka’s defence works and harbour

Along the top of both ramparts ran wooden embattle- but no evidence for any was found. Neither were any ments. The evidence for this includes post-holes, a burnt settlement traces found during excavation. Instead an worked wooden (pine) plank and hundreds of clinch additional three graves were discovered (Hedenstierna nails of the same type as previously found in the exten- Jonson 2000). In my opinion the hillfort construction sion of the town-rampart. The structure of this rampart holds no similarities with bishop’s forts, which impor- belongs to some extent to the native tradition as seen in tantly were connected to Christianity in a way that Bir- earlier hillforts but the wooden formations might imply ka never was. influence from e.g. the Slavic hillfort building tradition. Among contemporary north European trading plac- Results of 14C- and TL-dating together with the arte- es only Hedeby has a known hillfort close to its town. facts show that the earlier, lower, rampart was built at However it does not lie in direct contact with the town the same time as the town settlement was founded. It but rather 180 metres to the north. The hillfort in Hede- was burned down sometime in the 9th century and re- by has not been examined in the same intensive way as built to twice the size. The later rampart shows signs of its town, and as was the case with Birka, it has been the repeated burning towards the end of the 10th century, subject of various interpretations (cf. Kalmring in prep.). which might indicate that the fort was still in use after The Church chronicler Adam of Bremen proposed that the town had been abandoned. in the 10th century it may have been a Saxon A highly interesting find was a burial which existed fort under I. It has been suggested instead that before the rampart was built and into which it was in- during that century it was the seat of the “Swedish” corporated (Holmquist Olausson & Götherström 1998; . A third suggestion is that it was the fort where Fennö Muyingo 2000). The grave contained a skele- emperor Otto II stayed when he visited Hedeby in the ton which DNA analyses showed to be male. He was year 975. Lesser-known excavations were carried out on accompanied by a young stallion, 2-3 years old. Only the hillfort rampart in the 1930s and 1950s. The fort lies eighteen of Birka’s c. 1100 graves contain a horse buri- on a 26 metre-high plateau. The plateau is almost entire- al. The horse represents considerable financial value and ly encompassed by a roughly rectangular rampart, 225 manifests the high rank and social status of the buried x 45-70 metres long. It is remarkable that the rampart individual. 14C-dating of a tooth from the horse indi- does not surround the whole of the raised plateau, but cates that the burial took place in the first half of the leaves its northern point outside. Today the rampart is 8th century. The grave is therefore contemporary with 1.5 metres high. The rampart’s rather organic form is less the previously mentioned longhouse which was excavat- obvious in the northern part of the fort, where it is more ed close to the town rampart. This also means that the angular at a section where the ground is also higher. This grave is of the period of the founding of the town. - may indicate a secondary adjustment at some later date thermore, this provides a terminus post quem dating for when the hillfort had a different function. (Kalmring & the foundation of the first rampart. Holmquist 2010; 2012; Holmquist & Kalmring 2012) Another aim of the project was to understand the The hillforts of Birka and Hedeby differ in several function of the hillfort and whether it possibly held ways but also show similarities. They resemble one an- dwellings, storehouses or stables (Hedenstierna Jon- other in that neither can show any evidence for settle- son 2000). It has recently been proposed that a church ment within the ramparts. Both also have graves inside lay inside the hillfort, in line with bishop’s forts on the the enclosed space and just beyond. It seems appropri- Continent (Zachrisson 2012). In the Vita of Ansgar, ate that they should still be interpreted as defensive forts, monk and missionary, mention is made of a church in at least at an early stage, in the manner that Rimbert Birka which was erected by the town prefect, Hergeir, (cap. XIX) describes in his Life of Ansgar. on his land c. 830 AD (Rimbert, cap. XI). It is however debated what was meant by “his land” (cf. Holmquist Olausson 1993:35; Staecker 2009:320ff.). At the end The Garrison of the 19th century Hjalmar Stolpe excavated c.10 sites The Garrison lies on a slope down to the shore of Lake inside the hillfort. He interpreted all sites as unmarked Mälaren and is directly linked to the hillfort. The slope graves, lacking visible structures above ground. In order has been made less steep with the aid of stone-set ter- to see if it was possible to find traces of any settlement races which supported houses and other wooden struc- our project conducted several geophysical investigations tures. The Garrison is tactically placed to block the way

33 New Aspects on Viking-age Urbanism, c. 750-1100 from the shore up to the hillfort. Protected by rocky seats (thwarts). cliffs on either side, it commanded a good view of the Parts of shields and spears were also found along the surrounding waters. The adjoining area was further ramparts, where they may have been hung up or left protected by a wooden palisade which ran from the leaning. The Garrison produced a total of 400 weapons shoreline around the whole of the settlement (Holm- representing all Viking-Age Scandinavian types. A spe- quist-Olausson 2002) (fig.4) Just nearby and to the cial and unusual find was iron strips or lamellae from the north of the Garrison lies one of Björkö’s richest ceme- Eurasian Steppe. Many Byzantine paintings of warrior teries; with a large number of chamber graves, several of saints illustrate how the lamellae were linked together. which have been interpreted as of royal status by Inga Such armour signified great worth and was reserved for Hägg (2002). warriors of the highest rank. The amount of different Four of the five presently visible terraces have been lamellae types on the site suggests that there were sev- excavated fully or partly (Kitzler 1997; Hedenstierna eral sets of this protective wear in the Garrison (Stjerna Jonson et al. 1998; Holmquist Olausson & Kitzler Åh- 2004). This is the only find of such armour from -Vi feldt 2002; Bergström 2013:13 fig. 4). The settlement king-Age Scandinavia. Parts of chainmail have also been was dense and well-organized. Along with buildings, a found and indicate that this was also worn by the Birka wooden-covered waste chute and a water tank were also warriors. discovered. The exact origin of the lamellar armour is a matter for The biggest terrace contained a large building of hall- debate, but the Byzantine area seems most likely. Fre- like character, which was used for various types of cere- quent contact with Steppe nomads at the border areas of monies. It was here that the king met with his warriors. the Byzantine Empire would have familiarized the Byz- The building measured 19 x 9 metres. This does not co- antines with their advanced fighting techniques. Other incide with the usual length of large halls, but the avail- finds of weapons inside the Garrison point in the same able space did not allow a longer building. The aim was direction. Parts of bows of eastern type (Lundström et to build a house that was clearly visible and spacious. al. 2009), including the so-called oriental mount from The slope in the ground had been compensated for by a the case of a bow, as well as arrows, are further evi- build-up of soil for half the length of the building. This dence for this. The soldiers in the Garrison belonged to was a three-aisled longhouse with slightly bowed long an eastern-oriented sphere where possessions signalled sides. It had two entrances in the southern long rampart. symbolic values, and where the warrior belt carrying ori- The northern rampart lay so close to the rock that it was ental mounts dominated. Mounts from such a belt were impossible to have any on that side. The distribu- found in the Garrison. These are normally deposited in tion of the finds indicated subdivisions into differently burials and are the only examples to have come from a functioning rooms (Bengtzon 2001). The eastern sec- Viking-age settlement in Sweden (Jansson 1988; Holm- tion was interpreted as a store for weapons and tools. quist Olausson & Hedenstierna-Jonson 2006, Heden- In the northwestern section of the house a large num- stierna-Jonson 2006b). ber of high-status finds were uncovered, such as sherds A strong connection with cultic activities was indi- from glass vessels, oriental mounts and a Byzantine coin. cated in several ways, including different forms of ritual It is highly likely that this was where the high-seat was depositions in the hall building and other constructions. placed. Sacrificial offerings of spearheads were found in various The finds of Byzantine character are especially inter- places within the Garrison. The spear was the symbol of esting as they were otherwise unusual during the Vi- , the god of warriors (Kitzler Åhfeldt 2000). An- king Age. Besides the already mentioned coin, richly other offering had been deposited in a posthole belong- ornamented gilt bronze mounts from helmets were also ing to one of the roof-bearing posts in the hall building. found (Holmquist Olausson & Petrovski 2006). It contained several different types of objects including A great number of keys and locks, as well as mounts a chape in Jellinge style (Hedenstierna-Jonson 2006a), from chests, were found close to the long ramparts of the weapons, a tiny Thor’s hammer in bone as well as pieces house. In all probability chests belonging to the warriors of ornamented comb-cases. The cases had been sawn in stood here, containing their personal equipment (Beng- two and then burned. Attempts to reconstruct them in- tzon 2001). This might suggest that the warriors carried dicate there were at least 40 comb-cases. This could well them on their ships where they functioned as rowing coincide with the number of warriors in the Garrison.

34 Birka’s defence works and harbour

The hall was built in accordance with the 1000- gions that they also learned how to build defence works. year–old longhouse tradition. It became unfashionable It is well known that warriors from Scandinavia were during the 10th century and was replaced by other included in the of the Byzantine em- building types. The idea of erecting a hall in an archaic peror. Furthermore, Vikings are recorded to have twice style can be interpreted as a conscious striving to hark attacked Constantinople on their warring expeditions back to older traditions and to honour the old gods. (cf. e.g. the Russian Primary Chronicle). The final phase of the Garrison can be dated by stra- Within the scope of our now concluded project, tigraphy and finds to the second half of the 10th cen- the different components of Birka’s fortifications have tury. The various components of Birka’s defence works been examined, with the exception of the structures in were built in several phases. The oldest dating in the the harbour area and its waters. Together with the pile ramparts as well as in the Garrison, reaches back to the blockades which had been previously documented (In- late 8th century. gelman-Sundberg 1972; Olsson 2004), the defences en- The wooden structures were burnt down in the fi- compass the town. Over a series of years (2009–2012) nal phase of the Garrison and this, together with the University College Södertörn and the Maritime Muse- large number of spearheads found, indicates that the um undertook research in the waters of Birka’s harbour site was under attack. Everything points to this coming as well as minor excavations at the water’s edge (Han- from the water side and apparently in two stages (Lind- son et al. 2010; Olsson et al. 2011; Lindström 2012). bom 2009). After this the Garrison was abandoned This produced the remains of more complex structures and never rebuilt. Andreas Nordberg (2000:116) has which are considerably more extensive than the simple suggested that it might have become a sanctified area pile blockades previously known. for hero worship, i.e. a place to which one returned Influenced by findings at Hedeby (cf. Kalmring to experience its glory. The site could have been ded- 2010; 2011), it has now been suggested that there were icated to Odin before the attack or after the fight was jetties, up to 100 metres long, stretching out into the over and won. In any event, respect for Odin was very water; and these would have functioned as the town’s strong in the Garrison. market place (Dagens Nyheter, 2012-08-21). This can- not be proved, not least with regard to the fact that a large part of these structures would have existed where The Harbour/Conclusion there is dry land today. Around the year 1970 an exca- Birka’s fortifications were built during the second half vation was conducted in Birka’s harbour area (Ambro- of the 10th century as a direct result of a political drive siani, Arrhenius et al. 1973) but the investigating trench linked to its function as a trading place, and its net- did not extend down to the water. By excavating along work of contacts. the shoreline (Kalmring 2013) it was hoped to answer For these functions to be upheld the protection of questions regarding the harbour’s construction and stra- defence works was essential. There is strong evidence tigraphy. With this in mind a new project was begun in that these defences were already in place when the April 2015 entitled Birka’s Black Earth Harbour where town was founded, and remained in continuous use the focus was directed to a previously noted caisson lo- throughout its existence. They were constructed to cated on the Viking-Age shoreline. The aim is to exam- meet the fighting techniques and weaponry of the day, ine the continuation of the jetty in the form of wooden when archery and sea battles dominated (cf. Holm- constructions and to establish its date (fig. 5). quist Olausson 2002). Piece by piece, Birka is being investigated and Tactics concentrated on threat, and . In mapped. An overview is slowly beginning to emerge, a broader perspective Birka’s defences should be seen greatly aided by this new knowledge of the defence as part of a deep defence system which included pile works. But we need more information on its harbour, blockades and other structures controlling the inlet to which lay at the heart of all activities. Lake Mälaren (cf. Olausson & Lindström 2003). In Hedeby, the nearest counterpart to Birka, 2200 The town’s fortifications were defended by well- m2 of shoreline was laid dry and excavations revealed a trained and well-equipped warriors who had far-rang- great number of pilings (Schietzel 1984). Recently these ing connections, especially eastwards to Kievan Rus´ results have been digitally analysed by Sven Kalmring and the Byzantine Empire. Perhaps it was in these re- (2010; 2011) who was able to reconstruct a series of

35 New Aspects on Viking-age Urbanism, c. 750-1100 large jetties, c. 40 x 40 metres square, which are inter- preted as piers for the off-loading and selling of trade items. On completion of our project on Birka’s defence works, it seems clear that fuller investigation of the har- bour is perhaps the most pertinent topic for future re- search, both in order to understand how the town was designed and also to complete our picture of its fortifi- cations.

Corresponding Author Lena Holmquist Institutionen för arkeologi och antikens kultur, Stock- holm university Stockholm, Sweden [email protected]

36 riet, Stockholms universitet. Stockholm. References Fennö Muyingo, H. & Holmquist Olausson, L. 1995. Ambrosiani, B. 1975. Neue Ausgrabungen in Birka. Rapport från en arkeologisk undersökning vid Birkas In: H. Jahnkuhn, W. Schlesinger & H. Steuer (eds.) stadsvall 1995. Björkö. Adelsö socken, RAÄ 188. Vor- und Frühformen der europeischen Stadt im Mit Arkeologiska forskningslaboratoriet. Stockholms tellalter 2: Philologisch-Historische Klasse. 3. Folge. universitet. No. 84. Göttingen. 58 – 62. Gräslund, A-S. 1980. The Burial Customs. A study of – 1998. Birkas första vall. Nya fynd visar att Birkas the graves on Björkö. Birka IV. Kungl. Vitterhets His vall tillkom på vikingatida 800-talet. Populär torie och Antikvitets Akademien. Stockholm. Arkeologi 1998 (2):19 – 23. Hagland, J. R. & Sandnes, J. 1997. Bjarkyretten. Ni – 2013. Excavations in the Black Earth 1990-1995. daros eldste bylov. . Stratigraphy. Vol. 1. Part One: The Site and the Shore. Hansson, J., Lindström, J., Olsson, A. & Rönnby, J. Part Two: The Bronze Caster´s Workshop. Birka Stud 2010. Nya undersökningar i Birkas hamn. Populär ies 9. Stockholm. Arkeologi 2010 (3). 12 – 13. Ambrosiani, B., et al. 1973. Birka – Svarta Jordens ham Hedenstierna-Jonson, Ch. 2000. Borg. Arkeologisk un nområde. Arkeologisk undersökning 1970-1971. dersökning 1998/1999. Uppland, Adelsö sn, Björkö, Riksantikvarieämbetet rapport C1. Stockholm. RAÄ 34. Borgar och befästningsverk i Mellansverige Arbman, H. 1939. Birka. Sveriges äldsta handelsstad. 400-1100 e. Kr. Rapport 2. Arkeologiska forsk Från forntid och medeltid 1. Stockholm. ningslaboratoriet. Stockholms universitet. Stock Bengtzon, I. 2001. En funktionsanalys av Garnisons holm. huset på Björkö. Seminar paper. Arkeologiska forsk – 2006a. A group of Viking age chapes reflect ningslaboratoriet. Stockholms universitet. Stock ing the political geography of the time. Journal holm. of Nordic Archaeological 13. Stockholm. 103 Bergström, L. 2013. Hus och hantverk. Arkeologisk un – 112. dersökning av de övre terrasserna i Birkas Garnison, – 2006b. The Birka Warrior – the material culture of a RAÄ 173, Björkö, Adelsö sn, Uppland 2001-2004. martial society. Theses and papers in scientific arche Rapporter från Arkeologiska forskningslaboratoriet ology 8. Diss. Stockholm. 20. Stockholms universitet. Stockholm. Hedenstierna-Jonson, Ch., Kitzler, L. & Stjerna. N. Bäck, M. 2010. Mellan byn och staden – Birkas och 1998. Garnisonen II. Arkeologisk undersökning 1998. Björkös bebyggelse ur nya synvinklar. Amatörarkeologis Uppland, Adelsö socken, Björkö RAÄ 35. Birkas be ka undersökningar 2005-2007. Uppland, Adelsö sock fästning 5. Arkeologiska forskningslaboratoriet. en, Björkö by och Grindsbacka, RAÄ 171. UV- Stockholms universitet. Stockholm. Mitt. Rapport 2010:40. Stockholm. Holmquist Olausson, L. 1988. Preliminär redogörelse Callmer, J. 1994. Urbanisation in Scandinavia and the för 1987 års undersökningar i Birkas stadsvall. Lab Baltic region c. AD 700-1100. Trading Places, Cen orativ arkeologi 1988 (3). 57 – 63. tres and Early Urban Sites. In: B. Ambrosiani & H. – 1993. Aspects on Birka. Investigations and surveys Clarke (eds.). Development around the Baltic and 1976-1989. Theses and papers in Archeology B:3. North Sea in Viking Age. Birka Studies 3 = Twelfth Diss. Stockholm. Viking Congress Stockholm. 50 – 90. – 1997. Birkas Borg efter avslutad undersökning. In: Crumlin-Pedersen, Ole & Schietzel, K. 1989. Havnen A. Åkerlund (ed.) Till Gunborg – arkeologiska sam i Hedeby. Skalk 1980 (3). 4 – 10. tal. SAR 33. Stockholm. 399 – 405. Fennö Muyingo, H. 1998. Borgvallen, Arkeologisk – 2001. Birkas befästningsverk – resultat från de senas undersökning 1996. Uppland. Adelsö socken, Björkö, te årens utgrävningar. In: M. Olausson (ed.) Birkas RAÄ 34. Birkas befästning 2. Arkeologiska forskn krigare. 11 artiklar kring Birkas befästningsverk och ingslaboratoriet. Stockholm. dess krigare. Borgar och befästningsverk i Mellans – 2000. Borgvallen II. Utvidgad undersökning av verige 400-1100 e. Kr. Rapport 5. Stockholm. 9 – Borgvalen och underliggande grav 1997. Arkeolo 15. gisk undersökning. Uppland, Adelsö sn, RAÄ 34. Bir – 2002a. Patterns of settlement and defence at the Pto- kas befästning 3. Arkeologiska forskningslaborato town of Birka, Lake Mälar. In: J. Jesch, (ed.) The

37 New Aspects on Viking-age Urbanism, c. 750-1100

Scandinavians from the Vendel Period to the tenth Straraja Ladoga - Novgorod - Kiev. Handel und Han century. An ethnographic perspective. . delsverbindungen im südlichen und östlichen Ostseer 153– 167. aum während des frühen Mittelalters = Bericht der – 2002b. The fortifications of Birka. Interaction be Römisch-Germnischen Kommision 1988 (69). . tween land and sea. In: A. Nørgård Jørgensen, J. 564 – 647. Pind & B. Clausen (eds.) Maritime warfare in Kalmring, S. 2010. Der Hafen von Haithabu. Die Aus Northern Europe. Technology, orgnistion, logistics and grabungen in Haithabu 14. Neumünster. administration 500 BC - 1500 AD. Publications of – 2011. The Harbour of Hedeby. In: S. Sigmundsson the National Museum. Studies in Archaeology and (ed.) Viking Settlement & Viking Society. Papers from History 6. . 159 – 167. the Proceedings of the sixteenth Viking Congress. Holmquist Olausson, L & Götherström, A. 1998. Reykjavik. 245 – 259. Sex identification of the skeleton in a new cham – 2012. The Birka Proto-Town GIS. – A Source for ber-grave from Birka. Laborativ Arkeologi 1998 Comprehensive Studies of Björkö. Fornvännen 2012 (10–11). 105 – 108. (107). 253 – 265. Holmquist Olausson, L & Kitzler Åhfeldt, L. 2002. – 2013. Etwas Altes, etwas Neues, und etwas Blaues. Krigarnas hus. Arkeologisk undersökning av ett hallhus Die fehlende Bindeglied zu Birkas Svarta Jorden i Birkas Garnison, RAÄ 35, Björkö, Adelsö sn, Up Hafen. Jahresbericht Zentrum für Baltische und Skan pland 1998-2000. Borgar och befästningsverk i dinavische Archäologie 2013. 49. Mellansverige 400-1100 e. Kr. Rapport 4. Arkeolo – in prep. Hedeby – Hochburg von Haithabu – Theo giska forskningslaboratoriet. Stockholms universitet. ries, of Research and Dating. Offa. Stockholm. Kalmring, S. & Holmquist, L. 2010. Die Hochburg Holmquist Olausson, L. & Hedenstierna-Jonson, Ch. von Hedeby – Hypothese – Forschungsstand und 2006. The Oriental Mounts from Birka´s Garrison. Perspektiven. Jahresbericht Zentrum für Baltische und Anexpression of warrior rank and status. Antikvariskt Skandinavische Archäologie 2010. 80. Arkiv 81. Stockholm. – 2012. Deutsch-schwedische Ausgrabung auf der Holmquist Olausson, L. & Petrovski, S. 2007. Curious Hochburg von Hedeby. Jahresbericht Zentrum für Birds – Two helmet (?) Mounts with a Christian Baltische und Skandinavische Archäologie 2012. 36 – Motif from Birka´s Garrison. In: U. Fransson, M. 37. Svedin, S. Bergerbrant & F. Androshchuk (eds) Kitzler, L. 1997. Rapport från utgrävningen av Garni Cultural interaction between east and west. Archaeolo sonen på Björkö 1997. Arkeologisk undersökning gy, artefacts and human contacts in northern Europe. Stockholms , Ekerö kommun, Adelsö socken, Björkö, Stockholm Studies in Archaeology 44. Stockholm. Fornlämningsnummer 35. Garnisonen. Birkas befäst 231 – 237. ning 4. Arkeologiska forskningslaboratoriet. Stock Holmquist, L. & Kalmring, S. 2012. Stadslika handel holms universitet. Stockholm. splatser med borgar. Populär Arkeologi 2012 (3). 4 – Kitzler Åhfeldt, L. 2000. Odenssymbolik i Birkas Gar 8. nison. Fornvännen 2000 (95). 13 – 21. Hägg, I. 2002. Aussagen der Textilfunde zu den ge Kitzler Åhfeldt, L. & Holmquist Olausson, L. 2002. sellschaflichen und wirtschaftlichen Verhältnissen Krigarnas hus. Arkeologisk undersökning av ett hallhus frühstädischer Zentren in Nordeuropa – die i Birkas Garnison. Borgar och befästningsverk i Mellan Beispiele Haithabu und. Birka. In: K. Brandt, M. sverige 400-1100 e. Kr. Rapport 4. Arkeologiska Müller-Wille & Chr. Radtke (eds.) Haithabu und forskningslaboratoriet. Stockholms universitet. die frühe Stadtentwicklung in nördlichen Europa, Kyhlberg, O. 2012. Den långa järnåldern. Sociala strat Schriften des Archäologischen Landesmuseums. egier, normer, traditioner. OPIA 54. Uppsala. Band 8. Neumünster. 181 – 218. Larsson, G. 2007. Ship and Society: maritime archeolo Ingelman Sundberg, C. 1972. Undervattensarkeolo gy in Late Iron Age Sweden. 37. Uppsala gisk undersökning utanför Birka. Fornvännen 1972 Lindbom, P. 2009. The assult on Helgö and Birka and (67) Stockholm. 127 – 135. the end of the Iron Age. Situne Dei 2009. 83 – 101. Jansson, I. 1988. Wikingerzeitlichen orientalischer Lundström, F., Hedenstierna-Jonson, Ch. & Holm Import in Skandinavien. In: Oldenburg - Wolin - quist Olausson, L. 2009. Eastern Archery in Birka´s

38 Birka’s defence works and harbour

garrison. In: L. Holmquist Olausson, & M. Olauss Culture of the North 2. . 309 – 329. on (eds.) The Martial Society. Aspects of warriors, forti Söderling, F. Vikingar byggde en stad på vattnet. Da fications, and social change in Scandinavia. Theses gens Nyheter, 2012-08-21. and papers in Archaeology B:11. Stockholm. 105– Tracy, J.C. 2000. Introduction. To wall or not to wall 116. : Evidence from medieval Germany. In: J.C. Tracy Lundström, J. 2012. Maritima Birka – staden från (ed.) City walls. The urban in global perspe sjösidan. Ledungen 2012 (4). 10 – 12. tive. Cambridge. 71 – 87. Mansell, P. 2010. . Splendour in the Mediterra Trinks, I. & Larsson, L-I. 2007. Undersökning utan ut nean. London. grävning. Kartläggning av Birka. Populär Arkeologi Nordberg, A. 2003. Krigarna i sal. Dödsföreställn 2007 (3). 14 – 15. ingar och krigarkult i fornnordisk religion. Stockhol Trinks, I., Neubauer, W. & Hinterleitner, A. 2014. msuniversitet. Diss. Stockholm. First High-resolution GPR and Magnetic Archeo Olausson, M & Lindström, J. 2003. Fållnäs. Arkeolo logical Prospection at the Viking Age settlement of giska undersökningar. Del 1. En ränna, RAÄ 600, ett Birka in Sweden. Archaeological Prospection 2014 gravfält 680:2 samt en fornborg, Skansberget RAÄ (21). 185 – 199. 721 vid Fållnäs gård samt en pålspärr, RAÄ 488, Wåhlander, L. 1998. Birkas stadsvall och dess förlängn Stäket, Sorunda sn, Södermanland 2000-2002. Bor ing fram till Borg. Seminar paper. Stockholm Uni gar och befästningsverk i Mellansverige 400-1100 e. versity. Stockholm. Kr. 3. Arkeologiska forskningslaboratoriet. Stock Zachrisson, T. 2011. The Archaeology of Rimbert: the holms universitet. Stockholm churches of Hergeir and Gautbert and Borg in Bir Olsson, A. 2004. Sondering av vattenområden vid Bir ka. In: S. Sigmundsson (ed.) Viking Settlement ka. Uppland. Ekerö kommun, Björkö socken, Björkö & Viking Society. Papers from the Proceedings of the 1:5. Statens maritima museer. Teknisk rapport sixteenth Viking Congress. Reykjavik. 469 – 493. 2004. Arbetshandling. Stockholm. Olsson, A., Lindström, J. & Rönnby, J. 2011. Birka under vattnet. Populär Arkeologi 2011 (4). 26 – 25. Rimbert. Anskar, The Apostle of the North, 801-865, translated from the Vita Ansgari by Bishop Rimbert his fellow missionary and sucessor. Transl. C.H. Rob inson (1921). London. Russian Primary Chronicle: The Russian Primary Chro nicle: Laurentian Text. Transl. and ed. S.H. Cross & O.P. Sherbowitz-Wetzor (1953). Cambridge, MA. Schietzel, K. 1984. Hafenlagen von Haithabu. In: H. Jahnkuhn, K. Schietzel & H. Reichstein (eds.) Archäologische und naturwissenschaftlche Untersu chungen im deutschen Küstengebiet vom 5. Jahrhun dert v. Chr. bis zum 11. Jahrhundert n. Chr. Handel splätze des frühen und hohen Mittelalters. Acta Hu maniora. Wenheim. 184 – 191. Sindbæck, S.M. 2007. Networks and nodal points. The emergence of towns in Early Viking Age Scan dinavia. Antiquity 2007 (81). 119 – 132. Stjerna, N. 2004. En stäppnomadisk rustning från Birka. Fornvännen 2004 (99). 27 – 32. Staecker, J. 2009. The 9th-century Christian mission to the North. In: A. Englert & A. Trakdas (eds.) Wulfstan´s Voyage. The Baltic Sea region in the early Viking Age as seen from the shipboard. Maritime

39 New Aspects on Viking-age Urbanism, c. 750-1100

Birka in tradition and research

Björn Ambrosiani

In a recent overview I have attempted to incorporate date of Birka as Viking age. Archaeological excavations in Birka’s research history into a more general historical the 1870s and since 1969 have shown that Birka and the context (Ambrosiani 2013a). But it is necessary to delve Mälaren area, as Sweden’s earliest urban settlement and its deeper in order to better understand underlying policies hinterland, hold exceptional potential for research which of Birka research. In the following I will be discussing leads to deeper understand of the Viking age of Scandina- Birka as a local concept in early times and Birka’s role as via. a setting for developing innovative methods in Swedish and international archaeological research. Birka as a center of archaeological innovation Birka must have played an exceptional and unique role in Birka as a living tradition contemporary Viking age society. Also in modern time An interesting problem is how long Birka lived on as a Birka has been a place where significant development concept and when it became a part of history. Today, of archaeological methods has taken place. Hjalmar there is a general consensus among researchers that the Stolpe’s understand of Birka’s Black Earth as a cultural settlement at Birka must have ended during the layer quickly became a classic concept in understanding AD. The royal estate, Alsnö Hovgården, which was con- ancient monuments and remains (Almgren 1904). The temporary with that period, continued however to func- detailed techniques he used for investigating settlement tion as such up until the 1500s, although seldom as a remains and graves as well as his methodical documenta- royal residence after the early 1300s. The construction of tion, using scaled drawings on graph paper, became the the large brick and stone palace, Alsnö hus, in the 1270s standard for archaeology in Sweden and internationally is evidence to the fact that it still at that time was the real (Lindberg 1996; Erikson 2015). center of the Svear in the Mälar area. Investigations in Birka’s Black Earth during the 1900s Still in the mid-1300s Bjärköarätten continued to ex- were similarly innovative in archaeological methods. Me- ist as the legislation for Swedish towns with town priv- chanical dry-sieving combined with wet-sieving made it ileges. It is therefore reasonable to claim that Birka, to- possible to collect significantly greater amounts of highly gether with the royal estate, Alsnö Hovgården, existed as fragmented materials (Ambrosiani et al. 1973; Ambro- a concept during the 1300s. siani 1992:73). This led to the need of computerized doc- In contrast, much suggests that Birka indeed became umentation of find data related to the town morphology, an historical place during the course of the 1400s. Al- something which was employed on a large scale by the though the site of Birka was still known in the 1400s, its Birka Excavations Project, 1990-1995. Conservation of political role was obsolete. New towns and royal palaces finds in the field and the joint cooperation between -os replaced Birka as centers for administration, trade and teologists and archaeologists in these excavations is an- crafts. With few exceptions this tradition continued to other aspect of this development. Using the technique of live on until the time when Hjalmar Stolpe, using archae- single-context and computerized documentation has also ological methods, finally succeeded in establishing the contributed to developing the stratigraphical analysis of

40 Birka in tradition and research this site (Ambrosiani 2013b). The favorable conditions for dating the site and eval- uating the situation of the ancient and historical shore- lines of the island of Björkö have played a very import- ant part in better understanding aspects of the process of shore displacement in the Mälaren area (Lindqvist 1928; Ambrosiani 1982; 2013a:63ff). For many decades Björkö has also been a key setting for developing both theoretical and practical aspects of ancient monument preservation in Sweden (Gustawsson 1970; 1977). This continues today in current work with developing new non-evasive geophysical methods for documenting cultural layers and ancient construc- tions (Trinks et al. 2013) in joint cooperation between the Swedish National Heritage Board and the Ludwig Boltzmann Institute for Archaeological Prospection and Virtual Archaeology, using Birka and other sites as inves- tigative projects. In all of these contexts, Birka has played and will continue to play a key role.

Corresponding author Björn Ambrosiani Riksantikvarieämbetet Stockholm, Sweden [email protected]

41 New Aspects on Viking-age Urbanism, c. 750-1100

References Almgren, O. 1904. Sveriges fasta fornlämningar från hednatiden. Stockholm. Ambrosiani, B. 1982. Mälarstäderna och landhöjnin gen. Bebyggelsehistorisk Tidskrift 1982 (3). 71 – 81. – 1992. Excavations in the Black Earth Harbour 1969-71. In: B. Ambrosiani & H. Clarke (eds.) Investigations in the Black Earth. Birka Studies 1. Stockholm. 71–82. – 2013a. Birka 1680-2013. Ett långvarigt forsknings projekt. A Long-Term Research Project. Stockholm. – 2013b. Excavations at Birka 1990-1995. The Stratigra phy, Vol. 1. Part One: The Site and the Shore. Part Two: The Bronze Caster’s Workshop. Birka Studies 9. Riksantikvarieämbetet. Stockholm. Ambrosiani, B. & Arrhenius, B. et al. 1973. Birka. Svarta Jordens Hamnområde. Arkeologisk un dersökning 1970–1971. Riksantikvarieämbetet Rap port C1 1973. Stockholm. Erikson, B.G. 2015. Kungen av Birka. Hjalmar Stolpe – arkeolog och etnograf. En biografi. Stockholm Gustawsson, K.A. 1970. Fornminnesvård. Andra Upplagan. Stockholm. Gustawsson, K.A. 1977. Björköområdets skötsel. Forn vännen 1977 (72). 87 – 101. Lindberg, C. 1996. En instruktion för arkeologiska ut grävningar – om pionjärerna Hjalmar Stolpe och Carl V. Hartman. Fornvännen 1996 (91). 81 – 90. Lindqvist, S. 1928. Vattenståndet vid Birka på 900-talet. Fornvännen 1928 (23). 118 – 120. Trinks, I., Neubauer, W., Nau, E., Gabler, M., Wallner, A., Hinterleitner, A., Biwall, A., Doneus, M. & Pregesbaer, M. 2013. Archaeological Prospection of the UNESCO World Cultural Heritage Site Birka– Hovgården. In: W. Neubauer, I. Trinks, R.B. Salis bury & C. Einwögener (eds.) Procedings of the 10th International Conference on Archaeological Prospection Vienna, May 29th – June 2nd 2013. Ludwig Boltz mann Institute. Vienna. 39 – 40

42 Elites, networks and the Finnish connection in Birka

Elites, networks and the Finnish connection in Birka

Ingrid Gustin

Introduction In Viking Age research, urban centers are often stud- ied together with trade and long-distance contacts (e.g. Hodges 1982; Müller-Wille 1989; Callmer 1994; Sind- baek 2007). When it comes to Birka, the connection with the and the Frisians was em- phasized early by archaeologists and historians (Stjerna 1909; Schück 1926; Arbman 1937; also Bäck 2007:8 and ref. therein). Later research at Birka has shown that, in its earlier phase, the town had long-distance con- tacts with not only Western Europe and Dorestad, but also the West Slavic coast region in today´s Poland and . It has also been pointed out that the West European influence in Birka decreased significantly from the second half of the ninth century. Instead, the town’s eastern contacts seem to have intensified marked- ly. The contacts with the West Slavic region were main- Fig. 1. Penannular brooch with faceted and pegged knobs, an object tained during this later period in Birka and there were originating from or showing heavy influence from . This pen- also links to other Scandinavian regions (Ambrosiani annular brooch was found in Bj 1053B. (Photo: Eva Vedin, SHMM). 2002a; 2002b). The eastern connections of Scandinavia and, espe- of Finland have been retrieved in the Black Earth of cially, Central Sweden (mainly ) with what Birka, as well as in the cemeteries of the town. Howev- is today Russia, and have also right- er, the contacts indicated by these objects have seldom ly attracted much attention in Swedish archaeological been addressed by archaeologists in Sweden. research. The links to these eastern regions have often been addressed (e.g. Arne 1911, 1914; Arbman 1955; Objects from Finland found in Birka Jansson 1987, 1997; Callmer 2000a; 2000b; Heden- stierna-Jonson 2006). When it comes to Birka, the Imported pottery is a find group often used in archae- great interest in West and East European contacts seems ological reports to indicate contact with other regions. to have resulted in other regions being overshadowed From the Birka project excavations in the Black Earth, and forgotten when the connections of the town are the proportion discussed. This is evident when it comes to Birka´s con- of the imported pottery from the first four years has tacts with the Finnish mainland. Objects originating or been published thus far. According to the survey made showing heavy influences from the south-western parts by Mathias Bäck, the west European pottery amounts

43 New Aspects on Viking-age Urbanism, c. 750-1100 to c. 1.5% of all pottery retrieved during these four sea- sons of excavation. In contrast, the Baltic Finnish ware The duration and consistency of contacts be- amounts to c. 6%. When it comes to imported ware, tween Birka and Western Finland only Slavic pottery turned out to be more frequent. The Another reason to highlight the connection between Slavic pottery amounted to c. 15-17% (Bäck 1995). Birka and Western Finland is its long duration. Sell- The relatively high proportion of Baltic Finnish ware ing maintained that Baltic Finnish pottery in the Birka in the Black Earth is consistent with the pottery found graves ranged from the second half of the ninth century in the town´s cemeteries. According to Dagmar Sell- to the middle of the tenth century. However in a re- ing´s study of Viking Age and early Medieval Ceramics cent study the datings for other type of objects such as in Sweden, the Baltic Finnish ware in Birka´s cemeter- combs, oval brooches, round brooches, sets of beads, ies accounts for 6 % of all pottery found in the graves. and pendants, found in the same graves as Baltic Finn- In her study Selling also noted that Björkö, the island ish ware were compiled (Gustin in print). These datings where Birka was situated, is the site showing the largest show that the contacts between Birka and Western Fin- number of finds of this type of ware in Sweden (Selling land go back to at least 760, as seen in grave Bj 682.3.3 1955:140–144, 153–155, 218, 229). This indicates that Baltic Finnish pottery occurred in Studies of artefacts in the Birka graves originating from the Birka graves from early on. The youngest grave with or with heavy influences from Finland1 show that there Baltic Finnish ware, Bj 348, has a set of beads that can are altogether 13 such objects (Kivikoski 1938; Gustin be dated to c. 935–955. The conclusion is that there in print). The most frequent object was the penannular was contact between Birka and Western Finland more brooch with faceted and pegged knobs on the terminals or less throughout the whole existence of the town. (fig. 1). Besides this type of penannular brooch, there In the abovementioned study, the consistency of is also a round brooch, comb-shaped bronze pendant, a the contact between Birka and Western Finland was bird-shaped object (perhaps a pendant), a fire-steel with a also discussed. In this case the datable graves contain- bronze grip in the form of an animal head and a fire-steel ing objects from the Finnish mainland were divided in with a bronze handle in the form of a mounted figure.2 to three time periods: early Birka, middle Birka/tran- The objects and their find contexts have been discussed sition early–late Birka, and late Birka. Of the graves at length elsewhere (Gustin in print). containing Baltic Finnish ware possible to date within If these objects are added to the Baltic Finnish pot- the three periods mentioned above, two belong to early tery found in the Birka graves, a total of 46 graves of Birka, two to middle Birka/transition early–late Birka, the excavated 1100 graves at Birka contained material and nine to late Birka. A comparison can be made with culture originating from or displaying heavy influences the Birka graves where it has been possible to date the from the Finnish mainland. This means that roughly 4 sets of beads (Callmer 1977:155f.). For theses dateable % of the Birka graves held artefacts from that area (see graves, 15 belong to early Birka, 21 to middle Birka/ Gustin 2012; Gustin in print).

Early Birka Middle Birka/transition early– Late Birka late Birka Percentage of graves with 17% 25% 58% datable bead sets Percentage of datable 15% 15% 70% graves with Baltic Finnish ware

Tab. 1. The percentage of Birka graves per period dated on the basis of bead sets (Callmer 1977:155f.) and the percentage of datable graves with Baltic Finnish ware.

44 Elites, networks and the Finnish connection in Birka

Grave type Graves with Graves with Baltic Number of Distribution Total distri- costume details Finnish ware graves with of different bution of from Finland objects grave types different from Fin- containing grave types, land objects from based on Finland Gräslund 1980 Cremation 104, 197, 776, 9, 45A/B, 51, 28 61% 51% graves Seton VI 94A/B, 98B, 104, 155, 201, 226, 326, 329, 348, 352, 370, 415, 418, 428, 437, 458, 679, 682, 714, 811, 912, 1042 Coffin graves 58A, 478, 480, 531, 971, 1095 8 17% 38% and other 759, 1162 skeleton graves Chamber 644, 954, 1053B, 731, 834, 845, 855, 10 22% 11% graves 1074 860, 901 Total 13 34 46 100% 100%

Tab. 2. Distribution of costume details from Finland and Baltic Finnish ware between different grave types in Birka.

Fig. 2. Chamber grave Bj 644 with a fire-steel likely to have been brought to Birka from Finland (from Arbman 1943:223 pl. 183).

transition early–late Birka, and 49 to late Birka. The proportion of graves per period is shown in table 1. Tab. 1 shows that the proportion of graves with Bal- tic Finnish ware is much larger in late Birka than in earlier periods. However, this seems to apply to graves with datable bead sets as well, and might be due to the fact that many of the graves that cannot be dated are cremation graves, and presumably of an early date. Fur- thermore, the easily dateable chamber graves were more common during Late Birka (Gräslund 1980:29). Tab. 1 also shows that the percentage per period for graves with Baltic Finnish ware is roughly the same as for graves with datable bead sets. This indicates that con- tacts between Birka and Western Finland were virtually constant over time. It should be noted that the continu- ity displayed by Baltic Finnish ware differs from other

45 New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 3. Distribution map for cubo-octahedral weights. (From Steuer 2002:139).

eastern material in the Black Earth. The excavations in the 1990s showed that, while Graves in Birka with objects originating or there were a number of artefacts reflecting eastern con- showing heavy influences from Western Fin- tacts as early as the first phases of Birka, they were rare land in comparison to material from Western Europe and the Cremation graves amount to 51% of all graves in Birka. area south of the Baltic Sea. During the second half of As presented in the tab. 2 c., 61 % of the graves in Birka the ninth century, however, the situation changed, and displaying Finnish provenance or influence were crema- Scandinavian and eastern objects began to exceed west- tion graves. This means that the proportion of graves ern ones in number (Ambrosiani 2002a). carrying objects from mainland Finland can be said to

46 Elites, networks and the Finnish connection in Birka more or less follow the general proportion for this grave cal legitimation of the group that controls their acqui- type in general. sition, distribution and use (Helms 1988; 1993:165; A grave type in Birka that, on the other hand, is Lucy 2005:108 with ref.; Tsigaridas Glørstad 2012:36). over-represented when it comes to objects from main- Furthermore, items such as scales, weights, and mea- land Finland is the chamber grave. These graves consti- suring rods, as well as Islamic coins in the graves of tute c. 11% of all graves in Birka, but as many as c. 22% some of the persons in Birka buried with objects from of them carried objects originating or showing heavy Western Finland, show that these persons belonged to influences from Finland. The Birka chamber graves groups involved in the trade and exchange of goods. In have often been considered representative of an elite south-western Finland there are many finds of norma- consisting of warriors or merchants with long-distance tive contacts (e.g. Stenberger 1971:704; Gräslund 1980:80; and standardized weights of the same types as in Birka. Steuer 1987:119; Clarke & Ambrosiani 1991:69f., 137; The high amount of these weights in this region is Ringstedt 1997:111f., 115; Callmer 2008:123f.). clearly illustrated by the distribution map for cubo-oc- . The high proportion of chamber graves containing tahedral weights compiled by Heiko Steuer (fig. 3). The costume details from Western Finland and Baltic Finn- presence of same types of weights in Birka and Western ish ware indicates that the elite in Birka was over-rep- Finland is a clear indication that both were part of a com- resented when it came to burials with objects with mon supra-regional network in the Viking Age. Not only provenance or influence from the Finnish mainland. the objects from Finland found in Birka, but also objects It is particularly remarkable that one of the chamber from Scandinavian and especially Central Sweden found graves, Bj 644, a double grave, is among the graves on in Western Finland confirm that there has been a close Björkö with the greatest wealth of artefacts (cf. Ringst- connection between Eastern Scandinavia and Birka on edt 1997:101). In this grave there was a fire-steel with the one hand and south-western Finland on the other. a bronze grip in the form of an animal head. This type The Scandinavian finds in Finland have been discussed of fire-steel is known from Finland and the Perm re- by Ella Kivikoski and Ingmar Jansson. Their results will gion. Even if it is hard to establish where this type was be presented below. produced, it is likely that the fire-steel found in Birka came there via Finland (see Gustin in print). In grave Bj 644 there were also objects such as scales, weights, a Scandinavian objects found in Finland pouch, a , stirrups, an axe, shield boss, game piec- Viking Age artefacts originating from Scandinavia es, gilded oval brooches, shears, and comb beaters (fig. and Central Sweden found in Finland have been sur- 2). Likewise Bj 834 and 845, which contained Baltic veyed and described by Ella Kivikoski (1938:232–241; Finnish ware, were very rich in artefacts. In both graves 1964:214). The objects Kivikoski noted consist of there was a staff, often interpreted as a measuring rod about twenty equal-armed fibulae, some oval brooch- (Gustin 2004; Gustin 2010; but see Price 2002 for dif- es, bronze dress pins with barrel shaped head (Sw. med ferent interpret.). In chamber grave Bj 954 there was a tunnformigt huvud), arm rings with line ornamentation well-equipped woman buried with a penannular brooch (Sw. med våglinjedekor), some twenty arm rings with with faceted and pegged knobs. Furthermore, her grave interlace ornament (Sw. bandflätning), scabbard chapes contained two oval brooches, a trefoil brooch and other with animal style ornamentation, silver pendants with female dress details and accessories as well as a weight swirl motives (Sw. silverbleck med virvelmotiv), iron and Islamic coins. rings with amulets, mount to a fire steel, a needle case Lavishly furnished chamber graves in Birka like Bj etcetera. Line ornamented equal-armed fibulae of the 644, 834, 845 and 954 indicate that the elite in the Ljønes type found in Finland may also have come from town were part of a network that was directly or indi- Central Sweden, where they are numerous. According rectly connected to groups in Finland. It is likely that to Kivikoski, Scandinavian objects of types common in foreign objects, as the ones relating to the Finnish main- Central Sweden have above all been found in south- land, were used by groups in Birka to assert and increase west Finland, but, for example, oval brooches also occur status and social position. in the eastern and north-eastern parts of the country Several studies have shown that objects acquired (Kivikoski 1938).4 from remote regions are actively used in the ideologi- More recently Ingmar Jansson has studied the Scan-

47 New Aspects on Viking-age Urbanism, c. 750-1100 dinavian female ornaments found in Finland (Jansson 1990). Jansson´s study confirmed that the material chiefly consists of equal-armed brooches.5 The brooch- es belong to the ninth and tenth centuries, and some of them are of types typical for Eastern Sweden. The Scandinavian female ornaments have above all been found in Finland Proper and Satakunta, that is, in the south-west (fig. 4). According to Jansson, the Scandina- vian equal-armed brooches were retrieved from graves showing local features. Only rarely more than one ob- ject from Scandinavia was found in each grave. Accord- ing to Jansson, this shows that the Scandinavian objects found in Finland hardly derive from migration. Kivikoski pointed out that the objects from Sweden appear on most of the larger grave fields in Western Finland. However, ninth and tenth century objects are mainly found in Vakka Suomi, a region that, in her view, had close ties to Birka and central Sweden (Kivikoski 1964:215). In much later archaeological research in Fin- land, it has also been stated that Vakka Suomi had closer ties with Birka and Central Sweden, than the other settle- ment areas in south-western Finland (Edgren 1992:249; 2008:481). So what characterizes Vakka Suomi?

Vakka Suomi Fig. 4. Map showing the distribution of Scandinavian ornaments in Finland (from Lehtosalo-Hilander 1982:72 fig. 14). Vakka Suomi is a sub-region within Finland Proper and it is situated in the north-western part of the latter re- king Age. Kivikoski pointed out that the parishes of gion (fig. 5). In Vakka Suomi there are two , Laitila and Kalanti were especially wealthy and densely Laitila and Kalanti, which are considered to be core populated during the latter time period. According to areas of settlement in Iron Age Finland. From ceme- Kivikoski, the long distance contacts were still import- teries in these districts, finds have been retrieved which ant during the Viking Age and the region seems to have are unique to the archaeological record of this country. had permanent contacts, initially with , then Grave goods, non-local forms of burial as well as stray Birka and then again Gotland (Kivikoski 1969:55–58). finds show that the people in these districts had However, Vakka-Suomi was not very fertile; the soil was long distance contacts from the Early Roman Iron Age barren and rocky and there could hardly have been any onwards (Purhonen 1996:13–15). agricultural surplus during the Viking Age (Kivikoski According to Marianne Schaumann-Lönnqvist, 1964:217). So how come groups in Vakka Soumi were there appear to have been several small seats during involved in a network that supplied them with presti- the Finnish Merovingian period where petty noble- gious goods? What was the economic basis for groups men represented the power elite on a local level. Since living there? What kind of surplus products could have the tightest concentration of Merovingian Period no- been handled and distributed? The views of Finnish ar- bility cemeteries known to date lies in Vakka Soumi, chaeologists differ. Schaumann-Lönnqvist has stated that this was According to Ella Kivikoski, furs and fur trade were the leading community in Finland at the time (Schau- the general basis for the prosperity of Western Finland. mann-Lönnqvist 1996:133). The inhabitants in south-western Satakunta had excel- The region seems to have been wealthy not only lent opportunities in the north and those who during the Merovingian Period but also during the Vi-

48 Elites, networks and the Finnish connection in Birka

P. Taavitsainen has found an intensification in the util- isation of wilderness during the Viking Age, as seen for example in Häme and Satakunta. In these regions, as well as in Savo, Karelia and Northern Finland settle- ments were established which can be described as trap- per stations. Taavitsainen maintains that these regions experienced a change in commercial wilderness econo- my characterised by fur trade during the Viking Period (Taavitsainen 1990:112f.). Even though furs and fur trade are regarded as an economic basis for the trading peasants in Vakka Suomi, other products might also have played a role. Torsten Edgren and Pirko-Liisa Lethosalo-Hilander have suggested that part of the prosperity of Vakka Suomi during the Viking Age can be explained by wood working. They argue that the inhabitants in this dis- trict could already have been producing and exporting moulded splint vessels, the wooden containers known in Finnish as vakka, and in Swedish as svepask. These containers were a very profitable export product during much later centuries (Lethosalo-Hilander 1982c:75; 1983:297; Edgren 2008:481). Another product which possibly could have been of importance when it comes to export is iron. Edgren has pointed out that ready ac- cess to ore and unlimited supplies of charcoal for smelt- ing made Finland a strategic centre of North Baltic iron Fig. 5. Map of Western Finland showing places and areas mentioned in the text (reworked after Lehtosalo-Hilander 1982:75, fig. 19). production (Edgren 2000:110f.). However, it was probably not only the local products distributed by the elite in Vakka Suomi that attracted lived in Häme had vast hinterlands rich in fur-bearing people in Eastern Sweden and Birka. Seen from an animals. Kivikoski pointed out that the routes from Eastern Scandinavian point of view, the elites in West- south-western Satakunta probably went through Vak- ern Finland were intermediaries’ when it came to early ka Suomi on the way to the coast. She therefore main- contacts and networks that stretched further east. These tained that the reason for the prosperity of Vakka Suomi contacts are evident from the early Islamic coins that could be due to export trade with fur from the inland as have been retrieved in Vakka Suomi. well as to shipping, fishing and seal hunting. Kivikoski also suggests that groups in Vakka Suomi were involved in the extraction of furs from the Sami people (Kivikos- Finds of early Islamic coins in Vakka Suomi ki 1964:215–217). and the eastern link According to Torsten Edgren, Vakka Soumi´s west- In Finland, coins were scarce before the 10th century ern connections were based on the region’s geographical (Talvio 2002:130). An interesting phenomenon when it location and the fact that it constituted a link between comes to Vakka Suomi is that the district has some of the Baltic region and the prosperous inland districts of the earliest finds of Islamic coins in Finland and that the Kokemäki, Eura and Köyliö. Also trade routes from inhabitants were pioneers of coin import from the east. Häme to the coast may partly have gone through Vakka Among nine early Islamic coins from Vakka Suomi, dat- Suomi (Edgren 1992:249). It is possible that the im- ed to AD 650/703–807/808, four are from Vainionmäki portance of the northern and north-eastern trade routes in Laitila. These coins were found on a cremation ceme- leading through Vakka Soumi should be viewed in the tery from the Merovingian Period. light of an increasing trade with furs. In his research J. The four dirham fragments from the Vainionmäki

49 New Aspects on Viking-age Urbanism, c. 750-1100 cemetery have been described by Tuukka Talvio (Talvio the Kama River. Here rich women´s graves from the 1996; 1999; 2002:86). Two of the fragments were dated 7th and 8th centuries contain not only belts, but also to AD 705–15 and AD 737–39. They belonged to the artefacts from , and such Umayyad dynasty, which came to an end in AD 750. as silver coins, carnelian and crystal beads (Hirviluoto The other two were Abbasid coins dated to AD 758–65 1986:72f.). So even before the Viking Age, trade routes and AD 803–04. According to Talvio, these coins were along the Volga River and the steppe linked the Kama not associated with any dateable objects. Due to the region to civilizations in Western and Central Asia. dating of the cemetery to the Merovingian Period, and During the Viking Age it is known that furs and wal- the fact there were no finds dating later than c. AD 800, rus tusks from the north were exchanged for Oriental Talvio finds it likely that the earliest three of the Islamic luxuries on markets along the Kama and Volga rivers coins reached Finland before this time. If this assump- (Jansson 2000b:15–17; Roslund in print). tion is correct, the coins from the cemetery in Laitila Permian belts and their mounts have been found in belong to a small group of very early arrivals of Islamic male graves from the 7th century in Western Finland. coins that precede the large scale import that began in Between Western Finland and the Kama region there the early 9th century. Islamic coins dated so early are are only few known finds.6 There have been various sug- neither common in Scandinavia nor in Northern Rus- gestions as to how the link between Western Finland sia. Talvio points out that finds of this type are restricted and the upper Volga and Kama-region could be main- to one from Sweden (Tuna in Alsike, Uppland) and one tained (see Hirviluoto 1986:71f.; Ryabinin 1986:217f.). or two from north-western Russia (Talvio 1996 with Christian Carpelan has noticed a correspondence in ref.). time between Central Russian axes found in Finland Talvio’s conclusion about the coin finds from the and the Permian belts and he has therefore maintained Vainionmäki cemetery and other finds of Islamic coins that goods from the Kama area were first transported by with early datings from places like Mynämäki and Mi- traders to Central Russia. From there local traders from etonen in Finland Proper is that a small amount of coins the Volga-Oka region transported axes and other objects reached south-western Finland around AD 800 and such as belts towards the northwest as far as Finland. possibly even earlier (Talvio 1999). Other archaeolog- However, the groups from Western Finland involved in ical finds confirm that there were established contacts this long distance exchange network did not know that between south-western Finland and regions in Russia the impressive belts were originally the belts of women. before the Viking Age. In this case, the connections be- Therefore, the belts ended up in high status male graves tween south-western Finland and the Kama and upper in Finland (Carpelan 2002:179). That the groups on Volga region in Russia are of special interest. the Finnish mainland as well as the groups living in the Volga-Oka region belonged to the same language fam- ily and the same cultural community, the Finno-Ugric, Permian belts has most probably contributed to the upholding of the A find group showing early links between south-west- contacts (Callmer 2000b:54; Jansson 2000b:15–17). ern Finland and areas much further east in Russia is Due to these far-reaching contacts, groups in Finland belts. A considerable number of ornamental belts of became involved in the developing west-east trade on the so called Nevolino type reached Finland during the the Volga River in the 8th century (Callmer 1980:209 eight century. These belts have a wide distribution; they with references; Roslund in print). are found from Tomsk in the east to Gamla Uppsala in The graves containing Permian belts in Western Fin- the west. However, they appear most frequently in the land were well-furnished. The finds indicate that the basin of the Kama River and in Western Finland. The buried belonged to the local population. Interestingly number of finds from Finland amounts to 19 (Callmer enough, in some cases these burials also contained a 1980:209; Callmer 1989:22). sword of Scandinavian type with animal ornamenta- The belts, often referred to as Permian belts, are tion. The find combination thus shows that these per- thought to originate from the regions around the river sons belonged to families with connections not only Kama and its confluence with the Volga River west of with Finno-Ugric groups in the Permian region, but the Ural mountains. The Kama region was an interna- also with Scandinavian groups on the western side of tional hub as shown by graves on burial grounds along the Baltic Sea. Furthermore, two mounts from a Perm-

50 Elites, networks and the Finnish connection in Birka ian belt were retrieved from one of the graves, Gull- rectly or indirectly connected to groups on the Finnish högen, in Gamla Uppsala, Uppland, the religious and mainland. Besides Baltic Finnish pottery or dress details political center of the “Svear”. It is likely that Gullhö- from Finland, these graves also contained objects such gen was a large mound situated on Högåsengravfältet, as weights, scales and Islamic coins – items indicating south-west of the so-called Royal mounds. This mound an affiliation to groups involved in the trade and ex- was excavated in the middle of the 19th century. Unfor- change of goods. tunately, the documentation of the excavation is poor. Finds of standardized weights in south-western Fin- However, it seems clear that Gullhögen contained a land clearly indicate that this region was part of a su- fragment from a vessel from Gotland or the continent, pra-regional network in the Viking Age, just like Birka. a gold cloisonné fragment as well as two mounts from Objects with provenance or influence from Finland a Permian belt. The grave is difficult to date but it has found in Birka on the one hand and objects from Scan- been suggested that it originates from the early 7th cen- dinavian and Central Sweden found in Finland on the tury (Ljungkvist 2013:41–44). other confirm that there has been a close connection be- The find of the Permian belt mounts shows that tween the Mälaren region and south-western Finland. there were early links between the elite in Uppland, the Probably products such as furs distributed by the elite elite in Western Finland and the groups in the Perm- in Vakka Suomi attracted groups in Eastern Sweden and ian region. It has been pointed out that it probably was Birka. through these elite contacts between “Svear” and groups However, the elite in Birka might also have had in Western Finland, that the “Svear” found out about other reasons for maintaining contacts with groups the routes along the Finnish Gulf and how to travel in south-western Finland. As has been shown by the to Lake Ladoga and further east (Jansson 2000a:124f; early Islamic coins in Laitila, as well as by the finds of 2000b:16f.; Carpelan 2002:180). Permian belts in Vakka-Suomi, groups in south-western Finland were linked to routes and networks spanning much further east. The Birka artefacts deriving or show- Summary ing heavy influences from Western Finland thus express The study of trade and long distance contacts is essential access not only to products from the Finnish mainland, for understanding Viking Age urban centers. However, but also, in a wider sense, to contacts and routes further there has been an inclination to emphasize the impor- east and by extension also to the riches of the Caliphate. tance of long-distance contacts with some particular regions, while others have been much less discussed de- spite the rich material. Research on contacts with these overlooked regions would certainly bring new aspects in Acknowledgement the study of Viking Age urbanism to light. I am grateful to Johan Callmer, Mats Roslund and In research on Birka, a strong focus has since long Anna Wessman for discussions and for comments on been directed towards long-distance contacts with this article. Many thanks also to Ingmar Jansson for Western as well with Northeastern Europe (todays Rus- constructive comments on the paper presented at the sia, Belarus and Ukraine). The great attention paid to conference in Stockholm. I am also indebted to Riks- West and East European contacts has likely contribut- bankens Jubileumsfond for making this work possible ed contacts with Finland falling out of the discussions. by providing financial support However, as seen from studies of the pottery in Birka, the Baltic Finnish ware constitutes the second largest group when it comes to imports to the town. Corresponding Author Baltic Finnish ware has not only been retrieved from Ingrid Gustin town layers in Birka. There are finds of this type of ware Institutionen för arkeologi och antikens historia, Lund in the graves as well as of dress details deriving from university or showing heavy influence from the Finnish main- Lund, Sweden land. Some of these finds derive from the most lavishly [email protected] furnished chamber graves in Birka, indicating that the elite in the town were part of a network that was di-

51 New Aspects on Viking-age Urbanism, c. 750-1100

500– 900. Medieval Scandinavia 2000 (13). 7 – 63. References – 2000b. From West to East: The Penetration of Scan Ambrosiani, B. 2002a. Birka im 10. Jahrhundert unter dinavians into Eastern Europe ca. 500–900. In: M. besonderer Berücksichtigung der Ostverbindung. In: Kazanski, A. Nercessian & C. Zuckerman (eds.) Les J. Henning (ed.) Europa im 10. Jahrhundert: Archäolo Centres proto-urbains russes entre Scandinavie, By gie einer Aufbruchszeit. Mainz. 227 – 235. ance et Orient. Actes du Colloque International tenu – 2002b. Osten und Westen im Ostseehandel zur au Collège de en octobre 1997. . 45 – Wikingerzeit. In: K. Brandt, M. Müller-Wille & 94. Chr. Radtke (eds.) Haithabu und die frühe Stad – 2008. Herrschaftsbildung und Machtausübung: Die tentwicklung in nörd lichen Europa. Schriften des Anfänge der ar-Rus (Rus’) ca. 500–1000 n.Chr. In: Archäologischen Landesmuseums 8. Neumünster. O. Auge, F. Biermann, M. Müller & D. Schultze 339 – 348. (eds.) Bereit zum Konflikt. Strategien und Medien der Arbman, H. 1937. Schweden und das karolingische Re Konflikterzeugung und Konfliktbewältigung im eu ich. Studien zu den Handelsverbindungen des 9. ropäischen Mittelalter. Ostfildern. 103 – 130. Jahrhunderts. Kungl. Vitterhets Historie och An Carpelan, C. 2002. The Volga Route through the Mil tikvitets Akademiens handlingar 43. Stockholm. lenia – as seen from Finland. In: M.A. Usmanov, – 1940. Birka I. Die Gräber. Tafeln. Kungl. Vitterhets K.S. Ishchakov, G.S. Khuzha, R.S. Khakimov, L.M. Histo rie och Antikvitets Akademien. Stockholm. Drobizheva, L.N. Andreeva, R.R. Khayrutdinov – 1943. Birka I. Die Gräber. Text. Kungl. Vitterhets (eds.) The Great Volga Route. History, Formation and Historie och Antikvitets Akademien. Stockholm. Development. Papers from the conference 2001. Ka – 1955. Svear i österviking. Stockholm. zan. 174 – 180. Arne, T.J. 1911. Sveriges förbindelser med Östern un Clarke, H. & Ambrosiani, B. 1991. Vikingastäder. der vikingatiden. Fornvännen 1911 (6). Stockholm. Höganäs. 1 – 66. Edgren, T. 1992. Den förhistoriska tiden. In: M. Norr – T.J. 1914. La Suède et l’Orient. Uppsala. back (ed.) historia 11. Esbo. 9 – 270. Bäck, M. 1995. Importkeramiken i Birka. Meta 1995 – 2000. The Eastern Route. Finland in the Viking Age. (95):1. 4 – 21. In: W. Fitzhugh & E. Ward (eds.) Vikings. The – 2007. Mötets dynamik – tröghets- och motståndsk North Atlantic . Washington. 103 – 115. rafter i mellansvensk vikingatid. In: J. Anund (ed.) – 2008. The Viking Age in Finland. In: S. Brink & N. Gården, tingen, graven. Arkeologiska perspektiv på Price (eds.) The Viking World. London & New York. Mälardalen. Riksantikvarieämbetet arkeologiska un 470 – 484. dersökningar skrifter 7. Stockholm. 7 – 40. Gräslund, A.-S. 1980. The Burial Customs: A Study of Callmer, J. 1977. Trade Beads and Bead Trade in Scandi the Graves on Björkö. Birka IV. Kungl. Vitterhets navia ca. 800–1000. Acta Archaeologica Lundensia. Historie och Antikvitets Akademien. Stockholm. Series in 4°. No. 11. Lund. Gustin, I. 2004. Mellan gåva och marknad: Handel, tillit – 1980. Numismatics and Archaeology: some problems och materiell kultur under vikingatid. Lund Studies of the Viking Period. Fornvännen 1980 (75). 203 – in Medieval Archaeology 34. Stockholm. 212. – 2010. Of rods and roles. Three women in Birka’s – 1989. The beginning of the East European trade con cham ber graves. In: C. Theune, F. Biermann, R. nections of Scandinavia and the Baltic Region in Struwe & G.H. Jeute (eds.) Zwischen Fjorden und the eighth and ninth centuries A.D. A Woszinsky Steppe: Festschrift für Johan Callmer. Internationale Mór Muzeum Evkönyve 1989 (15). 19 – 51. Archäologie. Studia honoraria. Rhaden/Westfahlen. – 1994. Urbanization in Scandinavia and the Baltic Re 343 – 354. gion c. AD 700-1100: Trading Places, Centres and – 2012. Birkaborna med föremål från finska fastlandet. Early Urban Sites. In: B. Ambrosiani & H. Clarke Vilka ar de? In: C. Hedenstierna-Jonson (ed.) Birka (eds.) Developments around the Baltic and the North Nu. Pågående forskning om världsarvet Birka och Hov Sea in the Viking Age. The Twelfth Viking Congress = gården. The National Historical Museum. Studies Birka Studies 3. Stockholm. 50–90. 22. Stockholm. 95 – 110. – 2000a. The Archaeology of the Early Rus’ c. A.D. – in print. Contacts, identity and hybridity. Objects from

52 Elites, networks and the Finnish connection in Birka

the Finnish mainland in the Birka graves. In: J. – 1969. Esihistoriallinen aika. In: E. Kivikoski & O. Callmer, I. Gustin & M. Roslund (eds.) Identity Koivis to (eds.) Laitilan historia I. Vammala. 1 – 69. forma tion and Diversity in the Early Medieval Baltic Lehtosalo-Hilander, P.-L. 1982. Luistari III. A Buri and Beyond. Communicators and Communication. al-Ground Reflecting the Finnish Viking Age Society= The Northern World 75. Suomen muinaismuistoyhdistyksen aikakauskirja/Fins Hedenstierna-Jonson, Ch. 2006. The Birka warrior. The ka fornminnesforeningens tidskrift 1982 (82):3. material culture of a martial society. Theses and Papers sinki. in Scientific Archaeology 8. Stockholm. – 1983. Gutarnas förbindelser med Finland. In: I. Jans Helms, M. 1988. Ulysses´ sail: An ethnographic odyssey son (ed.) Gutar och vikingar. Stockholm. 288 – 305. of power, knowledge and geographical distance. Prince – 1992. Finland. In: E. Roesdahl (ed.) Viking og ton. Hvidekrist: Norden og Europa 800–1200. Copenha Helms, M. 1993. Craft and the kingly ideal: Art, trade gen. 62 – 71. and power. Austin. Ljungkvist, J. 2013. Monumentaliseringen av Gamla Hirviluoto, A.-L. 1986. Finland´s cultural ties with the Uppsala. In: O. Sundqvist & P. Vikstrand (eds.) Kama region in the late Iron Age especially in the Gamla Uppsala i ny belysning. Religionsvetenskapliga light of pottery finds. In: I. Lehtinen (ed.) Traces of studier från Gävle 9. Uppsala. 33 – 67. Central Asian Culture in the North. Helsinki. 71 – 80. Lucy, S. 2005. Ethnic and cultural identities. In: M. Hodges, R. 1982. Dark Age Economics. The origins of Díaz-Andreu (ed.) Archaeology of Identities: approach towns and trade A.D. 600–1000. London. es to gender, age, status, ethnicity and religion. London Jansson, I. 1987. Communications between Scandina & New York. 86 – 109. via and Eastern Europe in the Viking Age: The Ar Müller-Wille, M. (ed.) 1989. Oldenburg – Wolin – Sta chaeological Evidence. In: K. Düwel & I. Jansson raja Ladoga – Novgorod – Kiev. Handel und Han (eds.) Untersuchungen zu Handel und Verkehr der vor- delsverbindungen im südlichen und östlichen Ostseer und frühgeschichtlichen Zeit in Mittel- und Nordeuro aum während des frühen Mittelalters. Bericht der Rö pa. Teil IV: Der Handel der Karolinger- und Wikinger misch Germanischen Kommision 69. Mainz. zeit. Abhandlungen der Akademie der Wissen Price, N. 2002. . Religion and War in schaften zu Göttingen. Philologisch-Historische late Iron Age Scandinavia. AUN 31. Uppsala. Klasse. Dritte Folge 156. Göttingen. 773 – 807. Purhonen, P. 1996. Introduction. In: P. Purhonen (ed.) – 1990. Wikingerzeitliche skandinavischer Frauen Vainionmäki – a Merovingian Period Cemetery in schmuck in Lettland, Finnland und der Rus’ – ein Laitila, Finland. Helsinki. 10 – 18. Vergleich. In: Ē. Mugurēvičs, Z. Apala & I. Ronis Ryabinin, Y.A. 1986. Cultural links of Finno-ugric (eds.) Die Kontakte zwischen Ostbaltikum und Skan tribes in the Middle Ages. In: I. Lehtinen (ed.) Traces dinavien im frühen Mittelalter. Riga. 23 – 25. of Central Asian Culture in the North. Helsinki. 213 – – 1997. Warfare, Trade or Colonization? Some General 222. Remarks on the Eastern Expansion of the Scandi Ringstedt, N. 1997. The Birka Chamber-Graves: Eco navians in the Viking Period. In: P. Hansson (ed.) nomic and Social Aspects. An analysis based on quanti The Rural Viking in Russia and Sweden. Örebro. 9 – tative methods. Stockholm Archaeological Reports 32. 64. Stockholm. – 2000a. Östersjöländerna och vikingatiden. In: K. Roslund. M. in print. Bringing “the periphery” into fo Dahlbäck (ed.) Att förstå det mänskliga: Humanistisk cus. Social interaction between Baltic and the forskning vid Stockholms universitet. Stockholm. 109 Svear in the Viking Age and Crusade Period (c. 800- – 137. 1200). In: J. Callmer, I. Gustin & M. Roslund (eds.), – 2000b. Forna färder över Östersjön – vikingars och Identity formation and Diversity in the Early Medieval andra. In: G. Hoppe (ed.) Öster om Östersjön = Ymer Baltic and Beyond. Communicators and Communica 2000. Visby. 9 – 23. tion. The Northern World 75. Kivikoski, E. 1938. Studien zu Birkas Handel im östli Schauman-Lönnqvist, M. 1996. The Vainionmäki so chen Ostseegebiet. Acta Archaeologica 1937 (8). 229 ciety. In: P. Purhonen (ed.) Vainionmäki – A – 250. Merovingian Period Cemetery in Laitila, Finland. – 1964. Finlands förhistoria. Helsingfors. Helsinki. 131 – 135.

53 New Aspects on Viking-age Urbanism, c. 750-1100

Schück, A. 1926. Det svenska stadsväsendets uppkomst Finska fornminnesföreningens tidskrift 1997 (104). och äldsta utveckling. Stockholm. Helsinki. Selling, D. 1955. Wikingerzeitliche und Frühmit telalterliche Keramik in Schweden. Stockholm. Sindbæk, S. 2007. Networks and nodal points. The (Endnotes) emergence of towns in Early Viking Age Scandina 1 Finland Proper (Varsinais-Suomi) and Satakunta in via. Antiquity 2007 (81). 119 – 132. south-west Finland, along with Häme (Tavastland), Stenberger, M. 1971. Sten, brons, järn. Stockholm. were the areas of ancient settlement (Edgren 2008:470). Steuer, H. 1987. Der Handel der Wikingerzeit zwischen The settlements were mainly located along the sea coast Nord- und Westeuropa aufgrund archäologischer and inland waterways (Lehtosalo-Hilander 1992:62). Quellen. In: K. Düwel & I. Jansson (eds.) Untersu When the terms “Finnish” or “Finland” refer to a pre- chungen zu Handel und Verkehr der vor- und frühges historic context in this text they designate the populat- chichtlichen Zeit in Mittel- und Nordeuropa. Teil IV: ed areas in these districts. Der Handel der Karolinger- und Wikingerzeit. Abhan dlungen der Akademie der Wissenschaften zu Göt 2 Recent published findings of this type of fire-steel tingen. Philologisch-Historische Klasse. Dritte Folge from Russia show that they are more frequent in North- 156. Göttingen. 113 – 197. ern Russia than previously assumed. The Russian finds Steuer, H., Stern, W.B. & Goldenberg, G. 2002. Der might lead to renewed discussions about the origin and Wechsel von der Münzgeld- zur Gewichtsgeld place of manufacture for fire-steels with bronze handles wirtschaft in Haithabu um 900 und die Herkunft in the form of mounted figures. Even if the production des Münzsilbers im 9. und 10. Jahrhundert. In: K. area for these fire-steels will continue to be matter of Brandt, M. Müller-Wille & Chr. Radtke (eds.) debate, it is still likely that the fire-steel brought to Bir- Haithabu und die frühe Stadt entwicklung im nördlic ka came via Finland and that it thus represents contacts hen Europa. Schriften des Archäologischen with this region Landesmuseums 8. Neumünster. 133 – 167. Stjerna, K. 1909. Lund och Birka. Historisk tidskrift för 3 A large number of the Birka graves were excavated by Skåneland: Skåne, , och Hjalmar Stolpe in the late 19th century. They are gen- 1909 (3). 171 – 225. erally designated by Bj and a number as in the works by Taavitsainen, J.-P. 1990. Ancient Hillforts of Finland: Holger Arbman who later published the finds (Arbman Problems of Analysis, Chronology and Interpretation 1940; 1943). The letters “Bj” refer to Björkö, the island with Special Reference to the Hillfort of Kuhmoinen = where Birka is situated. Finska Fornminnesföreningens Tidskrift 1990 (94). Helsinki. 4 It should be added that some of the Viking Age Scan- Talvio, T. 1996. The Coins. In: P. Purhonen (ed) Vain dinavian artefacts found in Ladoga-Karelia and in East- ionmäki – A Merovingian period Cemetery in Laitila, ern Finland were probably mediated through encoun- Finland. National Board of Antiquities. Helsinki. 51 – ters between Scandinavians and the population of the 52. Karelian isthmus in the area around Lake Ladoga and – 1999. Early coin finds from Finland. In: P. Purhonen via the trading site of Staraja Ladoga (Uino 1997:113f., (ed.) Fenno-Ugri et Slavi 1997. Cultural Contacts in 179ff.; Uino pers. com.) the Area of the Gulf of Finland in the 9th-13th Centu ries. Helsinki. 94 – 100. 5 Local forms of equal-armed brooches were produced in – 2002. Coin and coin finds in Finland AD 800-1200. Vakka-Suomi. These brooches were based on types im- Iskos 12. Helsingfors. ported from Central Sweden and had a vast distribution Tsigaridas Glörstad, Z. 2012. Sign of the Times? The in the populated areas of Finland (Edgren 1992:242). Transferand Transformation of Penannular Brooches in Viking-Age . Norwegian Archaeological Re 6 One such find derive from Staraja Ladoga, the trading view 2012 (45):1. 28 – 51. center close to Lake Ladoga. Uino, P. 1997. Ancient Karelia: Archaeological Studies = Suomen Muinaismuistoyhdistyksen aikakauskirja/

54 Hedeby’s demise in the late Viking age and the shift to Schlesvig

Hedeby’s demise in the late Viking age and the shift to Schlesvig

Volker Hilberg

At the end of the 10th or the beginning of the 11th cen- ry of his félagi Erik, who lost his life when drengs were tury, the famous Viking period emporium of Hedeby besieging Hedeby; and he was a steersman, a very good was engaged in military campaigns as evidenced by dreng.“ (Lund 1991:128). by two rune-stones (fig. 1) erected in memory of two high-ranking warriors of King Sven’s retinue. King It is not certain if these military events were contem- Sven – apparently the Danish king Sven Forkbeard porary or could already be dated to the year 983 AD, (987–1014) and not Sven Estridsen (1047–1074) when the Danes apparently shook off the Ottonian su- (Stoklund 2000, 112)1 – dedicated a rune stone c. 158 premacy after the death of Emperor Otto II. (Roesdahl cm in height to the memory of Skarthi, a warrior of his &Wilson 2006:226–227; Kalmring 2010:48). They entourage who had accompanied the king to the West, could also be related to later events around c. 991/1000, but who was finally killed at Hedeby: when a Swedish king named Erik Segersäll apparently went on campaigns into Denmark (Roesdahl & Wilson (side A) : suin : kunukR : sati :│‌‌‌‌‌stin : uftiR : skarþa│sin 2006:226–227; Sawyer & Sawyer 2002:190–191). If : himþiga : ias : uas :│: farin : uestr : iąn : nu : these disputes are not known from any other written record, the upheaval of Sven against his father Harald (side B) : uarþ : tauþr : at : hiþa : bu (DR 3). Bluetooth (c. 958–987) may have also caused military disputes (cf. Roesdahl & Wilson 2006:227). But it “King Sven erected this stone in memory of his hemþægi seems to be clear that these battles had their impact on Skarde, who had gone west but now lost his life at Hede- Hedeby. In a seminal paper, Johan Callmer emphasised by.“ (Lund 1991:128). the apparently dramatic transformation in the second half of the 10th century, which developed from a re- Another high-ranking warrior of the royal retinue cession to a collapse of an older trade system with its named Thorulf also erected a rune-stone for his com- maritime trading places, the emporia2, around the Baltic rade-in-arms Erik, who was a commander of a ship in Sea (Callmer 1994:72–73 fig. 8). In the 960s or 970s, the king’s fleet and who was killed when the warriors Birka declined rapidly as a trading place (Callmer 1994: besieged Hedeby: 72; cf. Arbman 1939:130–132; Roslund 2015:45) and Hedeby also suffered “a very notable recession in the 3 (side A) x þurlf x risþi x stin x þąnsi x│x himþigi x and 990s.“ (Callmer 1994, 72) (fig. 2). In the suins x eftiR x │erik x filaga x sin x ias x uarþ estuary, the emporium of Truso seemed to van- ish around 1000 (Brather & Jagodziński 2012:296– (side B) : tauþr x þą x trekiaR│satu x um x haiþa x 302), but the important harbour of Wolin in the estuary experienced the arrival of the new century (Hatz bu│x ian : han : uas : sturi : matr : tregR x │x harþa : kuþr 1974:169; Wehner 2007:29–31; Filipowiak & Konop- x (DR 1). ka 2008). ”Thorulf, Sven’s hemþægi, erected this stone in memo-

55 New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 1. Rune-stones ded- icated to the memory of Skarthi (DR 3, left) and Erik (DR 1, right), who were killed in action at Hedeby in the late 10th or early 11th century (Archive of the State Ar- chaeological Museum Schleswig, Hedeby exca- vations).

Fig. 2. Hedeby’s shrinking im- portance from c. 950 – c. 1050 according to Johan Callmer’s model. Left: AD 950–1000: “Reorganization of the Urban Network and Political Tur- moil“, right: AD 1000–1050: “Collapse of the Old System and the Establishment of the first New-Model Urban Centres“. Hedeby is marked in purple (J. Callmer 1994, fig. 7 + 8).

56 Hedeby’s demise in the late Viking age and the shift to Schlesvig

This development is especially connected with the nu- of Schleswig would develop in the future (Jankuhn mismatic record from both places and from the whole 1937:199). But in these early stages of Hedeby research, of Scandinavia and the Baltic area: In the Viking period, the archaeologists were understandably more interested vast amounts of Islamic silver coins, so-called dirhams, in the questions that led to the foundation and early reached Scandinavia and Eastern Europe where silver development of the emporium, especially to archaeo- or related metal ores were not to be found or mined at logically testify to its presence in the early 9th century that time (Noonan 1994; Kilger 2008; Brather 2010). AD before the supposed “Swedish rule“ (cf. Jankuhn After an immense increase in the inflow of silver com- 1949:12–13). In a summarising article on the results ing from Samanid Central Asia from the late 9th cen- and problems of his excavations published in 1949, tury onwards, this inflow of silver ceased in the second Jankuhn pointed out that Hedeby’s latest phase after half of the 10th century, connected with a reduction and c. 983 AD seems to be very difficult to trace due to a shrinking of the trading and contact networks (Spufford lack of knowledge of archaeologically relevant materials 1988:65–73; Noonan 2000/01: esp. 198–199 , 217; and due to sparse or contradictory historic documen- Leimus 2007; Roslund 2015). The supply of silver to tation like Erik’s and Skarthi’s rune-stones (Jankuhn Scandinavia was subsequently replaced by a reorienta- 1949: 85–86). Relating to the find material, he was of tion towards West European ore deposits and therefore the opinion of a slow regression, which could perhaps much shorter silver-routes – especially from a Danish be related to the situation in Birka at the same time. perspective (Spufford 1988:74–79; Steuer 2004). With the sudden death of Emperor Otto II. in 983 AD, At the end of the 10th century AD, the political sce- Hedeby again became Danish after the ”Swedish rule“ nario in Scandinavia and the adjacent areas around the which was superseded in 934 by an Ottonian suprema- Baltic also changed markedly, with the beginning of cy. He maintained that no minting took place in Hede- state formation with unified kingdoms connected with by during Sven Forkbeard’s reign and that Denmark’s the conversion political interests reorientated towards England, which and Christianisation of the former pagan kings and did not support the emporium’s role in the transit with their peoples. New forms of urbanism also developed the Baltic: As ambiguous as the last decades remained with civitates as episcopal sees. This process started in also Hedeby’s end in the middle of the 11th century. Denmark around 960/65 with the baptism of King The apparent economic decline of Hedeby in the late and lasted up to the second half of 10th century was further strengthened in Jankuhn’s sev- the 11th century when the episcopal sees obtained more eral post-war editions of his Hedeby book, where new and more accurate outlines ( 2007). It has been results from the 1960s excavations were also included speculated that the Danish royal dynasty, the Gorm (e.g. Jankuhn 1956:198–199, 236–238; 1986:184– family from Jelling, favoured new urban centres in the 185, 198, 222–223). Here he argued that the minting eastern part of their realm, where towns like Roskilde of coins in Hedeby – the so-called ”Hedeby half-bracte- and Lund were founded in the late 10th century (in ge- ates“ or Dorestad imitations Malmer KG 7–9 – ceased nerell Hybel & Poulsen 2007:230–235; for Roskilde in the 980s and was not renewed during Sven Fork- cf. Birkebæk 1992; Christensen & Andersen 2000; for beard’s reign. Furthermore Jankuhn referred to the al- Lund cf. Carelli 2012:43–72). These new towns could most total lack of single coin finds during these decades apparently be interpreted as royal foundations after in German or Anglo-Saxon mints, which are very com- Western European archetypes and not as mere copies of mon in Scandinavian coin hoards from this time: The trading places, showing the presence of political power Anglo-Saxon pennies struck during Aethelred II. reign and of the church and its administraion from the begin- (978–1016) and the so-called Otto-Adelheid-pen- ning (Callmer (1994:73). nies struck in the Harz mountains. While the former According to the influential opinion of Herbert seemed to be complete absent of the latter only one sin- Jankuhn, who led large-scale excavations in Hedeby gle specimen was known from Hedeby (cf. Wiechmann between 1930 and 1939 and who had the best knowl- 2007:210, 260 cat. no. 55). Jankuhn concluded that edge of the findings and features unearthed, Hedeby Hedeby suffered a heavy economic decline from 990 existed until c. 1050 AD (Jankuhn 1943:58, 86–88; AD onwards, and nearly lost its economic importance idem 1949:18, 85–86) and was transferred afterwards save for a minor resurgence during King Cnut’s reign to the northern coast of the Schlei fjord where the town (1016/1019–1035), as evidenced by a new coinage and

57 New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 3. Hedeby’s decline in the late Viking age. Synoptic table of old interpretations and new insights.

58 Hedeby’s demise in the late Viking age and the shift to Schlesvig in some single coin finds found in Hedeby. In the mid- landscapes of marshes and lowlands in the west and the dle of the 11th century, coin finds vanished and Hedeby Schlei fjord in the east, narrowing it down to a strip reached its end and its functions were finally transferred only 15 km wide. Hedeby itself is located far inland (22 on the northern side of the Schlei fjord, where the town nautical miles, c. 40 km) at a small inlet at the inner end of Schleswig continued to develop. According to Chris- of the Schlei, the so-called Haddebyer Noor. Two traf- tian Radtke, however, this process had already started fic routes ran through the Schleswig isthmus and could with Hedeby’s decline from the early 11th century on- be controlled and protected by the fortifications of the wards (Radtke 2007; 2009:151–156). A continuity of /Danewerk – Denmark’s southern border in the settlement and of the trading activities has further the Middle Ages (Andersen 1998, 2004; Dobat 2008; been suggested based on the pottery finds from both Tummuscheit 2012; Tummuscheit & Witte 2014) – Hedeby and Schleswig during the 11th century and and by the settlement of Hedeby itself. A north-south H. Lüdtke, taking into consideration that a decrease axis was formed by the Hærvej/Ochsenweg, which ran of settlement and trade in the middle of the 11th cen- overland from Viborg in the north down to the river tury is less plausible, developed a model of a gradual in the south and which can be traced back at least shift of functions from Hedeby to Schleswig (Lüdtke to the Bronze Age. Even more important to trade was 1985:131–138; 2013:88–90). A co-existence of both the east-west axis between the Baltic and the North Sea, towns in the early 11th century has been assumed un- from the Schlei in the east, over the land bridge and til Hedeby’s final abandonment ( 2009:191 note down the rivers and to the . 19). Beyond this, the topographical situation was geostrate- Albeit J. Callmer emphasized Hedeby’s decline gically characterized by its position within the border- during the 980s and 990s, and he referred to the pos- land between the Danes in the North, the Frisians to sibility that the youngest settlement layers could have the West in the Wadden Sea, the Saxons and the Frank- subsequently been destroyed and eroded (Callmer ish elite (which later became Germans) to the South 1994:72). Chr. Radtke observed a sudden breaking off and the Slavonic tribe of the Obodrites in the Eastern in the coinage after 983/85 as an intentional interven- part of Holstein (Kalmring 2010:27–40). tion of the Danish kings Harald Bluetooth and Sven This geostrategic position is also the cause for the in- Forkbeard connected with a basic economic restructur- terest of the Carolingian and Ottonian emperors in the ing of the place, which resulted in far-reaching structur- Schleswig isthmus and in Hedeby. With this interface, al changes in the settlement’s inner organization to the an access to the Baltic was possible, and, furthermore, it point of a topographic reorientation at another place facilitated a connection to the trading networks and the (Radtke 2002: 385–386). Thus, the settlement’s reloca- flow of commodities (especially silver) to and from the tion from the south to the north shore of the Schlei fjord East (Hilberg 2014:183– was equated with the general shift from Viking period embankment and a parcelling of the area. The first Hedeby to medieval Schleswig. But the discrepancy be- known wooden structure was established at the end tween the archaeological record (Meyer 1998:111) and of the first third of the 9th century as a solid access to historical interpretation (Radtke 2002:399; 405–406) a landing site. The hitherto existing beachmarket was has not yet been resolved convincingly. transferred on top of the wooden jetties, which were used as the emporium’s marketplace and facilitated a Why and how did Hedeby survive the demise of the mooring of the ships alongside. A considerable building emporia at the end of the 10th century, a demise which boom at the harbour could be observed in the and apparently led to the abandonment of the central em- (Kalmring 2010:451–461). At that time Hedeby porium of Birka? This question could eventually be was a vivid hub in long-distance trade and manufac- connected with Hedeby’s geostrategical position. Deci- turing of all different kind of commodities (cf. Hilberg sive for the significance and development of the ear- 2009; Sindbæk 2012:29–30; Hilberg & Kalmring ly medieval emporia is their favourable geostrategical 2014). location (McCormick 2007). Hedeby is situated at a With the expansion of Scandinavian activities in most favourable position for trade and communication the hereupon developing dominion of the Old Rus, the at the narrowest point of the Cimbrian peninsula on influences from the Centralasiatic part of the Islamic the Schleswig isthmus, restricted by the natural barrier Khalifate coming via the Baltic to Scandinavia strength-

59 New Aspects on Viking-age Urbanism, c. 750-1100 ened from the late 9th century (Noonan 2000/01) on- wards. The fur and slave trade with the Khalifate in- creased, for commercial transactions a new system of silver-weight economy by using scales and standardized weights developed, and large amounts of late Abbasid and Samanid dirhams reached the Baltic (Steuer 2014). From the onwards the Southern Scandinavian bor- derland with Hedeby came to the fore of the East Frank- ish Kingdom of the Ottonian kings and emperors – even if the precise events could not be reconstructed reliably enough (Hoffmann 1984; Sawyer & Sawyer 2002:171– 173, 179–185; Stieldorf 2012:109–112, 432, 442, 487; different Hybel 2013). Through this pressure from the Ottonian Empire and the connected impact on their Northern neighbours, not only were forms of tribu- tary dependency or the establishment of the first bish- oprics in Denmark in 948 implemented, but also the state formation under the Jelling dynasty was gradually developed (Keller & Althoff 2008:178–180, 230–239; Bagge 2009:esp. 148). In AD 982, the failed campaign of Emperor Otto II in , at which several office-hold- ers also from the northern borderlands participated, and the disastrous defeat at led with Otto’s dead in AD 983 and the following turbulences in the succession to the to a spacious uprising of the western Slavs and also the Danes (Keller & Althoff 2008:262–279). These alterations during the second half of the 10th cen- tury also had their impact on Hedeby. The emporium Fig. 4. Hedeby 1967, central excavation area. Irregular and apparently was fortified by a semicircular rampart and the settle- ploughed-up packages of stones immediately underneath the surface. ment activities were concentrated inside.

As described preliminarily, this development did not (1) For a long time it has been argued that Hedeby’s changed abruptly from the 980s onwards. Hedeby also latest dendrodate from 1020 AD could be related succeeded in performing its role as a prominent centre of to its decline and its abandonment of settlement production and long-distance trade in the first half of the activities (Jankuhn 1986:223; Radtke 2009:151). th 11 century (fig . 3). Changes might have occurred very This dendrodate belongs to a wooden well (Schiet- well, but not all elements could be traced sufficiently due zel 1981:68–69; Schietzel 2014: 164–165), which to the preservation of archaeologically relevant settlement was dug into the ground and thereby in deeper wa- features and layers. The following remarks will deal with terlogged layers. In general the layers with wooden (1) the late settlement structures and (2) the use of the preservation could be dated to the 9th century (Schul- harbour and considers also (3) the problem of the young- tze 2008:68), normally the younger settlement layers est tombs and the abandonment of the graveyards. It then are not preserved in wood as they are too high, only discusses (4 ) Hedeby as a Late Viking period mint before wells or the of the stream running through the vast amount of (5) late Viking period single finds is Hedeby’s settlement core were put into the ground presented and finally (6) the archaeological evidence of and could be dendrodated into the 10th or 11th cen- trade and especially long-distance trade with a focus on turies (Schultze 2008:64–68; Schultze in von Car- th 11 century coins and standardized weights, summing up nap-Bornheim, Hilberg & Schultze 2014:236). The Hedeby’s functions until its abandonment in the middle upper and younger layers are decayed, in some parts of the . of the central settlement core, irregular and apparent-

60 Hedeby’s demise in the late Viking age and the shift to Schlesvig

Fig. 5. Hedeby 1908, sections of trial excavation trenches. The upper layer of 50-60 cm depth is nominated as “black earth” (“Schwarze Erde”). Scale 1:40.

ly ploughed-up packages of stones are found imme- an increasing draught of large cargo vessels in con- diately underneath the surface (fig. 4). The upper part nection with a further siltation of Hedeby’s harbour of the layers has been also denominated as Hedeby’s was one reason why the harbour operation became “black earth” consisting of decayed settlement layers increasingly impossible, and the harbour and its sur- (fig. 5). rounding town shifted finally to the north shore of the Schlei fjord (e.g. Sawyer & Sawyer 2002:273). (2) The youngest building activities in the harbour could This argument, however, has been invalidated via a be traced to the early 11th century, more precisely thorough evaluation of Hedeby’s harbour excavation only relatively to around 990 to 1010 according to (Kalmring 2010) and an examination of the large the sunken wreck 1 and the superposition- cargo ships from medieval Danish waters between c. ing of a younger wooden jetty (Kalmring 2010:237– 1000–1250 (Englert 2015). The estimated and cal- 239, 242–243, 454 fig. 324). In the harbour the culated draughts of the 11 known cargo ships over largest hitherto known trading vessel of the knorr- 14 m in length analysed by Anton Englert show that type also sunk, which was a large cargo ship built these vessels like Hedeby 3 (t.p.q. 1023), Roskilde 4 after 1023 with a calculated water displacement of (t.p.q. 1108/13) and Lynæs 1 (t.p.q. 1140) do not 75 tons and a cargo capacity of c. 60 tons (Kalm- exceed a draught of c. 1.5 m, for smaller vessels the ring 2010:122–125; Englert 2015:60 fig. 4:8). The draught is even smaller (Englert 2015: 277–278 fig. late use of the harbour and its facilities further ex- 9:5). In Hedeby’s harbour, the wooden harbour jet- plains the high amount of late Viking period pottery, ties were prolonged successfully to fit these demands especially of the Baltic ware, as observed by Lüdtke with another building phase dated “before between (Lüdtke 2013:60, 88–90). And argument has been 990 and 1010” AD (Kalmring 2010:242, 358, 454, put forth that the development in shipbuilding with fig. 275, 324; jetties structures IV.2, III.3 and V.2, cf.

61 New Aspects on Viking-age Urbanism, c. 750-1100

fig. 182). But now in the 11th century, these long and enlarged harbour jetties built in oak – the southern jetty was with c. 54 m the longest jetty, whereof c. 44 m were built into the water – needed to be main- tained furthermore and regularly to facilitate any har- bour operation! (3) With the beginning – when Harald Bluetooth converted in c. 960/65 AD - grave goods vanished from burials in Denmark, which hither- to made a dating of the interment possible (Eisen- schmidt 2004:137–138; Arents & Eisenschmidt 2010 [vol. 1]: 297–298; Pedersen 2014:172 –173). The youngest datable graves from Hedeby’s ceme- teries do exist for the late 10th century, it has been assumed that they do not exist for the 11th centu- ry (Arents & Eisenschmidt 2010: 297–298). In the large cemetery inside the rampart, one of the latest datable graves is grave 33 with a Otto-Adel- heid penny, which was struck between c. 990/95 and 1025/30 AD (Wiechmann 2007:210; 221; cat. no. 55). Also two deniers occur in this area: a Fig. 6. Hedeby 1905, exacavation cemetery. In black are marked the penny from the Salian emperor Konrad II. (1024– graves 45 and 44, both bottoms in a depth of c. 1 m. Above marked in 1039 AD) and a penny struck between c. 1055– yellow skeletal remains, iron nails and a coin (cf. arrow) of lower lying 1065 from the Norwegian king Harald Hardråde younger graves, which are ploughed up. Scale 1:40. (1047–1066 AD) (Wiechmann 2007:221, cat. no. 53, 128). They might derive from 11th century in- humations, which had been ploughed up later and (4) One of the most important arguments for Hede- th destroyed and according to their slight depth, they by’s decline already in the late 10 century was the are only documented by some fragmented bones breaking off of an own coinage with the end of the and iron nails (fig. 6)5. Besides both pennies could imitative Dorestad-coinage of group Malmer KG not be associated with younger settlement features 9 in the 980s AD, its minting had been attributed or materials, which are not yet known in this area. to Hedeby due to stylistic arguments based on the preponderance of this imitative Dorestad-coinage from the early 9th century onwards in Hedeby and

Fig. 7. Deniers of types Årstad 95 (left) and 96 (right), struck between c. AD 1010/15–1040 presumably in Hedeby. Wiechmann 2013: fig. 6+7.

62 Hedeby’s demise in the late Viking age and the shift to Schlesvig

Fig. 8. Hedeby, metal-detected small finds of the late Viking period and coins struck after c. AD 983.

its hinterland (cf. Jankuhn 1986:184–185, 198, 2013:145), where coins of this type were accepted in 222–223; Wiechmann 2007:185–188 fig. 1b; Hil- transactions besides the traditional Dorestad-coinage. berg 2011:210–214, fig. 10:1). Already in the middle With Harald Bluetooth’s death in 986 AD the cross of the 970s another Danish coinage was rapidly in- coinage was quickly vanishing again, and with his son troduced, Malmer KG 10 + 11 with the depiction of and successor Sven Forkbeard the minting of a na- a cross on one side, presumably struck by the already tional coinage started (Jonsson 1994:223–230). But Christianized king Harald Bluetooth. Coins of this before King Cnut, who became king in Denmark in type were not known in Hedeby and its surroundings 1019, Hedeby is not attested as a mint. A small series for a long time, but they were well known in hoard of light deniers weighing c. 0.75 g with blundered in- finds especially in Skåne and on Bornholm, and it is scriptions may have been minted there from 1026/28 assumed they were minted in Lund (cf. Wiechmann onwards (Jonsson 1994:223–227 fig. 11:8, 11). 2007:187; Moesgaard 2012:114–117). But newly found cross coins from settlement contexts in Recently, due to his analysis of the whole group, Ralf and Hedeby have changed this distribution and led to Wiechmann (2013) has argued that another imitative the assumption that the new cross types may have also cross coinage, named Årstad 95+96 after a Norwegian been struck in Hedeby (Hilberg & Moesgaard 2010; hoard find (Ilisch 1988), being stylistically related to Moesgaard 2012: 116–117, 128–132; Wiechmann Anglo-Scandinavian imitations and the last cross coin

63 New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 9. Chronological distribution of late Viking period coin finds from Hedeby struck after c. AD 983.

group KG 13 (struck from c. 995 onwards), could be enough precision, reinforced by their character as related with great probability to Hedeby as their mint.. being single finds from the ploughzone without any This coinage shows a blundered cross design on both further datable context. It is impossible to say if sev- sides, combining both Byzantine and Late Anglo-Saxon eral objects were used only early in the 11th century patterns (fig. 7). The average weight amounts to c. 0.6– or even later (e.g. horse harness fittings, continental 0.7 g and follows a South Scandinavian standard known brooches). The most recent assessment was rejected in Jutland, apparently the mint had a low degree of or- given that the lateViking period findings of the 11th ganisation due to quite a simple structure of the coinage. century could only be linked with a change in func- The minting of these coins could have started around tion of Hedeby from a centre of long-distance trade 1010/15 and lasted to c. 1030 or even later according towards a base of royal power.. The for to the known hoard finds. The only settlement finds are Skarthi and Erik, written records about a deploy- known from Hedeby, were the corpus of known coins ment of long ships with their naval forces after the has grown to 8 specimens of Årstad 95 and 6 specimens Danish conquest of England in 1013 AD, but also of Årstad 96. the new archaeological findings consisting mainly of equipment have been interpreted in this (5) Through the recent fieldwork, the picture of Hedeby’s direction (Radtke 2009:151–157; Lüdtke 2013:89; late phase has changed fundamentally. Due to sys- Kleingärtner 2014:180 with note 389). Indeed, the tematic metaldetecting, a new horizon of late Viking 11th century saw a significant increase in the use of period findings has emerged on a larger scale (fig. 8). riding equipment (Graham-Campbell 1991; Peder- A methodological problem seems to be the fact that sen 1996/97; Williams 1997; Hinton 2008:82–83), these small finds quite often could not be dated with but it was used more for the elite’s prestige and not

64 Hedeby’s demise in the late Viking age and the shift to Schlesvig

Fig. 10. Hedeby, distribution of Ottp-Adelheid-Pennies, types Hz III-IV, c. AD 990–1030/35.

Fig. 11. Hedeby, metal-detected truncated spheres accoring to Heiko Steuer’s . 65 New Aspects on Viking-age Urbanism, c. 750-1100

for a mounted combat (Bartlett 1996:106; Williams century, not only Otto-Adelheid-pennies or other coins 2008:111–112; Juraske 2010:403–404; Hjardar & from , but also larger amounts of coins from Vike 2012:87–89). Dress accessories like brooches Cologne and the lower reached and were used (of Scandinavian manufacture and also as Conti- and lost in Hedeby. They refer to the trade relations of nental imports), half fabricated artefacts and moulds the via the metropolis Cologne to Southern have also been registered. Scandinavia (fig. 9). Just as these coins are distribut- ed over Hedeby’s entire settlement area, there is also an The most important argument that Hedeby still exist- occurrence of equally widespread standardised weights ed as a centre of long-distance trade is the increasing demonstrating the continuing economic role of the evidence of trade visible in the loss of coins and stan- emporium in the 11th century: from 147 metaldetected dardized weights during the course of the 11th century. truncated spheres (“Kugelzonengewichte”), 91 speci- Whereas until 2002 at least 25 coin finds, which were mens could be identified and assigned to Heiko Steuer’s struck after 983 AD, were known, the amount of late typology (Steuer 1997:44–52, 285–320, esp. fig. 15, coins has significantly risen through metaldetecting 232; Steuer 2012). Here the overwhelming number be- (63 specimens until 2014). A smaller scale excavation long to Steuer’s types B1 middle, B2, B1 late, B3 and between 2005–2010 (4 specimens) revealed altogether B4 (fig. 11). Without any datable context, these weights 92 coins of the late Viking period after 983 AD (fig. do not permit a preciser chronological classification, 9), exceeding the amount of coin finds in the new ur- which is a phenomenon quite often observed with met- ban civitates like Lund and Roskilde respectively. Af- aldetected objects. But there could be no doubt left that ter the massive influx of Islamic coinage in the Baltic Hedeby is quite economically active in the 11th century. and to Scandinavia, West European deniers dominate the hoards from the 970s onwards (Leimus 2007) and From the 960s and 970s onwards, long-term growth en- the import of coinage from late Anglo-Saxon England compassed all areas of life and led to a new economic and the East Frankish-German realm of the Ottonian wealth in West and . This has been dis- kings and emperors is dated traditionally to the same cussed and pointed out in relation to the increasing use years (Hatz 1974:47–51; Jonsson 1990; Metcalf 1998; in the circulation of silver from new deposits (i. e. the Moesgaard 2006). But Danish hoards like Grisebjerg- Harz mountains), which led to a massive rise in the vol- gård (t.p.q. 942/3), Jyndevad (t.p.q. 954/5) and argu- ume of trade (Steuer 2004:138–140; Keller & Althoff ably also Terslev (t.p.q. 940) containing early deniers 2008:399–407; Sawyer 2013:98–110). from Saxony bearing the name OTTO or ODDO, so- Due to the results from new fieldwork, Hedeby also called “Sachsenpfennige”, show that this development participated in this development. The increasing ar- had already started in the middle of the 10th century chaeological testimony indicates more and more that (Jonsson 1990:139–141; Kilger & Moesgaard 2008). the model of a gradual shift of functions from Hedeby Consequently, this was before the early “Sachsenpfen- to Schleswig already in the first half of the 11th century nige” spread across the West Slavonic area in the Baltic (Radtke 2010:157; Lüdtke 2013:89–90) was not appli- and Sweden (Hatz 1985) around 970. One of these ear- cable. Hedeby’s function as a node of transit – its over- ly “Sachsenpfennige” with a fence-like arrangement of whelming position in the merchandise traffic between crosses under the church of the coin’s backside is also the North and Baltic Sea resp. (between Scandinavia known from a settlement context in Hedeby (Wiech- and the Continent) – may have been pursued and main- mann 2007:260 cat. no. 57). With the coinage of the tained without interruption by the civitas of Schleswig prolific Otto-Adelheid-pennies struck in the Harz area as its successor planned the shift during the reign of after 983 (Ilisch 2013), these coins come in large quan- King Svend Estridsen (1047–74), apparently around tities to Scandinavia and towards Hedeby, where they the year 1066. were in use until c. 1030/35 (Rundberg 2000:9–19). In the first half of the 11th century, Hedeby’s economic Alongside two already published specimens (Wiech- relations and its function as a main node of transit with mann 2007:260 cat. no. 54–55) 19 Otto-Adelheid-pen- Continental Europe may have flourished, according nies have been further registered in Hedeby in recent to the existing coinages especially with centres like Co- years from the last excavation’s metaldetecting between logne or the Harz area via , or Magde- 2005–10 (fig. 10). In the first four decades of the th11 burg. This vital role is documented in the high amount

66 Hedeby’s demise in the late Viking age and the shift to Schlesvig of coins, weights or small metalfinds. It is completed e.g. urban centres took place (cf. Keller & Althoff 2008:417– through the several glass vessel sherds of the Saint Sav- 428). The Christian civitas had to fit the needs of the in-type from Hedeby’s settlement (Steppuhn 1998:61– king and the bishop and their retinues. During the 11th 62 pl. 12:1–5; Simon-Hiernard & Gratuze 2011; century, the increasing professionalisation of long-dis- Baumgartner & Marti 2013). These could be regarded tance trade also developed new demands and required at- as a testimony for luxurious tableware coming via France tractive harbour facilities (Englert 2015:287–290). These or Germany, likely fom Italian workshops, and used for may have also been the reasons for the Danish king to private dining and public feasting in late Hedeby6. In dislocate the old and outworn emporium or portus Hede- late Viking period Scandinavia, this luxurious glass vessel by to the newly spreading civitas of Schleswig by transfer- ware is only known from the civitas of Sigtuna (Henric- ring its harbour – and the tolls and levies. son 2006:43 fig. 7 right). Despite the high amount of recently registered small Acknowledgements 11th century finds from Hedeby’s ploughzone, it is still This study was thankfully enabled through a research difficult to connect them with a more exact sequence of project of the Volkswagen Foundation in Hannover, Hedeby’s development because their manufacture and which analysed Hedeby’s final stage and the beginning loss could not be dated precisely enough. Regarding of early Schleswig in a common project of the Archae- the coin finds (fig. 9), it could be stated that only a ological Museum Schleswig and University from small number of coins date after c. 1050 AD. The lack 2012–14. of silver coins from East Frankish or German mints in Hedeby decreased distinctly from the onwards. Corresponding Author Peter Spufford (1988:95–96) has posited that the sup- ply of silver from the Harz was quite short-lived and Volker Hilberg could be dated between c. 991 and 1040. Afterwards Archäologisches Landesmuseum, Stiftung there was a rapid decline7 connected with the impover- Schleswig-Holsteinische Landesmuseen, Schloss ishment of Saxony in the second half of the 11th century Gottorf Schleswig, Germany and a lack of silver also in the areas to the south of Sax- [email protected] ony, which even reached formerly economically strong areas like the Rhineland and the valley. But in his monographic analysis of the German coins from hoards in Sweden, Gert Hatz (1974:185–186; 1987:91; cf. also Mäkeler 2005) pointed out that German coins reached the Baltic until c. 1125 and that the decreas- ing inflow was connected with the development of a national coinage esp. in Denmark and Norway. This increasingly replaced the foreign currency and started the regionalisation of the German economy and coin- age, which fit more with the needs of a regional market. Such a dramatic economic decline caused by a short- age of the silver supply from the Harz ores as stated by Spufford is observed by neither the historical nor the recently intensified mining archaeological research for the Harz mountains (Hauptmeyer 1997:1133–1137; Alper 2003:19–27; Bartels 2008; Bartels & Klappauf 2008:121–155, 164–179; Klappauf et al. 2008; Klap- pauf 2011; Asmus 2012:105–145). The shift from Hedeby to Schleswig after the recorded destructions from 1050 and 1066 was not only provoked by these events. Already from the 10th century onwards, new developments in other demands and perceptions of

67 New Aspects on Viking-age Urbanism, c. 750-1100

Bompaire, M. 1997. Généralisation et diversification de References l’essor économique (930-1180). V. Les problèmes Alper, G. 2003. „Johanneser Kurhaus“. Ein mittelalterli monétaires. In: Ph. Contamine (ed.) L’économie mé cher Blei-/Silbergewinnungsplatz bei Clausthal-Zeller diévale. Collection U. Histoire médiévale 361. Paris. feld im Oberharz. Materialhefte zur Ur- und Frühge 196 – 208. schichte Niederschsens 32. Rahden/Westf.. Brather, S. 2010. Silver, weights and scales around the Andersen, H.H. 1998. Danevirke og Kovirke. Arkæologi baltic, 8th to 11th centuries. In: B. Ludowici, H. ske undersøgelser 1861-1993. Højbjerg. Jöns, S. Kleingärtner, J. Scheschkewitz & M. Hardt – 2004. Til hele rigets Værn. Danevirkes arkæologi og hi (eds.) Trade and Communication Networks of the first storie. Højbjerg. Millennium AD in the northern part of Central Eu Arbman, H. 1939: Birka. Sveriges äldsta Handelsstad. rope: Central Places, Beach Markets, Landing Places Från Forntid och Medeltid I. Stockholm. and Trading Centers. Neue Studien zur Sachsenfor Arents, U. & Eisenschmidt, S. 2010: Die Gräber von schung 1: Hannover. 143– 164. Haitha bu. Die Ausgrabungen in Haithabu 15. Neu Brather, S. & Jagodziński, M.F. 2012. Der wikingerzeit münster. liche Seehandelsplatz von Janów (Truso). Geophysikali Asmus, B. 2012: Medieval Copper Smelting in the Harz sche, archäopedologische und archäologische Untersu Mountains, Germany. Montanregion Harz 10. Bo chungen 2004–2008. Zeitschrift für Archäologie des . Mittelalters. Beiheft 24. . Bagge, S. 2009: Early state formation in Scandinavia. Callmer, J. 1994. Urbanization in Scandinavia and the In: W. Pohl & V. Wieser (eds.) Der frühmittelalterli Baltic Region c. AD 700 – 1100: Trading Places, che Staat – europäische Perspektiven. Forschungen zur Centres and Early Urban Sites. In: B. Ambrosiani & Geschichte des Mittelalters 16. Wien. 145 – 154. H. Clarke (eds.) Developments around the Baltic and Bartels, Chr. 2008. Bergbau und Münzstätten im Harz. North Sea in the Viking Age. The Twelfth Viking Abhandlungen der Braunschweigischen Wissenschaftli Congress = Birka Studies 3. Stockholm. 50 – 90. chen Gesellschaft 2008 (60). 23 – 50. Carelli, P. 2012: Historia – staden och omlandet. Bartels, Chr. & Klappauf, L. 2008. Das Mittelalter. Der 1. Medeltiden. En metropol växer fram (990–1536). Aufschwung des Bergbaus unter den karolingische Lund. und ottonischen Herrschern, die mittelalterliche von Carnap-Bornheim, C., Hilberg, V. & Schultze, J. Blüte und der Abschwung bis zur Mitte des 14. Jahr 2014. Research in Hedeby. Obligations and Respon hunderts. In: Chr. Bartels & R. Slotta (eds.) Der alt sibilities. In: C. von Carnap-Bornheim (ed.) Quo europäische Bergbau.Von den Anfängen bis zur Mitte Vadis? Status quo and Future Perspectives of Long-term des 18. Jahrhunderts. Geschichte des deutschen Berg Excavations in Europe. Schriften des Archäologischen baus 1. Münster. 111 – 248. Landesmuseums. Ergänzungsreihe 10. Hamburg & Bartlett, R. 1996. Die Geburt Europas aus dem Geist der Neumünster: 225 – 248. Gewalt. Eroberung, Kolonisierung und kultureller Christensen, T. & Andersen, M. (red.) 2000. Civitas Wandel von 950 bis 1350. München. Roscald – fra byens begyndelse. Roskilde. Baumgartner, E. & Marti, R. 2013. Glasgeschirr. In: Dobat, A.S. 2008. Danevirke revisited: An investiga R. Marti, W. Meyer & J. Obrecht (eds.) Der Alten tion into military and socio-political organisation in berg bei Füllinsdorf. Eine Adelsburg des 11. Jahrhun South Scandinavia (c AD 700 to 1100). Medieval Ar derts. Schriften der Archäologie Baselland 50. . chaeology 2008 (52): 27 – 67. 181 – 186. DR = Jacobsen, L. & Moltke, E. 1942: Danmarks Ru Birkebæk, F.A. 1992. Fra handelsplads til metropol neindskrifter. København. 950–1080. In: F.A. Birkebæk, E. Verwohlt & M. Eisenschmidt, S. 2004. Kristendommen indtrængen i Høj (red.) Roskilde bys historie – tiden indtil 1536. Syddanmark. In: N. Lund (red.) Kristendommen i Roskilde. 37 – 119. Danmark før 1050. Et symposium i Roskilde den Bolton, T. 2009. The Empire of . Conquest 5.–7. februar 2003. Roskilde. 123 – 141. and the Consolidation of Power in Northern Europe in Englert, A. 2015. Large Cargo Ships in Danish Waters the Early Eleventh Century. The Northern World 40. 1000-1250. Evidence of specialised merchant seafaring & Boston. prior to the

68 Hedeby’s demise in the late Viking age and the shift to Schlesvig

Hanseatic Period. Ships and Boats of the North 7. The Baltic Sea region in the early Viking Age as seen Roskilde. from shipboard. Maritime Culture of the North 2. Filipowiak, W. & Konopka, M. 2008. The identity of a Roskilde. 79 – 113. town: Wolin, town-state – 9th–12th centuries. Quaes – 2011. Silver economies of the 9th and 10th centuries tiones Medii Aevi Novae 2008 (13). 243 – 288. AD in Hedeby. In: J. Graham-Campbell, S.M. Sind Gelting, M. 2007. The kingdom of Denmark. In: N. bæk & G. Williams (eds.) Silver Economies, Moneti Berend (ed.) Christianization and the Rise of Monar sation & Society in Scandinavia, 800-1100. . chy. Scandinavia, Central Europe and Rus’ c. 900– 203 – 225. 1200. Cambridge. 73–120 – 2014. Zwischen Innovation und Tradition. Der ka Gilomen, H.-J. 2014. Wirtschaftsgeschichte des Mit rolingische Einfluss auf das Münzwesen in Skandina telalters. München. vien. In: K. Hofmann, H. Kamp & M. Wemhoff Graham-Campbell, J. 1991. Anglo-Scandinavian (eds.) Die Wikinger und das Fränkische . Identi equestrian equipment in 11th century England. Ang täten zwischen Konfrontation und Annäherung. Mit lo-Norman Studies 1991 (14). 77 – 89. telalter Studien 29. . 133 – 215. Hatz, G. 1974. Handel und Verkehr zwischen dem Deut Hilberg, V. & Kalmring, S. 2014. Viking Age Hedeby schen Reich und Schweden in der späten Wikingerzeit. and Its Relations with and the North Atlan Die deutschen Münzen des 10. und 11. Jahrhunderts tic: Communication, Long-distance Trade and Pro in Schweden. Kungl. Vitterhets Historie och An duction. In: D. Zori & J. Byock (eds.) Viking Age tikvitets Akademien. Lund. Archaeology in Iceland: Mosfell Archaeological Project. – 1985. Die ersten Sachsenpfennige in Schweden. In: S. Cursor Mundi 20. Turnhout. 221 – 245. Kuczyński & S. Suchodolski (eds.) Nummus et Histo Hilberg, V. & Moesgaard, J.Chr. 2010. Opsigtsvækken ria. Pieniądz Europy średniowiecznej [Festschrift de fund af korsmønter. Er hypotesen om ”Tyskervæl Ryszard Kiersnowski]: 33–42, pl. I. Warszawa: Pol det” i Hedeby 974-983 forkert? Nordisk Numisma skie Towarzyst wo Archeologiczne i Numizmatyczne, tisk Unions Medlemsblad 2010 (4). 143 – 150. Komisja Numizmatyczna. Hinton, D.A. 2008. The Alfred Jewel and Other Late An – 1987. Der Handel in der späten Wikingerzeit zwi glo-Saxon Decorated Metalwork. Ashmolean Hand schen Nordeuropa (insbesondere Schweden) und books. Oxford. dem Deutschen Reich nach numismatischen Quel Hjardar, K. & Vike, V. 2012. Vikingar i krig. Stock len. In: K. Düwel, H. Jankuhn, H. Siems & D. holm. Timpe (eds.) Der Handel der Karolinger- und Wikin Hodges, R. 2008. Fifty years after Dunning: reflections gerzeit. Untersuchungen zu Handel und Verkehr der on emporia, their origin and development. In: T. vor- und frühgeschichtlichen Zeit in Mittel- und Abramson (ed.) Two Decades of Discovery. Studies in Nordeuropa IV. Göttingen. 86 – 112. Early Medieval Coinage 1. Woodbridge. 113 – 118. Hauptmeyer, C.-H. 1997. Niedersächsische Wirt Hoffmann, E. 1984. Beiträge zur Geschichte der Be schafts- und Sozialgeschichte im hohen und späten ziehungen zwischen dem deutschen und dem däni Mittelalter (1000–1500). In: E. Schubert (ed.) Ge schen Reich für die Zeit von 934 bis 1035. In: Chr. schichte Niedersachsens 2:1. Politik, Verfassung, Wirt Radtke & W. Körber (eds.) 850 Jahre St. –Petri–Dom schaft vom 9. bis zum ausgehenden 15. Jahrhundert. zu Schleswig 1134–1984. Schleswig. 105–132. Hannover. 1039 – 1319. Hybel, N. 2013. Marca und Feudum – Das dänische Henricson, L.G. 2006. Hålglas i Sigtuna – 300-tal till Königtum und der Kaiser, 842–1214. In: St.B. 1900-tal. Situne Die 2006. 37 – 53. Frandsen, M. Krieger & F. Lubowitz (eds.) 1200 Jah Herrmann, J. 1994/95. Frühe Seehandelsplätze am “äu re deutsch-dänische Grenze. Neumünster. 95 – 109. ßersten Ende des westlichen Ozeans”. Geschichtliche Hybel, N. & Poulsen, B. 2007. The Danish Resources c. Grundlagen, siedlungstopographische Strukturen 1000–1550. Growth and Recession. The Northern und ethnische Herkunft ihrer Bewohner. Acta Prae World 34. Leiden & Boston. historica et Archaeologica 1994/95 (26/27). 57 – 72. Ilisch, P. 1988: Tysk, nordisk eller slavisk? En gruppe Hilberg, V. 2009. Hedeby in Wulfstan’s days: a Danish møn timitationer fra 1000-tallet. Nordisk Numisma emporium of the viking Age between East and West. tisk Unions Medlemsblad 1988 (1). 9 – 14. In: A. Englert & A. Trakadas (eds.) Wulfstan’s Voyage. – 2013. De tidligste Otto Adelheid-mønter – foreløbige

69 New Aspects on Viking-age Urbanism, c. 750-1100

bemærkninger på grundlag af Strandby-skatten. Nor 9.–11. 9. 2010. Arbeits- und Forschungsberichte zur disk Numismatisk Unions Medlemblad 2013 (4). 100 sächsischen Boden denkmalpflege Beiheft 22. Dres – 107. den. 169 – 178. Jankuhn, H. 1937. Haithabu, eine germanische Stadt der Klappauf, L., Bartels, Chr., Linke, F.-A. & Asmus, B. Frühzeit. First edition. Neumünster. 2008. Das Montanwesen am und im – 1943. Die Ausgrabungen in Haithabu (1937–1939). Westharz. Historische und archäologische Quellen z Vorläuiger Grabungsbericht. . um 12. und 13. Jahrhundert. In: M. Brandt (ed.) – 1949. Ergebnisse und Probleme der Haithabugrabun Bild und Bestie. Hildesheimer Bronzen der Stauferzeit. gen 1930–1939. Zeitschrift der Gesellschaft für Schles Wiesbaden. 65 – 76. wig-Holsteinische Geschichte 1949 (73). 1 – 86. Kleingärtner, S. 2014: Die frühe Phase der Urbanisie – 1956. Haithabu, ein Handelsplatz der Wikingerzeit. rung ander südlichen Ostseeküste im ersten nachchrist Third edition. Neumünster. lichen Jahrtausend. Studien zur Siedlungsgeschichte – 1986. Haithabu, ein Handelsplatz der Wikingerzeit. und Archäologie der Ostseegebiete 13. Neumünster. Eighth edition. Neumünster. Leimus, I. 2007. Die letzte Welle des orientalischen Jonsson, K. 1990. The import of German coins to Den Münzsilbers im Norden. In: M. Andersen, H.W. mark and Sweden c. 920–990. In: K. Jonsson & B. Horsnæs & J.Chr. Moesgaard (eds.) Magister Mone Malmer (eds.) Sigtuna Papers. Proceedings of the tae. Studies in Honour of Jørgen Steen Jensen. Publica Sigtuna Symposium on Viking-Age Coinage 1–4 tions of the National Museum. Studies in Archaeolo June 1989. Commentationes de Nummis Saecu gy and History 13. Copenhagen. 111 – 125. lorum IX-XI in Suecia repertis.Nova Series 6. Stock Loveluck, Chr. 2013. Northwest Europe in the Early holm.139 – 143. Middle Ages, c. AD 600–1150. A Comparative Archa – 1994. The coinage of Cnut. In: A.R. Rumble (ed.) eology. Cambridge. The Reign of Cnut: King of England, Denmark and Lüdtke, H. 1985. Die mittelalterliche Keramik von Norway. London. 193 – 230. Schleswig Ausgrabung Schild 1971–1975. Ausgrabun Juraske, A. 2010. Steigbügel und Pflug. Das Zeitalter gen in Schleswig. Berichte und Studien 4. Neumüns der Panzerreiter du die Etablierung des Feudalismus. ter. In: Chr. Kaindel & A. Obenaus (eds.) Krieg im mit – 2013. Die einheimische Keramik aus dem Hafen von telalterlichen Abendland. Krieg und Gesellschaft 1. Haithabu. Die Ausgrabungen in Haithabu 17. Neu Wien. 393 – 410. münster. Kalmring, S. 2010. Der Hafen von Haithabu. Ausgra Lund, N. 1991. The Danish Perspective. In: D. Scragg bungen in Haithabu 14. Neumünster. (ed.) The AD 991. Oxford. 114 – Keller, H. & Althoff, G. 2008. Die Zeit der späten Ka 142. rolinger und der Ottonen 888–1024. Gebhardt Hand McCormick, M. 2007. Where do trading towns come buch der deutschen Geschichte 3. . from? Early medieval and the northern empo Kilger, Chr. 2008. Kaupang from Afar: Aspects of the ria. In: J. Henning (ed.) The Heirs of the Roman West. Interpretation of Dirham Finds in Northern and Post-Roman Towns and Trade in Europe, Byzantium Eastern Europe between the Late 8th and early 10th and the Near-East 1. Millenium-Studien zu Kultur Centuries. In: D. Skre (ed.) Means of Exchange. Deal und Geschichte des ersten Jahrtausends n.Chr. 5: ing with Silver in the Viking Age. Kaupang Excavation Mäkeler, H. 2005. Wikingerzeitlicher Geldumlauf Project Publication Series 2. Norske Oldfunn XXIII. im Ostseeraum – Neue Perspektiven. Quaestiones Aarhus. 199 – 252. Medii Aevi Novae 2005 (10). 121 – 149. Kilger, Chr. & Moesgaard, J.Chr. 2008: Sachsenpfen Metcalf, D.M. 1998. Viking-Age Numismatics 4. The nig i Grisebjerggårdfyndet (tpq 942/3) – En ny pus currency of German and Anglo-Saxon coins in the selbit i den tyska mynthistorien. Nordisk Numisma Northern Lands. The Numismatic Chronicle 1998 tisk Unions Medlemsblad 2008 (1). 15 – 20. (158). 345 – 371. Klappauf, L. 2011. Montanarchäologie im Westharz. Meyer, J. 1998: Keramik der Zeit um 1100 aus dem In: R. Smolnik (ed.) Aufbruch unter Tage. Stand und Hafen viertel von Schleswig. Archäologie in Schleswig/ Aufgaben der montanarchäologischen Forschung in Arkæologi i Slesvig 5. Haderslev & Kiel. 89 – 116. Sachsen. Internationale Fachtagung, Dippoldiswalde Moesgaard, J.Chr. 2006. The import of English coins

70 Hedeby’s demise in the late Viking age and the shift to Schlesvig

to the Northern Lands: some remarks on coin circu work. In: L. Larsson, F. Ekengren, B. Helgesson & B. lation in the Viking Age based on new evidence from Söderberg (eds.) Small things – wide horizons. Studies Denmark. In: B. Cook & G. Williams (eds.) Coinage in honour of Birgitta Hårdh. Oxford. 43 – 48. and History in the North Sea World c. 500–1250. Es Rundberg, J. 2000. Otto-Adelheid-Pfennige. Ett försök says in Honour of Marion Archibald. The Northern till revidering av en omdebatterad tyske myntgrupp world 19. Leiden. 389 – 433. från vikingatiden. C-uppsats i Arkeologi. Stockhol – 2012. Hedeby og den danske kongemagt i 900–tallet. ms Universitet. Arkeologiska Institutionen. Stock Mønternes udsagn. Kuml 2012. 111 – 136. holm. Noonan, Th.S. 1994. The Vikings in the east: Coins Sawyer, P. 2013. The wealth of Anglo-Saxon England. and Commerce. In: B. Ambrosiani & H. Clarke Oxford. (eds.) Developments around the Baltic and the North Sawyer, B. & Sawyer, P. 2002. Die Welt der Wikinger. Sea in the Viking Age. The Twelfth Viking Congress. Die Deutschen und das europäische Mittelalter. Ber = Birka Studies 3. Stockholm. 215 – 236. lin. – 2000/01. Volga Bulghāria’s tenth-century trade with Scheel, O. 1940. Haithabu. Welthandelsplatz des Nor Sāmānid Cetral Asia. Archivum Eurasiae Medii Aevi dens zwischen Asien und Europa im 9.–11. Jahrhun 2000/01 (11). 140 – 218. dert. Hamburg- der Nordischen Gesellschaft Pedersen, A. 1996/97. Riding gear from Late Viking-age 2. Hamburg. Denmark. Journal of Danish Archaeology 1996/97 Schietzel, K. 1981. Stand der siedlungsarchäologischen (13). 133 – 160. Forschung in Haithabu – Ergebnisse und Probleme. – 2014. Dead Warriors in Living Memory. A Study of Berichte über die Ausgrabungen in Haithabu 16. Weapon and Equestrian Burials in Viking-Age Den Neumünster. mark, AD 800-1000. Publications from the National – 2014: Spurensuche Haithabu. Dokumentation und Museum. Studies in Archaeology & History Vol. Chronik 1963–2013. Neumünster & Hamburg. 20:1. 1–2 Jelling Series. Odense. Schledermann, H. 1966. Slesvig/Hedebys tilblivelse. Raddatz, K. 1964/65. Bericht über die Probegrabung Sted navne og fund I. Sønderjyske Årbøger 1966. 1 – auf der Südsiedlung. Offa 1964/65 (21/22). 58 – 73. 65. Radtke, Chr. 2002. Schleswig im vorlübischen Geld- Schultze, J. 2008. Haithabu – Die Siedlungsgrabungen und Warenverkehr zwischen westlichem Kontinent 1. Methoden und Möglichkeiten der Auswertung. Die und Ostseeraum. In: K. Brandt, M. Müller-Wille & Ausgrabungen in Haithabu 13. Neumünster. Chr. Radtke (eds.) Haithabu und die frühe Stadtent Schwantes, G. 1932. Führer durch Haithabu. Schleswig. wicklung im nördlichen Europa. Schriften des Ar Simon-Hiernard, D. & Gratuze, B. 2011. Le vase de chäologischen Landesmuseums 8. Neumünster. 379 Saint-Savin en et les verres médiévaux – 429. bleu-cobalt à décors blancs. Bulletin de l’Association – 2007. Schleswig ca. 1000–1250. Systemtheoretische Française pour l’Archéologie du Verre 2011 (25). 69 – Skizzen eines Urbanisierungsprofils. In: S. Burmeis 73. ter, H.Derks & J. von Richthofen (eds.) Zweiund Sindbæk, S.M. 2012. Viking Disruptions or Growing vierzig. Festschrift für Michael Gebühr zum 65. Ge Integration? Contextualising Communication burtstag. Internationale Archäologie. Studia honora Networks in the 10th Century North Sea. In: S. ria 25. Rahden, Westf. 317 – 338. Kleingärtner & G. Zeilinger (eds.) Raumbildung – 2009. Haithabu. Perspektiven einer Stadtentwicklung durch Netzwerke? Der Ostseeraum zwischen Wikinger in drei Stationen – 800, 900, 1000. Zeitschrift für zeit und Spätmittelalter aus archäologischer und ge Archäologie des Mittelalters 2009 (37). 135 – 162. schichtswissenschaftlicher Perspektive. Beiträge des am Roesdahl, E. & Wilson, D.M. 2006. The Århus rune- 28. und 29. Oktober 2010 in Kiel ver stones. In: P. Gammeltoft & B. Jørgensen (eds.) anstalteten internationalen Workshops. Zeitschrift Names through the Looking-Glass. Festschrift in Hon für Archäologie des Mittelalters. Beiheft 23. Bonn. our of Gillian Fellow-Jensen. Navnestudier 39. Køben 19 – 38. havn. 208 – 229. Spufford, P. 1988.Money and its use in Medieval Europe. Roslund, M. 2015. At the end of the silver flow. Islamic Cambridge. dirhams in Sigtuna and the shrinking Viking net Steppuhn, P. 1998. Die Glasfunde von Haithabu. Be

71 New Aspects on Viking-age Urbanism, c. 750-1100

richte über die Ausgrabungen in Haithabu 32. Neu ham (ed.) The long eighth century: production, distri münster. bution and demand. The transformation of the Ro Steuer, H. 1974. Die Südsiedlung von Haithabu. Studien man world 11. Leiden, Boston & Köln. 105 – 120. zur frühmittelalterlichen Keramik im Nordseeküstenbe Wehner, D. 2007. Der frühgeschichtliche Seehandelsplatz reich und in Schleswig-Holstein. Ausgrabungen in Ha Wolin und sein Umland. Eine Studie zu Zentrum und ithabu 6. Neumünster. Peripherie. Studien zur Siedlungsgeschichte und Ar – 1997. Waagen und Gewichte aus dem mittelalterlichen chäologie der Ostseegebiete 8. Neumünster. Schleswig. Funde des 11. bis 13. Jahrhunderts aus Eu Wiechmann, R. 2007. Haithabu und sein Hinterland ropa Quellen zur Handels- und Währungsgeschich – ein lokaler numismatischer Raum? Münzen und te. Zeitschrift für Archäologie des Mittelalters. Bei Münzfunde aus Haithabu (bis zum Jahr 2002). In: heft 10. Köln & Bonn. K. Schietzel (ed.) Das archäologische Fundmaterial – 2004. Münzprägung, Silberströme und Bergbau um VIII. Berichte über die Ausgrabungen in Haithabu das Jahr 1000 in Europa – wirtschaftlicher Aufbruch 36. Neumünster. 182 – 278. und technische Innovation. In: A. Hubel & B. – 2013. Byzantinische Treppen und englische Kreuz – Schneidmüller (eds.) Aufbruch ins zweite Jahrtau Zwei Münztypen des 11. Jahrhunderts – geprägt in send. Innovation und Kontinuität in der Mitte des Haithabu? In: S. Kleingärtner, U. Müller & J. Mittelalters. Mittelalter-Forschungen 16. Ostfildern. Scheschkewitz (eds.) Kulturwandel im Spannungsfeld 117 – 149. von Tradition und Innovation. Festschrift für Michael – 2012. Waagen und Gewichte in Janów. In: S. Brather Müller-Wille. Neumünster. 145 – 160. & M.F. Jagodziński, Der wikingerzeitliche Seehandel Williams, A. 2008. The World before Domesday. The En splatz von Janów (Truso). Zeitschrift für Archäologie glish Aristocracy 900–1066. London. des Mittelalters. Beiheft 24. Bonn. 185 – 280. Williams, D. 1997. Late Saxon stirrup-strap mounts. A – 2014. Mittelasien und der wikingerzeitliche Norden. classification and catalogue. A contribution to the study In: K. Hofmann, H. Kamp & M. Wemhoff (eds.) of Late Saxon ornamental metalwork. CBA Research Die Wikinger und das Fränkische Reich. Identitäten Report 111. York. zwischen Konfrontation und Annäherung. Mittelalter Studien 29. Paderborn. 217 – 238. Stieldorf, A. 2012. Marken und Markgrafen. Studien zur (Endnotes) Grenzsicherung durch die fränkisch-deutschen Herr 1 With the same runic inscriptions for Erik and Skarthi scher. Monumenta Germaniae Historica. Schriften Otto Scheel, historians at Kiel University argued that 64. Hannover. these battles and Hedeby’s demise could be connected Stoklund, M. 2000. Die Inschriften von Ribe, Hedeby with King Sven Estridsen and Hedeby’s pillage by the und Schleswig und die Bedeutung der Schweden Norwegian king Harald Hardrade around 1050 (Scheel herrschaft. In: K. Düwel, E. Marold & Chr. Zim 1940:23f) – an opinion of which also Gustav Schwantes mermann (eds.) Von Thorsberg nach Schleswig. Sprac in 1932 (Schwantes 1932:28-30) or Helmuth Schleder- he und Schriftlichkeit eines Grenzgebietes im Wandel mann still in 1966 (Schledermann 1966:21, 46) were eines Jahrtausends. Ergänzungsbände RGA 25. Berlin persuaded of. & New York. 111 – 126. Tummuscheit, A. 2012. Das Tor im Danewerk bei Ro 2 For a definition see Verhulst 2000; Hodges 2008; thenkrug – Ausgrabungen 2010 und 2011. Gesell Loveluck 2013:180–212; Kleingärtner 2014: esp. schaft für Schleswig-Holsteinische Geschichte Mittei 67–73, 175–199; and S. Kalmring, this volume. lungen 2012 (83). 3 – 15. Tummuscheit, A. & Witte, F. 2014. „Der einzige Weg 3 Callmer related to Schietzel 1981:68f, where such a durchs Danewerk“: Zu den Ausgrabungen am Dane conclusion could not yet be found. werk im Jahr 2013. Archäologie in Schleswig/Arkæolo gi i Slesvig 15. Haderlsev & Kiel. 153 – 162. 4 According to Herrmann 1994/95:64f, farmsteads con- Verhulst, A. 2000. Roman , emporia and new stituted the basis of the settlement structure of early em- towns (sixth-ninth centuries). In: I. Lyse Hansen & poria of the 8th century. Chr. Wick

72 Hedeby’s demise in the late Viking age and the shift to Schlesvig

5 This stratigraphic situation is not considered appro- priately in the monograph on the Hedeby graves, cf. Arents/Eisenschmidt 2010:228, 297f, 321–324.

6 Recently Loveluck (2013:252) pointed out that “glass drinking vessels are all but absent on sites of the tenth to twelfth centuries” being replaced by drinking horns or metal vessels.

7 Spufford’s observation has been variously adopted, cf. e.g. Bompaire 1997:198; Gilomen 2014:79f.

73 New Aspects on Viking-age Urbanism, c. 750-1100

Geomagnetic survey at Hedeby Hochburg

Andreas Viberg & Sven Kalmring

in breadth. The south-eastern part of the rampart was Introduction and background destroyed early on by the open-pit extraction of The hillfort of the Hochburg is situated on a longish for brickyards. One gate seems to be preserved at the and slender plateau of a ridge approximately 180 m eastern corner of the rampart. Facing a forework-like north of the semi- of the proto-town terrace at the north-eastern narrow side of the rampart, Hedeby (fig. 1). The actual ridge on which the hillfort this section was additionally secured by a small in is situated is composed of post glacial clay, which gained front of the wall. Inside the rampart of the Hochburg, its form by sedimentation when the two dead ice holes in a beech and oak tree forest, there are numerous flat of the valley and Haddebyer Noor melted away burial mounds. A small additional group of mounds is (Gripp 1940:52, 62–64). The once deforested ridge at situated north-east of the rampart, a little below the tip the inlet of the Haddebyer Noor offered a superb con- of the forework-like terrace, and one possible mound is trolling view of the inner Schlei fjord. The plateau itself located at the descent towards the Noor. The mounds possesses a small earthen rampart, nowadays not more are not uniform, but vary in height between 0.2 to than 1.5 m high, that encloses an almost rectangular 1.0 m and differ considerably in diameter. Because of area of approximately 225 m in length and 45 to 70 m their limited elevation, they sometimes are hard to dis-

Fig. 1. Map showing the location of Hedeby Hochburg in present day Northern Germany (after Kalmring 2010:fig. 5)

74 Geomagnetic survey at Hedeby Hochburg

Fig. 2. Map showing the location of the magnetometer survey area in relationship to the known extension of the monument Hedeby Hochburg (after Loewe 1998:fig. 19 with additions). cern from natural bumps inside the dense forest of the burg does date back to the 16th century. The main lines Hochburg. Therefore, the data of the exact of interpretation were a construction as the seat of a number of graves sometimes varies greatly. In the pub- German margrave or as a refuge fort after the model lication on the archaeological survey of the admin- of the hillfort Borg at Birka in Lake Mälaren (Kalm- istrative district of Schleswig (Loewe 1998:51–53) ring & Holmquist 2010). While the historical classifi- 57 mounds are listed. Ten graves had previously been cation was a matter of considerable debate, the actual excavated, but no datable material was ever recovered archaeological evidence is still astonishingly sparse. Up (cf. Holmquist & Kalmring 2012:7–8). According to to today, neither the rampart nor the burials have been Arents & Eisenschmidt (2010:286) the construction of dated with any certainty. This also implies that neither the graves imply that they most likely originate from their chronological succession nor the hillfort’s relation- the 8th–9th century AD. ship to Hedeby is known. At this point, the survey of Hedeby Hochburg has always been the neglected a Swedish-German excavation team began its work in cousin of Hedeby’s archaeological research. However, June 2012. Apart from the main trench trough the ram- a comprehensive academic discussion of the Hoch- part and an adjacent mound, the odd circumstance that

75 New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 3. Magnetometer survey in progress at Hedeby Hochburg. The photograph is taken towards the south and depicts the northernmost part of the Hochburg rampart in the background. The Hochburg plateau is covered with large trees and in some areas very dense vegetation making geophysical surveys challenging. (Photo: Antje Wendt 2012).

the very tip of the plateau – the forework-like terrace – ditches and pits (Aspinall et al. 2008) which enable should have remained omitted from the fortification led their detection with a magnetometer. The method has to complementary geomagnetic surveys in that area (fig. been used extensively within German archaeology since 1–2). the 1950s (Clark 1996:19 and references cited therein), but has only been used sporadically for the detection of archaeological remains in Sweden until 2005 (Viberg Method 2012:87). Large-scale geophysical surveys have also A magnetometer is a passive geophysical instrument successfully been carried out within the semi-circular that makes detailed measurements of the Earth’s mag- rampart of the proto-town Hedeby, producing evidence netic field. Since 1958, the method has successfully of its spatial layout (e.g. Neubauer et al. 2003; Hilberg been used on archaeological sites worldwide (e.g. Aitken 2007). 1958; Becker & Fassbinder 1999a; 1999b; Kvamme The survey at Hedeby Hochburg was primarily car- 2008; Gaffney et al. 2012; Neubauer et al. 2013). The ried out with the aim of detecting buried archaeological Earth’s magnetic field is vital to archaeological magne- remains that might light on any activities carried tometer prospecting. This field can be locally distorted out in the area in protohistoric times. We also hoped by buried archaeological remains, such as hearths, kilns, to find an explanation as to why the tip of the plateau

76 Geomagnetic survey at Hedeby Hochburg

Fig. 4. Results from gradiometer measurements at Hedeby Hochburg. Grey scale: -1 nT white to +1 nT black (left). Grey scale: -4 nT white to +6 nT black (right).

Fig. 5. Map showing the correlation between the strong magnetic anomaly and the footpath crossing the survey area, the probable location of the 1959 excavation trench and the ditch with nodular pits (left). Results from gradiometer measurements at Hedeby Hochburg and their correlation with recently discovered mounds at the site (blue). Possible additional mounds suggested by the magnetometer data are also presented (purple) (right).

77 New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 6. 78 Detail of the architectural drawing for the planned Socialist thing assembly site with the co-called “sanctuary” surrounded by a groove (courtesy Ute Drews, Wikinger Museum Haithabu, Schleswig) superimposed on the hexagonal anomaly indicated by the magnetometer data (state archive Schleswig, div. 309,35.770). Note its position on the highest point on the very tip of the Hochburg-ridge. Geomagnetic survey at Hedeby Hochburg is the only part not included in the defensive works of situated within the survey area (blue colour in fig. 5). the hill fort (Kalmring & Holmquist 2012:64). The se- These mounds are most ikely graves similar to those sit- lected survey area would, by normal standards, not be uated within the main fortification and below the tip of considered suitable for geophysical surveys as the tip the terrace. Two additional areas of increased magnet- of the plateau is covered by very dense vegetation (cf. ic susceptibility might show the location of additional Holmquist & Kalmring 2012:7 fig.3). Dense vegeta- graves not visible above ground surface (purple colour tion and geophysical surveys in forested areas are time in fig. 5). A strong linear bipolar anomaly is also visi- consuming and data collection in large enough areas ble in the middle of the survey area and partly seems or in rectangular grids is often difficult. As time and to follow the foot path running through the area (fig. cost for geophysical surveys are intimately connected, 5). However, as the bipolar anomaly and the foot path surveys in forests are often not considered worthwhile don’t coincide completely, it is likely that the anoma- as they would be too expensive and the quality of the ly is caused by something other than this. The strength collected data too poor. In Sweden, for example, where of the anomaly and its direction of magnetisation yet 53% of the total area is covered by forests (Statistiska suggest that it could have a modern origin and as its Centralbyrån 2005), the use of geophysical methods curving very much follows the spurs’ 26 metre contour would, according to such a standard, be unsuitable in line (see fig. 4 & 5). It therefore might be identified as more than half the country. An underlying purpose was the remnants of a former fence with iron cramps and therefore also to consider whether it could be worth- wire. The slightly weakened magnetic susceptibility of while carrying out surveys in similar environments in the along the ditch in front of the rampart, in the Germany and abroad. Previous attempts to carry out southern part of the survey area, might be explained by geomagnetic measurements in densely forested areas in the removal of magnetically enhanced topsoil from the Sweden have unfortunately not produced convincing area when constructing the rampart. For a discussion on data and most surveys have not been carried out in the enhancement of magnetic susceptibility of top soil large enough grids to provide any valuable information see, for example, Le Borgne (1955; 1960) and Aspinall (see for example George 2011). et al. (2008). A likewise clear negative magnetic suscep- During the surveys at Hedeby Hochburg, a single tibility can be noticed in the southern corner of the sur- probe Foerster Ferex 4.032 fluxgate gradiometer was vey area starting off at the trail and leading up towards used measuring the vertical gradient of the Earth’s the rampart (fig. 4–5). With its rectangular shape and at magnetic field. The survey area measured 40x28m a width of two meters it can be identified as Jankuhns’ with an additional smaller extension of 20x11m to lost trench from 1959 (Jankuhn w/o year). the north (fig. 2). The crossline sampling distance for With a starting point at the gate of the hillfort, the the measurements was 0.5m and the inline sampling distribution of the mounds on the terrace seems almost distance was 0.1m with data collected in zigzag for- to be bifurcated, omitting an area towards the north. mation. The post-survey data management and fil- Here on its highest point, a somewhat circular anomaly tering was carried out using the software Data2Line can be observed that, as Jankuhn’s excavation trenches by Foerster. The results are destaggered, corrected for or the , has a lower magnetic susceptibility than Zero mean Traverse (ZMT) and the remaining stripy the surrounding soil matrix (fig. 4–5). Thus it can be pattern, caused by heading errors, introduced as the assumed that the structure, of about eight metres in survey was conducted in a densely forested area, were diameter, is a dug-down circular ditch. In addition partly removed using a 2D-FFT filter. The magnetic the anomaly seems to be accompanied by six nodular results were subsequently interpreted and turned into pits measuring about two meters in diameter. It seems georeferenced maps using the GIS ArcMap 9.3. to coincidence with the hexagonal “sanctuary” of the planned National Socialist Thing assembly site from the mid-1930ies (state archive Schleswig, div. 309,35.770) Results and interpretation of which one hitherto believed only the left wing of The results from the geomagnetic measurements (fig. 4) the open air theatre was realised before the plans where show several areas of elevated magnetic susceptibility. soon abandoned (fig.6; Haßmann & Jantzen 1994:15). These anomalies coincide with the location of addition- al mounds identified during the 2012 field campaign

79 New Aspects on Viking-age Urbanism, c. 750-1100

Discussion It is interesting to note that the areas inside and outside Acknowledgements the rampart are indeed very similar in terms of utilisa- tion. Obviously the whole of the plateau was used as We would like to thank Charlotte Hedenstierna-Jon- a burial ground and the currently registered 57 graves son, Laila Kitzler Åhfeldt, Joakim Schultzén, Lena should, at least, be increased by the additional eight Holmquist, Antje Wendt, Björn Rauchfuss, Martin to ten graves discovered during the field campaign in Segschneider and Timo Ibsen for their kind support 2012. Moreover it can be assumed that the present day and assistance during field work. spatial layout of the rampart most likely isn’t its original course as this would leave the northernmost part of the Corresponding authors Hochburg plateau undefended. This would imply that the present day north-easternmost part of the rampart Sven Kalmring has to be a later addition. The14 C-dating from the from Centre for Baltic and Scandinavian Archaeology, the 2012-main trench trough the rampart and an adja- Schloss Gottorf cent mound showed that that the burial mound dates as Schleswig, Germany early as the second half of the 7th century while the mu- [email protected] li-phased embankment itself was first raised in the course of the 9th century (Kalmring 2016). Andreas Viberg The results furthermore show that geomagnetic mea- Archaeological research laboratory, Stockholm Univer- surement could be a useful tool even in a densely forest- sity ed area, and the correlation between above ground vis- Stockholm, Sweden ible mounds and the increased magnetic susceptibility, [email protected] evident in the magnetic data, show that it may be used as a complement to other, more traditional archaeolog- ical field inventory methods. The data must, however, be collected in a rigorous manner with the aid of guide ropes and the surveyed grids need to be properly geo- referenced using either a total station or an RTK-GPS. The sampling density also needs to be properly adapted to the sizes of the expected features in order to provide a proper resolution. It is also important to survey a large enough area in order to assess background levels and to avoid unnecessary interpretational errors (cf. e.g. Gaff- ney & Gater 2003:92)

Conclusion Geomagnetic measurements at Hedeby Hochburg have provided evidence of burials on the apparently non-fortified part of the Hochburg plateau. As a result, the number of registered graves should be increased by at least eight to ten graves. Moreover the construction of the National Socialist Thing assembly site obviously had advanced further than previously assumed. The re- sults also show that if proper time is devoted to the geo- physical survey, meaningful results in areas not usually deemed suitable for such surveys might provide import- ant information on subsurface features.

80 Geomagnetic survey at Hedeby Hochburg

187 – 203. References Holmquist, L. & Kalmring, S. 2012. Hedeby och Birka Aitken, M.J. 1958. Magnetic prospecting. I – The Wa Stadslika handelsplatser med borgar. Populär Arkeolo ter Newton Survey. Archaeometry 1958 (1):1. Ox gi 2012:3. 4 – 8. ford. 24 –29. Jankuhn, H. w/o year. Wallschnitt 1959 auf der Hoch Arents, U. & Eisenschmidt, S. 2010. Die Gräber von burg von Haithabu. Unpublished excavation-report. Haithabu 1: Text, Literatur. Ausgrabungen in Hait Hedeby Archive; drawer “Gräber 1930–38; 1959– habu 15. Neumünster. 60“; folder “Hb 1933“. Aspinall, A., Gaffney, C. & Schmidt, A. 2008. Magne Kalmring, S. 2010. Der Hafen von Haithabu. Ausgra tometry for Archaeologists. Lanham. bungen in Haithabu 14. Neumünster. Becker, H. & Fassbinder, J.W.E. 1999a. In search for – 2016. Hedeby Hochburg – Theories, State of Research Piramesse - the lost capital of Ramses II, in the Nile and Dating. Offa 2014/15 (71/72). Neumünster. Delta () by caesium magnetometry. In: J.W.E. Kalmring, S. & Holmquist, L. 2010. Die Hochburg von Fassbinder & W.E. Irlinger (eds.) Archaeological Pro Haithabu – Hypothese, Forschungsstand und Pers spection, Bayerischen Landesamt für Denkmalpflege. pektiven. Jahresbericht Zentrum für Baltische und München. 146 – 150. Skandinavische Archäologie. Schleswig. 80. – 1999b. Magnetometry of a Scythian Settlement in – 2012. Deutsch-schwedische Ausgrabung auf der Hoch Siberia near Cicah in the Baraba Steppe 1999. Fass burg von Haithabu. Jahresbericht Zentrum für Balti binder, J.W.E. & W.E. Irlinger, (eds.) Archaeolo gi sche und Skandinavische Archäologie. Schleswig. 36 – cal Prospection, Bayerischen Landesamt für Denk 37. malpflege, München. 168 – 172. Kvamme, K.L. 2008. Archaeological Prospecting at the Clark, A. 1996. Seeing beneath the soil. Prospecting meth Double Ditch State Historic Site, North Dakota, ods in archaeology. Oxon. USA. Archaeological Prospection 2008 (15) :1. New Gaffney, C., Gaffney, V., Neubauer, W., Baldwin, E., York. 62 –79. Chapman, H., Garwood, P., Moulden, H., Sparrow, Le Borgne, E. 1955. Susceptibilité magnétique anor T., Bates, R., Löcker, K., Hinterleitner, A., Trinks, I., male du sol superficiel. Annales de Géophysique 1955 Nau, E., Zitz, T., Floery, S., Verhoeven, G. & Done (11). Paris. 399 – 419. us, M. 2012. The Stonehenge Hidden Landscapes – 1960. Influence du feu sur les propriétés magnétiques Project. Archaeological Prospection 2012 (19):2. New du sol et sur celles du schiste et du . Annales de York. 147 – 155. Géophysique 1960 (16). Paris. 159 – 195. Gaffney, C. & Gater, J. 2003. Revealing the buried past. Loewe, G. 1998. Kreis Schleswig (seit 1974 Kreis Schles Geophysics for archaeologists. Gloucestershire. wig-). Archäologische Denkmäler Schles George, O. 2011. Arkeologisk forskningsundersökning av wig- VIII. Veröffentlichungen des Archäo stenåldersboplatsen Raä 51:1 i Ramsele sn, Ångerman logischen Landesamtes Schleswig-Holstein in Schles land., Murberget. Länsmuseet Västernorrland. Här wig. Neumünster. nösand. Neubauer, W., Eder-Hinterleitner, A., Seren, S., Becker, Gripp, K. 1940. Die Entstehung der Landschaft um H. & Fassbinder, J.W.E. 2003. Magnetic survey of Haithabu. Offa 1940 (5). Neumünster. 37 – 64. the Viking Age settlement of Haithabu, Germany. Haßmann, H. & Jantzen, D. 1994: „Die deutsche Vor In: T. Herbich (ed.) Archaeologia Polona (41). Special geschichte – eine nationale Wissenschaft.“ Das Kie theme: Archaeological prospection. Wroclaw. 239 – ler Museum vorgeschichtlicher Altertümer im Drit 241. ten Reich. Offa 1994 (51). Neumünster. 9 – 23. Neubauer, W., Trinks, I., Salisbury, R.B. & Einwögerer, Hilberg, V. 2007. Haithabu im 11. Jahrhundert. Auf C. 2013. Archaeological Prospection. Proceedings of the der Suche nach dem Niedergang eines dänischen 10th International Conference. Vienna. emporiums der Wikingerzeit. In: M. Posselt, B. Viberg, A. 2012. Remnant echoes of the past. Geophysical Zickgraf & C. Dobiat (eds.) Geophysik und Ausgra archaeological prospection in Sweden. Theses and pa bung. Einsatz und Auswertung zerstörungsfreier Pros pers in Scientific Archaeology 13. Archaeological Re pektion in der Archäologie. Internationale Archäolo search Laboratory, Stockholm University. Stockholm. gie – Naturwissenschaft und Technologie 6. Rahden.

81 New Aspects on Viking-age Urbanism, c. 750-1100

Kosel-ost: a Viking-age settlement in the hinterland of Hedeby

Tobias Schade

Fig. 1. Kosel in Schwansen. Also indicated the Schlei fjord and the location of Hedeby. Map data (c) ‘OpenStreetMap’ and contributors, CC-BY-SA.

From 1975 to 1993 two Viking Age settlements and one Bronze Age to the Middle Ages in the landscapes Viking Age burial ground have been excavated in Kosel, Angeln and Schwansen (Meier 1994:26). The site of Kr. -Eckernförde, in Northern Germany situated Kosel itself is located at the foot of the Schwansen roughly in between the Schlei fjord and Eckernförder bay peninsular. Kosel presumably was a ”specific“ with the inlet of the Windebyer Noor ca. 13 km east of settlement in the hinterland of Hedeby. While its Hedeby (fig. 1).The excavations eastwards of the present function and structure is still ambiguous, some village of Kosel were part of the „Angeln-Schwansen“-project other points are eye-catching. Kosel is situated on of the German Research Foundation, which started in 1982. the stream „Koseler Au“ which gives access to the The objective was to analyse the settlement patterns from the Schlei fjord, which is 3 km distant and it is located

82 Kosel-ost: a Viking-age settlement in the hinterland of Hedeby

Fig. 2. Kosel. The present village and the west- and the east-settlement (Meier 1994:17, fig. 2). Excavation areas of the east-settlement marked.

in the hinterland of Hedeby. Hence it was connected bones and 42 inhumed individuals had been identified to major trade routes. As well its closeness to the a few (fig. 3). The features and artefacts can be dated in the kilometers southwards lying „Osterwall“, a part of the 9th and 10th century. While the analysis of the settlement Danewerk, and its Saxon and Slavonic borderland are is not completed yet, the analysed inhumation burials favorable. Generally there is a certain continuity of date the burial ground to 930-975 AD. Presumably settlement occupation in the Kosel-area stretching from it is possible to identify an older but highly disturbed the Roman Iron Age and the Migration Period to the cremation burial ground. Burial rites and grave goods Viking Age – with two settlements – and the Middle illustrate both a religious and social grouping and Ages. indicate a transformation from a pagan burial ground to a Christian influenced burial ground (Schade 2010/11). In contrast to the west-settlements (Ger. Kosel-West) Further analyses should clarify if the social diversification from the 9th/10th and 11th century, which were published shown in the structure of the graveyard can be found in 1994 (Meier 1994), the east-settlement (Ger. Kosel- in the settlement as well or if they show us differences Ost) until now was only presented in preliminary reports between the living and the dead. In this context it is also (fig. 2). The burial ground will be published in 2014. interesting if the settlement and the burial ground were At Kosel-Ost in an area of nearly 40,000 m² at least abandoned contemporarily after the religious shift or if three to four three-aisled longhouses and 53 pit-houses the settlement was used for a longer period while the as well as a contemporaneous burial ground with 40 Christians stated to bury their dead on a new consecrated inhumation burials and one cremation burial had been graveyard. This suggested early Christian cemetery is not excavated. Thereby three concentrations of cremated localised yet, but it could lie near the Romanic church of

83 New Aspects on Viking-age Urbanism, c. 750-1100

Kosel from the , which is situated near the could be connected with actual documented features. The west-settlement from the 11th/12th century. subsequent values are applying to these 7972 fragments only. This compound can be differentiated in 11% rim A second objective is to compare Kosel on the basis of sherds, 6% bottom sherds, 71% body sherds, 11% very archaeological, archaeobotanical and archaeozoological strong fragmented sherds and 1% sherds, which present a analyses with contemporaneous sites in Denmark to contamination from different periods (, Roman recognize transformation processes and supra-regional Iron Age, Medieval). The majority of the sherds ended structures as well as Kosels’ role in the 10th century local up in the filling of the abandoned pit-houses. setting. It shall be clarified whether the people of Kosel- The rims of the domestic Viking Age wares were Ost had a focus on a special production, participated documented after the model of Kosel-West in 1994. in trade and how they were linked to Hedeby. It could Meier used a system that was first introduced by be assumed that they based their economy on animal Steuer in his analysis of the south-settlement (Ger. husbandry or cultivation of grain, which they delivered Südsiedlung) of Hedeby (Steuer 1974:60; addendum to Hedeby. Remains from mainly horses and cattle, 1-2). Following Meier (1994), the rims were categorised but also from pigs, and goats had been found in the sections A, B, C and D. Rims which did not fit in particularly in the pit-houses and pits. However, these categories were analysed on the basis of Kempkes’ the bones weight is only about 5 kg and just a few (1984) report on the Slavonic pottery in Starigrad/ individuals could be identified (unpublished analysis Oldenburg.. C. Wustrow, Kiel). It is vague, whether this reflects actual circumstances of the former husbandry or rather More than two-thirds of the sherds are belonging to is a result of the soil conditions which also destroyed the domestic Viking-age pottery. More precisely this a lot of the human remains in the burial ground. It pottery can be divided into two groups. On the one must be pointed out that this is only a first review of hand there is a large amount of ca. 40% of rim sherds, the zoological data. So far, the archaeobotanical data which belong to the undecorated Viking Age jars with is only analysed partially, too, but six pit-houses were inward-turning rim (A). On the other hand there is a examined in terms of macro-botanical analyses (Kroll smaller percentage of 28% of rim sherds, which can be 1986). Rye, barley and oats as well as weeds illustrate identified as domestic pottery with mainly out-bending an agricultural producing community. But apart from and abrupt rims (B and C) (fig. 4). 14% of the rim the small amount of examined features only, which are sherds have to be excluded since they are either from in no way representative, the few botanical analyses do other periods or indeterminable fragmented. The body not proof a surplus of production. However, the high of Viking Age pottery in Kosel-Ost is dominated of percentage of peas and bread wheat is noticeable (Kroll undecorated jars with inward-turning rim of the group 1986:450). Bread wheat is rare in Hedeby and peas A (fig. 4). They are handmade, thick-walled and mostly are absent (Kroll 1986:451). Yet this is a preliminary irregularly burned. They are typical for Danish Viking- assessment as well. age settlements and also called „halvkuglekar“. But contrary to the most parts of Denmark, the jars from The character of the settlement as a place of production Kosel-Ost are mostly flat bottomed like similar jars is being emphasized by the amount of found pit- found in Kosel-West (Meier 1994:130–131), houses. Nearly 51 kg of loom weights show that textiles (Madsen 1991:222) and especially the Slavonic region, production was important for the inhabitants. And even where a potter’s wheel was already used (Kempke if the weight of the slag is not very high with almost 3 kg, 2001:213–214). Therefore the flat bottomed vessels of it is a proof for ironworking on site. Besides of personal Kosel indicate some Slavonic influence, but only 4.5% objects, a few weapons, glass beads and other jewelry, the of the bottom sherds could be attached to fragments, finds are mainly composed of the aforementioned pottery, which had a mark from a potter’s wheel on it (Ger. slag, loom weights, and lava. One can assume that Achsabdruck) and thus were definitely of Slavonic the most important key for clarifying Kosels’ function origin. are the ceramic sherds. In the following the first results are being presented. Nearly 133 kg of pottery consisting C. 3.8% of the 7972 sherds belonging to documented of 8067 sherds were analysed. As only 7972 potsherds features show Slavonic elements. Although the value is

84 Kosel-ost: a Viking-age settlement in the hinterland of Hedeby

Fig. 3: Kosel-Ost. Excavation plan. Post buildings (green); pit-houses (red); postholes (grey); pits (orange); burial ground (darkgrey). Based on documentations of H. Dieterich, Kiel.

Fig. 4. Kosel-Ost. Percentages of domestic rims and decorated sherds as well as Frankish import and Slavonic sherds (without whole vessels).

85 New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 5. Kosel-Ost. Distribution of Slavonic sherds. Discerned are rims, bottom sherds with stamp and decorated sherds according to their quantity. not that high, it still should be pointed out that on the with stamp of which the decorated pottery seems to one hand Early and Middle Slavonic vessels and on the be dominating (fig. 4). However this is only caused by other hand the Viking-age domestic vessels are both weak the many decorated specimen among the body sherds. and irregular burnt earthenware and differ only in the Looking only at the rims another impression is gained. use of a potter’s wheel and in decoration. In conclusion 18% (n=164) of the rim sherds can be defined as Slavonic true Slavonic ceramic can mainly be discerned on the influenced pottery and thus undecorated vessels seem to basis of its rim shape, a decorated zone in the upper be dominating in the Slavonic inventory. The decorated part of the vessel or a flat bottom with a stamp. The vessels can mainly be linked to the Menkendorf-type. assemblage of possible Slavonic sherds in Kosel-Ost Only 0.2% of the rim sherds are belonging to Western are composed of undecorated rim sherds, decorated Import originating from at least one vessel of the Badorf rim sherds, decorated body sherds and bottom sherds type. Other western pottery could not be identified.

86 Kosel-ost: a Viking-age settlement in the hinterland of Hedeby

Even if single sherds were found in pits and postholes, the borderland. It is however still dubious how this the majority were found in the pit-houses. Yet some contacts looked like. In contrast to this the settlement pithouses yielded no ceramic at all. Ca. 50% of the pit- itself, grave goods and burial rites of the 10th century houses had Slavonic-influenced sherds in their filling do not show any Slavonic influence. But the burial they primary appear in the west part of the settlement ground as well as some artefacts from the settlement and are related to a longhouse as well as the central zone - e.g. the example of the Badorf vessel – show that of the settled area. Half of all Slavonic sherds are gathered some people in Kosel participated in trade and had in three pit-houses only, lying in a line in the central access to rare and prestigious goods. The family of the area of the settlement. Each of them is situated near a woman in chamber grave 2 for example participated post-building (fig. 5). The amount of Slavonic sherds in a supra-regional lifestyle and buried her with a cart in Kosel-Ost (4%)1 in comparison to the percentage body, two horses and jewelry as e.g. a Terslev in Kosel-West of the same time period in the 9th/10th (Schade 2010/11). Especially the Terslev fibula is a century with less than 1% – a few rim sherds and about symbol for the connection between Kosel and Hedeby. a dozen decorated sherds (Meier 1994:132–133) – is Not only are these fibulas found in chamber-graves in significant. The results of Kosel-Ost are not typical Hedeby, it is also assumed by Kleingärtner (2004:306, for contemporaneous rural settlements. According to 309) that these fibulas were distributed from Hedeby Kempke (2001:251) many Danish and Swedish rural to women with a certain social status. Although the settlements had no Slavonic pottery in their inventory analyses of the settlement are not completed yet and before the 10th century. some interpretations related to the archaeobotanical Another impression is gained when comparing with and archaezoological data could change, it is obviously proto-urban centres as Hedeby (Kempke 2001:251). that this rural settlement had an importance and the Based on the excavations from H. Jankuhn in 1959 people of Kosel-Ost had to offer something to Hedeby Hübener published his monograph on the pottery of in exchange for the ”imported goods“ found on-site. Hedeby from within the the semi-circular rampart. He counted about ca 200 rims, which he defined as Baltic Ware. While the percentage of this type in Hedeby of Corresponding Author the 9th and 10th century is comparably low with 1%, Tobias Schade this amount is rising up to ca. 5% in the 10th and Institut für Ur- und Frühgeschichte, Kiel university 11th century (Hübener 1959:170, 173). Based on K. Kiel, Germany Schietzels excavations of the years 1962–1969 Hahn [email protected] (1977:8) reported on ca. 80,000 new ceramic fragments and confirmed Hübeners’ results with a slight correction to 4.76% rims of Baltic Ware. Steuer (1974:163), who analysed the settlement southwards the semi-circular rampart (“Südsiedlung”) however counted 15% of Early and Middle Slavonic sherds in the 9th century. He mainly refers to rims, but his data includes also a few bottom sherds as well as decorated sherds. Comparing these results it can be pointed out that Kosel-Ost and its comparably high percentage of Slavonic rim sherds can be related best to the south-settlement of Hedeby. It is remarkable that both settlements have a higher percentage than the settlement in the semi-circular rampart itself. Further analyses of the pottery of Kosel- Ost regarding to the distribution in the settlement, technological attributes or utilisation are not completed yet and further results are expected. The artefacts and features show that the people had contacts to Hedeby and to the Slavonic neighbours in

87 New Aspects on Viking-age Urbanism, c. 750-1100

References Hahn, K.-D. 1977. Die Einheimische Keramik von Haithabu. Dissertation Kiel University. Hübener, W. 1959. Die Keramik von Haithabu. Ausgrabungen in Haithabu 2. Neumünster. Madsen, H.J. 1991. Vikingetidens keramik som historisk kilde. In: P. Mortensen & B.M. Rasmussen (eds.) Fra Stamme til Stat i Danmark – 2. Høvdingesamfund og Kongemagt. Aarhus. Meier, D. 1994. Die wikingerzeitliche Siedlung von Kosel (Kosel-West), Kreis Rendsburg-Eckernförde Siedlungs archäologische Untersuchungen in Angeln und Schwansen 3. Offa-Bücher 76. Neumünster. Kempke, Th. 1984. Die Keramik des 8. -12. Jahrhunderts. Starigard/Oldenburg. Hauptburg der Slawen in Wagrien II.Untersuchungen aus dem Schleswig- Holsteinischen Landesmuseum für Vor- und Frühgeschichte. Offa-Bücher 53. Neumünster. – 2001. Slawische Keramik. In: H. Lüdtke & K. Schietzel (eds.) Handbuch zur mittelalterlichen Keramik in Nordeuropa 1. Schriften des Archäologischen Landesmuseums 6. Neumünster. 209 – 256. Kleingärtner, S. 2004. Fibeln und Anhänger vom Typ Terslev und ihre gegossenen Imitationen. In: M. Müller-Wille (ed.) Zwischen Tier und Kreuz – Untersuchungen zur wikingerzeitlichen Ornamentik im Ostseeraum. Neumünster. 205 – 376. Kroll, H. 1986. Zur Bearbeitung der Pflanzenfunde der Ausgrabungen von Kosel, Schwansen (Kr. Rendsburg-Eckernförde). Bericht der Römisch- Germanischen Kommission 1986 (67). 445 – 453. Schade, T. 2010/11. Das wikingerzeitliche Gräberfeld von Kosel-Ost. Offa 2010/11 (67/68). 203 – 321. Steuer, H. 1974. Die Südsiedlung von Haithabu. Ausgrabungen in Haithabu 6. Neumünster. Wustrow, Ch. 1992. Die Tierknochen aus Kosel, einem Siedlungsplatz der römischen Kaiserzeit, der Wikingerzeit und des hohen Mittelalters in Schwansen (unpublished report).

(Endnotes) 1 The value of 18% in Kosel-Ost is only referring to Slavonic rim sherds and Slavonic decorated rims, but not to Slavonic bottom sherds or decorated body sherds. Referring to all these attributes only 4% of all sherds are Slavonic influenced.

88 Husby and the equestrian graves in Angeln and Schwansen

Husby and the equestrian graves in Angeln and Schwansen – different chronological stages in the development of a royal ad- ministration? Thorsten Lemm

practice from cremation graves to inhumation graves Introduction was completed. Then, in the 10th century, suddenly While the equestrian and weapon graves since long male and female inhumation graves with relatively rich are understood as the last resting-places of warriors of grave furnishings appeared (Eisenschmidt 2004:293ff). certain military ranks, who played an important role In the case of the male burials, the ones that shall be in the process of state formation in Denmark, the focused on in this article, the grave goods consisted Danish Huseby-places have attracted far less attention. of several weapons and/or riding gear together with Nevertheless these farms or hamlets, which in Sweden other objects like food or drinking vessels and artefacts and Norway are interpreted to have a royal background, connected with pastimes or sports (Pedersen 2012:48). might also have contributed their share to the formation Some male burials were lacking the riding equipment, or rather the administration of the Danish kingdom. In but whether this absence speaks for a lower social status the following it shall therefore be discussed, whether the of the deceased or merely a different burial custom is an equestrian graves and the Huseby-places could indicate open question. different chronological stages in the development of a More than 50 % of the equestrian graves were royal administration in Denmark. discovered and unearthed in the 19th century. Very often the grave goods are poorly preserved and incomplete and the documentations do not suffice modern requirements Graves in south Schleswig (Pedersen 1997a:175; 2012:48). Therefore an From Angeln and Schwansen, the two regions that interpretation as chamber graves is arguable in many cases. are relevant in the following, approximately 130 On the other hand equestrian graves archaeologically archaeological finds and sites from the Viking Age are excavated in the 20th century all turned out to be registered (fig. 1). But still the state of research leaves chamber graves. Therefore according to Eisenschmidt a lot to be desired. In general this applies especially to (2004:69) it can be assumed that all equestrian graves some areas further away from the archaeological centre with a complete weapon and riding furnishing had been in the city of Schleswig and in particular this concerns chamber graves. mainly the settlement-archaeological research. Therefore Apart from ship graves the chamber graves are the the picture of Viking Age Angeln and Schwansen is far most laborious grave constructions known from the from being complete. Among these finds and sites there Viking Age. Their most characteristic feature is the casing are 9 grave fields and further 46 graves or single finds of the grave-pit with woodwork that created a small thought to derive from graves from the 8th to the 10th grave-chamber of several square metres in which the dead century. often was buried in a coffin (Eisenschmidt 2004:67f). With very few exceptions like for instance the boat- Another characteristic feature is their covering by a small chamber grave and a few other chamber graves in Hedeby mound or their insertion into an existing mound from from the 9th century the burials from the period prior prehistoric times in order to create visual monuments to the 10th century were mostly cremation graves with in the landscape. The latter might indicate tradition and none or only a few grave goods (Arents & Eisenschmidt continuity within a local area and strong local bonds 2010:267ff). Around 900 AD a change in the burial between the family of the deceased and that area (Pedersen

89 New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 1. Angeln and Schwansen. Viking Age finds and sites and reconstructed roads (topographic data: Landesvermessungsamt Schleswig-Holste

Fig. 2. Equestrian graves in Viking Age Denmark. Dot = stirrups with/without spurs, triangle = spurs only (after Pedersen 2002:85, fig. 4

90 Husby and the equestrian graves in Angeln and Schwansen

1997b:126, 201:48; Eisenschmidt 2004:233f). For example chamber grave no. 4 from a small necropolis located on the so called Thorsberg in Süderbrarup with a size of 2,9 x 2,1 m was covered by a small mound and contained a long winged head, a pair of spurs and furthermore in the southeast corner a gaming board, a pair of stirrups of the Farsø type, several pieces belonging to the horse furnishing and a silver coin (Aner 1952:70ff, fig. 7–11; Eisenschmidt 2004:552ff, 816ff, pl. 115–118). All this dates the grave probably to the third quarter of the 10th century. Other equestrian graves contained swords instead of . Although there is a slight variation in the weapon furnishing, a striking similarity among the properly excavated equestrian burials can be observed. Naturally, the equestrian and weapon burials from Schleswig must be interpreted in front of a Danish background.

Equestrian and weapon graves in 10th century Denmark Till this day ca. 45 burials consisting of horse furnishings and weapons, another 15 containing horse furnishings and/or spurs alone and about 180 weapon graves that are lacking any form of riding gear are known from Viking Age Denmark including Skåne and Schleswig (fig. 2) (Pedersen 1997a:173; 2012:48). The Fig. 3. Equestrian chamber graves from Rosenlund KR, Fyn (1), crucial factor for the differentiation between the two Süderbrarup IV, Kr. Schleswig-Flensburg (2), Grimstrup, Jylland (3) th main burial groups of the 10 century, equestrian and and Thumby-Bienebek 37A, Kr. Rendsburg-Eckernförde (4) with very weapon graves, therefore is the presence or absence of similar internal layouts and furnishing with grave goods – 1. sword, riding gear (Pedersen 1997b:126). 2. axe, 3. spearhead, 4. shield boss, 5. stirrups, 6. spurs, 7. bucket, 8. According to Randsborg (1980:127) the equestrian gaming board (after Pedersen 1997b:131, fig. 8). graves represented men with certain military duties towards the king, who in return for their services might have received land. This view is partly based on the Eisenschmidt (2004) pass on a lot more source criticism, distribution of the burials in a zone around Jelling, because variations in the grave goods of equestrian and the presumed centre of Harald Bluetooth´s unification weapon burials, and even within these two groups might of the Danish kingdom. To Randsborg´s mind these be expressions of “social and regional differences as well graves, along with the Trelleborg fortifications and as chronological changes within the burial customs and the Danevirke, indicate the boundaries of the core of the religious beliefs” (Pedersen the kingdom. Differences in the grave furnishing he 1997b:124). Another factor might also be the essential perceived as reflections of the different military positions requirement of the higher social classes to visualize their and social statuses of the deceased (Randsborg 1980:129, authority and power as well as their social status and 131ff). military rank through richly furnished graves within Pedersen (1997a; 1997b; 2012) as well as their local communities (Pedersen 2012:45f).

91 New Aspects on Viking-age Urbanism, c. 750-1100

On the basis of changing combinations of Besides Angeln, Schwansen and north Jutland the certain types of stirrups, bridle bits and weapons the equestrian graves are mainly located on , Fyn tenth century equestrian graves show a significant and both sides of the Storebælt, while there has not a chronological depth and are surprisingly precisely single grave with an equivalent rich furnishing been found datable (Pedersen 1997a:179; 1997b:127; 2012:53). in eastern and Scania. According to Pedersen These datings suggest that the custom of depositing this distribution mainly reflects different regional burial riding equipment was first practised in the south of customs (Pedersen 1997b:124; 2012:53f). Beyond Denmark, where the earliest graves are to be found, that the contrast between east and west might even be and then moved northwards, as the latest burials of based on “deep-rooted structural and social differences this kind are known from northern Jutland, especially between these two regions” (Pedersen 2012:55). The in the western region (Pedersen 1997a:180; Storebælt appears to represent a cultural border with 1997b:129). equestrian graves in the west and prominent settlements, The custom probably came into practice around 925 the so-called productive sites, in the east (Jørgensen AD, before the rise of the so called Jelling dynasty 2003:177; Pedersen 2012:55). This border has also been known from the written sources, and was carried out carved out by Sindbæk (2008:195, fig. 6) by mapping approximately until the end of the century (Pedersen different sorts of find material. In the east there was 1997a:178; 1997b:128, fig. 5). Throughout the same probably a rather stably structured society and therefore time and even a little longer than riding gear weapons no need for the display of military/social rank through had been placed in graves (Pedersen 1997b:129f). richly furnished graves, whereas in the west “a socially During the progress of change within the combinations mobile society undergoing change in the tenth century” of grave goods the axe seems to have become the (Pedersen 2012:60) might have felt the urge to do so. prevalent weapon of choice to be deposited in both The low number of equestrian graves in middle and types of burials (Pedersen 2012:52f). western Jutland appears to rather have been caused by

Fig. 4. Quern-Scheersberg. Grave from 1928. Stirrups c. 25.5 cm long and 12.5 cm wide with streaky damascened silver and copper inlays and several pieces of the horse´s equipment.

92 Husby and the equestrian graves in Angeln and Schwansen deviating burial customs or a lower and loosely scattered population than being an expression of a stably structured Traditional and contemporary background society, as it fits very well with the distribution of Viking In Angeln the ties of the Viking Age elite to its local Age grave finds in general (cf. Näsman 1991:166, fig.1; community are probably visible at Süderbrarup and Pedersen 1997a:177). Quern, because at both locations the dominance of Some of the graves furnished with riding equipment certain families in the 10th century might have been may actually have had a connection to the royal power based on centuries long traditions. Süderbrarup was of manifested in monuments from the middle and the great importance in Angeln long before the Viking Age, second half of the 10th century, like the north mound as a huge grave field, which had been in use from the at Jelling (958/959 AD), construction phases of the pre-Roman Iron Age to the 6th century, and the famous Danevirke (connection wall 965/968 AD; Kovirke c. war booty offerings from the younger Roman Iron 980 AD), the Ravning Enge bridge (979/980 AD) and Age in the Thorsberg bog just north of the Viking Age the Trelleborg fortresses that were built around 980 chamber graves suggest (Raddatz 1987; Bantelmann AD (Pedersen 1997a:178f; Andersen 1998:131, 167; 1988; Gebühr 1998:53). A similar situation can Roesdahl 2002:98, 103). But the burial custom had be observed in Quern, where a grave field from the come into practice earlier and was probably still carried Roman Iron Age has been discovered close to the 10th out after King Harald´s death in c. 987 AD (Pedersen century equestrian burials on the Scheersberg (fig. 4) 1997a:179). (Röschmann 1963:492ff; Eisenschmidt 2004:520ff). Nevertheless, it might be assumed that those men Among these graves one was furnished with gold and buried in wooden chambers with weapons, riding silver objects making it the richest grave from the gear and other grave goods at least in the second half older Roman Iron Age in Schleswig-Holstein (Schmidt of the 10th century were provided with military ranks 2006:117f). and played an important role in the formation of the During the younger Roman Iron Age the population Danish kingdom under Harald Bluetooth. But this of Angeln started to decrease continuously in several would probably also apply to the members of the elite stages, which resulted in a decline of the settlements and in eastern Denmark, who did not express their social a concentration of the burials on a few large grave fields, rank through richly furnished graves with weapons the one at Süderbrarup being one of them (Gebühr and riding equipment. In spite of regional differences 1998:52, 55). Apart from only a few known exceptions amongst the c. 45 Danish equestrian graves only certain the depopulation apparently lead to a settlement gap in types of swords, spears, axes, bridle bits and stirrups with the 6th and 7th century. However, together with Willroth fixed combinations were chosen as grave goods. This (1992:477), Gebühr (1998:55) and Dobat (2008:87; speaks for highly standardized and widely established 2010:188, 206) one has to assume that some of those burial practices that probably have been connected with localities, where the remaining rest of the population a military background of the deceased and rights of concentrated at the end of the Roman Iron Age, their families in the society (Pedersen 1997a:175, 181; outlasted the time of the settlement gap. Most probably 1997b:130; 2012:57). Süderbrarup and maybe also the vicinity of Quern were Besides the same narrow range of chosen objects among those localities. also the burial chambers had been constructed with As mentioned at the beginning the state of research the same dimensions of more or less 3 x 2.5 m (fig. concerning the Viking Age is far from being perfect, 3). Furthermore the placement of the deceased and as well. Therefore a look at the equestrian graves from the goods show almost identical layouts of the graves, Süderbrarup and Quern-Scheersberg in relation to although there were long distances between them. contemporary archaeological finds and sites does not According to these observations it is likely that the reveal much. No settlements have been excavated and buried men and their families belonged to “a social elite no interesting objects have been found in the vicinities with contacts reaching far beyond the local community of the graves, except for maybe one: In Sörup, 5,5 km and joined by common values and ritual practices” southwest of the Scheersberg (fig. 1), there has been (Pedersen 2012:48). On the other hand those families found a golden arm ring, which according to Wendt were deeply rooted within their communities (Pedersen (2012:58ff) might be seen in connection with the 2012:57f, 61). Viking Age graves on the Scheersberg. But whether

93 New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 5. Husby in Angeln. Archaeological finds and sites as well as place-names in the vicinity. PRIA=pre-Roman Iron Age, MP=Migration Period, YGIA=younger Germanic Iron Age, VA=Viking Age, MA=high Middle Ages, BA=Bronze Age. The exact location of the thing-site in the surroundings of the church is unknown. this exceptional find can emphasize the predominant to Dobat (2011:51f, 56ff) the distribution of karle- position of the family using the Scheersberg as their place-names in Denmark is rather similar to that of the burial ground or whether it rather indicates other equestrian burials and 10th and 11th century runestones members of an elite resident in or nearby Sörup remains in Denmark, which could suggest a dating of the place- an open question. name to the end of the 10th or the 11th century. 2 km north of the Scheersberg the modern village of Although concrete statements on a contemporary Kalleby is situated. In the context with men, whose settlement context and on the situation from the military ranks were expressed through grave goods in the Migration Period to the Viking Age are not possible at form of weapons and riding gear, this place-name could the moment, the observations mentioned above could be of high interest. It might derive from the Viking Age speak for settlement continuity and might support place-name Karleby, which includes the word karlar as the interpretation that those families represented by a term for warriors or retainers, who could follow the the equestrian graves in Süderbrarup and Quern- king as his lið or hirð (Brink 1999:424f). According Scheersberg had distinct local ties and were embedded

94 Husby and the equestrian graves in Angeln and Schwansen in traditional structures. as well as the highly frequented use of the place name The distribution of equestrian graves in Angeln and Huseby (or similar) and the distribution of these place- Schwansen is somewhat irregular. They were mainly names indicate that the great majority of the Husebyer located in southern Angeln and in Schwansen, which seems to have a royal background (Brink 2000b:276f; might be explained by the Schlei fjord as an important Westerdahl & Stylegar 2004:114). The standardized waterway, either in an economic or a military sense. *húsabýr-appellation, which most probably can be The graves in southern Angeln additionally surrounded regarded as a technical or administrative term denoting probable waterways of the stream system of the “royal farm or manor”, has in each case replaced Füsinger Au. Furthermore most of the equestrian the earlier name of a settlement (Brink 2000a:285; graves seem to have been connected to the ancient road 2000b:276f). However, the precise dating of a presumed system, reconstructed on the basis of rows of Stone and Huseby-system is still not certain. According to Brink Bronze Age grave mounds and historical maps from the (2000b:276f) one of the most plausible 17th and 18th centuries. Some of the graves were even of Huseby is “the with the king´s storehouses”. located in the close vicinities of crossroads. Especially Westerdahl & Stylegar (2004:105) perceive the meaning the graves at on the hærvej (the “army road”) of Huseby as “a place with many houses” and argue and Süderbrarup mark important strategic positions that it must have been buildings with many different in the landscape (fig. 1). These observations suggest functions. that some of the families represented by the equestrian In the surroundings of Husebyer very often graves might have achieved their social status through impressive archaeological objects and sites from the control of those strategic positions in the landscape younger Germanic Iron Age/Vendel Period or the and thereby playing a decisive role in the distribution Viking Age are found; like for instance spurs made of goods within Angeln (cf. Pedersen 2012:61). During of gold in Huseby in Rygge or such as large burial the unification process of the Danish kingdom in the mounds and/or large cemeteries (Hyenstrand 1974; second half of the 10th century they had probably been Brink 2000a:288; 2000b:277f; Westerdahl & Stylegar of importance for the royal power in order to gain or 2004:109.1 In terms of indications of central places (cf. maintain control over the territory. Fabech & Ringtved 1995:19ff, fig. 7) these sites and finds indicate a former importance of those hamlets or manors. This impression is supported by excavated Husby in Angeln hall buildings at Huseby in Tjølling, , Norway With regard to the reconstructed ancient road system as well as Husaby in Glanshammar, Närke and another strategic position in the landscape is taken by Huseby in Värend, Småland, both Sweden (Ekman the village Husby in the north of Angeln c. 7 km south 2000:19ff; Brink 2000b:278; Skre 2007:231–243), of the Flensborg fjord (fig. 1). This location is passed by as this type of building probably represented lordship several long-distance roads and may represent one of and power. Moreover very often early churches are the more important nodal points in Angeln. So far no located at Husebyer in Norway and Sweden; among equestrian grave has been discovered there, but the place these two of the 14 medieval Norwegian royal name itself points to a settlement of some importance (Brink 2000b:278; Westerdahl & Stylegar 2004:119). in the late Viking Age. Husby, mentioned in the written According to their locations on roads and waterways sources as “in Husebuy” for the first time in 1319 (Laur and therefore at strategic positions within the landscape 1992:355), is the southernmost of more than 130 many Husebyer might have served the purpose of hamlets or farms in Scandinavia and the Orkneys known controlling transport and communication (Ekman by the name Husaby, Huseby, Husby, Hoseby, Husebø 2000:12; Pettersson 2000:59; Brink 2000b:278; etc. They all derive from an Old Norse *húsabýr (Brink Westerdahl & Stylegar 2004:112). And finally, especially 2000b:274). While the Huseby-places are relatively well in Sweden significant place-names very often occur in studied archaeologically and onomastically in Sweden the vicinities of Husebyer, indicating lordship, , and Norway, they have not attracted much attention in handcraft and pagan cult, as Brink (1996; 1997; 1999) Denmark, so far. was able to point out. The Husebyer seem to have been References in Swedish provincial law-codes and a embedded within these structures of older central place diploma from 1233 by the Danish king Valdemar II complexes which functions probably at least partly were

95 New Aspects on Viking-age Urbanism, c. 750-1100 still intact. Although the observations mentioned above characterize a high number of Husebyer as important settlements, nevertheless, there are several that represent merely small farms located on the periphery. As it appears today, not all Husebyer seem to have had the same quality or background (Brink 2000b:278; Westerdahl & Stylegar 2004:109).

The archaeological background of Husby Almost all of the Huseby-characteristics mentioned above can also be found at Husby in Angeln. In the late Pre-Roman Iron Age, around 100 BC, a very important man was buried in a small stone chamber together with the iron remnants of a wagon and a large imported bronze vessel. This cremation grave, which is one of only three wagon graves in south Scandinavia, represents the richest burial from the Pre-Roman Iron Age in Schleswig-Holstein (Raddatz 1967; Gebühr 2000:279; Schovsbo 2010:22ff). The burial belongs to the oldest graves of a cemetery, which then continuously had been in use until the 5th century (fig. 5). The high amount of more than 1300 exclusively male (!) graves allows the conclusion to be drawn that there has been a high density of population in this area, especially in the older Roman Iron Age (Gebühr 2000:279; Rau 2010:201, fig. 77, 142). Just like in Süderbrarup and Quern also in Husby there is a gap within the archaeological find material from the Migration Period to the Viking Age like it appears with little variation everywhere in Schleswig- Fig. 6. Husby, LA 50. Collection of finds from the younger Germanic Holstein. This phenomenon, which is partly based Iron Age/Vendel Period and Viking Age. 1 rectangular fibula, gilded on the state of research, is a particular archaeological bronze; 2–3 fragments of rectangular fibulae, bronze; 4 fragment of an problem that has to be dealt with in the future. The next oval brooch, gilded bronze; 5 “Kreuz-Emaille-Scheibenfibel”, bronze, secure find from the close vicinity of Husby used to be remnants of red enamel; 6 three-dimensional disc brooch, Borre-style, gilded bronze; 7 fibula, Urnes-style, lead-tin. Scale 1:1. a hoard of more than 300 silver coins of the Dorestad- type that had been minted in Hedeby between 950 and 980 AD (Wiechmann 1996:141f, 269). This hoard church in an exposed location on a sandy height might have represented a respectable value and may called Hunbarg/Hundebjerg, which would have been be regarded as an outstanding find that could indicate visible from a long distance, if the vegetation had been a certain degree of importance of Husby in the late relatively low (fig. 5). Directly east of the height there Viking Age (Lemm 2010:61). Further archaeological is an old road that was known by the name kongevej finds from the first millennium have not been known (King´s highway) in the Medieval period and leads from Husby until a few years ago. from the Holnis peninsula on the Flensborg inlet in the In September 2011 systematic metal detector surveys North to Schleswig respectively the inner Schlei-region 2 revealed a settlement site (Husby, LA 50) , which seems in the South (fig. 1; fig. 5) (Röschmann 1963:106). The to have the potential to solve some of the problems favourable location at the crossroads of long distance related to the Scandinavian Huseby phenomenon. connections probably represented one determinant The site is situated c. 300 m south of the Romanesque factor fort the choice of place and for the importance

96 Husby and the equestrian graves in Angeln and Schwansen

II, more precisely the south Scandinavian style C (after Ørsnes 1966), depicts five intertwined creatures and suggests a dating to the second half of the 7th century (cf. Høilund Nielsen 1991:130; Jørgensen & Nørgård Jørgensen1997:28–31, fig. 18, 35, fig. 24). To the same period date two other bronze rectangular fibulae that are fragmented and less sophisticatedly decorated in the same art style (fig. 6:2–3). A little younger, from the beginning of the 8th century (cf. Høilund Nielsen 1991:130; Rundkvist 2010:135f, 157 tab. 10), is a hacked oval brooch of an early type, which shows the rest of gilding in the background of its complex ornamentation (fig. 6:4) (probably variant N Id after Rundkvist 2010). While the fragmented objects could be thought of as elder pieces of scrap metal, such an assumption is rather unlikely for the gilded rectangular fibula. Because of its not fragmented, but complete preservation and its high quality it seems acceptable to regard this magnificent piece of craftsmanship as proof for the actual presence of persons of high social standing at this place in the 7th century. According to the prevalent view of research members of the elite wore animal style objects in order to present their high social status (Høilund Nielsen 1998:2ff, 10f; Pesch 2011:273ff; 2012:660). The find material from the Viking Age consists of a wide range of objects (cf. Lemm 2014). Among the fibulae and brooches there is Fig. 7. Husby, LA 50. Collection of finds from the Viking Age. 1 a common continental type, a “Kreuz-Emaille-Fibel”, Christiana-Religio-denarius; 2, 4–5 fragments of dirhams; as well as typical Scandinavian examples in Borre and 3, 7 fragments of silver bracelets; 6 hacked fragment of a silver bar; Urnes style and fragments of trefoil and rectangular 8 fragment of a silver ingot; 9, 13 truncated-spherical weights, brass coated iron; 10–12 cubo-octahedral weights, bronze; 14 cubic weight, fibulae (fig. 7). Fragmented Arabian dirhams, a Frankish lead-tin. Scale 1:1. Christiana-Religio-Denar, pieces of silver bracelets and silver ingots can be regarded as means of payment or of Husby (Lemm 2010:61). East of the height and the as casting material. Different weights of the truncated- kongevej there is a large, formerly impassable wetland spherical, the cubo-octahedral and the cubic type can area, which created a sort of a pass-situation at the be associated with trade or the production of metal foot of the settlement and thereby provided it with a objects. The latter is also suggested by many remains strategic position. of melted copper-alloy and lead. Unfortunately the The find material recovered until the end of 2012 production of metal objects at this point of time cannot ranges chronologically from the second half of the 7th be precisely dated to the younger Germanic Iron Age/ century to the medieval period. Among the finds the Vendel Period, the Viking Age or the Medieval Period. finest object is a nearly completely preserved gilded While the magnificent rectangular fibula of the th bronze rectangular fibula of the highest quality with 7 and perhaps also the gilded oval brooch from the th a size of 5 x 2.5 cm (fig. 6:1). The decoration of this early 8 century give reason to assume a local elite, magnificent piece belongs to the Germanic animal style outstanding artefacts from the Viking Age are yet to be

97 New Aspects on Viking-age Urbanism, c. 750-1100 found amongst the archaeological material. Currently to the West-Norse horgr. The place-name Helleager it is just the hoard of more than 300 silver coins found meaning “holy acre” and being located about 1 km to about 2 kilometres further north, which might point the east could relate to the presumable horgr or indicate to the presence of members of elite circles of society at another cultic place in this vicinity (Hald 1948:8, 208; the end of the 10th century. Nevertheless, with all due Laur 2001:207). caution the quantity and the compilation of the finds As a result of the advanced Christianization as well as their chronological range set the site in the presumably a stave church had been built in Husby in light of those settlements that are neutrally described the 11th century, which around 1200 AD was replaced as “metal-rich places” or “productive sites” and in all by a Romanesque church made of granite ashlars. So likelihood adopted different central location functions. far, there is no archaeological evidence on hand for a This site probably has to be regarded as “the hamlet wooden church, but most probably it is depicted on the with the king´s storehouses” respectively “the place medieval seal of the Husby-Harde (Hinz 1981). The with many houses” (cf. Brink 2000b:276; Westerdahl wooden church would in a way have superseded the & Stylegar 2004:105). sanctuaries discussed above as cultic component in the ritual landscape. The “metal-rich” settlement site from the th7 to Indications of „power and cult“ in Husby approximately the 12th/13th century, the silver hoard In the Medieval Period Husby was the centre of the from the end of the 10th century, the conclusion of formerly largest Harde3 in Angeln, which probably an early wooden church from the late Viking Age, the originally covered the whole of northern part of the strategic location at a crossroads and last but not least territory (Laur 1992:355). Close to the church in the theophoric place-names create the image of an Husby the thing was held (Klauder 1932:26; Hamer important settlement in political, cultic and probably 1994:345). Place-names like Galgehøj (gallows hill), also economical respects, both before and after its Galgemose (gallows moor) and Galgemark (gallows promotion to the Huseby-status. This outstanding field) clearly point to the fact that justice had been importance provided Husby with a special role in the spoken here. Although the earliest mention of the northern part of Angeln in the younger Germanic Iron thing in the written sources dates only from 1400 („aa Age/Vendel Period, the Viking Age and the Medieval Husby thing“; Laur 1992:355), it is quite likely that this Period. If the gap in the archaeological find material institution at the site was based on an older tradition. between the 5th and the 7th century can be filled in About 1.8 km north-west of Husby there is a now the future and thereby a settlement continuity made silted lake of approximately 700 x 300 m extension provable, the continuous importance of Husby might bearing the name Wonsdamm respectively Wonskjær/ even be traced all the way back to the great man, who Vonskær (fig. 5), most certainly denoting a holy lake found his last resting place in the famous wagon grave or bog that was once dedicated to the god Odin (Hald from the time around 100 BC4. 1948:29f, 184; Laur 1954:114; 2001:108 f). A height c. 500 m south-west of the Romanesque church in Husby bears the name Wieh, which according to The king´s hirð in Angeln and Schwansen Hald (1948:192) and Laur (1954:115) probably is The custom of depositing riding equipment in graves an equivalent to the Scandinavian ví, meaning sacred seems to have been carried out in Denmark mainly place or sanctuary (Vikstrand 2001:298). Place-names between c. 925 and 975 AD (Pedersen 1997b:128). Thus like Trolddel and Troldmose in the immediate vicinity it had come into practice before the rise of the Jelling referring to trolls as creatures of the medieval Nordic dynasty and was probably still carried out after King mythology as well as the direct proximity to the thing Harald had died in c. 987 AD (Pedersen 1997a:179). and execution site of the later Husby-Harde might The phenomenon of richly furnished equestrian graves support this interpretation (Laur 2001:209). About therefore cannot be directly or rather not exclusively 3 km south of Husby the place-names Harresbjerg, explained through impulses of the Jelling dynasty. At Harreshøj, Harreskrog and Helleager are to be found. least not in the form of warriors of the higher classes The first three contain the word *Harre, which probably being stationed on the borders of the mid-tenth century derives from an Old-. *hargh and corresponds Jelling .

98 Husby and the equestrian graves in Angeln and Schwansen

Fig. 8. Angeln and Schwansen. Finds of a certain type of bridle mounts in Borre-style in Quern (grave from 1928, 3 pieces), Langballigau (mound 1, 2 pieces), Thumby-Bienebek (grave K 37A, 3 identical pieces and a slightly different one), Hedeby and (after: Müller-Wille 1973:106, fig. 29; 1977:45, fig. 3,2–3; 1987, pl. 76; Meinhardt & Siegloff 2014:4). For an explanation of the symbols see fig. 1. Scale 1:3.

The equestrian graves might be regarded as differences in the burial customs, the quantity of weapons particularly strong expressions of pagan burial customs in the graves, the presence respectively the absence of as reaction to the progressing Christianization of riding equipment, as well as the amount and quality of Denmark (cf. Pedersen 2002:89f). These graves might the grave goods all taken together lead Pedersen to the also have been the result of intensified contacts with the conclusion that the burials express graduations in the East-Frankish Empire, where as warriors had rank and the social position of the buried and their kin. become increasingly important in the fighting against Considering the highly standardized characteristics of the and the Slavonic tribes since the second many graves “the use of weapons was probably governed quarter of the 10th century. On the one hand these by conventions and regulations. This again suggests the customs may have simply ceased towards the end of the possibility of military offices and an understanding 10th century, but on the other hand the social elite may of military organization” (Pedersen 2012:60); offices, just as well have satisfied their needs for display rank which as , dreng and are known from 10th and power by burying their dead in or next to churches century runic inscriptions (Pedersen 2012:46). (Pedersen 1997b:130, 132; 2002:90; 2011:63f). These individuals at the top of “a locally defined In spite of obvious regional and chronological social and possibly military hierarchy” (Pedersen

99 New Aspects on Viking-age Urbanism, c. 750-1100

2012:61) with supraregional contacts most definitely course of the 11th century were royal property (Callmer played an important role in the formation of the Danish 2002:77). According to Callmer the regional rulers had kingdom under Harald Bluetooth. They served the king played out their roles towards the end of the 10th century possibly in exchange for land, as Randsborg (1980:127) and new structures were formed that were more closely suggested or because they received valuable objects like linked to an apparatus of power under direct control of perhaps the arm ring from Sörup mentioned above or the king. Referring to the title of this publication this was riding equipment (Pedersen 1997b:132). The latter the time of a restart in Denmark. might be observed in the Schlei region (fig. 8), for there The new territorial order and administrative structures were found sets of very similar though not in all cases were no longer merely based on military force and identical bridle mounts in three graves in Langballigau networks of the local elites. The bands of allied (mound 1: 2 pieces), Quern-Scheersberg (grave from chieftains and the relatively small number of the king´s 1928: 3 pieces) and Thumby Bienebek (grave K 37A: retainers had probably not been sufficient to oppose 4 pieces; Eisenschmidt 2004:157). These gilded bronze effectively the political and military pressure evoked by items that have to be regarded as mass products in Borre- the powerful East-Frankish Empire, which had been style and therefore do not represent precious objects the reason and the triggering factor for a new policy of could nevertheless indicate some early kind of uniform concentrating power in Denmark (Callmer 2002:77). or insignia of the equestrian warriors in this region. A Nor would the rather questionable enforced or bought very similar piece has been found at a newly discovered loyalties of chieftains have been the ideal choice of “metal-rich” settlement site in Ellingstedt, situated political structures for a long-term unification of the almost halfway on the road between and kingdom. Therefore the new administrative structures Hedeby in the middle of the so-called Schleswiger were instead based on proto-urban centres like Roskilde, Landenge, the shortest distance to cross the Jutland Viborg, Aalborg, Odense, Lund and later Schleswig peninsula (Meinhardt & Siegloff 2014). One could very (Callmer 1994:72ff; 2002:77f; 2007:247, 265; Böhme well imagine that one of the king´s high ranking officers 2001:492ff; Sindbæk 2009:105). They were also based controlled this settlement, the important road and the on ecclesiastical administration and probably also in Danevirke, which lies only a about one kilometre to the Denmark based on Huseby-places at some time in the south. One and a half objects of this type from Hedeby 11th century. might suggest that these mounts have been produced Most Huseby-places in Denmark seem to have been there in a central workshop and then were given out located near the coast, although not on the coast, and to retainers on behalf of the king. Other than in south might therefore have been connected to sea-routes (fig. Schleswig one exemplar of this type of bridle mounts 9). Husby in Angeln was situated at the crossing of has been found at a “productive site” in Vester Egesborg important long-distance roads. But what administrative in the south of Sjælland (Ulriksen 2006:193, fig. 9). units did they refer to? Due to their regional and local distribution Steinnes (1955:120) linked them with administrative and Time of a restart military functions. Both Steinnes (1955:117ff) and Together with the Trelleborg-fortresses, the Ravning Enge Andrén (1983:50f, fig. 11) point to the apparent clear bridge and some parts of the Danevirke the equestrian geographic connection between the Husebyer and the warriors represent aspects of the strongly military based historic -division, where there is at most one Huseby formation of the Danish kingdom, which was established per syssel5. Although some sysler are lacking Husebyer, soon after 980 AD (cf. Roesdahl 2002; Pedersen the Huseby-organisation according to Andrén could 2012:46). Around this time the custom of depositing correspond to the syssel-based system of the kongelev6 riding gear and/or weapons in burials was abandoned. that had been set out in writing in king Valdemar´s II Roughly at the same time central places like for instance jordebog in 1231. Andrén (1983:50) considers that the Tissø, Lejre, Uppåkra, Järrestad and in the Schlei region Husebyer have to be regarded as remnants of an older Füsing lost their importance or were completely given up administrative system; from a time when the property and were superseded by proto-urban centres at nearby connected to the royal office had not been explicitly locations. Judging by the few written sources all of the separated from the other belongings of the royal new central places that developed into towns in the family. Referring to the chronological and geographical

100 Husby and the equestrian graves in Angeln and Schwansen distribution of the Danish rune stones by Randsborg in Jellingsyssel has recently been identified as a fairly (1980:80) Andrén suggests a dating of the Old-Danish young settlement from the end of the 18th or the early Husebyer to the late 10th or the early 11th century. 19th century (Christensen 2014). One could argue that Braathen (1989:158ff) also regards the Husebyer in Jelling as royal seat would not have required a Huseby connection with the medieval syssel-division. Likewise in its syssel, but according to Braathen´s model there he takes into consideration the 10th century equestrian should be either a Huseby or equestrian graves. It graves compiled by Brøndsted (1936), who did not is questionable in the first place, whether the syssel- include the graves south of the Danish-German border. division is the right background for an interpretation of Like the equestrian graves by Randsborg (1980:129) the Husebyer´s distribution, as it only was established and the Husebyer by Steinnes (1955:120f), both are in Jutland, but not in Fyn, Sjælland or Blekinge, where regarded by Braathen in connection to the royal seat there are Husebyer as well. in Jelling. To his mind their distribution reveals such In Sweden and Norway the Huseby-places a systematically conducted, almost geometrically built agglomerate in those that are not assumed pattern of either a Huseby or equestrian graves in each to have been the point of origin for the particular of the sysler that they probably characterize one and the unification of each of the two (Westerdahl same sphere of interest. Both Husebyer and equestrian & Stylegar 2004:116ff). For Westerdahl & Stylegar graves are supposed to have been placed geographically (2004:116, 125) the opposite seems to be the case in by deliberate planning (Braathen 1989:158ff). Denmark, as the concentration of Husebyer is to be At first sight Braathen´s map of burials with riding found in Jutland and around Jelling. Therefore they gear in a zone between 75 and 120 km radiuses around put up for discussion, whether the Husebyer, which Jelling is striking (Braathen 1989:159, fig. 46), even if they comprehend against a political and economic there are a few that do not lie on his “belt of equestrian background, represent a by origin Danish phenomenon, graves”. But just as Randsborg´s also Braathen´s dealing which was closely linked with the unification process. with these graves is lacking the different source critical Another interpretation of the Danish Husebyer has factors that were outlined above. Moreover it seems recently been yielded by Iversen (2011:233f), who as if he perceives the graves as intentionally placed correlates the geographical distribution of the Husebyer symbols to mark the royal sphere of power. By all with the earliest borders (fig. 9). The Huseby in means the funerals of the elite also served the purpose Hardsyssel situated close to the Jutish west coast and to demonstrate and perhaps justify power (Pedersen on the northern edge of the Ribe diocese could to his 2012:45f), but this was meant to prove the authority mind once have belonged to the Viborg diocese, where of the deceased and his family and not the king´s. The otherwise no Huseby is found; in each other diocese king himself naturally needed followers among the there was only one Huseby (Iversen 2011:234)7. local elites, but they were most certainly also situated in It is hardly astonishing that the administrative regions without equestrian burials, including the area division of Late Viking Age Denmark might have around Jelling. been connected to ecclesiastical borders. When Harald Furthermore several observations, some of them Bluetooth converted to Christianity he did this hardly already made by Braathen (1989:158) himself, give out of religious conviction. He probably did it to be reason to doubt his theory on a close connection recognized by the ruler of the East-Frankish Empire between the equestrian graves and the Husebyer: and the west European kings as an equal sovereign ruler Besides Hardsyssel, mentioned by (cf. Hoffmann 1984:116f; Pedersen 2002:81). At the Braathen, a combination of a Huseby and equestrian same time the Christian belief freely based on the motto graves is also present in Istedsyssel, namely in Angeln “one king, one god – one king through god” assured and Schwansen. On the other hand there are neither his position among the Danish people and provided Husebyer nor equestrian burials in Vendsyssel and the basis for a long-term unification of Denmark. Barvedsyssel. One may also bring into question, Furthermore the ability to write and the centuries-long whether one grave each in Hardsyssel and in Vardesyssel experience in administrative work of the were are enough to back Braathen´s theory, especially as vital factors in order to govern the kingdom in the 11th the grave from Grimstrup is not located within his century (cf. Westerdahl & Stylegar 2004:116). “belt of equestrian graves”. Husby in Grejs parish Ecclesiastical administrative structures may have

101 New Aspects on Viking-age Urbanism, c. 750-1100 brought forward or even were required for the installation of political administrative structures in the Danish kingdom (cf. Callmer 2007:265). It is quite possible that the documented by the written sources were identical with secular administrative units. It is also possible that these units had already been older pre-Christian divisions that later were used as diocese borders. Within the new structures of the 11th century, according to Frode Iversen, the Husebyer might have functioned as “storage facilities for the king´s income from these regions, […] where the bishop played a major role collecting taxes and revenues on behalf of the king” (Iversen 2011:234). It is probable that the Husebyer as well as the proto-urban centres and the ecclesiastical institutions made a contribution to the administration of the Danish kingdom, which in the course of the 11th century was stepping closer to the feudal order of the Medieval Period. But certainly more research on the Danish Husebyer has to be done in order to support or challenge this hypothesis. In the absence of archaeological excavations/ investigations and detailed context analyses the current state of research does not allow solid conclusions concerning most of the old-Danish Huseby-places. But without doubt Husby in Angeln has to be regarded in a row with important Huseby-places in Sweden and Norway. Assuming that this might also apply to the other old-Danish Husebyer these places might have played a key role in the administration of the kingdom in the 11th century.

Corresponding Author Thorsten Lemm Centre for Baltic and Scandinavian Archaeology (ZBSA), Schloss Gottorf Schleswig, Germany [email protected]

102 Husby and the equestrian graves in Angeln and Schwansen

Baltic Region c. AD 700–1100: Trading Places, References Centres and early Urban Sites. In: B. Ambrosiani & Andersen, H.H. 1998. Danevirke og Kovirke. H. Clarke (eds.) Developments Around the Baltic and Arkæologiske undersøgelser 1861–1993. Jysk the North Sea in the Viking Age. The Twelfth Viking Arkæologisk Selskab. Congress = Birka Studies 3. Stockholm. 50 – 90. Andrén, A. 1983. Städer och kungamakt – en studie i – 2002. Machtzentren des 10. Jahrhunderts und der Danmarks politiska geografi före 1230. Scandia Zeit um 1000 in Südskandinavien. In: J. Henning 1983 (49). 31 – 76. (ed.) Europa im 10. Jahrhundert. Archäologie einer Aner, E. 1952. Die wikingerzeitlichen Kammergräber Aufbruchszeit. Internationale Tagung in Vorbereitung am Thorsberger Moor. Offa 1952 (11). 60 – 78. der Ausstellung Otto der Große, und Arents, U. & Eisenschmidt, S. 2010. Die Gräber von Europa“. Mainz. 65 – 79. Haithabu 1. Die Ausgrabungen in Haithabu 15. – 2007. Urbanisation in Northern and Eastern Europe, Neumünster. ca. AD 700–1100. In: J. Henning (ed.) The Heirs Bantelmann, N. 1988. Süderbrarup. Ein Gräberfeld der of the Roman West. Post-Roman Towns and Trade in römischen Kaiserzeit und Völkerwanderungszeit Europe, Byzantium and the Near-East 1. Millenium- in Angeln 1. Archäologische Untersuchungen. Studien zu Kultur und Geschichte des ersten Urnenfriedhöfe Schleswig-Holsteins 11:1. Offa Jahrtausends n. Chr. 5. Berlin. 233 – 270. Bücher 63. Neumünster. Christensen, L. & Eilersgaard, 2014. Stednavne og Böhme, H.W. 2001. Gedanken zu den frühen Markt- arkæologi omkring Jelling i Østjylland. In: B. Eggert und Handelsplätzen in Südskandinavien. In: E. & R. Olesen Steenholt (eds.) Navne og skel. Rapport Pohl, U. Recker, & C. Theune (eds.) Archäologisches fra den femtende nordiske navneforskerkongres i Askov Zellwerk. Beiträge zur Kulturgeschichte in Europa und den 6.-9. juni 2012. Uppsala. In press. Asien. Festschrift Helmuth Roth zum 60. Geburtstag. Dobat, A. S. 2008. angulus non desertus! II. Erste Internationale Archäologie. Studia honoria 16. Ergebnisse der Suche nach Siedlungen des 6. und 7. Rahden/Westf. 483 – 498. Jahrhunderts in Südschleswig. Arkæologi i Slesvig/ Braathen, H. 1989. Ryttergraver. Politiske strukturer i Archäologie in Schleswig 11. Neumünster. 87 – 94. eldre rikssamlingstid. Varia 19. Oslo. – 2010. Füsing. Ein frühmittelalterlicher Zentralplatz im Brink, S. 1996. Political and social structures in early Umfeld von Haithabu / Schleswig. Bericht über die Scandinavia. A settlement-historical Pre-study of the Ergebnisse der Prospektionen 2003–2005. Studien zu Central Place. Tor 1996 (28). 235 – 282. Haithabu und Füsing. Die Ausgrabungen in – 1997. Political and social structures in early Haithabu 16. Neumünster. Scandinavia 2. Aspects of space and territoriality – – 2011. Mapping social order. Place-names containing the settlement district. rink or karl and the development of political Tor 1997 (29). 389 – 437. administration in 10th century South-Scandinavia. – 1999. Social order in the early Scandinavian landscape. In: L. Eilersgaard Christensen & B. Jørgensen (eds.) In: C. Fabech & J. Ringtved (eds.) Settlement and Navnemiljøer og samfund i jernalder og vikingetid. Landscape. Proceedings of a conference in Århus, NORNA-Rapporter 86. Uppsala. 41 – 73. Denmark May 4–7 1998. Højbjerg. 423 – 439. Eisenschmidt, S. 2004. Grabfunde des 8. bis 11. – 2000a. Nordens husabyar – unga eller gamla? In: I. Jahrhunderts zwischen Kongeå und Eider. Zur Fuglestvedt, T. Gansum & A. Opedal (eds.) Et Bestattungssitte der Wikingerzeit im südlichen hus med mange rom. Vennebok til Bjørn Myhre på Altdänemark. Studien zur Siedlungsgeschichte und 60-årsdagen. Stavanger. 283 – 291. Archäologie der Ostseegebiete 5:1. Neumünster. – 2000b. Headword ”Husby”. In: J. Hoops (ed.) Ekman, T. 2000. Item Husaby in Niericia. In: M. Reallexikon der Germanischen Altertumskunde 15. Olausson (ed.) En bok om Husbyar. Uppsala. 9 – 38. Berlin & New York. 274 – 278. Fabech, C. & Ringtved, J. 1995. Magtens geografi i Brøndsted, J. 1936. Danish inhumation graves of the Sydskandinavien – om kulturlandskab, produktion Viking Age. A Survey. Acta Archaeologica 1936 (7). og bebyggelsesmønster. In: H. G. Resi (ed.) 81 – 228. Produksjon og Samfunn. Beretning fra 2. nordiske Callmer, J. 1994. Urbanization in Scandinavia and the jernaldersymposium på Granavolden Gjæstgiveri 7.–10

103 New Aspects on Viking-age Urbanism, c. 750-1100

mai 1992. Universitetets Oldsaksamling. Varia 30. Serie B:14. Det kongelige nordiske oldskriftselskab. Oslo. 11 – 37. København. Gebühr, M. 1998. Angulus desertus? Studien zur Klauder, O. 1932. Die Husbyharde in ältester Zeit. Sachsenforschung 1998 (11). 43 – 85. Jahrbuch des Angler Heimatvereins 3. 20 – 26. – 2000. Headword ”Husby“. In: J. Hoops (ed.) Laur, W. 1954. Theophore Ortsnamen in der Landschaft Reallexikon der Germanischen Altertumskunde 15. Angeln. Jahrbuch des Angler Heimatvereins 18. 106 – Berlin & New York. 278 – 280. 118. Hald, K. 1948. Nordangel og Flensborg. Sydslesvigs – 1992. Historisches Ortsnamenlexikon von Schleswig- Stednavne 7. København. Holstein. Veröffentlichungen des Schleswig- Hamer, B. 1994. Topographie der Landschaft Angeln 1. Holsteinischen Landesarchivs 28. Neumünster. . – 2001. Germanische Heiligtümer und Religion im Hinz, H. 1981. Stabkirchen im Landesteil Schleswig. Spiegel der Ortsnamen. Kieler Beiträge zur deutschen Offa 1981 (38). 349 – 355. Sprachgeschichte 21. Neumünster. Hoffmann, E. 1984. Beiträge zur Geschichte der Lemm, Th. 2010. Husby i Angeln. Fornvännen 2010 Beziehungen zwischen dem deutschen und dem (105):1. 161–163. dänischen Reich für die Zeit von 943 bis 1035. In: – 2014. Husby in Angeln – Ein königlicher Hof der Ch. Radtke & W. Körber (eds.) 850 Jahre St. Petri- späten Wikingerzeit? Praehistorische Zeitschrift 2014 Dom zu Schleswig 1134–1984. Schleswig. 105 – (89):2. 371 – 389. 132. Meinhardt, S. & Siegloff, E.2014. Das liegt doch auf Hyenstrand, Å. 1974. Centralbygd-Randbygd. dem Weg! Reiche wikingerzeitliche Siedlung. Strukturella, ekonomiska och administrativa Archäologie in Deutschland 2014 (1). 4 – 5. huvudlinjer i mellansvensk yngre järnålder. Acta Müller-Wille, M. 1973. Das jüngereisenzeitliche Universitatis Stockholmiensis. Studies in North- Gräberfeld. In: F. Voss, M. Müller-Wille & E.-W. European Archaeology 5. Stockholm. Raabe (eds.) Das Höftland von Langballigau an der Høilund Nielsen, K. 1991. Territoriale studier af dyrestil Flensburger Förde. Offa 1973 (30). 82 – 123. og kvindesmykker i yngre germansk jernalder i Syd- – 1977. Krieger und Reiter im Spiegel früh- und hoch og Østskandinavien. In: P. Mortensen & B. M. mittelalterlicher Funde Schleswig-Holsteins. Offa Rasmussen (eds.) Fra Stamme til Stat i Danmark 2. 1977 (34). 40 – 74. Jysk Arkæologisk Selskabs Skrifter XXII:2. Aarhus. – 1987. Das wikingerzeitliche Gräberfeld von Thumby- 127 – 154. Bienebek (Kr. Rendsburg-Eckernförde) II. Offa-Bücher – 1998. Animal Style – A symbol of Might and Myth. 62. Neumünster. Salin´s Style II in a European Context. Acta Näsman, U. 1991. Grav og økse. Mammen og den Archaeologica 1998 danske vikingetids våbengrave. In: M. Iversen (red.) (69). 1 – 52. Mammen. Grav, kunst og samfund i vikingetid. Jysk Iversen, F. 2011. The of Bona Regalia and the Arkæologisk Selskabs Skrifter XXVIII. Højbjerg. Growth of Supra-regional Powers in Scandinavia. In: 163 – 180. S. Sigmundsson (ed.) Viking Settlements and Viking Pedersen, A. 1997a. Similar Finds – Different Society. Papers from the Proceedings of the Sixteenth Meanings? In: C. K. Jensen & K. Høilund Nielsen Viking Congress. Reykjavík and Reykholt, 16-23 (eds.) Burial & Society The Chronological and August 2009. Reykjavík. 225 – 244. Social Analysis of Archaeological Burial Data.. Jørgensen, L. 2003. Manor and Market at Lake Tissø in Aarhus. 171 – 183. the Sixth to Eleventh Centuries: The Danish – 1997b. Weapons and riding gear in burials – evidence „Productive” Sites. In: T. Pestell & K. Ulmschneider of military and social rank in 10th century Denmark? (eds.) Markets in Early Medieval Europe. Trading and In: A. Nørgård Jørgensen & B.L. Clausen (eds.) “Productive” Sites, 650–850. Macclesfield. 175 – Military Aspects of Scandinavian Society in a European 207. Perspective, AD 1–1300. Papers from an International Jørgensen, L. & Nørgård Jørgensen, A. 1997. Nørre Research Seminar at the Danish National Museum, Sandegård Vest – A Cemetery from the 6th–8th Copenhagen, 2–4 May 1996. Copenhagen. 123 – Centuries on Bornholm. Nordiske Fortidsminder 135.

104 Husby and the equestrian graves in Angeln and Schwansen

– 2002. Prachtgräber des 10. Jahrhunderts in Rundkvist, M. 2010. Domed oblong brooches of Südskandinavien – Tradition und Erneuerung. In: J. Vendel Period Scandinavia. Ørsnes types N & O Henning (ed.) Europa im 10. Jahrhundert. Archäologie and similar brooches, including transitional types einer Aufbruchszeit. Internationale Tagung in surviving into the Early Viking Period. In: B. Hårdh Vorbereitung der Ausstellung „Otto der Große, (ed.) Från romartida skalpeller till senvikingatida Magdeburg und Europa“. Mainz. 81 – 94. urnesspännen. Nya materialstudier från Uppåkra. – 2011. Nabo, fjende og forbillede – Danernes forhold Uppåkrastudier 11. Kristianstad. 127 – 199. til Tyskland i det arkæologiske fundbillede. In: N. Schmidt, J.P. 2006. Goldberlocken – Meisterwerke Lund N (ed.) Beretning fra det tredivte tværfaglige kaiserzeitlicher Goldschmiedekunst. In: B. Bleile vikingesymposium. Danmark og Tyskland i det 10. (ed.) Magischer Glanz. Gold aus archäologischen århundrede. Højbjerg. 55 – 82. Sammlungen Norddeutschlands. Weding. 116 – 121. Pesch, A. 2011. Netzwerk der Zentralplätze. Schovsbo, P.O. 2010. Dejbjergvognene. Keltiske impulser Elitenkontakte und Zusammen­arbeit früh i førromersk jernalder. Jysk Arkæologisk Selskabs mittelalterlicher Reichtumszentren im Spiegel der Skrifter 74. Højbjerg. Goldbrakteaten. In: W. Heizmann & M. Axboe Sindbæk, S.M. 2008. The Lands of Denemearc: (eds.) Die Goldbrakteaten der Völkerwanderungs­zeit Cultural Difference and Social Networks of the – Auswertung und Neufunde. Ergänzungsbände Viking Age in South Scandinavia. Viking and RGA 40. Berlin & New York. 231 – 277. Medieval Scandinavia 2008 (4). 169 – 208. – 2012. Fallstricke und Glatteis: Die germanische – 2009. Open Access, Nodal Points and Central Places. Tierornamentik. In: H. Beck, G. Geuenich & H. Maritime communication and locational principles Steuer (eds.) Altertumskunde – Altertumswissenschaft for coastal sites in south Scandinavia, c. AD – Kulturwissenschaft. Erträge und Perspektiven 400.1200. Estonian Journal of Archaeology 2009 nach 40 Jahren Reallexikon der germanischen (13):2. 96 – 109. Altertumskunde. Ergänzungsbände RGA 77. Berlin Skre, D. 2007. Excavations of the Hall at Huseby. In: & Boston. 633 – 687. D. Skre (ed.) Kaupang in . Kaupang Pettersson, J. 2000. Husabyarna – en kritisk Excavation Project Publications Series 1. Aarhus. forskningsöversikt. In: M. Olausson (ed.) En bok om 223 – 247. Husbyar. Uppsala. 49 – 63. Steinnes, A. 1955. Husebyar. Oslo. Raddatz, K. 1967. Das Wagengrab der jüngeren Ulriksen, J. 2006. Vester Egesborg – a coastal settlement vorrömischen Eisenzeit von Husby. Offa-Bücher 20. from the Late Iron Age on Zealand. Journal of Danish Neumünster. Archaeology 2006 (14). 187 – 199. – 1987. Der Thorsberger Moorfund. Katalog: Teile von Vikstrand, P. 2001. Gudarnas platser. Förkristna sakrala Waffen und Pferdegeschirr, sonstige Fundstücke aus ortnamn i Mälarlandskapen. Acta Academiae Regiae Metall und Glas, Ton- und Holzgefäße, Steingeräte. Gustavi Adolphi LXXVII. Studier till en svensk Offa-Bücher 65. Neumünster. ortnamnsatlas 17. Uppsala. Randsborg, K. 1980. The Viking Age in Denmark – The Wendt, A. 2012. Wikingerzeitliche Goldringe. Eine formation of a state. London. Fundgruppe ohne Kontext? In: P. Gammeltoft & Rau, A. 2010. Nydam Mose 1. Die personengebundenen V. Hilberg (eds.) Beretning fra niogtyvende tværfaglige Gegenstände. Grabungen 1989–1999. Jysk vikingesymposium. Højbjerg. 50 – 68. Arkæologisk Selskabs Skrifter 72:1. Aarhus. Westerdahl, C. & Stylegar, F.-A. 2004. Husebyene i Roesdahl, E. 2002. Harald Blauzahn – ein dänischer Norden. Viking 2004 (67). 101 – 138. Wikingerkönig aus archäologischer Sicht. In: J. Wiechmann, R. 1996. Edelmetalldepots der Wikingerzeit Henning (ed.) Europa im 10. Jahrhundert. Archäologie in Schleswig-Holstein. Offa-Bücher 77. Neumünster. einer Aufbruchszeit. Internationale Tagung in Willroth, K.-H. 1992. Untersuchungen zur Vorbereitung der Ausstellung „Otto der Große, Besiedlungsgeschichte der Landschaften Angeln und Magdeburg und Europa“. Mainz. 95– 108. Schwansen von der älteren Bronzezeit bis zum frühen Röschmann, J. 1963. Vorgeschichte des Kreises Flensburg. Mittelalter. Offa-Bücher 72. Neumünster. Vor- und Frühgeschichtliche Denkmäler und Funde Ørsnes, M. 1966. Form og Stil i Sydskandinaviens yngre in Schleswig-Holstein 6. Neumünster. germansk jernalder. Nationalmuseets Skrifter.

105 New Aspects on Viking-age Urbanism, c. 750-1100

Arkæologisk-historisk række XL. København.

(Endnotes) 1 Several examples for big grave mounds, large grave fields and remarkable single finds in the surroundings of Swedish and Norwegian Husebyer are compiled by Brink (2000a:288f; 2000b:277f) and Westerdahl & Stylegar (2004:107ff) with further reading.

2 At this point I would like to thank the voluntary members of the metal detector group of the archaeological state department of Schleswig-Holstein for their great assistance and the families Nissen, Jacobsen and Gondesen for the permission to conduct surveys on their properties.

3 A Harde is a medieval, maybe already late Viking Age administrative unit; Dan. herred, Sw. härad (Laur 1992:308).

4 So far, a Byzantine copper coin from the time of Justinus I. (518–527 AD) or Justinus II. (565–578 AD) found in Maasbüllfeld about 1.5 km west of Husby is the only find from this period (Röschmann 1963:408); but as the coin is now lost and its find context is unknown, it cannot be taken into account as proof for a continuous population in this area. It might just as well have ended up in the soil a lot later.

5 As administrative district in Jutland in the Middle Ages a syssel represented the next higher administrative unit above the herred (cf. note 4; Laur 1992:637).

6 In king Valdemar´s liber census Daniae (kong Valdemars jordebog) several places are listed as belongings to the kongelev. In contrast to the dynastic property of the royal family (patrimonium) the kongelev represented the property of the crown, which was bound to royal office (Andrén 1983:33f; Iversen 2011:225).

7 Christensen´s (2014) revealing Husby in Grejs parish a few kilometres from Jelling not to be a “true” Huseby supports Iversen´s interpretation of one Huseby in each diocese. Otherwise after allocating Husby in Hardsyssel to the Viborg diocese there would still have been two Husebyer in the Ribe diocese.

106 Sigtuna: royal site and Christian town and the regional perspective, c. 980-1100

Sigtuna: royal site and Christian town and the regional perspective, c. 980-1100

Sten Tesch

Fig. 1a.. The Sigtuna penisula from south (arial photo by Leif Gren).

today and Lake Mälaren formed part of the Bal- Introduction tic Sea. The only place suitable for settlement and Sigtuna is the oldest town still extant in Sweden. It was building was a narrow flat strip of virgin land on founded c. 9801, almost two hundred years before urban- the tip of the island, about 100–150 metres wide, ization, the real expansion of towns, took place in medieval between the shore and a steep rocky ridge. Topo- Sweden. Sigtuna is situated on Lake Mälaren in East Central graphically, a hilly area separated the town from the Sweden, c. 35 kilometres north of Birka and 35 kilometres rural settlements on the island. It was also distinct south of Gamla Uppsala (Old Uppsala). The settlement area from the rural settlements on the opposite side of stretches along the southern tip of what once was a large the waterway, as these did not regularly face the island (7 x 4 kilometres), now the Sigtuna peninsula, located water. Strategically, however, Sigtuna was centrally between the waterways leading to Uppsala (fig. 1). When the located at an intersection of important waterways, town was founded sea level was c. 4–5 metres higher than which facilitated access to large parts of the Mälar

107 New Aspects on Viking-age Urbanism, c. 750-1100 region by boat or by ice in the winter. In other words, this location seems to have been very deliberately cho- sen: in the midst of, but separate from, the surrounding society. The first part of this article will focus on the physical structures that existed in early Sigtuna - houses, plots, town yards, cemeteries and churches along with the town plan. I will also discuss why and by whom the town was founded, the staging of the town as a setting for political power; who the inhabitants were and how a Christian Mass could be celebrated in the absence of regular church buildings. This is followed by a dis- cussion on whether or not Sigtuna had an interactive relationship with the aristocrats in the Lake Mälaren region. The article ends with a comparative perspective on Scandinavian and other European urbanization fo- cussing on town plans.

Sigtuna Research The late Viking and medieval settlement area in Sigtu- na, includes the “Black Earth”, covering just over 10 hectares, plus c. 20 hectares of surrounding graveyards and churchyards (for the distinction between these, see below). The deposits along the main street, Stora Gatan, are up to three metres thick. More than 5% have been excavated. The preserved layers date mainly from the late 10th to the mid 13th centuries. Fig. 1b. Map of with important places encircled. Scientific archaeological research on Sigtuna began almost a hundred years ago (1915), when the histori- an Olof Palme (“the older”, 1884-1918) started exca- Down to the present, more than 400 excavations have 2 vations on a plot in the very centre of the town. The been undertaken in Sigtuna, resulting in an extensive following year a town museum was founded on the archaeological source material and comprehensive on- 3 same site, also on the initiative of Palme. After Palme´s going research activity that have both changed and en- untimely early death in the , research riched our knowledge and interpretation of the early on Sigtuna in the 1920s and 1930s was continued by town.4 archaeologists such as Otto Janse, Erik Floderus and Bengt Thordeman. In the 1940s, they were succeeded in turn by Wilhelm Holmqvist and Holger Arbman. In Birka and Sigtuna – Continuity or not? the early 1950s Holmqvist became involved in the Hel- For a long time, indeed since the 17th century, Sigtuna gö excavations and Arbman, who was also committed was generally supposed to have been the direct succes- to research on Birka had, since 1945, become professor sor of Birka (fig. 1). Discussion has mainly focused on of archaeology in Lund. This left less and less time for chronological and functional continuity. Even Stock- work on Sigtuna, with the result that Sigtuna research holm has been included, as Sigtuna’s successor, in the stagnated (Tesch 2006a). This inactive period lasted proposed functional sequence: Birka – Sigtuna – Stock- for some decades until the middle of the 1980s (Tesch holm (Ambrosiani 1957; 1996:71). The most frequent 1989). The “Sigtuna Excavations 1988-90”, examining explanations for why Birka was abandoned, and why the Trädgårdsmästaren site, contributed significantly it was necessary to move the town and its inhabitants to this new start (Tesch 1990; 2001b; 2006; 2007b). to a new place such as Sigtuna, are changes in trading

108 Sigtuna: royal site and Christian town and the regional perspective, c. 980-1100

Fig. 2. The Trädgårdsmästaren site in the middle of the 11th century. Reconstruction by Björn Petterson & Mats Vänehem (after Tesch 1996). Complemented with a detail drawing of one townyard showing the four functional zones.

109 New Aspects on Viking-age Urbanism, c. 750-1100 routes, raids and warfare. The continued use of the royal account.6 This gives rise to the possibility of a local or Birka/Adelsö dynasty names – Erik, Olof and Anund regional initiative behind the founding of the town. In – in Sigtuna from its foundation, has been claimed to earlier research it was assumed that the new town was signify that both towns were ruled by the same dynasty, founded on the outfields of Fornsigtuna (e.g. Floderus which in turn has been interpreted as evidence of direct 1941:11ff). The latter is situated only a few kilome- continuity (Carlsson 1997:83ff; Edberg 2006:102). All tres WNW of Sigtuna, on the opposite side of the bay notions of a direct functional continuity between Birka (fig. 1). The history of Fornsigtuna goes back to the and Sigtuna usually presuppose that trade was the pri- time of the birth of Christ and the site was, accord- mary function of both sites. ing to 14C-datings, abandoned around the time as, or Over the last forty years, however, research on Scan- slightly after, the appearance of the town of Sigtuna dinavian urbanization has rather favoured discontinua- (Damell1991b:84, 91). The site was partly excavated in tion (Cinthio 1972; Andersson 1996), which has led to 1984-88 and the excavated remains included the Viking a more fruitful and in-depth discussion. Trade is no lon- Age foundations of two impressive halls, 28 x 9.5 me- ger the prime function of every town. However ques- tres and 40 x 10.5 metres, of which the latter has been tions of continuity have proved to be more complex. In interpreted as having functioned as a garrison (compare order to discuss the topic in a more profound way, we Birka). To reinforce the manifestation of power, the first need to understand the role of each site, both in its hall-buildings were built on artificial platforms made own sphere and in a wider context. up of thick layers of stones and clay. Along the former This means that, even if there is a chronological con- shoreline there are remains of smaller buildings together nection between these two sites, it does not follow that with stone foundations for jetties (Hedman 1991:58ff). there has been a direct transfer of people and function Finally, the hill, crested by a huge mound with flattened from the one location to the other. Rather, the downfall top, has been assumed to have functioned as a Thing of Birka and the rise of Sigtuna have been interpreted mound (Sjösvärd 1991:38ff). This elaborate scenogra- as a consequence of a political, economic and religious phy suggests that the settlement should be interpreted change which affected the entire Baltic region during as an administrative and judicial centre of a pre-Chris- the second half of the 10th century, putting an end to tian Viking Age royal estate contemporary with Bir- the system of long distance trading centres (Tesch 1989; ka-Adelsö.7 The site was most probably not inhabited 1990; 2008b).5 all year round, as very few graves can be linked to it, and The current thinking is that the growth of Sigtuna the finds from the excavations are few. was a result of the town’s political and religious position It has also been discussed whether the place name, as the base for a super-regional royal power and a Chris- Sigtuna, indicates continuity with Fornsigtuna or tian centre, in a region where royal power was weak. This whether the name was adopted by new rulers to legit- would explain why archaeological traces of trade and imize its foundation. A hostile takeover has even been craft are sparse in the early layers of Sigtuna (e.g. Tesch suggested, with the aim partly of benefitting from its 2007a; b). On the other hand modern research on Birka namesake’s sacred and prestigious nature; and partly to has argued that the economic importance of craft pro- clarify that new rulers were now in power (Tesch 2007a: duction, long-distance trade and exchange in the town 240f). has been overestimated, and that its significance was The existence of a further early medieval estate, Vi- of a more symbolic nature (e.g. Hed-Jakobsson 2003). by-Venngarn, situated on the former island, in Sigtuna´s Recently, however, the international trade-network that immediate , increases the complexity of involved the people of Birka has been highlighted as a a possible local origin (fig. 1). crucial political and economic force that characterized late Birka society (Hedenstierna-Jonsson 2012). So, al- though the establishment of Sigtuna seems to reflect a ROYAL AND CHRISTIAN TOWN fundamental structural change in society, the question of continuity or not is clearly a very complex issue. Plots, Houses and Townyards When discussing a potential forerunner to Sigtuna, The Trädgårdsmästaren site is still the key site for in- and the nature of its founding, a nearby site named terpreting and reconstructing Sigtuna’s town plan and Fornsigtuna (Ancient Sigtuna) must also be taken into

110

New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 3. The Trädgårdsmästaren site. A compilation of the ten phases documented (after Söderberg 2011, compilation by the author).

townyards. More recent investigations elsewhere in the original plots, a part of a fifth and an alley at right an- town have however increased our archaeological knowl- gles to the main street. edge of Sigtuna, both on a general level and in detail, The excavation revealed a very regular and stable pat- and, consequently, new questions have been raised and tern of plots with two “townyards” (Sw. stadsgård)8 shar- new interpretations made. ing one narrow passage/courtyard and with an eaves- drip gap in between each pair of townyards, from west to east as follows: townyard/ alley/ townyard/ passage/ The Trädgårdsmästaren site townyard/ eavesdrip gap/ townyard/ passage/ townyard/ The “Sigtuna Excavations 1988-90” – at the Trädgårds- eavesdrip gap. The fact that house types and functions mästaren site – covered 1,140 square metres (38 x 32 usually correlated on both sides of the passage/court- metres) on a centrally located plot and lasted for two yard, led to the interpretation that each row of hous- years. Unlike most other contemporary towns the orig- es constituted one household and one townyard.9 The inal town plan in Sigtuna has never been replaced by a width of the wood-paved main street was half the width grid plan, which means that the excavation trenches will of the street today, which is the reason why the houses usually comply with the original plot structures. The adjacent to today’s main street were not fully excavated Trädgårdsmästaren excavation covered no less than four (Tesch 1990 and later) (fig. 2).

112 Sigtuna: royal site and Christian town and the regional perspective, c. 980-1100

The chronological development on the site has been out any special division according to function, to long divided into ten phases covering almost three hundred plots containing dwelling-houses at the rear of the space years, c. 985 to 1260 (fig. 3). In total about two hun- and workshops close to the main street. dred “houses” (the term is used to describe all types of Main phase I corresponds with the late 10th century; buildings) were investigated (Petterson 1990; 1995; Sö- main phase II with the 11th and main phase III with the derberg 2011a). Only four of these have been identified 12th (Pettersson 1995:71ff) (fig. 2–3). Unfortunately as buildings for (small) animals. Stables and barns are there are no preserved culture layers from the Trädgårds- completely missing.10 To one side of the four town- mästaren site later than the first half of the 13th century; yards lies one of the very few alleys so far document- and the preserved culture layers from the last phases are ed in Sigtuna as crossing the main street.11 Over time more like small islands in an ocean of layers, destroyed the plots became longer and finally reached beyond the by modern ground works. On the whole, well-pre- limits of the excavation area (fig. 3). It is important to served culture layers from the 13th century and later are note also that the town was densely built up from the very rare in Sigtuna. One exception is the Professorn beginning (phase 1). With the exception of the main 1 site (c. 250 square metres excavated 1999-2000) al- street and passages, there were very few open spaces. most 20 metres east of the Trädgårdsmästaren site, but situated on the south side of the main street and thus complementary to the archaeological evidence from the Main phases, townyards and house types Trädgårdsmästaren site. On the Professorn 1 site parts The ten phases have been grouped into three “main of three medieval townyards were investigated, consist- phases” based on changes in the structure of the town- ing of the area with the workshop and multipurpose yards and on construction methods and functions of buildings close to the main street, along with small parts the buildings. These range from short townyards with- of dwelling houses. A very important result was the dis-

Fig. 4. Sigtuna in the early 11th century. Reconstruction by Sten Tesch & Jacques Vincent (after Tesch & Vincent 2003).

113 New Aspects on Viking-age Urbanism, c. 750-1100 covery that the small workshop (main phase III) dwelling, one with an open hearth or oven in a corner facing the street were replaced with two-storey timber of the house (zone 3) and the other, at the rear end of buildings with lofts (main phase IV, 13th century) with the plot, with an open hearth on the central axis (zone the long side adjoining the street. Finds of weights and 4). scales indicate that a merchant occupied the lower floor. The first house-type was usually a cross-timbered Five cloth-seals of lead found in the same zone under- building, 6–9 metres long. The floor was made of clay line this assumption12 However, only main phases I and or gravel and the roof was normally made of sods. II fall within the time-frame of this article. There was only one room, a combined cooking and liv- Main phase I consists of only one sub-phase, (phase ing room, with low wide (c. 1.4 m) benches of timber 1, c. 980/(985)–1000) with two or three houses in a fixed to the walls. The house also functioned as a sleep- row on each plot, making the built area relatively short, ing-room for the townyard inhabitants. Looms were c. 20 metres (fig. 3). Both passages and eavesdrip gaps common in the houses of zones 3 and 4. were relatively wide. Most buildings in this phase are The second house-type wasc . 8–13 metres long, with interpreted as dwelling houses with walls made of wattle low benches along the walls. The walls were construct- and daub and posts dug directly into the ground; clay ed in a number of different ways; with timber framing floors and central hearths. The smaller houses had no forming panels filled with upright or horizontal planks; hearths at all. or of wattle and daub, on rows of sill stones. The roofs, Main phase II – consists of five sub-phases (phases 2–6), made of thatch, reed or planks, had a more acute angle covering the time span c. 1000–1125 (fig. 2–3 and note than the roofs of the cross-timber houses. The houses 10). The fully built-up townyard now consisted of a have been interpreted as “halls”, but they are not truly narrow wood-paved entrance from the street, a paved such, being quite small compared with contemporary passage/courtyard, and four or five houses per plot, in- hall buildings found in rural aristocratic contexts. Still, creasing the built area to at least 30 metres. The paving they both have ceremonial and representative functions of the narrow courtyard, c. 2–3 metres wide, normally in common. comprised logs and planks. The continuity during this main phase is evident from the fact that the houses, along with the hearths, were usually rebuilt in the same The original town plan place for over a century (fig. 3). Evidence from other archaeological excavations in Sigtuna has clarified that the long narrow plots and the four-zone division of the townyards, revealed in the House types and functional zones Trädgårdsmästaren site, was a consistent pattern over Each townyard seems to have been divided into four most of the town (fig. 4–5). This being the case it was functional zones, delineated by their particular artefacts, rather easy to make a first hypothetical reconstruction activities and house-types (fig. 2). The workshops lay of the original town plan (Tesch 1990:35 fig. 10).14 closest to the main street (zone 1); then came the multi- The town was laid out on virgin land according to purpose buildings (zone 2), then two types of dwelling a preconceived design. The central thoroughfare, the houses (zones 3 and 4) (Pettersson 1995:71ff). The floor “high street”, corresponding to today’s Stora Gatan, layers in the workshops contained waste products from was the main centre of activity. Along both sides of the several crafts, but in particular from bronze casting and street were a hundred long narrow plots, about 8 me- blacksmithing. The amount of waste increased after the tres wide,15 with the short side of the plot adjoining the middle of the 11th century, and the artefacts found sug- street. The town turned its back on Lake Mälaren, in gest that the different craftsmen were becoming more a manner of speaking! Narrow lanes at right angles to specialized; but they were still itinerant and worked to the main street were few. It has been suggested that a order for the townyard owner.13 The workshops and the royal hall was located on a large irregular plot in the multipurpose buildings were smaller and less solid than town centre, on a slightly raised and rocky promontory, the dwellings. making it easily visible to those approaching Sigtuna by The domestic zone, containing the dwellings had, in boat (Tesch 1990 and later).16 contrast to main phase I, been moved to the back of the Various court workshops surrounded the king´s plot (fig. 2–3). Each townyard usually had two types of yard – a decentralised “landscape of royal workshops”

114 Sigtuna: royal site and Christian town and the regional perspective, c. 980-1100

Fig. 5. The 11th century town plan (after Tesch 1998 and 2014). a. Early 11th century. b. Late 11th century. Red ovals = graveyards and visible markings above ground. Green oval = graveyards with mounds and stone settings. The open crown marks the king’s plot. Tingsplats? = possible thing site.

(Söderberg 2011b:17ff. fig.11; Tesch 2015a:22ff) (fig. plan. The general picture is still valid but the new results 4a). In the mid 11th century, the king’s yard was aban- have made it more complex and problematical. Here I doned and a new royal manor was most likely built in will highlight only a few significant examples of con- the western part of the town area, where some twenty flicting plot and townyard patterns. townyards had been evicted. The former royal plot was An early problem arose in the Urmakaren 1 site ex- most probably donated to the newly established Episco- cavated in 1990-91, where “King Olof’s mint” was situ- pal See and a wooden church with churchyard was soon ated (fig. 5a:1).17 Instead of a single row of houses, this built there (Tesch 2006; 2014). townyard formed two rows with a narrow courtyard in between. Accordingly, this plot is both shorter and twice Some deviating patterns as wide as a regular plot (Ros 2009:79, fig 121; Söder- Over time the town plan and townyards were repeated- berg 2013:52ff). Most likely, this “anomaly” is due ei- ly subjected to both major and minor changes. A large ther to insufficient space for the usual lengthy plot or it number of excavations have been carried out since the had a special purpose linking it to the king´s yard on the Trädgårdsmästaren project, and with each excavation other side of the alleyway.18 There is reason to assume new details have come to light, which either confirm or that all plots along this section, adjacent to the king´s lead to new interpretations of parts of the reconstructed yard, might have been townyards of the same kind as the

115 New Aspects on Viking-age Urbanism, c. 750-1100 one documented in the Urmakaren 1 site. main street. The distance between the road and the main Another example of a deviating pattern was revealed street was c. 45 metres, that is 10 metres longer than in the Humlegården site in 2006 (fig. 5a:2). Here plots the contemporary 11th century plots in the Trädgårds- on both sides of a road parallel to the main street were mästaren site (Ljung & Wikström 2008:85, 94 fig. 31). excavated. On the south side of the road lay several long There are at least two possible explanations. Either each and narrow plots with zone 4 houses. These plots lay as plot in this part of the town was divided transversely usual with their short sides facing the road. On the other into two sections, one of normal length facing the main side of the road, however, was a long and narrow plot of street and one short and broad, reminiscent of the Ur- more than 30 metres, with three or four houses in a row, makaren plot mentioned above, facing the road behind placed long side to the road, as opposed to the usual plot the settlement area. Or, in this part of the town, the pattern. The house constructions and the findings from plots were somewhat longer than those in the centre, this plot suggest that it had connections with a contem- where the Trädgårdsmästaren site was situated,20 with an porary court workshop belonging to “King Olof´s mint” extra couple of houses in each townyard. (Ljung & Wikström 2008:60ff; Söderberg 2008:119ff; Some of the earlier interpretations have also been 2011b:16f; 2013:60). During the late 11th century, a revised. One example is from the first or foundation Romanesque church (with associated churchyard) was phase. This phase differs from subsequent phases in built on the same spot, above the remains of the old the Trädgårdsmästaren site in that there were only a buildings (Ljung & Wikström 2008:70ff), which also few houses on each plot. Also, the walls were in wattle indicates that the location held a certain significance to and daub without wooden sills. Another difference is the town.19 that the zone 4 house in main phase II with a centrally The houses or “halls” on the south side of the road placed open hearth, is located facing the street (fig. 3). all had central hearths similar to the zone 4 houses fea- As I have previously suggested this may be the result tured in the model of the Trädgårdsmästaren site (fig. of an original town plan of a single row of townyards 2), and were likewise situated farthest away from the along the shoreline, with the same four-zone divisions

Fig.. 6. An alternative interpretation of the foundation phases and townyard development from late 10th century until the 14th century (Tesch 2007b).

116 Sigtuna: royal site and Christian town and the regional perspective, c. 980-1100 as in later phases. Unlike the later phases, where the most of them are only c. 20 metres apart. Most have townyards on both sides of the street faced the main been found below the flat surface, without any traces street, these townyards faced the water (fig. 6) (Tesch of markings above ground. But a few were situated in 2007a:245ff; Tesch 2007b:88ff).21 a more hilly terrain behind the other graveyards, with This interpretation of the foundation phase, with small mounds and stone settings reminiscent of ceme- only one row of townyards along a coastal road paral- teries in the countryside. Regardless of this, all excavat- lel to the shoreline, clearly differs from previous recon- ed graves contained inhumation burials with a roughly structions of the earliest town plan (fig. 4–6). What one east-west orientation.23 The deceased was normally bur- can say is that the revised interpretation reflects a more ied in a wooden coffin with little or no accompanying dynamic development. It implies that all buildings had small objects (Tesch 2006; 2007b:102ff; 2014:107ff).24 been demolished when the main street was laid out Three graveyards have been sufficiently investigated across the original townyards, to enable reconstruction to exclude the presence of a church on the site.25 This on both sides of the new main street. Such a major and is due to the absence of an area empty of graves in the rather drastic change might tentatively be interpreted as centre of the graveyard. Nor are there any physical trac- a re-founding of the town initiated by a new ruler. It is es of a wooden church, such as postholes. Moreover, the possible that the original town (phase 1) was directly or graves’ diverse and divergent orientation from the east- indirectly established by the Danish king Harald Blue- west axis shows that gravediggers had no church build- tooth and the revised town (phase 2) by the Swedish ing to relate to. However, it cannot be ruled out that king Erik Segersäll (Tesch 2007a:245ff; 2007b:88ff). In there may have been a church in another more centrally any event, this is an idea worthy of further analysis. located place, such as the king´s plot in the centre of Parallel ditches, or “foundation ditches”, revealed in town (Tesch 2007b:103; 2014).26 There are no physical the bottom layers of many early towns in Scandina- traces either of boundaries surrounding the graveyards. via have been interpreted as primary plot boundaries As a cemetery consecrated by a bishop would have been giving ultimate proof of a planned town founded on clearly defined, with deep ditches for example, it is open virgin land (discussed by e.g. Schia 1987:201ff; Tesch for discussion whether the graveyards in Sigtuna were 1990:30ff; and Christophersen 1994). However, I have properly consecrated. However, it is possible that each recently (Tesch 2007b:90ff) questioned whether the grave had been consecrated separately at the time of the plot-boundary ditches in the Trädgårdsmästaren site funeral (Tesch 2014; 2015b). 27 were a part of the original scheme in Sigtuna (fig. 3), or The total number of graves from the graveyard period whether they were first dug c. 1020 at the start of phase can be estimated as up to a thousand. Equal numbers of 3 (main phase II). Perhaps, on the whole, these ditch- men women, and children, were buried there, in con- es should not be interpreted as marking the borders of trast to the later churchyards, where there is a marked plots at all, and were just drainage ditches laid adjacent predominance of males (Kjellström 2005:36ff).28 A to and along the actual, no longer visible, boundaries.22 possible explanation for this difference in gender dis- tribution may be that graveyards reflect the household members of the -farmers who had been moved from the countryside to tend to their townyards, while CEMETERIES AND CHURCHES churchyards represent a later more urban demographic Graveyards pattern. Christian burial standards were in place in Sigtuna from the outset. On the other hand, there is no clear archaeological evidence of church buildings until nearly Portable altars - portable churches a hundred years later, and up to many centuries after The graves suggest that the townspeople in Sigtuna were that in rural areas in the Lake Mälaren region (Tesch Christians already from the time the town was founded. 2014:110ff). About 25 early Christian “graveyards” How then could Mass be celebrated before there were (Sw. gravgård), have been identified and partly exca- any properly consecrated churches? The answer is by vated (fig. 5a and 8) in the town. They are all situated the use of portable altars. In the Christian church the in a zone just north of the urban settlement area and Mass, its most sacred rite, is celebrated at the altar. A

117 New Aspects on Viking-age Urbanism, c. 750-1100 church without an altar is nothing, but with a portable tigation in the Magistern site in 1998, situated at the structure large enough to support a chalice and a paten, eastern edge of the town area (fig. 5b). A total of 28 Mass could be attended anywhere and in any building. early medieval graves were excavated (Wikström & A portable altar may in its simplest form consist of a Kjellström 2010). The regular burial pattern and a con- small wooden board, with or without a cavity for the sistent east-west orientation indicate they were aligned relics (sepulcrum), and an inserted altar stone (Sw. sep- on a church in a “churchyard” (Sw. kyrkogård). The ab- ulkralsten), made of some kind of exclusive rock such as sence of lime mortar building waste in the grave filling Greek green porphyry.29 In order to operate in a litur- suggests that the church must have been made of wood. gical context the altar stone had to be consecrated by a The church itself was probably located north of the ex- bishop. A consecrated altar stone alone could be used, cavated graves. It was built sometime during the second without a wooden board or frame (Tesch 2007c and half of the 11th century and abandoned a few decades references cited). later. Other secondary archaeological evidence has been Nine small flat stones interpreted as altar stones interpreted as connected to Bishop Adalvard’s wooden have been found in Sigtuna. Eight of these were made church situated in the middle of the town, on the for- of Greek green porphyry and one of serpentine (Tesch mer king’s yard. Early graves have also been found in 2007c:45ff fig. 1–2; 2008).30 The Sigtuna slabs are par- another three cemeteries, but it is uncertain whether ticularly interesting as they were not found in connec- these burials should be associated with a graveyard or tion with churches and church plots, but when excavat- a churchyard with a wooden church (Tesch 2000:17f; ing the settlement area (fig. 7).31 I have interpreted the 2001a:30; 2014:110ff). Towards the end of the 11th distribution of these finds to indicate that Mass, before century the wooden church assumed to be that of the there were any churches, may have been celebrated in bishop, was replaced with a stone church, probably the the small hall at the rear of each plot, the zone 4 house first of its kind in the Mälar region. that was the townyard owner’s representative building (Tesch 2007c, 2008a). From there it is a short step to the assumption that Mass could also be celebrated with A Townscape of Power Politics the same liturgical equipment in the halls on magnate The regulated town plan in Sigtuna and the fact that farms in the Lake Mälaren region.32 The consecrated the settlement was built in one fell swoop, indicates a portable altar, and any relics that were in it, could turn powerful initiative (Tesch 1990:29ff; 2007b:240ff). In a former pre-Christian environment like the chieftains’ Norway, and in Denmark for that matter, it is consid- halls into a Christian space representing both Christ ered quite obvious that the early towns were founded by and the saints (Tesch 2007c; Dengsö Jessen 2010). kings (Andrén 1983; Nordeide 2007; Carelli 2012) and the History of the kings of Norway, written down in the 13th century, are quite clear on that point. According to Churches was founded by King Olov The graveyards were abandoned in the course of the Tryggvasson in 997. Oslo, and Sarpsborg are second half of the 11th century. Some may have been also among the Norwegian towns which Snorre assumes replaced by wooden churches surrounded by church- were founded by a royal power. Although the credibility yards on the same spot - a likely consequence of the of these written sources is open to question, it is felt that establishment of an espiscopal see by the Hamburg-Bre- the archaeological record has in many ways confirmed men archdiocese in the early 1060’s (fig. 5b). However their view. Sigtuna is mentioned in the History of the there is no clear archaeological evidence for any such kings of Norway only when some of these kings, or fu- churches (Tesch 2014) (fig. 4b). In any case, the period ture kings, pause in the town on their way to and from with wooden churches would have been limited, as the Rus’; which at least tells us something about the town’s erection of Romanesque stone churches in the town, a status (Tesch 1990; 2007a; b). total of six, had already commenced towards the end of The supra-regional kingship of Sigtuna needed Chris- the 11th century. tianity as an ideological tool to establish the idea of a The most obvious secondary evidence for a wooden hierarchic system of power, and to turn the town into church in Sigtuna is to be found in an early medieval a royal and ecclesiastical administrative centre. Sigtu- cemetery discovered during an archaeological inves- na therefore, as with the early Danish and Norwegian

118 Sigtuna: royal site and Christian town and the regional perspective, c. 980-1100

Fig.. 7. A schematic drawing of the11th century staged royal power political townscape in Sigtuna.

towns was not just a centre of power politics but also built as they were in Sigtuna. In that respect the royal a central place for the establishment of the Christian initiative and investment in Sigtuna was an immediate Church in the region – a “port of Faith”. This fact may success. On the other hand, the medieval be one of the reasons why the new towns were usually never became much larger than the area comprising the built on virgin land in a neutral environment (Nordeide original town plan. 2007). Sigtuna’s oldest town design does not appear to be Sigtuna is distinguished from most of the other towns “practical” in nature, but rather a staged “royal town- that emerged in Scandinavia around the year 1000 in scape”, or in the words of Ola Kyhlberg “a narrative that the expansion was more rapid. According to the mindscape” (2012:43) (fig. 8).33 All the townyards, in- evidence from the Trädgårdsmästaren site there were al- cluding the coastal ones, were oriented towards the ready two or three houses on all plots in main phase I main street’s middle axis. The king’s yard was situated (fig. 3). In the Norwegian towns, although virgin land in the very centre of the town and was surrounded by a was regulated into plots, very few were occupied. And it number of court workshops.34 Even the fact that most took a long time before these townyards were as densely plots were roughly the same width, and the consistent

119 New Aspects on Viking-age Urbanism, c. 750-1100 division of the townyards into four functional zones, characterized as heterarchical or stratified, which means indicates a authoritative town planner. The graveyards that the power structure, based on networks of relatives, were all located behind the settlement area and further friends and confidants, was variable. These networks up, behind the cemeteries, was a zone with rune stones and alliances were created and maintained with the help standing along an assumed road that led up to a Thing of rich gifts/dependencies and marriage relations. In a site situated on a steep hill, immediately behind the town stratified society like this there were multiple levels of area (Tesch 2001; 2007a; Tesch & Vincent 2002; Kyhl- people with power and social status (Hermansson 2000; berg 2012:42f) (fig. 8). More than thirty rune stones Kyhlberg 2012:139). However, a crucial question is the are known from Sigtuna. Most of them are fragmentary extent to which Sigtuna was integrated into the rural hi- and have been found in secondary contexts. Two stones erarchy? In the History of the Kings of Norway, it is said (U391 and U393) however, dated to the first half of that kings gave town plots to farmers and others that the 11th century, consist of carvings on boulders located they favoured (e.g. in Trondheim). “Farmers” should along the assumed road. Another stone (U392) stands in this context be understood as the regional or local at a crossroad between a road to Uppsala and the road aristocracy. Sigtuna’s relationship with the surrounding leading to the Thing site. The number of rune stones inhabitants has been explained in a similar way (Tesch along this track might originally have been much larger. 1990:29ff; 2007b:240ff), and is now an accepted part The Thing site is not known from any written sources of the so called ”established story” concerning the early except the Saga of . Here it is stated town and its interaction with its rural neighbours (e.g. that Astrid Olofsdotter35 has convened a large meeting Zachrisson 1998; Kyhlberg 2012:249ff). at the Thing in Sigtuna at a place calledHangrar (Hau- Since the townyard buildings, as we have seen, con- gar, mounds) (Kyhlberg 2012:42).36 sisted almost exclusively of dwellings, small workshops Later on, towards the end of the 11th century, the town and multipurpose buildings; the households must have was remodelled into a religious townscape, with a been highly dependent on the local/regional society stone-paved processional road connecting all the stone to get food and other supplies.37 It is therefore reason- churches then under construction (fig. 7). Some decades able to believe that these deliveries were made from the later the churches were left unfinished and the diocese magnate farms in the countryside, from those aristo- closed down. crats who also disposed of, leased or owned a plot in the town.38 While it was strategically important to be based in the town, these probably preferred to live on the ancestral farm in the countryside, where A REGIONAL PERSPECTIVE they would also be buried. The farm was the centre of Relationship with the Regional Society the estate while the townyard constituted an . In the town, they had a more or less permanent house- Sigtuna has for a long time been perceived as a bridge- hold, with subordinates, including artisans and slaves, head for the Christian mission in the Mälar region. It under the supervision of a . He would also receive has also been considered that the Christianization pro- and house visitors, artisans, merchants, priests, etc., that cess, different in many ways here from other regions, might stay for long or short periods.39 And since Chris- was prolonged due to the strong resistance from the tianity includes everyone, it is reasonable to suppose centre of the pagan cult in Gamla Uppsala. This the- that the imported townyard residents – those who had ory is supported both by the gradual transformation of no ancestors to relate to and no inheritance rights to the the rural Viking Age cemeteries, which were still used cultivated land (oðal law) – were buried in the grave- for Christian burials well into the 12th century, and the yards in Sigtuna. But likewise others, unconnected to lack of evidence for wooden churches. The continuing the magnate farms in the countryside, who died while erection of rune stones here for a longer period and staying in Sigtuna, may have been interred there as well more intensely than elsewhere in Scandinavia, has also (Tesch 2007a:102ff). been interpreted as a sign of a society in crisis – wavering There are however other diverging interpretations between and Christianity. of Sigtuna’s interaction with the Mälar region. Thor- The social composition in the Lake Mälaren region ough studies by Mats Roslund of the ceramic culture in at the time of the foundation of Sigtuna can best be Sigtuna and the Mälar area as a whole, have shown that

120 Sigtuna: royal site and Christian town and the regional perspective, c. 980-1100 the so-called Baltic ware,40 that dominated in the Sigtu- important source material for an analysis of the relation- na cultural layers during the 10th and 11th centuries, did ship and interaction between Sigtuna and the inhabi- not have any impact on the countryside where the local tants of the Mälar region (see below). Viking Age pottery tradition was maintained. This was also the case when it came to the rich material of exclu- Various Cemeteries, Separate Worlds sive items from Rus’ and Byzantium, which is so abun- I feel this issue is too complicated for an easy answer. I dant in the Sigtuna layers. Very little of this material have, therefore, recently used the information from the seems to have reached the area’s aristocratic households. early Christian cemeteries in the Mälar region, to de- Roslund has interpreted this disparity as an expression termine whether it is possible to distinguish different of a cultural barrier that existed between Sigtuna and attitudes towards the newly founded town among the the rest of the Mälar region; the town was an enclave various strata in the society (Tesch 2014). that was not primarily involved with its immediate Sigtuna and the rest of the area exhibit a large num- surroundings. The exotic finds in Sigtuna should rath- ber of excavated Viking Age/early medieval cemeter- er be seen as an expression of the town´s international ies which are suitable for such a study. Furthermore, a character, revealing a with ambitions to play large part of the unexcavated sites are preserved and still a role in a wider European context, and therefore also visible in the landscape. This extensive archaeological serving as an international trading and meeting place material provides a unique opportunity to study the in- for merchants and craftsmen (Roslund 2001:229f). Ro- troduction of Christianity from the perspective of the slund’s view of the early town has much in common residents. with the way that Birka has been thought to function in There are wide variations in burial practice within its final stage during the 10th century – when the real each excavated burial ground and between the various power was held by wealthy “trading families” (proba- graveyards themselves. How then to deal with such a bly non-landowning) with wide international networks complex material? I have discovered that there seem to (Hedenstierna-Jonson 2012:214f). If so, it rather indi- be some crucial distinctions between at least two groups cates that Sigtuna should be regarded as a continuation of early Christian cemeteries, which can be applied to of Birka. This view also implies that there must have this issue (Tesch 2014). been people, other than those who had their roots in the countryside, who inhabited the townyards and buried 1. On one hand the rural population, despite being their dead in the graveyards in Sigtuna.41 Christian, continued to bury their dead beside their pa- However, in my opinion, the late Viking /early me- gan ancestors in Viking Age pagan cemeteries. The prac- dieval magnate farms in the Mälar region have not been tice of using the pre-Christian cemeteries for Christian sufficiently excavated to claim decisively that the find burials lasted well into the 12th century, or until the first material was different from that found in the town. On stone churches were built. There are even examples of the few large farms that have been investigated, I think its use into the 13th century in the northern part of the there are enough examples of findings that suggest the Mälar region. opposite view. That is to say, there does seem to have been some degree of communication between the town 2. On the other hand newly established cemeteries and certain magnate farms in the surrounding land- contained almost exclusively Christian inhumations. scape.42 In the case of the ceramic culture, it may simply This group occurs in an urban context like Sigtuna from be that vessels were valued primarily as storage contain- c. 980 and in rural areas somewhat later. The urban ers rather than commodities, which could explain why graveyards are however more uniform. the Baltic-ware pots stayed in Sigtuna where they were produced (Åqvist 1998:224ff; 2004:59). Containers of The first group of burials does not follow any Chris- wood and pans of iron may also have been used along- tian pattern except that they are inhumation graves, side the simple rural pottery. I also think that differenc- with the bodies usually placed in vaguely east-west ori- es in material culture may have existed independently ented wooden coffins. Otherwise the burial practice of other more important associations, either social, eco- can simply be described as “anything goes”. This fact is nomic or religious. There are also finds such as Slavonic usually attributed to the absence of ecclesiastical organi- temple rings 43 from Christian graves in the rural area zation in rural areas. But I think this hybrid Christian

121 New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 8a. The green rectangles marks the distribution of sepulcral stones found in Sigtuna, plotted on the reconstructed 12th century town plan. Red oval = possible wooden churches and/or graveyards (red oval with brokeb lines). Plan after Tesch (2014). practice could also be interpreted as an expression of a above the ground surface. The common denomina- more or less conscious resistance, not explicitly directed tor is that they are all inhumation graves. Unlike the against the Christian faith and Christianity as such, but first group, these cemeteries were not always situated against all the threats and changes to the existing social in the immediate vicinity of the farm. order that people saw coming in its wake. It was most A hybrid element in the burial practice is espe- likely in opposition to forces perceived as threats to oðal cially evident when it comes to the Christian burials ideology, which maintained an inherited right to the in the pagan cemeteries. At first glance it may look as cultivated land, in contrast to the individual ownership if these people were “less Christian” than those who of land according to the Canonical Law 44. buried their dead in the newly established Christian The second graveyard group is more uniform than churchyards. But the situation is more complex than the first. A characteristic and sometimes dominating that. feature of many of these cemeteries is a cluster of rect- An important contributing factor to this religious angular stone settings often lying side by side, like inhu- hybridity may have been that during the 11th centu- mation graves in modern cemeteries. But some ceme- ry in the Mälar region most people were not baptized teries have only round stone settings and others, similar until the end of their lives. Written sources suggest to the graveyards in Sigtuna, have no visible markers that many people only took on “white clothes” and

122 Sigtuna: royal site and Christian town and the regional perspective, c. 980-1100

Fig. 8b. A sepulcral stone (measuring 7,4x3,2x0,8-1,0 cm) found in the Professorn 2 site 1996. (Photo: Gabriel Hildbrand, Swedish National Heritage Board).

were Christianized with a sign of the cross on the fore- The first social , connected to the earlier head (Sw. primsigning); which may explain why Adam cemeteries, was indifferent or even hostile to the grow- of Bremen and others characterized so-called Christian ing town and what it stood for. Now Christian, they residents in the Mälar area as pagans (pagani) (Janson belonged to the old prehistoric elite, the oðal farmers 2004:228ff.). who used hybridity as a strategy to resist the threats There is also considerable evidence that many people to the existing social order, embodied in the super re- in the Mälar area practiced a Christian faith, which in gional kingship and the Church, that they saw in new part may have been influenced by the Eastern Church. Sigtuna. These influences may also have contributed to the reli- The other stratum, represented in the second group gious hybridity. There are many similarities in the early of Christian cemeteries, belonged to a social level be- Christian burial traditions between the Mälar region and low the uppermost elite. They saw the town and its Kiev Rus’. An especially interesting feature is that Chris- Christian ideals as an opportunity to climb both so- tian burials in Kiev Rus’, as mentioned earlier, seem to cially and economically. Consequently they had a have been consecrated separately, which may also have more interactive relationship with the town and the been the case in the Mälar area (Jansson 2005:55ff). and so were allocated townyards in Sigtu- The study of religious identity and practice within na. There seems to be a clear geographical covariation these two groups of burials in the Mälar region, has also between the occurrence of newly established Christian revealed two different strata in the 11th and 12th centu- cemeteries and Early Medieval churches in rural areas ry society, which contend, in completely opposite ways (Blennå & Söderström 1995; Bäck 2012; Tesch 2014). and with varied strategies, with both Christianity and It is therefore a reasonable conclusion that the same the rise of the super regional kingship; and ultimate- people who owned the townyards in Sigtuna financed ly with their relationship with Sigtuna (Tesch 2014; in the building of the Romanesque churches in the coun- print). tryside (Tesch 2014).

123 New Aspects on Viking-age Urbanism, c. 750-1100

and 1980s large-scale excavations in York and Dublin re- THE TOWN PLAN – A EUROPEAN vealed synchronized 10th century layout patterns, which PERSPECTIVE are in some way similar to the above plans in Scandina- New towns like Sigtuna in Sweden, Lund in Denmark via. In the Coppergate excavation in York, parts of four and Trondheim in Norway, seem to have been set up in long narrow plots 5-6 metres wide were investigated order to centralize political and ideological control in (Hall 1984). At Fishamble Street in Dublin the site was regions where the full rule of the supra-regional king- divided into fourteen contiguous plots of approximately dom was weak. Thus it was not only a shift of centres, the same width as those in York (Wallace 1987). Since but also a shift of power. The early coinage and the early the plot arrangement in these “Scandinavian” towns is churches in these towns underline the regal and reli- clearly older than in Sigtuna, it is tempting to imagine gious intentions (Andrén 1983). king Erik Segersäll, for example, “on the spot”, drawing New towns, with new social and political status, also inspiration for his town. Or could the plans have been require a change in their physical structures, i.e. the lay- Scandinavian from the outset? Investigations in Ribe, out of plots, town yards and buildings. The town plan the 8th century seasonal trading place in Denmark (Jen- of Sigtuna, with its centrally placed royal manor and its sen 1991,) have revealed a layout of plots remarkably long narrow plots branching out along the main street similar to the town plan of late 10th century Sigtuna. The was certainly something new; but at the same time very Ribe design could in turn have been influenced by that similar to plans for contemporary Norwegian towns found in Dorestad. On the whole it seems like Sigtuna’s such as Trondheim, Oslo/Gamlebyn and Bergen, re- long narrow plots on both sides of a main street parallel vealed through large-scale archaeological excavations. to the shore line, were typical not only for Scandinavia, The arrangement of the townyards was almost the same but also for the and the North Sea Rhine with four-zone functional division and similar types of area (Clarke & Ambrosiani 1991). house construction (Christophersen 1994; Schia 1987; The settlement in Rus´, however, seems to have been Molaug 2001; Hansen 2007). As in other late Viking planned along different lines than those in Sigtuna. Age towns there were few traces of extensive craft pro- In Novgorod the majority of plots had a rectangular duction in the earliest phases. or square shape. The plots were made up of an open However, Sigtuna differs from the early Norwegian space/courtyard surrounded by buildings for various towns in that the whole urban area and all the plots purposes. The buildings were, as a rule, placed inside were inhabited from the start (see above). On the other and along the fence that separated the plot from the hand this area never became much larger than the orig- street (Sorokin 2002:620). The archaeological evidence inal. In Sigtuna the general plot width is much more from Kiev seems to suggest a similar practice (Sahaidak uniform than in the other towns and remained around 1991). In Staraja Ladoga, however, explorations have eight metres. Similar uniform plot widths have also revealed plots of the same type and date as the plots in been revealed through archaeological excavations in late Ribe. These plots were 6 metres wide and were sepa- 12th early 13th century towns like Söderköping (Tesch rated by deep ditches measuring 1.5 metres wide and 1987) and Nyköping (Norberg & Pettersson 2008). 0.5 metres down, placed at right angles to the shoreline Might one suggest that there may have been one ur- of the Ladosjka river. The excavators believe the layout ban design typical for East Central Sweden in the Early can be attributed to the mythical prince and his Middle Ages? descendants, and therefore may have been of Scandina- The impulses behind the comparable town plan pat- vian origin (Kirpitsjnikov & Nazarenko 1992). terns in Sigtuna and the Norwegian towns must have Accordingly, the physical structures in most early arisen either from a common source, or were influenced medieval towns in Rus’ seem to have been quite differ- by contacts within the Scandinavian social and polit- ent from those of the Scandinavian towns. The question ical network of royal families. Without benefit of the is why? Does it mean that something other than the archaeological evidence, Adolf Schück, “the father of general process of state formation has affected this de- medieval town research in Sweden”, pointed out that velopment? One exception is the cross-timbered house the prototype of the early medieval town plan in Scan- structures, which, while common in both Rus´ and dinavia could be found in the Danish-Norwegian towns Scandinavia, were definitely older in Rus’. It is interest- in the British Isles (Schück 1926). During the 1970s ing that cross-timbered buildings first appeared in main

124 Sigtuna: royal site and Christian town and the regional perspective, c. 980-1100 phase II in Sigtuna. Could this be a consequence of the tral axis of the main street, along with the ephemeral excavation of exclusively zone 4 houses in main phase1 and elusive harbour installations. A densely built and (see above), or might the houses in main phase I have populated town, certainly, but it might well be defined, been inspired by the wattle and daub houses in Dublin at least in the 11th century, as a considerable royal ma- and York or even Denmark? norial pinnacle site – “a special purpose settlement”45 Sigtuna´s general affiliations in this period certainly – that attracted the aristocratic section of the surround- appear to lean more to the East. These contacts had also ing society, who looked upon Sigtuna and the Christian been maintained in the area for more than two hun- church as a means of ascending the social ladder. dred years before the foundation of the town. The dy- From the middle of the 11th century Sigtuna be- nastic relations in early Sigtuna help us imagine these came an ecclesiastical centre with a sacred townscape contacts. Many 11th century runic stones in the Mälar although church buildings were initially absent (as were region were erected in memory of (väringar) most of the known ). Finally, at the height of its who had returned, either alive or for burial, from service religious power while the construction of six Roman- in the guard of the Russian prince or the Byzantian em- esque churches was in full swing, and church proces- peror, underlining the region‘s connection with Rus’. sions honoured the patron saints chosen from all over The numerous Scandinavian finds found in Rus’ speak the Christian world, a setback took hold, and a new era for themselves. Furthermore, large amounts of objects began with Sigtuna eventually losing its special status of Rus’ and Byzantian origin have been found in Sigtu- and becoming a town amongst other towns. na, more than in any other contemporary site in Scan- dinavia. It is therefore reasonable to believe that this intense interaction with Rus’ might have left a distinct Corresponding Author mark on the cultural life of the Swedish town (Duczko Sten Tesch 1997; Roslund 1997; 2001; Tesch 2007a). Uppsala, Sweden So why did these obvious contacts not affect [email protected] the physical structures in the plots of the town? I think it could depend on the strong influence of Byzans on the Rus’ society, which for instance brought about scat- tered urban settlements with several centres, unfamil- iar in a Scandinavian context. In addition, the Roman in England and Germany had great weight in Scandinavia, in contrast to the influence of the Greek-Orthodox Church in Rus’. As mentioned above, this issue is even more complex since the early Christian burial practice in the Mälar region seems to have shown influences from the Eastern Church (Jans- son 2005:55ff; Tesch 2014). Once again –the finds from Sigtuna reveal much about its material culture and connections, but very little about the ideas and beliefs that were circulating then all over the known world.

THE EARLY TOWN – A SUMMARY PORTRAIT The foundation and growth of Sigtuna were, until the middle of the 11th century, largely characterized by a supra-regional Christian kingship. The ground plan of the town, coordinated from the start, is indicative of a strong originator. In many ways the design is directed inwards, manifested by the central location of the king’s yard, the court workshops surrounding it and the cen-

125 New Aspects on Viking-age Urbanism, c. 750-1100

Clareus, C. 1976. Arkeologisk undersökning 1971–72. References Del av fornlämning 55. Vikingatida gravfält Stav Ambrosiani, B. 1957. Birka, Sigtuna, Stockholm. Ett 10:5, Färentuna sn, Uppland. Riksantikvarieämbetet diskussionsinlägg. Tor 1957. 148 – 158. Rapport 1976 B16. Stockholm. Ambrosiani, B. & Eriksson, B.G. 1996. Vikingastaden Cinthio, E. 1972. Variationsmuster in dem frühmittel lever upp igen i TV:s modell av 800-talets Birka! Birka alterlichen Städtewesen Schonens. In: H. Hinz (ed.) Vikingastaden 5. Stockholm. Frühe Städte im westlichen Ostseeraum, Kiel Papers Andersson, H. 1996. Tendencies in the Field of Archae `72. Neumünster. 57 – 64. ology Concerning the Early Medieval Town in Scan Cinthio, M. 2003. De första stadsborna: medeltida gra dinavia. Debates and Discussion. In: L. Nilsson & S. var och människor i Lund. Eslöv. Lilja, (eds.) The Emergence of Towns: Archaeology and Clarke, H & Ambrosiani, B. 1991. Towns in the Viking Early Urbanization in Non-Roman, North-West Eu Age. Leicester & London. rope. The Swedish Institute of Urban History 75 Damell, D. 1991 (ed.). Fornsigtuna. En kungsgårds his years anniversary symposium. Studies in Urban His toria. Stiftelsen Upplands-Bro Fornforskning. Up tory 14. Stockholm. 95 – 112. plands-Bro. Andrén, A. 1984. Städer och kungamakt – en studie i Duczko, W. 1995. Kungar, thegnar, tegnebyar, juveler Danmarks politiska geografi före 1230. Scandia och silverskatter. Om danskt inflytande under sen 1983 (49):1. 30 – 76. vikingatid. Tor 1995 (27):2. 625 – 662. Bhabha, H. 1994. The location of culture. London. – 1998. Byzantine Presence in Viking Age Sweden, Ar Blennå, I. & Söderström, E. 1995. Från Montelius till chaeological Finds and their Interpretation. In: M. Läroplan 94. Kristnandeprocessen sedd ur två perspek Müller-Wille (ed.) Rom und Byzanz im Norden II. tiv: den arkeologiska forskningen och skolundervisni Mission und Glaubenswechsel im Ostseeraum während gen. Arkeologiska institutionen, Uppsala universitet, des 8.-14. Jahrhunderts. Abhandlungen der Geistes- CD-uppsats. Uppsala. und Sozialwissenschaftlichen Klasse. Akademie der Bäck, M. 2012. På andra sidan Birka. Södra Björkös Wissenschaften und der Literatur in Mainz 3:1. arkeologiska potential. In: Hedenstierna-Johnson, Stuttgart. 291 – 311. Ch. (red.), Birka nu. Pågående forskning om världsar Edberg, R. 2006. Folk och rövare i Sigtuna stad. Situne vet Birka och Hovgården. The National Historical Dei 2006. 95 – 129. Museum, Stockholm. Studies 22. Stockholm. 45 – – 2013. Subterranean Maritime Archaeology in Sig 68. tuna, Sweden: Excavated evidence of Viking Age Carlsson, A. 1997. Birka kungsgård på Adelsö och boat building and repair. The international Journal of Svearnas Fornsigtuna - två aristokratiska miljöer i Nautical Archaeology 2006 (42):1. 196 – 204. Mälardalen. In: J. Callmer & E. Rosengren (red.) ”... Ersgård, L. 2006. Dödens berg och Guds hus – förfäder gick Grendel att söka det höga huset...” Arkeologiska skult, kristnande och klostret i Alvastra i den tidiga källor till aristokratiska miljöer i Skandinavien under medeltidens Östergötland. In: L. Ersgård (red.) I hel yngre järnålder. GOTARC Serie C. Arkeologiska gonets boning. Studier från forskningsprojektet ”Det ti skrifter 17. Halmstad. 83 – 88. digmedeltida Alvastra”. Lund Studies in Historical Carelli, P. 2012. Lunds historia – staden och omlandet. Archaeology 5. Lund. 23 – 140 1. Medeltiden. En metropol växer fram. Lund. Floderus, F. 1941. Sigtuna. Vår äldsta medeltidsstad. Christophersen, A. 1990. Dwelling Houses, Work Stockholm. shops and Storehouses. Functional aspects of the de Hall, R. 1984. The Viking Dig. The excavations at York. velopment of wooden urban buildings in Trondheim London. from c. A.D.1000 to A.D.1400. Acta Archaeologica Hansen, G. 2005. Bergen c 800 - c 1170. The Emergence 1989 (60). 101 – 129. of a town. The papers. Main Series 6. – 1994. Gård og grunn. Mot syntesen. In: A. Chris Bergen. tophersen, & S.W. Nordeide (eds.) Kaupangen ved Hed Jakobsson, A. 2003. Smältdeglars härskare och Je Nidelva, 1000 års byhistorie belyst gjennom de rusalems tillskyndare. Berättelser om vikingatid och ti arkeologiske undersøkelsene på Folkebibliotekstom dig medeltid. Stockholm Studies in Archaeology 25. ten i Trondheim 1973–1985. Trondheim. 113–212. Stockholm.

126 Sigtuna: royal site and Christian town and the regional perspective, c. 980-1100

Hed Jakobsson, A, Runer, J. & Svensson, K. 2013. Vä från Sigtuna Museum 33. Sigtuna. 165 – 215. gen till dejan och bryten. En berättelse om den vikin Kyhlberg, O. 2012. Den långa järnåldern. Sociala strat gatida gården i Hjulsta. Skrifter från Arkeologikon egier, normer, traditioner. Occasional Papers in Ar sult 4. Upplands Väsby. chaeology 54. Uppsala. Hedenstierna-Jonson, Ch. 2012. Birkafolket. In: Heden Ljung, C., & Wikström, A. 2008. Fasbeskrivningar. In: stierna-Jonson, Ch. (ed.) Birka nu. Pågående forsk A. Wikström (red.) På väg mot Paradiset – arkeolo ning om världsarvet Birka och Hovgården. The Na gisk undersökning i kvarteret Humlegården 3 i Sigtuna tional Historical Museum, Stockholm. Studies 22. 2006. Rapport Arkeologisk Undersökning. Med Stockholm. 213 – 226. delanden och Rapporter från Sigtuna Museum 33. Hedman, A. 1991. Platåhusen. In: D. Damell (ed.) Forn Sigtuna. 47 – 74. sigtuna. En kungsgårds historia. Stiftelsen Up Molaug, P. 2001. Medieval House Building in Oslo. In: plands-Bro Fornforskning. Upplands-Bro. 58 – 74. M. Gläser (ed.) Der Hausbau. Lübecker Kolloquium Hermanson, L. 2000. Släkt, vänner och makt. En studie zur Stadtarchäologie III. Lübeck. 765 – 782. av elitens politiska struktur i 1100-talets Danmark. Nordeide , S.W. 2007. The Christianization of Norway. Avhandlingar från Historiska Institutionen. L’Europe en mouvement. Medieval Europe, Paris. Göteborgs universitet 24. Landskrona. 4e Congrès international d’Archéologie Médiévale et Janson, H. 2004. Konfliktlinjer i tidig nordeuropeisk Moderne, 3–8 septembre 2007, Paris (France). Paris. kyrkoorganisation. In: N. Lund (ed.) Kristendomen i http://medieval-europe-paris-2007.univ-paris1.fr/S. Danmark før 1050. Ett symposium i Roskilde den Nordeide.pdf 5.-7 februar 2003. Roskilde. 215 – 234. Petterson, B. 1990. Stadsgården - hus och hemmiljö. Jansson, I. 2005. Situationen i Norden och Östeuropa In: S. Tesch (ed.) Makt och människor i kungens Sigtu för 1000 år sedan: en arkeologs synpunkter på frågan na. Sigtunautgrävningen 1988-90. Sigtuna museer. om östkristna inflytanden under missionstiden. In: Sigtuna. 38 – 52. H. Janson (ed.) Från Bysans till Norden – östkyrkliga – 1995. Stratigraphic analysis and settlement stratigra kyrkoinfluenser under vikingatid och tidigmedeltid. phy in Early Medieval Sigtuna. Methods and prelim Skellefteå. 37 – 95. inary results. Laborativ Arkeologi. Journal of Nordic Jensen, S. 1991. Ribes vikinger. Ribe. Archaeological Science 8. 65 – 77. Johansson, C. 2009. Visby visuellt. Föreställningar om en Ros, J. 2009. Stad och gård. Sigtuna under vikingatid plats med utgångspunkt i bilder och kulturarv. och tidig medeltid. Uppsala. Linköping Studies in Arts and Science 484. Institu Roslund, M. 2001. Gäster i huset. Kulturell överföring tionen för studier av samhällsutveckling och kultur. mellan slaver och skandinaver 900 till 1300. Skrifter Tema Kultur och samhälle. Linköping. utgivna av Vetenskapssocieten i Lund 92. Lund. Kirpitsjnikov, A. & Nazarenko, V. 1992. Parseller fra – 2007. Crumbs from the Rich Man’s Table. Byzantine vikingtiden i Staraja Ladoga. SPOR 1992 (2). 16 – Finds in Lund and Sigtuna, c. 980-1260. In: A. An 17. dersson, P. Carelli, & L. Ersgård. (eds.) Visions of the Kjellström, A. 2005. The Urban Farmer. Osteoarchaeo Past. Trends and Traditions in Swedish Medieval Ar logical Analysis of Skeletons from Medieval Sigtuna In chaeology. Lund Studies in Medieval Archaeology 19 terpreted in a Socionomic Perspective. Theses and Pa = Riksantikvarieämbetet. Arkeologiska un pers in Osteoarchaeology 2. Stockholm. dersökningar Skrifter 24. Stockholm. 239 – 298. Kjellström, A., Tesch, S. & Wikström, A. 2005. Inhab Sahaidak, M.A. 1991. Davn’okyïvs’kyi Podil. Problemy itants of a sacred townscape. An archaeological and topohrafiï, stratyhrafiï, khronolohiï, Kiev. Osteological Analysis of Skeletal Remains From Late Sorokin, A.N. 2001. Domestic Architecture in Medi Viking Age and Medieval Sigtuna, Sweden. Acta Ar eval Novgorod. In: M. Gläser (ed.) Der Hausbau. Lü chaeologica 2005 (76):2. 87 – 110. becker Kolloquium zur Stadtarchäologie III. Lübeck. Kjellström, A. & Wikström, A. 2008. Kyrkogården och 605 – 626. det sakrala stadsrummet. In: A. Wikström (red.) På Schia, E. 1987. Bygården. In: E. Schia, E. (ed.) De arke väg mot Paradiset – arkeologisk undersökning i kvarter ologiske utgravninger i Gamlebyen Oslo 3. Övre Ervik. et Humlegården 3 i Sigtuna 2006. Rapport Arkeolo 191 – 228. gisk Undersökning. Meddelanden och Rapporter Schück, A. 1926. Studier rörande det svenska stadsväsen

127 New Aspects on Viking-age Urbanism, c. 750-1100

dets uppkomst och äldsta utveckling. Uppsala. 9 – 44. Sjösvärd, L. 1991. Signhilds kulle. In: D. Damell (ed.) – 2001b. Houses, Town Yards and Town Planning in Fornsigtuna. En kungsgårds historia. Stiftelsen Upp Late Viking Age and Medieval Sigtuna, Sweden. In: lands-Bro Fornforskning. Upplands-Bro. 38 – 47. M. Gläser (ed.) Der Hausbau. Lübecker Kolloquium Skre, D. 2006. Kaupang i Skiringssal. Kaupang Exca zur Stadtar chäologie III. Lübeck. 723 – 741. vation Project Publication Series 1. Norske Oldfunn – 2006a. Situne Dei – återkomsten. Situne dei 2006. XXII. Aarhus. 5 – 12. Söderberg, A. 2008. Metall- och glashantverk. In: A. – 2006b. På fast grund. Om det äldsta stenkyrkobyg Wikström (red.) På väg mot Paradiset – arkeologisk gandet i Sigtuna. Nordiskt kyrkoarkeologiskt möte undersökning i kvarteret Humlegården 3 i Sigtuna 2004, Vest- (Norge). Kirkearkœologi i Norden 2006. Rapport Arkeologisk Undersökning. Med 8 = Hikuin 2006 (33). 201 − 222. delanden och Rapporter från Sigtuna Museum 33. – 2007a. Kungen, Kristus och Sigtuna – platsen där Sigtuna. 97 – 130. människor och guld möttes. In: I. Lindgren (red.) – 2011a. Huvudfaser, stadsgårdar och zonindelning & Kult, guld och makt. Symposium i Götene 18-20 au Fas, tomt- och husbeskrivningar. In: A. Wikström gusti 2006. Historieforum Västra Götaland. . (ed.) Fem stadsgårdar – arkeologisk undersökning i kv. 233 – 257. Trädgårdsmästaren 9 & 10 i Sigtuna 1988–90. Med – 2007b. Sigtuna – det maktpolitiska och sakrala stads delanden och Rapporter från Sigtuna Museum 52. rummet under sen vikingatid och tidig medeltid (ca 23 – 140. 980 – 1200). ). In: A. Perlinge (ed.) Människors rum – 2011b. Eyvind Skáldaspillir’s silver – refing and stan och människors möten. Kulturhistoriska skisser. Berit dards in pre-monetary economies in the light of finds Wallenbergs Stiftelse 50 år. Vetenskapligt sympo from Sigtuna and Gotland. Situne Dei 2011. 5 – 34. sium på 14 november 2005. Stock – 2013. Om stratigrafi, tomtmönster och hantverk i holm. 71 – 122. kvarteret Urmakaren, Sigtuna. Situne Dei 2013. 47 – – 2007c. Heliga stenar från Köln. Om sepulkralstenar 70. och hallkult i 1000-talets Sigtuna. Situne Dei 2007. Tesch, S. 1989 (ed.). AVSTAMP - för en ny Sigtunaforsk 45 – 68. ning. 18 forskare om Sigtuna. Heldagsseminarium – 2008a. Laddade stenar. In: A. Wikström (ed.) På väg kring Sigtunaforskning den 26 november 1987. mot Paradiset. Rapport arkeologisk undersökning 2006. Gröna Ladan, Sigtuna. Kommittén för Sigtunaforsk Meddelanden och Rapporter från Sigtuna Museum ning. Sigtuna Museer. Sigtuna. 33. Sigtuna. 137–143. – 1990. Stad och stadsplan. In: S. Tesch (ed.) Makt och – 2008b. Sigtuna ca 980-1200. Det maktpolitiska och människor i kungens Sigtuna. Sigtunautgrävningen sakrala stadsrummet. In: H. Andersson, G. Hansen, 1988-90. Sigtuna museer. 23 – 37. & I. Øye (eds.) De första tvåhundra åren – nytt blikk – 1996. Sigtuna – Royal Stronghold and Early Town. på 27 skandinaviske middelalderbyer. Universitetet i In: L. Nilsson & S. Lilja, (eds.) The Emergence of Bergen Arkeologiske Skrifter 5. Bergen. 323–340. Towns: Archaeology and Early Urbanization in – 2014. Skilda gravar, skilda världar - tidigkristna gra Non-Roman, North-West Europe. The Swedish Insti var, kyrkor, stadsgårdar och storgårdar i Sigtuna och tute of Urban History 75 years anniversary sympo Mälarområdet. In: O. Karsvall & K. Jupiter (red.). sium. Studies in Urban History 14. Stockholm. 113 Medeltida storgårdar. 15 uppsatser om ett tvär – 126. vetenskapligt forskningsproblem. Acta Academiae Re – 1997 (ed.). Mariakyrkan i Sigtuna. Dominikankon giae Gustavi Adolphi 131. Uppsala. 101 – 130. vent och församlingskyrka 1247. Sigtuna Museers – 2015. Att parera eller paradera. Ett tusenårigt svärd skriftserie 7. Sigtuna. shjalt från Sigtuna och dess europeiska förbindelser. – 2000. Det sakrala stadsrummet. Den medeltida kyrko Situne Dei 2015. 14 – 27. topografin i Sigtuna. META 2000 (l). 3 – 26. – in print. A lost world? – Religious identity and prac – 2001a. Olof Palme, S:ta Gertrud och Sigtunas me tice during the introduction of Christianity in the deltida kyrkotopografi. In: S. Tesch, & R. Edberg Lake Mälaren region. In: Dying Gods – Traditional 2001 (red.) Biskopen i museets trädgård. En arkeolo Beliefs in Northern and Eastern Europe in the Time of gisk gåta. Sigtuna Museers skriftserie 9. Sigtuna. Christianisation.Paderborn, 7.–11.9.2013. Neue

128 Sigtuna: royal site and Christian town and the regional perspective, c. 980-1100

Studien zur Sachsenforschung. Altertumskommissi Museum since 1966. on für Westfalen. Niedersächsisches Landesmuseum Hannover. (https://independent.academia.edu/ 4 Other large-scale excavations were conducted at: Ur- StenTesch/Talks). makaren 1 in 1990-91, Professorn 2 and 4 in 1995- Tesch, S. & Vincent, J. 2002. Vyer från medeltidens 96, Professorn 1 in 1999-2000 and Humlegården 3 in Sigtuna. Sigtuna Museers skriftserie 10. Sigtuna. 2006. Wallace, P.F. 1987. The Layout of Later Viking Age Dublin: Indications of its regulation and problems of 5 Topographically, however, there is some correspon- continuity. In: J. E. Knirk (ed.) Proceedings of the dence between Birka and Sigtuna. Each site is situat- Tenth Viking Congress. Universitetets Oldsakssam ed on what was once an island. While the waterways lings Skrifter. Ny rekke 9. Oslo. 271 – 286. surrounding Sigtuna’s island were much narrower, Wikström, A. & Kjellström, A. 2010. Magistern 2, both places were located between key routes. Sigtuna 1998. Röda skolan, 1992. Sigtuna, Up. Rapport arke lay in the innermost archipelago (fig. 1), in contrast to ologisk undersökning. Meddelanden och Rapporter 11th/12th century centres and medieval towns such as Ös- från Sigtuna Museum 46. Sigtuna. tra Aros (Uppsala town), Enköping and Västerås, which Zachrisson, T. 1998. Gård, gräns, gravfält. Sammanhang were usually located at the apex of a river mouth, at the kring ädelmetalldepåer och runstenar från vikingatid junction of land routes (eskers) and sea/river routes, to och tidigmedeltid i Uppland och Gästrikland. Stock serve the reloading and transportation of goods. holm Studies in Archaeology 15. Stockholm. Åqvist, C. 1998. Gården på landet – en spegling av sta 6 Fornsigtuna has been called since the den. In: P. Bratt (red.) Forntid i ny dager. Arkeologi i 17th century. Stockholmstrakten. Stockholms läns museum. Stock holm. 222 – 230. 7 According to the , the first part of Snor- – 2004. Sanda – en gård i södra Uppland. Bebyggelse från ri Sturlusson’s “Chronicles of the Kings of Norway”, vendeltid till 1600-tal. Uppland, Fresta socken, Sanda Fornsigtuna was founded by the god Woden. 1:1, RAÄ 147. Riksantikvarieämbetet, Avdelningen för arkeologiska undersökningar. Rapport UV-Mitt 8 It is rather difficult to find a good English expression 2004:15. Stockholm. for the Swedish term “stadsgård” or the Norwegian term “bygård”. The Norwegian archaeologist Erik Schia dis- (Endnotes) cussed the problem in 1987. The term “townyard” was eventually chosen after discussion with the English ar- 1 Due to very few and uncertain dendrochronological chaeologist Clifford D. Long. A Swedish and Norwe- datings from the Trädgårdsmästaren site the foundation gian townyard consists of a parcel of land, a plot, with has been variously estimated over the years to c. 975, several buildings with various specialized functions and the 970s, c. 980 and now, in the Trädgårdsmästaren re- a courtyard: a densely built-up area of relatively limited port, to c. 985 (Söderberg 2011a:19ff). I have, however, dimensions compared with the agricultural farm (Sw. chosen to adhere to the dating c. 980, which I find is a gård) (Schia 1987:228). moderately round number that provides lea way both backwards and forwards … and the last word (concern- 9 The plot boundaries between the two townyards went ing dating) has not yet been heard (cf. e.g. Söderberg lengthwise in the centre of the courtyard, usually in the 2013:52). Poor preservation conditions for organic ma- form of a ditch concealed by a wooden paving and open terials in Sigtuna in general have resulted in very few ditches in the eavesdrip gaps. secure dendrochronological datings, even from the old- est layers. 10 Two byres (houses 303 and 334a), one pigsty (house 341) and one henhouse (house 295) are all from the 2 Olof Palme was an uncle of his more recently famous same plot (V), phase (3, 1020-1050) and zone (2 = the namesake, the late Swedish Prime Minister. zone with multi-purpose buildings except for house 334a and zone 1 = the workshop zone). House 303 is 3 The Sigtuna Fornhem Foundation known as Sigtuna replaced by house 295 (Söderberg 2011a). Thus, plot V,

129 New Aspects on Viking-age Urbanism, c. 750-1100 phase 3 is the exception that confirms the rule. It also carving walrus and cutting coin dies (Söderberg differs by the presence of five houses on the plot, instead 2013:62ff). of four, which was usual during the 11th century on the Trädgårdsmästaren site. These four buildings are very 18 The alley was a continuation of the alley documented small (12 to 15 square metres), more like sheds, so they in the Trädgårdsmästaren site. cannot be characterized as stables or barns. 19 Cf. the centrally located king´s yard, where a wood- 11 The alley was stone-paved c. 1125 (Söderberg 2011a). en church, interpreted as the bishop’s church, and its This change in pavement is probably due to the fact churchyard, were probably established in the 1060´s that the alley was then used to transport heavy building and replaced some decades later by a stone church materials from the harbour area up to the stone church- (Tesch 2014). es that were under construction. The stone-pavement 20 Some townyards had two zone 3 houses or two zone 4 may also have had a more symbolic function, as part houses. In the latter case, could one be rectangular and of a procession road connecting all the churches into a the other to the rear be square, reminiscent of a Swedish “sacred townscape” Tesch 2000; 2006b). If this inter- “eldhus” or ”stekhus”, which was a house used for cook- pretation is correct, the stone-paved street cannot have ing, baking and brewing? been constructed later than c. 1100. The main street was, however, paved with transverse logs and lengthwise 21 Without other comparisons, like the reconstruction planks throughout the 12th and 13th centuries. of Viking Age Kaupang in Norway (Skre 2006) and the townyards in Birka (Ambrosiani & Eriksson 1996). 12 Archaeological evidence from Trondheim (Chris- tophersen 1990; 1994) and Oslo (Schia 1987: 211) in- 22 Later discussed by Söderberg (2011a:35ff). dicate a similar change, that is to say, small workshops replaced by large buildings with a hearth on the upper 23 With the exception of one cremation grave excavated floor. A shop-board facing the street could be opened in in a cemetery in the hilly terrain (Hallström 1923:23). order to sell products. 24 Carbon in some of the graves may indicate influences 13 A general view indicates that craftwork during the from English (Cinthio 2003:54). first phases, except for the court workshops/the gold smith´s workshops, was at a very low level and primarily 25 Graves 29, 49 and 41 (Kjellström 2005:13 fig. 2.2). served the household’s own needs. Production for strict- ly mercantile purposes does not occur until c. 1100 (Sö- 26 New 14C-datings of some graves indicate the existence derberg 2013:59ff). of graves that are contemporary with the king´s yard (Söderberg 2013:52) 14 New advanced reconstructions illustrate the develop- ment of the town plan between the 11th and 14th centu- 27 Which was a common practice in the Eastern Church ries (Tesch 1997:14f; 2001a:122ff fig. 10-13; Tesch & (Cinthio 2003; Jansson 2005). Vincent 2002:10f, 17). 28 The relation to each other of the different graveyards 15 The Trädgårdsmästaren plots, c. 7– 8.5 metres wide. and the cult community each graveyard represented, is still an open question. 16 The archaeological material from the plot of the as- sumed king´s hall is however fragmentary and ambig- 29 Green porphyry was quarried during the Roman peri- uous. od in the province of Laconia in southern . Small slabs of green porphyry were used as wall decorations or 17 A goldsmith´s workshop, possibly belonging to a floor tiles in Roman palaces. In the Early Middle Ages money lender? This workshop handled precious met- porphyry slabs, spolia, were scavenged from the remains als, both gold and silver, and was engaged in forging of Roman buildings (e.g. in Cologne, a former Roman jewellery and other status items; finishing-off weights, province capital) to be reused for this holy purpose

130 Sigtuna: royal site and Christian town and the regional perspective, c. 980-1100

(Tesch 2007c). town´s population with food and fuel. It has been es- timated that 500 people would need about 180,000 kg 30 Portable altars of this kind exist primarily within ar- food per year and that 100 households as in Sigtuna eas belonging to the Roman Catholic Church. In the would require about 2000 cubic metres (c. 800,000 kg) Eastern Church portable altars in the form of collaps- of firewood per year (Edberg 2013:198). ible triptychs made of wood, metal or ivory were used instead. A door from a small Greek Orthodox porta- 38 It is questionable whether the possession of a town- ble altar of bronze has been found in Sigtuna (Tesch yard was a temporary benefit granted to the magnates 2007c:52 fig. 9). who declared their loyalty to the town leader, whether the leasehold was to be confirmed indefinitely, or wheth- 31 As is the case in other early towns like Hede- er it could be inherited within the family. It is reasonable by, Schleswig, Roskilde, Bergen and Dublin (Tesch to believe that the acquired freehold changed over the 2007c:58f; 2014). course of the century and that it could eventually be taxed and sold. 32 So far, however, no such stones have been found in the countryside. In Scandinavia in general, occasional 39 Including Slavs and Saami. The latter group lived as finds have been made, but mainly in aristocratic envi- far south as north Uppland and it was probably they ronments (Tesch 2007c:58; 2008a:147; in print). who brought elk antler to Sigtuna for use in comb-mak- ing. 33 About the concept mindscape see also Johansson 2009:50ff. 40 A smaller part of the Baltic ware was imported from Rus’, but the major part was produced in Sigtuna by 34 A decentralised “landscape of royal workshops” sur- Slavonic women and domestic potters. rounding the king’s yard (Söderberg 2011b:17 and fig. 11), with workshops producing coins, weights and 41 Also discussed by Ersgård 2006:70ff, 100f. goods (high status jewellery) connected with exchange and prestige for the court. The control of weight and 42 One such example is Sanda in Fresta parish c. 25 kilo- silver content indicates an aspect of trade security un- metres SE of Sigtuna (Åqvist 1998:224ff). Other exam- der royal supervision. The production also included ples are Hjulsta in Spånga parish 40 kilometres SSE of exclusive objects in Hiddensee and Mammen style usu- Sigtuna and Mälby in Tillinge parish 40 kilometres W ally connected to royal Danish workshops (Söderberg of Sigtuna (Hed Jakobsson et al. 2013:11ff, 45ff). 2011a:15f; Tesch 2007b:96; Tesch 2015:22ff). 43 Temple rings are a clear ethnic marker included in 35 Astrid was the illegitimate daughter of the Swedish the female Slavic dress. Several such rings have been king Olof Eriksson (skötkonung) and his Obodritian found in Sigtuna, both in graves (on four occasions) concubine Edla. In the year 1019 Astrid was married (Wikström & Kjellström 2010:9) and in cultural layers to King Olav Haraldsson (S. Olav) of Norway. After (see, e.g. Roslund 1990:53ff). Recently a temple ring Olav’s death (1030) she was allied with Olav’s illegiti- was found in a grave in a cemetery in Hjulsta, in Spånga mate son Magnus (the Good), in his struggle to reclaim parish. The cemetery has been defined as belonging to a Norway from the Danes. magnate’s farm (Hed Jakobsson et. al. 2013:55ff, 122). Another two temple rings came from a grave in an early 36 The location may possibly be identified as the raised Christian cemetery in Stav in Färingtuna parish (Cla- hill, Klockbacken, where today a bell-tower is situat- reus 1976:4). ed (Tesch & Vincent 2002:24f). A sheltered place on top of the height might have been the actual Thing site. 44 The theory of hybridity is an important analytical During the 1900’s, this was the site of an outdoor dance concept in modern humanities research; for example in pavilion with a café and tombola stalls. the case of the identity of minority groups and as an expression of cultural resistance in a colonial situation 37 Considerable resources were required to supply a (Bhabha 1994). Most burials during the transition peri-

131 New Aspects on Viking-age Urbanism, c. 750-1100 od in the Mälar region, especially in rural areas, cannot be labeled as Christian or non-Christian, because they seem rather to express a religious identity with charac- teristics of both. The concept of hybriditychallenges essentialism – the idea that there were two such clear- ly defined identity options. Hybridity includes both. Within written sources, such a hybrid religious identity is of course inconceivable, due to Christianity’s exclu- sivist claims. From the point of view of the pre-Chris- tian faith, which was not as systematically constructed as Christianity, hybridity may have come across as a completely natural solution to the challenges presented by the introduction of Christianity in the Mälar region (Tesch 2014; in print).

45 See discussion Kyhlberg 2012:248ff.

132 Sigtuna and excavations at the Urmakaren and Trädgårdsmästaren sites

Sigtuna and excavations at the Urmakaren and Trädgårdsmästaren sites

Jonas Ros

Fig.. 1. The town of Sigtuna with the present division into blocks and streets. The excavated areas at the Urmakaren (1) and Trädgårdsmästaren (2) sites are marked. The church on the Sigtuna museum site (3), St Per (4), St Olov (5), St Nikolai (6) and Virgin Mary (7) are marked.

that a was established in Sigtuna, but the Introduction scholars have not given the letter properly The town of Sigtuna was founded c. 980 AD and took attention. Two major excavations carried out in the over many of Birkas urban functions. Among Swedish Sigtuna block Trädgårdsmästaren and Urmakaren scholars there are different opinions about Sigtuna, give us knowledge of two different plot types and the royal institution, and how the state formation function of the townhouses. One of the excavated took place. The reason for this is that there are few houses was Olof Skötkonungs minting-house. written sources concerning these state of affairs. The sources could be interpreted as the Viking period Svea The founding of Sigtuna and the royal Kings were leaders and also temporarily institution overlords over other provinces. There is no written or archaeological evidence for a royal residence in the town The town of Sigtuna was founded c. 980 AD, by of Sigtuna, but there was one in Fornsigtuna. There which time Birka ceased to have urban functions. were many churches in Sigtuna. During c. 1060–1130 There is no written source that tells who founded Sigtuna was a bishop see, but we do not know which the town of Sigtuna, but it is generally agreed that it church was the bishops church. In 1221 the wrote was Erik Segersäll, the Victorious, who ruled c. 970–

133 New Aspects on Viking-age Urbanism, c. 750-1100

994/95. The early towns were founded by kings and to But Uppland did not become one region with one law understand the foundation and function of towns we until 1296. In the Danelaw the earldoms were typically need to characterize the royal institution. named after their “ towns”, for example In the Adam of Bremen talks about Sueonum had three subdistricts, ridings, and was named after regnum and imperio Sueonum (Adam of Bremen: lib. II York. cap. 35; lib. IV cap. 16) which means the Svea people’s Adam of Bremen calls Sigtuna civitas magna Sictone domination. There is reason to believe that during the and Sictonia civitate. Sigtuna was a missionary centre Viking period the Svea kings were federation leaders and a bishopric (Adam of Bremen: lib. IV cap. 25–26; and also temporarily overlords over other provinces. In lib. III cap. 15) c. 1060–1130. We do not know which Adam of Bremen’s description of (Old) Uppsala he says church was the bishops church, the church on the that the reges, i.e. kings, and the populi, i.e. peoples, had Sigtuna-museum site, St Lars, St Per and St Olov (figure to send their gifts to (Old) Uppsala (Adam of Bremen: 1) have all been suggested. There are strong arguments lib. IV cap. 27). Adam speaks about kings in plural and that the church of St Olov was the Episcopal church (cf. it supports the interpretation that the king’s hegemony Bonnier 1996:193), and it might also have served as a involved many lands administered by Sub-kings or mother church/minster. Some of the bishops of Sigtuna . The Anglo-Saxon kingdoms with their overlords were Benedictine monks (Adam of Bremen: lib. III called or High King could be seen as a model cap. 15; lib. IV cap. 29), and there could have been a for the conditions of Scandinavia during the Viking Benedictine chapter at St Olov. The church is unique period. Many overlordships in England were short (Ros in having a three-aisled , with room for several 2001:24–30 with refs.; 2002:170; Sawyer 1991:6–10; monks. The building work of St Olov was discontinued Yorke 1990:157–62). Administration through earls and the explanation was probably that the bishopric was offered more stable conditions. King Olof Skötkonungs closed down. The church had a wooden nave during half brother, Cnut the Great, divided England into four the medieval period. The present nave was built in the provinces with an over each of them, but later on 16th century (Ros 2009:208–214). However, we cannot Cnut had 11 earls in England (Larson 1912:114–6). The exclude the possibility that St Per was bishop church territorial outlines of the earldoms varied over time and during a period. St Olov could in that case have been some were created for individuals. We know that the a Benedictine monastery. In the a new bishop’s Viking period Svea Kings had earls (Adam of Bremen, see subordinated the in Lund was planted lib. III cap. 12; cap. 53). in (Old) Uppsala. Sigtuna Episcopal sees territory was There are different opinions as to why Sigtuna was connected there (Lovén 2010:14–15). founded. I am of the opinion that the King decided to abandon Birka and found a new town called Sigtuna in Sigtuna’s Viking Age predecessor Birka was an emporium a central place close to the borders between the three and there was a royal residence close to it at Adelsö. provinces , Attundaland and Fjärdundaland. Sigtuna took over Birkas function of being a port of On the Sigtuna-coins Olof Skötkonung entitles himself long distance trade. There were many foreign merchants Rex svevorum and Rex Situne and Rex an Situn i.e. “Svea in Sigtuna and there are strong arguments that St King” and “King in Sigtuna”. Earlier many Viking Nikolai (fig. 1:6) was an orthodox church belonging Age Svea Kings styled themselves Uppsalakonungr, i.e. to a Russian trading company. A prerequisite for trade Uppsalakings (eg. Sundqvist 2002:108–112). Sigtuna- was that the Crown gave the foreign merchants ships King was a new title and a new royal identity. From my military protection. It was a royal right to found cities point of view the king’s intention by founding Sigtuna and the towns yielded revenue. Merchants who came to might have been to combine the three provinces into one the town were liable to pay charges for protection and region with three subdistricts: ridings (tredingar), this for privileges to trade (Ros 2009:201; 2001:72–73). could be seen as a transforming process from kingdom Sigtuna was also a centre for craft production and a to territorial unit of administration. Olof Skötkonung market for domestic trade for the town and hinterland. might have appointed en earl in Sigtuna. Sigtuna was Sigtuna was also a meeting place for the elite and a a name of the town, but my hypothesis is that Sigtuna centre for the Crown. also might have referred to the tree provinces, now Some runic stones from Uppland are of importance known as Uppland (Ros 2001:25–29; 2009:186–187). and of interest when discussing Sigtuna: From Eds parish

134 Sigtuna and excavations at the Urmakaren and Trädgårdsmästaren sites in Uppland a runic stone tells: runa rista * lit * rahnualtr called Bjärköarätt, a name originating from Björkö, i.e. * huar a × griklanti * uas * lis * forunki * (U112), it can Birka. The king who founded Sigtuna likely transferred be read: Ragnvald had the carved. (He) was in the Bjärköarätt från Birka to Sigtuna. Thus, Sigtuna Greece, leader of a fleet (lids forungi). A runic inscription and its hinterland, with its hundreds, had different from Frösunda parish tells that an • ati bo • i þorsulmi laws. The town law was applied within a defined area • ok • i rolstam • - kibliþ (U348, cf. U347), it can be called takmark, and in the laws it is referred to stadz read: he owned a estate in Torsholma and a ship-soke mark, meaning town ground or town boundary (skiplith) in Rolsta. Further on retainers (huskarlar) are (Bjärköarätten: cap. 19; Ros 2009:27–29 and there cited mentioned on runic stones (U330; U335). Thus, the references). By founding Sigtuna the king probably gave runic inscriptions tell that individuals in Uppland could the town privileges related to trading rights and right to be leader of a fleet and that they could own ship-sokes keep markets. In the 1350s it was stated that markets and have retainers. should be held on Mondays in Sigtuna (Svenskt A well known runic stone from parish in Diplomatarium: no. 6010). During Medieval times the Valentuna hundred tells that A: × iarlibaki × lit × raisa Crown strove to locate the trade to towns and skilled × stan + þina × a … kuan + han × ati ain × tabu × alan craftsmen were obliged to move to towns, for example × - ... ont hans + B: × iarlabaki × lit raisa × stain × þin- at Sigtuna (Svenskt Diplomatarium: no. 2691). sik kuikuan × auk × þinkstaþ × þina × karþi + auk × ain ati + alt hu-tari × þita +, it can be read: A: Jarlabanki had this stone raised in memory of himself while alive. He Bryte and royal residence alone owned all of Täby. God help his soul. B: Jarlabanki In Birka the King had an official, named prefect by had this stone raised in memory of himself while alive, Rimbert (Vita Anskarii: cap. 11). In Sigtuna a prefect and made this thing-/assembly-place, and alone he is mentioned in 1274 (Svenskt Diplomatarium: no. owned all of this Hundred (atti alt hundari thetta). Thus, 937), and there was a need for an official from the very an individual could possess hundreds, control hundred beginning. In everyday life the official was called bryte or courts and levy fines as Jarlabanke probably did (U212; fogde. Thanks to a runic inscription from Adelsö where Andersson 1973). In return the individual might have Kings bryte (bry[t]i) in Roden is mentioned (U11) we done military service for the king. In Upplandslagen know that kings-officials could be called bryte. fiefs on service are mentioned (Upplandslagen: § III), Sten Tesch has assumed that there was a royal and the fiefs are hundreds and ship-sokes. By founding residence in Sigtuna and that this moved around Sigtuna the king could create lasting bounds with within the town. One of the residences is supposed to chieftains by granting plots in the town to them and have been on the Sigtuna-museum site (Tesch 2003). also lordship, i.e. jurisdiction over hundreds and ship- Originally Adolf Schück proposed that there was a royal sokes. And chieftains in the king’s service, could control residence in the site (1926:129). However there was no and man warships. This means that the Crown and the big plot in Sigtuna and there is no archaeological or individual noble men interacted in the civil and military written evidence for a royal residence in the town. The administration. Thus, there are strong arguments that royal residences are usually mentioned in the written it existed fiefs in the Mälar region before the ordinance sources. The Sigtuna-museum site is quite small c. of Alsnö 1280 (Ros 2008; 2009:184–185; cf. Andrae 40×40 meters (fig. 1:3) and was divided into parcels. 1960:69–72, compare Lindkvist 1990:50 who has a There is not room for representative hall-buildings in different view). It is important to emphasize that by the that area. The king did not show himself in the town. time Sigtuna was founded the Svea-king royal institution And the king had to think about his security and he also was old and the king had no need to show himself in the distanced himself from the townspeople by setting up town. residence outside the town. There was a royal residence on the other side of Sigtunafjärden in Fornsigtuna (Old Town law Sigtuna). Snorre Sturlusson wrote that Old Sigtuna was an old royal residence (Ros 2001:78–81; 186–192). In the cramped early towns people needed a town Fornsigtuna was granted to the bishop a period around law to solve conflicts occurring during everyday life 1130 (Svenskt Diplomatarium: no. 225). In 1299 king and trade. In Scandinavia the medieval town law was Birger Magnusson put his seal to a letter in Fornsigtuna.

135 New Aspects on Viking-age Urbanism, c. 750-1100

In Rannsakningarna, from the years 1667 and 1672, it is mentioned that at Fornsigtuna there was a walled The Urmakaren site courtyard, probably the remains of the medieval royal In 1990–91 I directed a large scale excavation in the residence. The ancient remains of the late Viking Age block Urmakaren 1 (figure 1:1). About 500 square royal residence in Fornsigtuna covers an area of c. meters were examined. A total of 74 houses were 300×200 meters, and there were two hall buildings, one excavated and sorted in to eight main phases (Ros of them 10×40 meters in size. There were at least eight 2009). The excavated area was divided by a small street other buildings. The King donated Fornsigtuna in 1627 (figure 2). An earlier hypothesis was that there was a (Svenskt Diplomatarium: no. 1855; Damell 1991a; royal residence on the area on one side of the street. 1991b; Ros 2009: 192–200 and ref. cited therein). We However, that area was divided into parcels, and parts of find the same structural similarities in Birka, where four plots were excavated there. The plots on that side of the royal residence was on the other side of the water the street had a width of 4.60–5.50 metres. On the other at Adelsö. Other royal residences close to Sigtuna was side of the street, parts of two parcels and two complete Håtuna and Rölunda (Svenskt Diplomatarium no. parcels were excavated. During phase 1, the two whole 2420; no. 990). parcels II and III were single yards with no passage in The kings´ household was quite large and many between them (fig. 2). The two parcels had two houses people were attached to the king. According to Snorre each. It is noteworthy that the houses on those parcels Sturluson King Olav Haraldsson had 60 warriors were built withdrawn from the street. During phase 2, (hirdmän) 30 messengers and traders (gäster), 30 the two whole parcels that were withdrawn from the retainers (huskarlar) working at the royal residence, a street during phase 1 had become one yard with five, total of 120 people. The king also had many unfree 3+2, houses on it. On the yard, a passage had been persons (Olav den Heliges saga chap. 57). There is created between the houses. It can be demonstrated that reason to believe that Svea-king had about the same the houses represent zones with different functions (fig. number of people attached to him. It is important to 2). There was a loft with multiple functions, a house emphasize that the Viking Age and Early Medieval with cooking functions, i.e. a kitchen, a storage house, kings with households were itinerant and travelled to a dwelling house and a hall, i.e. a house with a primarily royal manors and consumed food. In Uppland there social function. There was continuity of the function were royal manors called Husaby and they belonged to of the houses during the different phases. The plots Uppsala öd (eg. Brink 1990:58). The king had the right had a width of 5.10–6.10 metres and were built with to receive compulsory hospitality, gästning, and food structures up to 15.60 metres long. rents and taxes were gathered at the royal estates. Outside the excavated area, there were parcels with similar plot structures. Town plan and excavations The building techniques in the excavated area changed over time. The oldest houses were fixed to the Sigtuna has an S-shaped street running alongside Lake ground and stabilised with posts dug into the earth (fig. Mälaren. There were plots on each side of the street. The 2). The building techniques wattle-and-daub, vertical- cultural layer covers an area of c. 1100×650 metres, up planking and horizontal-planking were used. The first to 3,5 meters thick. In Sigtuna Storagatan functioned use of sill-stones appears during phase 2. Later on, all as market street and on the plots along the street, houses were built on sill-stones and horizontal-planking there were bodar i.e. shops, used for craft production became the dominating dominant building technique. and trade. Shops, gatubodar, are well known from It is remarkable that none of the houses were built using later medieval periods (eg. Svenskt Diplomatarium: the cross-jointing technique. no. 3219). Trade also took place at the harbour by One of the houses excavated was Olof Skötkonungs the water, but we do not know where the harbour was minting-house. It was 6.4×4.8 m and built with located. More than 400 archaeological excavations have horizontal-planking technique and was a loft (fig. 2). been carried out in Sigtuna (cf Wikström 2005). I will Houses with similar plans can be found in Norway, present two excavations carried out by Sigtuna museum one example is a house from Mykstu yard, Veggli in (see figure 1). Numedal in Norway (Hauglid 1980:184, fig. 3).

136 Sigtuna and excavations at the Urmakaren and Trädgårdsmästaren sites

Fig.. 2. Building remains from the Urmakaren site in Sigtuna. To the left you see the oldest phase, 1, dating c. 980 AD, with parcel I–VII. The houses on parcel II and III were built withdrawn from the street. The houses were simple and fixed to the ground with posts. To the right you see phase 3, which dates back to around the year 1000 AD. The houses represent zones with different functions. The two parcels II and III represent one yard. The minting-house was a loft (1a), and on the plot, a kitchen (3), a storage house (2), a dwelling house (1b) and a small hall-building (4) (after Ros 2009) were also located.

137 New Aspects on Viking-age Urbanism, c. 750-1100

Fig.. 4. In the minting-house, a piece of lead with the impression of a coin-die was found (after Malmer 1991).

Fig.. 3. A house from Mykstu yard, Veggli in Numedal in Norway (after Hauglid 1980, originally after drawing by Johan Meyer). The minting- house in Sigtuna had a similar plan, but was built with the horizontal- planking technique.

In the minting-house, the earliest Swedish coins were made. On the floor we found a piece of lead with an impression of a coin die (fig. 4). The coin production was very large. The average silver coin had a weight of 1,86 grams, and, in total, 200 different dies were used, and at the most 10,000 coins could be struck with one die (Malmer 1989:36; 2010:45–46.). That could, at the most, represent a coin production of 2 million coins or 3720 kilos of coins. Only seven coins struck in Sigtuna were found there (Malmer 2010:72), and one explanation for that is that the coins were handled with special care. When the coins had been used for some time they were changed to new ones and the old coins were melted down and new coins were fashioned from the same silver. Among Swedish scholars, how the early coins were used has frequently been discussed. The coins may have been used for payment, trade, taxes, rents on the coins also expressed the symbolic value of the and fines. During some periods they might have been Christian faith (Malmer 1989:36; Ros 2002 and there counted, during others they might have been used by cited references). weight. Not least the king’s warriors would appreciate Olof Eriksson Skötkonung had the first Swedish coins standardized payment in the form of coins. The cross made. The name Skötkonung was first recorded in the

138 Sigtuna and excavations at the Urmakaren and Trädgårdsmästaren sites

13th century and refers to the word “coin” or “tax”. culture layer that predates the graveyard could indicate The Anglo-Saxon word “sceatta,” for example, may that the plots at the site were laid out in the 10th century mean coin, means of payment, tribute or rent. Olof (Ros 2014). However, the finds from Sigtuna does not Skötkonung’s means, literally, “the coin king” support an interpretation that the plots were laid out or the “tax-collecting king” (Ros 2001:95–98). during the beginning of the 10th century. The 14C–dates do probably not show the time when the individuals died. It is probable that the early 14C–dates are a result St Per, the church in the Sigtuna museum site of the reservoir effect. The reservoir effect can result in and graves in Urmakaren site anomalously old radiocarbon ages. Old carbon could The church of St Per, located west of the town (fig. 1:4), be stored for example in fishes and individuals who ate could originally have been the king’s church (Ros 2009). such fish could accumulate old carbon in the skeleton We know that the Dean of Attundaland had the church and it could result in 14C–dates that are too old. of St Per at his disposal around 1219/1220 (Bonnier In 1220, two Dominican brothers came to Sigtuna. 1989:11). The deanery is known from the 1160s, but They had been in Bologna with the Dean of St Per of was probably established in the after the papal Sigtuna. But the brothers got into trouble when a new legate, Nicolaus of Breakspears, visited Sweden. The archbishop was appointed 1224; they left Sigtuna and king could have given the church of St Per to the Dean went on to Lund (Redelius 1975:112–114). In 1221, in the 1150s. The king then needed a new the pope gave indulgences to anyone who visited the royal and he could have built the stone church at the recently built monastery in Sigtuna, consecrated to Sigtuna Museum site (fig. 1:3). I have earlier proposed the Virgin Mary. It was built with much labour and that the church was a chapel royal and built in the at great expense (Svenskt Diplomatarium: no. 403). 1150s (Ros 2001; 2009:202–208). However, among Christian Lovén has proposed that these were two scholars, the church position in the Sigtuna-museum different events. An unsuccessful attempt to found a site is discussed. Sten Tesch has proposed that it was Dominican monastery and the Pope’s letter may refer the Episcopal Church (Tesch 2003; cf. Svensson 1987). to the establishment of a Benedictine monastery. It may The plot structure in the Urmakaren site was have been (Old) Uppsala’s chapter that moved stable from c. 980 until c. 1230/40. A graveyard was to Sigtuna (2001). A possibility is that the Benedictine established on one side of the street, site with 67 monastery was given to the church at the Sigtuna graves excavated in the site. The graves belong to the Museum site to use at their disposal (Ros 2014). The church in the Sigtuna-museum site. The thickness of church has a unique west end and I have proposed that the cultural layer, which predates the graveyard in the it served as the entrance hall, chapel and baptistery. Sigtuna Museum site, was 1.24 m and it is comprised The church’s plan could be compared with Askeby of 10 phases. Compared with a similar plot type and church in Östergötland and Dalby in Skåne, which zone from the Trädgårdsmästaren site, the thickness both functioned as monastery churches during various of the cultural layer indicates that the graveyard was periods. When the Sigtuna-museum site was excavated established c. 176 years after the layout of the plots. in 1993, it could be demonstrated that the church was I have therefore previously dated the establishment of extended on the south side (Ros 2001:181) and I mean the graveyard to the middle of the 12th century (Ros that it might be the result of the church beginning to 2001:186–194; 2009). function as a monastery church. During the excavation, Anna Kjellström (et al. 2009) had a great number of we found an individual buried with a crosier. Worth excavated skeletal material 14C–dated from the Sigtuna noting was that the grave lay south of the church wall. Museum site and the Urmakaren site. Anders Söderberg It has been suggested that the individual was a bishop, has therefore suggested that there were occasional but it may also have been a Benedictine or prior. burials in the Urmakaren site as early as the beginning Possibly, wooden chapter house was on the south side of of the 11th century, while there were houses on the the stone wall; used to be buried in the chapter neighbouring plots (2013). The14 C–dates indicates that houses. We do not know when the monastery was corpses were buried at the Sigtuna Museum site during closed. In 1237, Dominican brothers started arriving at the beginning of the 11th century (calibrated 2 sigma), Sigtuna probably having been granted the church at the during King Olof Skötkonung´s reign. The 1.24 m thick Sigtuna Museum site. The internments ceased in the

139 New Aspects on Viking-age Urbanism, c. 750-1100

13th century. The Dominican monastery (Holmqvist place at the block Trädgårdsmästaren 9–10 (fig. 1:2) 1947) may have moved to the present site, Virgin and around 1140 square metres were investigated. The Mary (fig. 1:7), in the 1250´s, after which the church excavation was directed by Björn Petterson and Mats on the Sigtuna museum site was demolished, or begun Roslund. Four complete and parts of a fifth plot were to be used as a chapel. Archbishop Jarlers tomb in the excavated. It can be demonstrated that the plot structure chancel’s north wall dates back to 1255, and I maintain developed after the foundation of the town. About 180 that it dates the chancel and the church. houses were excavated and sorted into III main phases. The plots had a length of 30–40 metres and were 7–8 Discussion about ownership of the metres wide. During main phase I, there were two or three Urmakaren plots houses on the plots. During main phase II, the functions were distributed between different houses. In main phase There are no written sources that can tell who owned III, pairs of workshops were erected along the street. the plots in Sigtuna during the town’s oldest centuries. I During main phase II and III it can be demonstrated that interpret the excavated parcels in the Urmakaren site as there was a functional division with four different zones being part of the large number of plots that were owned (fig. 5). In zone I, by the street, there were workshops with by the Crown. The plots could have been possessed economic functions for craft production and trade. The by people attached or employed by the king, such as houses in zone II had multiple functions such as storage. moneyers, skilled craftsmen and warriors. Thus, the The houses in zone III primarily had dwelling functions excavated plots might have been official residences for since they had fireplaces in one corner. In zone IV, at the people in the king’s service, as well as his family. rear of the plot, there was a hall building with a fireplace The fact that the plot structure changed during the in the middle of the floor. Its function seems primarily to th 13 century might indicate that there was a new owner of have been a social one (Petterson 1995; Wikström 2011). the area. The oldest mention of a plot in Sigtuna is from The buildings during the oldest period were simple 1282 when King Magnus Ladulås gave a plot, fundum, and fixed to the ground by posts or simple frames. to the Dominican brothers (Svenskt Diplomatarium: Later on, all of the houses were built on sill-stones. no. 1215). At the Urmakaren site, a hypocaust dating The building techniques used were wattle-and-daub, 1360±40 AD was excavated and it superimposed the vertical-planking, horizontal-planking, and cross- graveyard. The dating shows the last time the hypocaust jointing. was heated (Ros 2001; 2009:143:202–204).There was a hypocaust in the Dominican convent in Sigtuna and there is a possibility that also the hypocaust in Discussion on the Trädgårdsmästaren site the Urmakaren site was built by the Dominicans, but Compared with finds from other excavations it could as well have been built by the . from the Middle Ages, many of the finds from the Hypocausts are quite rare and would usually be found Trädgårdsmästaren site are extraordinary and exclusive, in and . I would like to propose that indicating that the plots were owned by members of it might have been the big plot in the Urmakaren site the elite. On one of the plots, a rib-bone with runic that King Magnus gave to the Dominicans in 1282. inscription on both sides was found; side A: kunukr: iaR We know that he made donations to monasteries, and : matr : bestr : han : a f mest : han iaR : þekili B: mari • selti that Magnus Ladulås built castles, one of them being • rif • a fe mest • aff asa (Gustafsson 1992:165–167). It Alsnöhus. The ordinance of Alsnö was an act of Magnus can be read: A: The king is most generous with food. He th Ladulås. During the 13 century, there were major owns the most. He is full of favours. B: Mar(r)e (?) gave changes, and the older royal administrative system the rib-bone. He owns the most…Asa. The inscription based on Husabyar ceased to be in function. A new on the B side tells that a wealthy person named Mar(r) administration based on castles was established (Fritz e had a feast. On the A side, the king is glorified as 1972-73). most generous with food. Only people attached to the king received the king’s hospitality. The king’s subjects The Trädgårdsmästaren site were obliged to contribute to the king’s household’s compulsory hospitality. Thus, the runic-bone supports During the years 1988–90, a large scale excavation took the theory that the plot in the Trädgårsmästaren site was

140 Sigtuna and excavations at the Urmakaren and Trädgårdsmästaren sites

Fig.. 5. The main phases of the building development on the plots in the Trädgårdsmästaren 9 and 10 block in Sigtuna. In main phase I, there are two or three houses. In main phase II, the functions have been distributed between different houses. In main phase III, pairs of workshops have been erected along the street. The buildings on the plots were divided up by function in main phase II and III and have been grouped into four different zones. By the street there were shops. The houses in the next zone had multi-purpose buildings and storage functions. Behind them, dwellings with fireplaces were found in one corner. At the back of the plot there is a hall with a central fireplace on the floor (after Petterson 1995).

inhabited by a person attached to the king. shops that were leased. People who rented houses in During the oldest period, the Crown could have the towns could be called husmän, (Magnus Erikssons owned the plots and the plots could have been granted to Stadslag: §101) and they were craftsmen, merchants and possessed by ecclesiastical institutions, clerics, lords and krämare, i.e. traders. The craftsmen could have and chieftains of whom many were the king’s men. Later paid their rents in artefacts to the one who controlled on, they may have been given ownership of their plots. or owned the plots. The fact that the shops lay along It is possible that the possession of plots by that time the street shows that they had economic functions and was associated with the obligation to pay charges. Lords they were used for market production and trade. The and clerics may have resided in the town-houses or have street was a Market Street. The craftsmen and traders had agents attached to them, i.e. retainers, huskarlar, sold their products through opened or gaps and servants in the town. Along the street there were in the wall along the street. Debris from antler is very

141 New Aspects on Viking-age Urbanism, c. 750-1100 extensive in the shops along the street. It is remarkable that elk antler was the most common raw material for Importance of the excavations and outlooks on comb production. Bone from elk is very rare from late some Viking Age towns Iron Age settlements in the Mälaren area. Extensive elk The archaeological excavations in the Urmakaren and hunting had rendered the species extinct in the Mälaren the Trädgårdsmästaren sites are of importance since basin by that time. The raw material elk antler was entire plots were excavated, increasing the knowledge imported to Sigtuna from the provinces to the north. of the functions of the houses. The type of plot Bronze casting and iron forging are examples of other excavated in the Trädgårdsmästaren site was the most crafts that took place in the shops. The craftsmen in common plot type in Sigtuna. In the Urmakaren site, the town produced goods and sold products, kramvaror, we can see that the plots could be of another type than such as combs, needles, dice, buckles and shoes for those in the Trädgårdsmästaren site. The find of Olof use in the town and the surrounding countryside. In Skötkonung’s minting-house is important because we can the oldest layers, debris from craft was fairly limited archaeologically prove that King Olof had coins made and during that period the craftsmen were itinerant, in Sigtuna. Previously, it has been theorized that these periodically and seasonally active in the town-houses coins were made in Västergötland. The archaeological (Ros 2001:109–112; 2009). excavations are also of great importance since the vast During the 12th century there were pairs of shops number of excavated houses, not less than about 254, on the plots along the street, and at the rear of one of enable us to study the building techniques between the plots, there were two pairs of dwelling houses (fig. 980–1230 AD. On the countryside, in the middle of 5). By that time, there were no longer representative Sweden, quite a few houses with sill-stones have been hall building on the plot. This might indicate that the found from that period. craftsmen and traders that were active in the shops along The two Sigtuna excavations are also of importance the street lived in the houses on the plots. It is possible when discussing urbanism and comparing different types that the craftsmen and the traders by that time rented of Viking Age plots. The plots in the Trädgårdsmästaren parts of plots with dwellings and shops along the street. site display considerable similarities to plots excavated Possession of parts of plots with access to the street is in Trondheim and Oslo. Generally, plots in early Viking known from the Middle Ages. Earlier, the craftsmen Age towns, such as Birka (Ambrosiani 2013), Ribe could have lived outside of the plots or slept in the shops. (Feveile 2006), Dublin (Wallace 1992) and York (Hall During the Medieval period, ecclesiastical institutions, 1986), were relatively small and there could be one lords and other individuals received ownership of a great or a couple of houses with many functions; dwelling numbers of urban plots, which they rented out. Another houses as well as some used for goods storage. Even craft possibility is that the craftsmen and traders in Sigtuna production took place there. There is no evidence of acquired ownership of plots during the 12th century. hall buildings on the plots and that indicates that those One of the explanations of the decreased occurrence of who had the plots at their disposal did not have the representative urban hall building on the plots in Sigtuna need or the social position­ to build such houses. Much during the 12th century might be that new towns were indicates that those who used the buildings­ rented them founded, and lords and clerics might have acquired new and/or were dependent on those who owned the houses urban residences. The towns of Enköping and Uppsala and the plots. The early towns were craft and trading were founded in the adjacent regions of Fjädrundaland places, emporiae, or, as in England and , also and Tiundaland. Attempts were also made to establish royal strongholds for soldiers to assemble, as well as to a town, villulam forensem, at a thing assembly site exercise control over regions (Ros 2009:257). called Folklandstingstad, in Attundaland during the In late Viking Age towns, founded around 1000 14th century. The trade and craft activities that took AD, Sigtuna, Trondheim (Christophersen & Nordeide place there were, however, opposed by the king and 1994) and Oslo (Schia 1987), the plots were bigger and the activities seem to have ceased in the 1350s (Svenskt there were a number of houses with different functions; Diplomatarium: no. 2388; no. 6010). e.g. shops, storage houses, kitchens, dwelling houses and hall buildings (fig. 2; 5). On the plots many people were active; house-owners and their families that sometimes resided in the halls, retainers, serfs preparing food and

142 Sigtuna and excavations at the Urmakaren and Trädgårdsmästaren sites craftsmen and traders renting shops along the street. In many cases, the plots were comparably narrow and the reason for this was that it was important to hold many plots with access to the street. The model for the plots excavated in Sigtuna, Trondheim and Oslo might have been the English towns, but also Birka and Ribe. They might have been models that were modified and developed, with more houses with differentiated functions being added (Ros 2009:256). During the Iron Age, the three-aisled longhouse was a common house type. The houses on the plots during the 11th century, for example in the Trädgårdsmästaren site, could be compared with longhouses split up into different rooms.

Corresponding Author Jonas Ros Stiftelsen Kulturmiljövård Västerås, Sweden [email protected]

143 New Aspects on Viking-age Urbanism, c. 750-1100

References Gustafsson, J.H., Snædal, T. & Åhlén, T. 1992. Runfynd Adam of Bremen. Historien om Hamburgstiftet och dess 1989 och 1990. Fornvännen 1992 (87). 153 – 174. biskopar. Översatt av E. Svenberg 1984. Med Hall, R. 1986. Archaeological aspects. In: P.V. Addyman kommentarer av C. F. Hallencreutz, K. Johanneson, (ed.) Post-Roman Coins from York Excavations 1971- T. Nyberg och A. Piltz. Stockholm. 81. The Coins. The Archaeology of York 18. York. Ambrosiani, B. 2013. Excavations in the Black Earth 15 – 25. Hauglid, R. 1980. Laftekunst, Laftehusets 1990-1995. Stratigraphy 1. The site and the shore. opprinnelse og eldste historie. Oslo. Birka studies 9. Stockholm. Holmqvist, W. 1947. Mariakyrkan och klostret. In: H. Andersson, T. 1973. Iarlabanki atti alt hundari thetta. Arbman (ed.) Sigtuna Mariakyrka 1247-1947. Till tolkningen av Jarlabanke-ristningen i Vallentuna. Sigtuna: 7 – 44. Svenska studier från runtid till nutid tillägnade Carl Kjellström, A., Storå, J., Possnert, G. & Linderholm, A. Ivar Ståhle på 60-årsdagen den 27 juni 1973. 2009. Dietary patterns and social structures in Nämnden för svensk språkvård 48. Stockholm. 16 – medieval Sigtuna, Sweden, as reflected in stable 29. isotope values in Andræ, C.G. 1960. Kyrka och frälse i Sverige under äldre medeltid. Studia historica Upsaliensia 4. Uppsala. Bjärköarätten. Äldre västgötalagen, Yngre västgötalagen, Smålandslagens kyrkobalk och Bjärköarätten. Tolkade och förklarade för nutidens Svenskar. Svenska landskapslagar 5. Transl. & eds. Å. Holmbäck & E. Wessén 1946. Stockholm. Bonnier. A.C. 1989. Sigtuna och kyrkorna. In: S. Tesch (ed.) Avstamp - för en ny Sigtunaforskning. 18 forskar om Sigtuna, heldagsseminarium kring Sigtunaforskning den 26 november 1987 Gröna Ladan, Sigtuna. Sigtuna. 9 – 15. Bonnier, A.C. 1996. Kyrkor, dopfuntar och gravmonument In: B. Nilsson (ed.) Kristnandet i Sverige. Gamla källor och nya perspektiv. Uppsala. 181 – 214. Brink, S. 1990. Sockenbildning och sockennamn. Studier i äldre territoriell indelning i Norden. Uppsala. Christophersen, A. & Nordeide, S.W. 1994. Kaupangen ved Nidelva. 1000 års byhistorie belyst gjennom de arkeologiske undersøkelsene på Folkebibliotekstomten i Trondheim 1973–1985. Trondheim. Damell, D. 1991a. Utgrävningens metodik. In: A. Allerstav (ed.) Fornsigtuna. En kungsgårds historia. Upplands-Bro. 32 – 33. – 1991b. C14- och termoluminicensdateringar. In: A. Allerstav (ed.) Fornsigtuna. En kungsgårds historia. Upplands-Bro. 83 – 85. Feveile, C. (ed.) 2006. Det ældste Ribe: udgravningen på nordsiden af Ribe Å 1984-2000. Ribe studier 1:1–2. Jysk Arkæologisk Selskabs skrifter 51. Højbjerg. Fritz, B. 1972–73. Hus, land och län. Förvaltningen i Sverige 1250-1434. Acta Universitatis Stockholmensis 12. Stockholm studies in History 16 and 18. Stockholm.

144 No kingdom without a town

No Kingdom without a town. Anund Olofsson’s policy for national independence and its materiality

Rune Edberg

Has a Danish king reigned in Sigtuna? The answer is the Scandinavian Viking coins should not be read liter- yes according to some scholars such as Thomas Lind- ally. Sigtuna coins “…in some cases do not provide the kvist in “The Viking World”, a British Stand der For- name and the title of the real issuer of the coins, but a schung collection of texts by leading scholars, covering mixture of real names and titles and names and titles of numerous aspects of the Scandinavian Viking Age. persons found on the prototypes”. He added: “When Lindkvist wrote in his article, entitled “The Emergence understood in the context of the nature of coin legends of Sweden”: “In the middle of the 1020’s in this period, it becomes clear that the legend CNVT was defeated by King Canute the Great of Denmark, REX SW is almost certainly a meaningless idiom rather who for a period was in control over Sigtuna” (Lind- than a statement of fact” (Jonsson 2007:274). kvist 2008:670). Anund Jacob, also known as Anund In her work, Brita Malmer was able to show that Olofsson, (Anund, for short here, the only name he all inscriptions of CNUT REX SW derive from one used on his Sigtuna coins) was the son and successor and the same obverse, which is part of a Danish stamp of Olof Skötkonung, Sweden’s supposedly first Christian chain. She concluded that its die was probably manu- ruler. Canute (in Scandinavia known as Knud or Knut) factured in Lund (Malmer 2010:86). was an Anglo-Danish conqueror king, the son of Sven Forkbeard and grandson of Harold Bluetooth, of Jelling fame. Canute’s England was the jewel in Canute’s crown. After having subjugated this rich country and then stamped out re- Canute’s coins sistance on the home turf, he had, in the early 1020’s, Among the many designs of coins minted in Sigtuna in become a major European player. Consequently, when the first decades of the 11th century AD, some bear the he went to to assist at the papal of inscription CNVT REX SW. These coins were known the German emperor Conrad, he was received there to scholars already in the 19th century and have been with great honours and treated by the other attending crucial for the claim that Canute had had brief political dignitaries as their peer – at least according to his own control of the Lake Mälaren area of central east Swe- accord, in the so called Proclamation of 1027 (Laws of den, where Sigtuna then was the only town (Jonsson the Kings 146–153). 2007:272f has the full research history). Canute’s aspirations also extended north, to the Scan- However, it has recently been pointed out that this dinavian Peninsula, and east, to the Vendish areas of evidence is quite unfounded, first by Kenneth Jonsson, the southern Baltic coast. First and foremost he aimed who published an article on the matter, and later by at Norway, which had previously been under Danish Brita Malmer, who in a major work published and com- , in its key region of Trøndelag through vas- mented on all known coins struck in Sigtuna (Jonsson sals, the earls of Lade. Now and then, anti-Danish pre- 2007; Malmer 2010). The two numismatics both re- tenders had succeeded to oust these earls and come to jected the notion that these coins may be taken as proof power themselves. While Canute was busy conquering of any Canute’s minting in Sigtuna. England, a Viking adventurer, Olav Haraldsson, had Kenneth Jonsson explained that the inscriptions on been recognized as their king by the Norwegian mag-

145 New Aspects on Viking-age Urbanism, c. 750-1100 nates (Lund 2008:665f; Sawyer & Sawyer 1993:58f). Chronicle is called the ”Holy River”, usually under- Consequently, after having firmly established his stood as the Helgeå estuary in east Scania. The event , Canute planned to make short work took place c. 1026 AD. Canute’s troops, both Danes of Olav just as his father Sven had a few decades earlier and Englishmen, confronted Anund’s and Olav Har- (at the Svold three-king sea-battle, 999 or 1000 AD) aldsson’s joint land and sea forces. (Anglo-Saxon Chron- made with another rebellious Norwegian namesake and icle: 157; Helgeå is also mentioned in skaldic verses and upstart, Olav Tryggvason. other sources, cf. Moberg 1942; 1985; 1987; 1995 and references therein; also Gräslund 1986:211f; cf. Gahrn 1988:131) This is the event that Thomas Lindkvist al- Aspiration for Sweden ludes a defeat for Anund. In the previous generation, Olof Skötkonung had assisted According to Ove Moberg, a leading scholar on the Sven Forkbeard with a naval force when Olav Tryggva- Helgeå subject, there are surprisingly many sources on son was disposed of, and Swedish troops under Danish this military campaign, considering that we are dealing supreme command had also taken part in the extreme- with events in Scandinavia’s historically very lean Vi- ly rewarding, repeated pillaging of England. Some of king Age (Moberg 1942; 1985; 1987; 1995.) Despite these warriors are known by name from a number of this, it is paradoxically not clear who defeated whom at contemporary runic stones, in Uppland and elsewhere Helgeå. (cf. Edberg 2006, 2007:90f, 2011). But these mutually One annal, the Anglo-Saxon chronicle (Anglo-Saxon profitable relations were now history. Chronicle: 157) states that “The had possession In the Proclamation of 1027, Canute named himself of the place of slaughter” after the battle. Some scholars as king of England, Denmark, the Norwegians and also (e.g. Axelson 1955:49) interpret this as if Anund (and ”partis suanorum”, i.e. a part of the Swedes, (Laws of the Olav) did in fact defeat Canute. But as noted above, Kings:147; cf. Axelson 1955:36) an expression which others, using different evidence, conclude that Canute has been given diverging interpretations by scholars. prevailed. But regardless of the immediate military out- Some have seen it as Canute’s tribute to the Swedish come, the long-term consequences are quite clear. For units in his victorious armies (e.g. Lund 2008:667). even if Canute claimed to have carried the day at Hel- Others have regarded it as a sort of propaganda (e.g. geå, he never was to challenge Anund again. It was, at Axelson ibid.; Malmer 2010:22f) while still others have most, a Pyrrhic victory. considered it as reflecting a real situation (e.g. Gräslund 1986:217), and hence supporting the alleged Sigtuna coin evidence. No occupation The Danish ambition to conquer Sweden was, in As Canute’s coins now can be left out of consideration, fact, nothing new. It had been experienced on several the idea of a Danish occupation or supremacy of Sigtu- previous occasions. The half-legendary battle of Fyris- na in the 1020’s must be rejected. The circumstantial vall (near Uppsala) in about 985 AD is an example. evidence, previously used to support the idea, is not That was a failed military attempt with a Swedish royal enough. This applies not only to Canute’s alleged Hel- nephew, Styrbjörn, as point man and Danish puppet- geå victory but also e.g. to a number of coeval archae- to-be. ological findings of fine jewellery in the Lake Mälar- en area, associated with the Danish royalty (Duczko 1995:644f). These artefacts can as well be explained as Pyrrhic victory originating from returning warriors, the exchange of After Anund had rejected Canute’s proposals (accompa- prestige gifts in general or even by local craftsmanship. nied by bribes) to stay neutral while the Dane took care Also, Bo Gräslund’s clever attempt to relocate Helgeå of Norway, war was unavoidable. And rather than wait- from Scania to Uppland can hardly be maintained, ing to be overrun one at a time, Sweden’s Anund and first and foremost because of he took Canute’s Sigtu- Norway’s Olav joined in opening hostilities far away na (now dismissed) minting for undisputed (Gräslund from their centres, the towns Sigtuna and Trondheim 1986:211f). respectively. They got a military settlement on Can- Regarding Sweden, the outcome of the Helgeå bat- ute’s hereditary dominions, at what in the Anglo-Saxon tle kept Anglo-Danish imperialism at bay. However,

146 No kingdom without a town

Fig. 1. Coin struck for King Anund in Sigtuna c. 1030 AD. “Pointed helmet”-type. One example of the king’s material manifestations and part of the town’s still surviving legacy. Photo courtesy of Kenneth Jonsson.

Norway was not so fortunate. In a new campaign, Ca- of Bremen summed it up (Adam of Bremen: lib. II cap. nute chased Olav Haraldsson in exile and later, with 57). And only a few years after Olav’s martyrdom, the the help of a host of local supporters, he even succeed- pre-Christian opposition in Norway had surrendered ed in having him put to death (at the battle, and converted. Canute’s death in 1035 AD also helped. 1030 AD). But this victory soon also proved Pyrrhic Norway was then to stay politically independent for as it gave the Norwegians a national hero, very soon 300 years(Sawyer & Sawyer 1993:57f). after his death hailed as S:t Olav and soon to become Ulf Näsman, researching the pre-history of Den- very popular all over Northern Europe(cf. Steinsland mark, has emphasized that there is nothing given by 1990:158f). The Norwegians fought for freedom, the nature in the formation of the three Scandinavian king- Danes for power, as late 11th century chronicler Adam doms. They are the result of specific historical circum-

147 New Aspects on Viking-age Urbanism, c. 750-1100 stances (Näsman 2006:207f). One might add that even from interfering with the pagan shrine at Uppsala de- single events may decide the life and death of nations. spite pressure from the bishops (Adam of Bremen: lib. Many examples could be given from crucial moments II cap. 58; lib. IV cap. 39). It has long intrigued scholars in Swedish – or any other nation’s – history. It will suf- why the kings were forced to make this concession to fice to mention the ’s crossing of the fro- the pagans – or for their own personal security, even zen Belts in 1658 after which the panic-strucken Danes at some instances leave Uppland for a safe haven in ceded Scania to Sweden (cf. Weibull 1949:149– 185). Västergötland, where, at Skara, Olof Skötkonung and the The battle of Helgeå may well be looked upon among Hamburg-Bremen diocese had established an episcopal those. see. Also, even when a see eventually was established in Sigtuna in the 1060’s it was unoccupied for long pe- riods of time, highlighting this domestic religious and No collaborators political tug-of-war (Hallencreutz 1996:246f; Lindkvist If accepting this interpretation so far, we may look at 2006:89f; cf. Gahrn 1988:111f). the issue of how Anund’s success can be explained in During this turbulent era, in counterbalance to the view of his serious social trouble at the home front. religious rift, cohesive social forces apparently also ex- For it is acknowledged that parts of the Uppland elite isted and prevailed. As a suggestion, it was the pagans’ and peasantry refused to rally behind his Europeanized part in defending the common entity, which explains the kingship, with its foothold in Sigtuna. Instead they kept tolerance they were shown. After the Helgeå battle, the to the pre-Christian Pantheon with its trusted rituals, Uppsala temple and cult would persist in this manner handed down for many generations. for nearly 100 years, without the aspiring Swedish na- In principle, any baptized king would have entered tion being shattered politically. Anund and his imme- into a fellowship with universal claims and with cen- diate successors seem to have simply waited out their trally determined, compelling sacraments. He must cut pagan antagonists, even accepting stints of yielding some all ties with the old cult, where he had enjoyed a sacred power, certain that the pagans’ time was running out. An status, as patronage of the rituals and political power exception proves the rule: a short-lived pagan restoration were two sides of the same coin. And to retain his su- headed by a king who goes by the name of Blot-Sven perior position, his people must also follow in his foot- (“Sven the Sacrifier”) is said have occurred in the 1080’s steps and be baptized. Leaving them unbaptized would (cf. Hyenstrand 1989:20f; Lindkvist 2006:93f; Zachris- be tantamount to risking a split of society (Steinsland son 1998:154f). 2000:85f; cf. Lindkvist 2006:84f). Reasonably, if such a thing had occurred, it would have greatly weakened Anund’s power to take military Steel and pillars action. But this split did not take place, as far as we The struggle for political support and societal unity in know. Also, for the non-Christian Upplanders to have the 11th century was surely fought not only in the hearts strengthened their hand by bypassing Anund and seek- and of people, but also with material culture. ing alliance with Canute would not have been surpris- For, as recently pointed out by Bjørnar Olsen, “societies ing. But no Swedish counterparts to the Norwegian and nation-states are not cognitive sketches resting in collaborators, notably the (also pagan) earls of Lade, are the minds of people; they are real entities solidly build mentioned in any sources. The Uppland non-Christians and well tied together”. And Olsen continues: “Rather thus appear to have been loyal to the dynasty, or at least than ‘revealing’ entities as constructed and made up, neutral, even though it now had embraced Christian- our attention should be devoted to analysing how these ity (cf. Hyenstrand 1996:70). As a matter of fact, no entities (e.g. societies and cultures) are put together and tradition of a king who converted his people (e.g. like the real building materials–the concrete and steel, re- in Russia, , Denmark etc.) ever developed in bars and pillars–involved in their construction.” Olsen Sweden (Lindkvist 2006:83f). concludes: “We should pay far more attention to the material components that constitute the very condi- tion of possibility for those features we associate with Peaceful tug-of-war social order, structural durability, and power…” (Olsen During the 11th century, the Christian kings refrained 2010:5).

148 No kingdom without a town

The foundation of Sigtuna may be looked upon as an lievers alike may be interpreted in ethnical terms (more initiative by the ruling dynasty to adapt to the new era below) (cf. Zachrisson 1998:136f). and towards nation building. It may very well be seen At the same time, the implicit meaning of the runic as a necessary construct and requisite for establishing stones, the erection of which manifested support to the and maintaining Sweden, a new, aspiring but still fragile new Christian order, would not have evaded anyone (cf. nation-state in the European periphery. The town itself Lönnroth 1982:16f; Sawyer & Sawyer 1993:14f). became the most prominent materiality in the region at the time. Indeed, the expansion of the town and its fa- cilities (cemeteries, dwellings, streets, workshops, mar- Sacred king kets and churches) may stratigraphically be followed Anund’s personal standing must have been very strong. from Erik Segersäll (“the Victorious”), his son Olof Although Christian (if only second generation), he de- Skötkonung, grandson Anund and on. Today Sigtuna’s scended from kings with legendary roots tracing back to “Black Earth”, with its hundreds of tons of waste, its the pagan gods. The tradition around Anund’s double thousands of burials and its many ruins remind us of name (Adam of Bremen: lib. II cap. 59) with Anund this epoch. marking continuity with the pagan naming of the Yn- A kingdom worthy of its salt, however, must not glingar royal house and Jacob (“James”) being a pure- only have a town, but also a mint. ly Christian name, may also be put into this context. The Ynglingar, claiming divine origin, were known as lords of the emporium and necropolis of Birka in the Coins and rune stones 9th and 10th centuries (cf. Carlsson 1997:87; Malmer The Sigtuna mint, the coins featuring the Cross and the 1996:99f). King and initiated by Olof Skötkonung c. 995, was cru- The king’s sacred nature may itself be seen as a cata- cial to establishing the Christian stronghold. However, lyst in the conversion process, and it has been pointed during Anund’s reign the silver supply from the Viking out that Christ himself came to Scandinavia as warrior fleecing of England and Saxony went more or less dry king (cf. Steinsland 2000:90f). and Anund’s minting petered out in the 1030’s. (Figure Anund’s family alliances with the royal and princely 1) houses around the Baltic were also a major aspect in se- Instead, the town become a node of another diffus- curing his position. His sister Ingegerd had wed Prince ible materiality, the runic stones, a huge wave of which Yaroslav of Novgorod, later to become Grand Prince of from the early 1000’s, infiltrating the countryside with Kiev; his stepsister Astrid married Norway’s king and the promise of resurrection and paradise (cf. Sawyer & saint-to-be Olav Haraldsson. Another very important Sawyer 1993:14f: Zachrisson 1998:127f, 149f). Sand ally was the Danish pretender Sven Estridsson. Sven, a stone fragments, some with traces of runic ornaments, mortal enemy of Canute, bided his time in exile with have recently been identified in its occupation layers. Anund, later to return with might to Denmark (cf. Hal- This may be traces of a workshop for rune stones or lencreutz 1996:246f). Norwegian kings-to-be Magnus early Christian grave monuments, or both (Söderberg Olavsson and Harald Hard-ruler both made friendly 2013:67f). stops at Anund’s Sigtuna when they travelled home Incidentally, the word infiltrating here seems very from Russian exile to Norway (c. 1030 and 1044 AD, appropriate, as no fixed border existed between to two respectively). camps, which lived alongside each other throughout the 11th century, even in the same . Analyses of grave customs have been central in this discussion, but it has Svethiudh and Sverige also involved the distribution of silver hoards and other Åke Hyenstrand once stated: “A major weakness of the factors (cf. Broberg 1990:31f,70f; Lindkvist 1996:224f; discussion has been the mix-up of the Swedes and the Nilsson 1996:354f; Sawyer & Sawyer 1993:104; Zach- Swedish kingdom. These should be seen as two separate risson 1989:235f). concepts, two structures of society that must be kept The rune stone messages’ using both the old, well- apart if the discussion is to be touched up” (Hyenstrand known heathen image symbolism and the common 1989:15, my transl.). language to appeal to believers and non-be- Even if there is no agreement among scholars about

149 New Aspects on Viking-age Urbanism, c. 750-1100 the details, Sweden’s formation to a nation-state, Sverige (“Svea-kingdom”) was late, compared to e.g. Denmark. However, Sverige’s roots can no doubt be found in the ethnic entity formed during the late Iron Age, which under the name Svethiudh (“Svea-people”) is known from several older sources (cf. Hjärne 1952:128f; Gahrn 1988:38f, 1993:25f). Christianization and nation building in Sweden was an internal process. Sweden never suffered from any sword’s mission (cf. Lindkvist 1996:219f). I suggest that the Swedish resistance to Danish and Anglo-Dan- ish imperialism may be interpreted in ethnic and pro- to-national terms, and that the lengthy transition from pagan to Christian order in society never escalated to a civil war may be looked upon against the same back- ground, these two conditions being intertwined. In its turn, waging war against the Danes was part of what constituted the Swedish identity during many centu- ries, until the death of Charles XII in 1718.

New urbanism In the Tiundaland part of Uppland the religious show- down appears to have lasted till the 1120’s. Until then, the building of a stately church could not be safely initi- ated in Uppsala, on, or close to, the very spot of Svethi- udh’s main pre-Christian cult centre. And in the 1160’s Sweden became an ecclesiastic province, also signify- ing the final political unification of the Svealand and Götaland regions (Sawyer & Sawyer 1993:60f; Lind- kvist 2008:671). Also, under king Knut Eriksson, Swe- den eventually produced its own national martyr-saint, namely Knut’s father (nota bene assassinated by a Dan- ish prince). Saint Erik became a powerful symbol for Uppsala in its new role as the new national Christian shrine, where the saint’s holy relics were, and still are, kept (Lindkvist 2006:83f; Nilsson 1996:408f; Sawyer & Sawyer 1993:58f).

Corresponding Author Rune Edberg Sigtuna, Sweden [email protected]

150 No kingdom without a town

legend show that Cnut the Great ruled at Sigtuna? References In: U. Fransson, M. Svedin, S. Bergerbrant & F. Adam of Bremen. Historien om Hamburgstiftet och dess Androshchuk (eds.) Cultural interaction between biskopar. Transl. E. Svenberg. Comments by C.F. east and west. Archaeology,artefacts and human Hallen creutz et al. (1984). Stockholm. contacts in northern Europe. Stockholm Studies in Anglo-Saxon Chronicle. Transl. with an introduction Archaeology 44. Stockholm. 272 – 278. by G.N. Garmonsway (1967). London. Laws of the Kings = The Laws of the Kings of England Axelson, S. 1955. Sverige i utländsk annalistik 900– from Edmund to Henry I. Ed. & transl. A.J. Robert 1400. Uppsala. son (1925). Cambridge. – 1956. Sverige i dansk annalistik 900–1400. Uppsala. Lindkvist, Th. 1996. Kungamakt, kristnande, stats Broberg, A. 1991. Religionsskifte och sockenbild bildning. In: B. Nilsson (ed.) Kristnandet i Sverige. ning i Norduppland. O. Ferm (ed.) Kyrka och socken Gamla källor och nya perspektiv. Uppsala. 217 – i medeltidens Sverige. Stockholm. 49 – 79. 241. Carlsson, A. 1997. Birkas kungsgård på Adelsö och – 2006. Att skapa ett kungarike. Maktlegitimering, svearnas Fornsigtuna. Två aristokratiska miljöer i regional variation och framväxten av ett kristet kun Mälardalen. In: Callmer & E. Rosengren (eds.) gadöme i Sverige. Saga och sed 2006. 83 – 98. “-gick Grendel att söka det höga huset-”: arkeologiska - 2008. The Emergence of Sweden. In: S. Brink & N. källor till aristokratiska miljöer i Skandinavien under Price (eds.) The Viking World. London & New York. yngre järnålder: rapport från ett seminarium i Falken 668 - 674. berg 16-17 november 1995. Halmstad. 83 - 88. Lund, N. 2008. Cnut the Great and his Empire. In: S. Duczko, W. 1995. Kungar, thegnar, tegnebyar, juveler Brink & N. Price (eds.) The Viking World. London och silverskatter. Om danskt inflytande i Sverige & New York. 665 – 667. under sen vikingatid. Tor 1995 (27):2. 625 – 662. Lönnroth, E. 1982. Administration och samhälle i Edberg, R. 2006. Folk och rövare i Sigtuna stad. Da 1000-talets Sverige. Bebyggelsehistorisk tidskrift 1982 nagäldernas roll vid samhällsomvälvningen omkring (4). 10 – 23. år 1000. Situne Dei 2006. 95 – 129. Malmer, B. 1996. Sigtunamyntningen som källa tilL - 2007. Sigtunaleden och mysteriet med de saknade Sveriges kristnande. In: B. Nilsson (ed.), Kristnandet vikingaskeppen. Situne Dei 2007. 79 – 97. i Sverige. Gamla källor och nya perspektiv. Uppsala. – 2011. Exit Knut. Ett apropå till de nya rönen om 85 – 113. Sigtunamyntningen. Situne Dei 2001. 107 – 111. – 2010. Den svenska mynthistorien. Vikingatiden ca Gahrn, L. 1988. Sveariket i källor och historieskrivning. 995–1030. Stockholm. Göteborg. Moberg, O. 1941. Olav Haraldsson, Knut den store och – 1993. Svearna i de uppländska folklanden och det Sverige. Lund. svenska rikets uppkomst. Kärnhuset i riksäpplet = – 1985. Knut den stores motståndare i slaget vid Hel Årsboken Uppland 1993. 9 – 74. geå. Scandia 1985 (51):1–2. 7 – 17. Gräslund, B. 1986. Knut den store och sveariket. Slag - 1987. Slaget vid Helgeå och dess följder. Scandia et vid Helgeå i ny belysning. Scandia 1986 (52):2. 1987 (53):1. 175 – 184. 211 – 238. – 1995. Trekungaslaget vid Helgeå. Ale 1995 (4). 16 Hallencreutz, C.F. 1996. Riksidentitet, stiftsidentitet – 24. och den vidare Europagemenskapen. In: B. Nilsson Nilsson, B. 1996. Från gravfält till kyrkogård. Föränd (ed.) Kristnandet i Sverige. Gamla källor och nya per ringar och variation i gravskicket. In: B. Nilsson spektiv. Uppsala. 243 - 268. (ed.) Kristnandet i Sverige. Gamla källor och nya per Hjärne, E. 1952. Svethiudh. Namn och Bygd 1952. spektiv. Uppsala. 349 – 385. 91-183. Näsman, U. 2006. Danerne og det danske kongeriges Hyenstrand, Å. 1989. Sverige 989. Makt och herravälde opkomst. Om forskningsprogrammet ”Fra Stamme I. Stockholm. til Stat i Danmark”. Kuml 2006. 205 – 237. - 1996. Lejonet, draken och korset. Sverige 500–1000. Olsen, B. 2010. In defence of things. Archaeology and Lund. the ontology of objects. Lanham, MD. Jonsson, K. 2007. Does the CNVT REX SW coin Sawyer, B. & Sawyer, P. 1993. Medieval Scandinavia.

151 New Aspects on Viking-age Urbanism, c. 750-1100

From Conversion to circa 800–1500. Minneapolis & London. Steinsland, G. 2000. Den hellige kongen. Om religion og herskermakt fra vikingtid til middelalder. Oslo. Söderberg, A. 2013. Om stratigrafi, tomtmönster och hantverk i kvarteret Urmakaren, Sigtuna. Situne Dei 2013. 47 – 70. Weibull, C. 1949. Händelser och utvecklingslinjer. Lund. Zachrisson, T. 1989 … tog Knuts gäld i England… Förekomst och frånvaro av vikingatida ädelmetall depåer i Uppland och Södermanland. In: M. Bur ström (red.) Mänsklighet genom millennier. En vän bok till Åke Hyenstrand. Stockholm. 235 – 242. - 1998. Gård, gräns, gravfält. Om sammanhang kring ädel metalldepåer och runstenar från vikingatid och tidigmedeltid i Uppland och Gästrikland. Stockholm.

152 The Schleswig waterfront

The Schleswig waterfront - a place of major significance for the emergence of the town?

Felix Rösch

Fig. 1: The inner Schlei fjord with the settle- ment areas (hatched) of Hedeby at the Hadde- byer Noor (South) and the town of Schleswig on the northern shore of the Schlei (North). The black lines mark the semicircular ram- part of Hedeby and the Danevirke (Map: M. Schimmer, modified by F. Rösch).

On the inner end of the Schlei, two of the most im- Schleswig’s earliest inhabitants. By the fortunate cir- portant trading places in Northern Europe are known cumstance that only a few deep-foundation buildings during Viking and High Middle Ages. On the one were recently erected in the Old Town of Schleswig, a hand, the well-known Viking emporium of Hedeby, ex- significant increase of the groundwater level has taken isting from the late 8th century to the 11th century on place over centuries. Younger layers create an airtight the west bank of the Haddebyer Noor, and on the other, seal, providing excellent preservation conditions for or- the town of Schleswig, which is located two kilometres ganic materials. Thus, the first settlement phases of the to the north on the banks of the Schlei and succeed- majoritarian wooden city could often be documented. ed Hedeby in the course of the 11th century (fig. 1). Among them, for example, a cemetery under the present While research in Hedeby has an over 100-year histo- marketplace (Rathausmarkt) occupied from the ry1, Schleswig, meanwhile, did not come into archaeo- onward, in which the buried still showed hair growth logical focus until the late 1960s. In the 1970s and 80s and funeral garments (Lüdtke 1997) (fig. 2:4). To date, especially, numerous excavations were carried out in the nearly 10 percent of the 12-hectare-large Old Town of course of reorganisation measures in the Old Town of Schleswig has been investigated archaeologically, which Schleswig (Vogel 1983) (fig. 2). Thereby one hit upon is an outstanding amount, higher than the one from layers several metres thick with the material remains of the Hanseatic city of Lübeck, which is, after all, con-

153 New Aspects on Viking-age Urbanism, c. 750-1100 sidered to be one of the best studied medieval cities in no such time gap at the inner Schlei. Quite the con- Europe2. Despite spectacular results so far, only a frac- trary, because of them, we are witness to a number of tion of the Schleswig excavations were processed and significant events. In these transmissions it is import- published3. Recently, however, some excavations have ant to note that people had different names for Hede- come back into focus. As part of a project funded by the by, and later Schleswig, in different languages (Hilberg Volkswagen-Foundation, the transition from Hedeby 2009:80). Among other things, the chronicler Adam of to Schleswig in the 11th century is being investigated4. Bremen reported in his Church , One important aspect that constitutes the evaluation written down between 1072–1075/76, that Sliaswig is are the excavations north of the present Schlei shore also known under the name Heidiba. Therefore it is un- (the so-called Plessenstraße 83/3, Hafenstraße 13 and clear, at least because of the patchy time span due to the Hafengang 11), whose features have always been linked lack of archaeological sources, whether it is still Hedeby with the medieval port and the waterfront of Schleswig or already Schleswig mentioned in the historical writ- (fig. 2:1–3). The intermediate results of these studies ings. In the year 1042, we learn that the King of Den- and what they can tell us about the place’s emergence in mark, Magnus the Good (1042–1047) met Archbishop terms of urbanization will be presented here. Bezelin of Bremen and other dignitaries of church and politics in Sliaswig. The same thing happened ten years

th later, this time under King Svend Estridsen (1047– The Eventful 11 Century 1074), while the Bishop of Schleswig is mentioned as a Both the end of the Viking Age city Hedeby, as well as host. In the year 1063, the Archbishop of Bremen even the foundation of Schleswig, occurred in the 11th centu- plans the General Synod of all Nordic Bishops in Slias- ry. Not only because of the inner end of the Schlei, but wig, which, however, does not come about because of also through processes such as the Christianization of disagreements between him and the King of Denmark. Scandinavia, this period universally applies in the his- In the years 1050 and 1066, Sliaswig is target of enemy torical as an era of upheaval. At the same time attacks, the first by the Norwegian king Harald III. He this century marked the end of Viking Age and Early looted the city and it was then destroyed by Slavonic Middle Ages and the beginning of the High Middle Abodrites, a tribal community settled in the area be- Ages (Müller & Schneider 2010; Mescoli 2008; Boshof tween the present-day cities of Kiel and . From 2008; Brink & Price 2008:609f.). How the transition the years 1085 and 1086, we learn that the Danish King from Hedeby to Schleswig took place exactly is still un- Knut the Holy (1080–1086) stayed in Hethebi for con- known. Considering the dendro dates of timbers used sultations. The place is described as a locus celeberrimus for construction purposes, a gap of almost 50 years and portus, i.e. as a highly frequented place and a port. exists between the two places. The last archaeological- These statements highlight its importance. When polit- ly recorded construction activity in Hedeby, the place- ical issues between Denmark and the ment of a box-shaped well in the year 1020 and a high had to be addressed or greater concerns of the Scan- sea cargo carrier built in 1023 (which later sunk in the dinavian Christianization had to be bargained (already harbour), provide the latest data from this settlement in 948 Hedeby/Schleswig became a bishopric) Hede- (Radtke 2007a:96; Crumlin-Pedersen 1997:99–104). by/Schleswig generally played a central role. During Schleswig’s first dated settlement activities are known these occasions, it was usually the Danish king and the from 1071. It is the erection of a wooden house, Archbishop of Bremen, whose Archdiocese included which was uncovered in the “Schild” excavation (Vogel the bishopric of Schleswig, who were present (Radtke 1991:269) (fig. 2:5). However, there are reasons to be- 2007b:320–321; 2009:151f.; Mescoli 2008:70f.). lieve that Hedeby, until the late 11th or even early 12th century, was still frequented, while the same is true in reverse for Schleswig, where earlier settlement activities Centre of Power, Faith and Trade are very likely. Especially important here is the spread From the outset, Schleswig’s significance was constitut- of certain ceramic types (Hübener 1959:146; Vogel ed by three major aspects. The first is Schleswig as a 1983:23; Lüdtke 1985:134f.; Janssen 1987:141–142; centre of power. For the 11th century, this can be dis- Meyer 1996:196f). cerned from the historically warranted royal visits and However, if one consults the written sources, there is the earliest mention of a royal palace in 1134. Until

154 The Schleswig waterfront

Fig. 2. The old town of Schleswig and the western part of the Holm island with the most important excavations taken out until 2012 and the historical shoreline of the 11th/12th century (hatched = extension of the excavations is not known). Excavations mentioned: 1. Plessenstraße 83/3, 2. Hafenstraße 13, 3. Hafengang 11, 4. Rathausmarkt, 5. Schild, 6. Graukloster (Map is based on © Open- StreetMap contributors; historical shoreline after Vogel 1999, fig. 1).

1218, important meetings of the Danish kingdom latinate buildings known from Paderborn or Goslar, so took place there. From 1110 the city of Schleswig is it can be considered as pretty sure that this construction also the capital of the , the also fulfilled the function of a Palatinate. It filled the later Schleswig (Radtke 2007b:326). The factor king’s need for a befitting representation building and of “power” also finds its equivalent in the excavated ma- served him during his temporary visits as a residence. terial culture. Since its establishment was quite late, one or more pre- After a major refurbishment of the Schleswig decessors are likely (Radtke 1977; 2007b:326–327; Town Hall (the so-called Graukloster “Grey ” ac- Vogel 1983:27–31; 1989:34–43). Schleswig had a rep- cording to the Franciscan order also called gråbröder or resentative building since at least the reign of Sven Es- gråmunkar) in the 1980s, a floor plan from the second tridsen, because without one, the mentioned meetings half of the 12th century was detected. It showed a hall are hard to imagine. Besides that, hints to the location building with annexes, built of field stone and bricks of the ducal residence can also be found. It is the St. (fig. 2:6). In its layout it is strongly reminiscent of Pa- Jürgens , which is also mentioned the first time in

155 New Aspects on Viking-age Urbanism, c. 750-1100

1134, and very likely located on the Möweninsel (Seagull Island) in the Schlei in front of Schleswig, where some remains of ramparts are still visible. A small excavation in 1950 uncovered foundations of boulders and bricks, but wasn’t able to date them or to gain any information about the castles gen- eral layout. During low water levels, pile stumps of a former bridge leading from the Old Town to the island could also be documented, but not yet be dated. Although historians set the construction of the castle in the 11th century, it has not yet been succeeded in verifying this with archaeological methods (Vogel 1983:34–35; 1989:70–71). A tra- ditional picture of the castle can be found on the first Schleswig town seal of 1250, which still graces the city arms (Stoob 1973). The second aspect is Schleswig as a cen- tre of Christian faith. As mentioned, Hedeby was already a bishopric in 948, which moves in the th course of the 11 century to Schleswig. At the lat- Fig. 3. Selection of finds from the Hafengang 11 excavation which illustrate est in 1052/53, when the Bishop of Schleswig, Ratolf, the long-distance trade connections of Schleswig. From the left to right: is named as host of the festivities, it is to be expected (Rhineland), Paffrath ware (Rhineland), whetstone (Norway), porphyry to find appropriate facilities like an Episcopal church (Rhineland), amphora sherd (Byzantium), horseshoe brooch (Northeastern Baltic), Pingsdorf ware (Rhineland), Ardennes ware (Rhineland), axe amulet at the inner Schlei. From Hedeby we know so far ar- (Russia, Baltic States) (Photo: M. Schimmer). chaeologically of neither churches nor representative buildings (Radtke 2007b:320f.). Only an indirect in- dication for a church has yet been found. A bell dating th to the 8/9 century and therefore one of the oldest bells the St. Michael’s Chapel. That faith was a significant known from north of the , which has been found in factor is not least demonstrated by a source from the the harbour (Kalmring 2010:440–441; for the discus- late 12th century, where no less than seven churches are sion of St. Andreas church as successor to a stave church mentioned for Schleswig (Radtke 2007b:320f.; Vogel cf. ibid. note 19). The Schleswig cathedral was not built 1989:52–60). until around 1120/30 and first mentioned in 1134 and Schleswig’s third constituting aspect is its role because of that, there are several theories dealing with as a supraregional economic and trade centre. Until a necessarily existing predecessor. On the one hand, a the 13th century, the city is one of the most important suitable construction is discussed lying under the cathe- trading places in northern Europe, since it has taken dral and on the other hand, the first Episcopal church of over this function from Hedeby in the course of the 11th Schleswig might have also been located on the present century. Due to its favourable topographical location at marketplace (fig. 2:4). Besides the cemetery, the exca- the narrowest point of the Cimbric peninsula, the Dan- vations at the marketplace revealed the foundations of ish-German border and the Hærvejen, it is the central a church built around 1120, which is supposed to be hub in the exchange networks between the north and linked to the patronage of St. . Since the cem- Baltic Sea as well as the Continent and Scandinavia. etery was occupied no later than the 1080s, an earlier, From the Baltic States, furs, wax and honey especially possibly wooden sacral building must have stood here were delivered, while from the west, textiles, gold jew- - the first cathedral of Schleswig? Besides this, a kind of ellery, swords, high quality ceramics and, in particular, bishop’s palace and an associated monastery as seat of silver in the form of coins came to Schleswig. Silver was canon convent is also expected. The latter is assumed for the universal currency equivalent at the time. While in the remains of a rotunda north of the Old Town, which the west payment with coinage prevails, the Baltic trade was unearthed in the 1970s during the construction of is dominated up to the 12th century by the weight of

156 The Schleswig waterfront

Fig. 4. The completely digitalized timbers of the Plessenstraße 83/3 excavation. Highlighted by colours are the plots marked by wattle wood fences on the former shore (green), all main corduroy streets (blue) and the split-plank walls of the shore revetments and dams (red). The arrows with dates are located on the levels of the dam construction phases. The numbers one to three mark the areas with proof of buildings (1), the hypothetical shipyard with the supporting structures (2) and the broad plot with the corresponding dam, discussed as marketplace (3; Map: F. Rösch).

157 New Aspects on Viking-age Urbanism, c. 750-1100 money, in which not the nominal value of the coin, but play, which indicate the structure of the early water- only the pure silver weight is crucial. Thus Schleswig front and therefore a part of the oldest town layout of represents also the central point of conversion between Schleswig. Having a clearer picture of the early water- the two monetary systems. It is assumed that the ma- front thus makes it possible to answer questions con- jority of all coins minted in the west from 1000–1130, cerning the earliest stages of urbanization. and which have later been deposited in the countless hoards around the Baltic, got there by passing through Hedeby/Schleswig. The role of Schleswig as a hub is Excavations and Analysis also reflected in a number of archaeological finds that During the large settlement excavations in the old town were made during the numerous excavations. Thus ce- of Schleswig in the 1970s and 1980s, it was the stat- ramics were found from Slavonic areas of the southern ed goal to document the extensive wooden structures Baltic coast, the Rhineland and even from Byzantium, effectively. Therefore the excavators harked back to a axe-shaped amulets from the Baltics, grindstones, mill- methodology that has been developed by Kurt Schietzel stones and soapstone vessels from Norway, cloth from for the well-known settlement excavations in Hedeby in and England as well as an extensive amount the late 1960s (Schietzel 1969; Schultze 2008:19–23). of additional material (fig. 3). The wide and not nearly Following the method, the excavation area was divided complete spectrum presented here illustrates the nu- into rectangular areas, then dug down in artificial layers merous contacts of the location. By contrast, many im- of about 15 cm, the resulting levels drawn in large-for- portant trade goods didn’t leave any archaeological re- mat plans on a scale of 1:20. In addition to the draw- cord behind. These include the already mentioned furs ings, further information about soil, timbers and height and honey, but also salt, grains, spices, and slaves. values was also registered on the plans. This process Their importance can sometimes be inferred from writ- was repeated until the archaeological cultural layer was ten sources. They tell us that the Schleswig furriers act completely erased. In the end, the port-related excava- a significant part, which argues for an intense fur trade. tions Plessenstraße 83/3 and Hafenstraße 13 resulted in We also learn about the merchant Alwig, the first inhab- no less than 270 large-format plans, which documented itant of Schleswig known by name, who travels between an area of 2,4 hectares in total. Along with 70 profile the years 1112 and 1120 twice to St. Gilles in southern drawings they represent the main source basis for the France to sell sables from Russia (Radtke 2007b:321– evaluation, supplemented only by a series of photos 324; Vogel 2002). Finally worth mentioning are the and the dendrochronological dates of nearly 1000 tim- finds of scales and standardized weights, which prove bers. A few years ago it was still virtually impossible to trading activities (Steuer 1997). analyse such extensive documented excavations in total. From the early 13th century onwards, after But today electronic data processing tools are available about 150 years of thriving commerce, Schleswig in- that make detailed analysis possible. In addition to da- creasingly loses its importance as a trading centre. This tabase programs, the especially geographic information depends to a large degree on the rise of Lübeck and the systems (GIS) should be mentioned, which are able to founding of the , which soon domi- visualize spatial data. In our case, this means that the nates on an axis with Hamburg the trade in the North scanned plans can be loaded into the GIS and set in and Baltic Sea. Due to this process Schleswig’s econom- the correct spatial relationship to one another. Then the ic function breaks down to a place of small scale local timbers documented on the plans can be digitally drawn trade. But in terms of a residence for bishop and duke it and the information written on the plans associated still plays a role as a centre of power and faith for centu- with each individual timber-drawing can be recorded in ries (Jahnke 2006; Radtke 2007b:331–332). a linked table. Finally, the digitalized timbers can then However, during its heyday as an outstanding be displayed in different colours or even 3-dimensional trading place, Schleswig required an appropriate infra- and queries can be run on the plenty attributes listed in structure that sufficed for this highly frequented loca- the table (fig. 4). So coherent structures across multiple tion. In a time when long-distance trade was mainly layers can be identified and precisely analysed5. ship-based, this infrastructure could only be the har- bour. At this point, the extensive excavations on the south side of the Old Town peninsula are brought into

158 The Schleswig waterfront

Fig. 5: Isometric drawing of a supporting structure from the western area of the Plessenstraße 83/3 excavation (W. Karrasch, Archäologisches Landesmuseum Schloss Gottorf in der Stiftung Schleswig-Holsteinische Landesmu- seen).

to be the same. On the other hand, this is contrasted The Waterfront by comparable structures, where the plots usually are In the 11th century, the Old Town peninsula of of more or less the same size. This, for instance, is the Schleswig was of an entirely different nature than today. case in Dorestad (van Es & Verwers 2009:100–103), Aside from a languet in the north, it was surrounded by Ribe (Feveile 2010:144–145, fig. 117), Sigtuna (Tesch water on three sides forming an almost bursiform pen- 1996:116–118; Tesch & Vincent 2005:16–17) and insular. The Holm in the south-east, through landfills in Haithabu (Schietzel 1981:82; Vogel 1986:259). Possibly the 19th century connected with the old town, was still due to the changing wet-dry limnic environment near an island. The northern shore of the Schlei was also not the shore, which was due to the strong and often fluc- on the level of today’s marina, but about 100 m further tuating water levels of the Schlei, the wattle fences run- inland, which was determined via the trench profiles ning parallel to the shore had to be renewed frequently. and a number of drill cores (fig. 2) (Vogel 1999:187). So in some cases up to seven renewal phases have been On this shore in the area of the Plessenstraße, one got preserved in the subsurface. According the function of hold of the first settlement activities at the waterfront. the plots the studies weren’t able to provide any precise Around the middle of the 11th century, at least five par- information yet. Against the background of the large allel plots, rectangular to the Schlei and with bound- number of excavated plots and some written sources it aries marked by wattle fences, were been established. is much likely, however, that the plots were established They were separated by clearances or simple corduroy by a plan and associated with specific rights and also du- streets (fig. 4). So far, none of the plots could be dated ties. In medieval Trondheim, for instance, it is known precisely. But since they mainly showed up in layers that from the written sources that the local residents were are below those with the timber used from the 1070s individually responsible for the maintenance of the cor- onwards, it is expected that they have been established duroy street section in front of their plots. This was con- previously. The longitudinal expansion to the north firmed by excavation, where the exposed sections varied could not be documented. Their widths vary between greatly in quality (Christophersen 1999:165–166). Di- seven and 20 meters, so they do not seem to have been rections in terms of building activities and neighbour- of a standardized size. On the one hand, in places like hood are also known from the so called Sachsenspiegel, Lübeck (Legant 2010:122–129) or even Viking Age which was written down between 1220 and 1235 (cf. Trondheim (Christophersen 1999:161–162), it seems Fansa 1995). From Dorestad we learn, according to the Carolingian regesta imperii (Mühlbacher 1908:No. 578), that in 815 there were certain plots belonging to 159 New Aspects on Viking-age Urbanism, c. 750-1100 the church and used by merchants. On these plots for- “building boom”. Within a few years, in 1091, 1094/95 eign merchants were allowed to enter without paying and 1098, more plots follow this example while existing tax to the king (Rohwer 1937:51–52). Generally, plots dams are extended step by step further into the Schlei are known in seasonal use as part of annual markets as (fig. 4). The gaps between the dams remained free or well as those in urban centres with constant develop- have, at least in one case, been covered with a north- ment. They served different functions. There is proof of south running corduroy street. This development didn’t resident craftsmen and merchants, who stored and sold only take place in the Plessenstraße, but could also be goods, as well as for habitation (Feveile 2010:143f.; van followed in the Hafenstraße 13 (Vogel 1999:189f., fig. Es & Verwers 2009:289). Some well-known examples 2) and Hafengang 11 (von Carnap-Bornheim & Lüth of such plots can be found at the places already men- 2008; Hilberg et al. 2012) as well as in numerous small- tioned above. scale rescue excavations6, located in between these sites. Around 1079/80, the water-facing sides of two In practical terms, this means that around the year plots were reinforced with of planks. In the 1100, on the southern shore of the Old Town penin- years 1081 and 1084, two more plots follow, so that sula on a width of at least 200 m, dams up to 50 m ultimately a continuous plank wall at the level of the in length existed. Therefore, within only a decade, large former shore gradient exists in the north-eastern part of amounts of material and labour were invested. Both the the Plessenstraße. Simultaneously, or even a little earlier, dimensionality and the short construction period are a corduroy street was installed at the front end of these not yet known from any High Medieval or Viking Age four plots (fig. 4). This boardwalk was through pairs waterfront excavation in Northern Europe. This raises of piles connected with yoke beams and specifically the question of the purpose of these constructions and worked joists substantially more complex and massive the intention that lead to their establishment. The first than their counterparts between the plots. Since this idea that came to mind when imaging these huge dams surface laid clearly above the ground, it was most likely is that they served as jetties for ships and this has been established to allow a smooth traffic flow regardless of the most common interpretation up to now (cf. Vogel the varying water levels. It is within this context that the 1989; 1997; 1999). However, this question could not plank walls are to be understood, which probably served be answered without having taken a closer look at the as flood barriers. At the level of the 20-metre-wide plot historical water level. in the western part of the Plessenstraße site, this de- The 11th century Schlei has so far been deter- velopment, however, couldn’t be recorded. During the mined only roughly (Voss 1967; Labes 2005; Kalmring years around 1080, neither a shore revetment was built, 2010:289–307). However, more accurate data is known nor the corduroy street continued. On the contrary, it for the late 9th century. During his research on Hedeby ends just east of the plot. This relates possibly to a cer- harbour, S. Kalmring (2010:303) was able to figure out tain circumstance, which will be discussed in more de- that due to datable timbers in a drift line layer, the water tail below. level was most likely about 80 cm below the present one. For 1087, the next major construction activi- In Schleswig, however, the north-south profiles, which ties have been recorded. In front of the broad plot and run through the shoreline, are able to provide a series of its eastern neighbour, earthen dam-like structures were first clues. They show layers of limnic sands which end erected, which extend into the shallow water of the Schlei at an average height of 30 cm below the present water fjord. They were constructed with split planks, driven level. Because they are covered by the artificial layers of into the ground like a bulkhead in the shape of a big U. brushwood of the first dam construction phases, they Inside they are filled with layers of brushwood several can be considered as an indicator of the late 11th century decimetres thick as well as dung and soil. Remarkably, water level. This would mean that the average water level these massive structures follow exactly the widths of the in the inner Schlei rose within 200 years by as much as plots on the shore, which suggests that they are directly 50 cm. That this increase is not a regional phenomenon, related – a pattern that also could become observed in but also occurred on the southern coast of the Baltic Sea, Hedeby (Kalmring 2010:255f.), Dorestad and Trond- the North Sea and the North German inland, is proven heim (Christophersen 1999:163f.) - but it did not stop by a series of corresponding research (Bleile 2008:17– at these two dams. The year 1087 seems to have been 20, 114–117; Dörfler et al. 2009). But what does this the start-up for what can best be described with the term mean for the interpretation of the dams? When putting

160 The Schleswig waterfront the limnic sand layers, the determined water level and skeletons from the cemetery below the Rathausmarkt the southern sides of the dams in relation to each other, (fig. 2:4). In the early stages of its occupancy, new-borns as a result, depending on the stage of expansion, differ- were breastfed considerably less than two and not, as ent images appear. The water in front of the first dams generally common, two to three years. This result sug- built in 1087 was only about 30 cm deep, while those gests a higher birth rate, because long lactations delay the from 1094/95 reached into depths between 70 cm and ovulation and thus a new pregnancy (Grupe et al. 2013). 100 cm depending on their location. Greater depths are As further support for the hypothesis of land reclamation to be expected only for subsequent dam sections. This the numerous relics of houses can be used. They have means that the dams were not primarily built as jetties, been found on the plots as well as on the dams in the enabling larger vessels to berth while still floating. By eastern part of the Plessenstraße (fig. 4:1). Together with taking a look at the predominant vessel types in the 11th the remains of fireplaces, barrel wells and possibly even and 12th century, we learn that even medium-sized car- for small animals, it is likely that they served riers with a payload of 10 - 25 t, as known from sever- as dwellings. Some of the houses were built shortly after al Scandinavian sites and also from the Möweninsel in the dams, sometimes even directly to their head, as an front of Schleswig, had a drought of about one meter example from the Hafengang 11 demonstrates (Schul- (Crumlin-Pedersen 1999; Englert 2000:135–136, tab. tze 2012). This situation can’t be considered as a hint 45; Belasus 2004; 2009). This means that aside from for loading and unloading goods at all. In this context local transport vessels with little drought (cf. Brandt & the composition of the dams must be discussed again. Kühn 2004), a function as jetty was first realistic when Their construction, thick layers of brushwood inside and the dams reached into deeper water. In the Hafenstraße intervening spaces, allows optimal water drainage after 13 area, where the shoreline formed a small bay-like flood events. In addition, houses built on them are less situation, it was probably the case already in 1095 (cf. susceptible to the enormous forces of ice, as it would be fig. 2). As a result these dams were way longer and built the case with piled platforms. This would result in a rap- into deeper water from the beginning. There, the larg- id destruction of the building (cf. Kalmring 2010:307). est known cargo ships, with a draught of about 1,5 m However, this approach can’t be applied to all areas of the and a transport capacity up to 60 t, could berth float- Plessenstraße covered with dams. In the extreme western ing. The oldest ship of this type is known as Hedeby part of the site at the front end of a dam three staggered Wreck III and has already been mentioned above. It was support structures that were possibly part of a shipyard built in southern Denmark and later it sank in front of have been found (fig. 4:2). They could have served to the harbour in the Haddebyer Noor (Crumlin-Peders- hold the stem of a ship in place for the duration of con- en 1997:99–104). Another ship of such type, but from struction (fig. 13). However, this interpretation must the 12th century, is the so-called Karschauwrack found remain hypothetical, because the structures besides one in the Schlei in front of Germany’s smallest town Arnis example from Hedeby totally lack of parallels. That ships (Kühn 2009:81–84). But when the dams haven’t been were built in Schleswig is beyond question. Apart from erected as jetties in the first case there must have been several findings connected with seafaring, shipbuilding another intention. One point might be the small size of (Siegloff 2006; 2012) and the fact that early Schleswig Schleswig’s Old Town peninsula compared to the inner was a major port, we also know of matching written surface of the semi-circular rampart of Hedeby, which sources. The Miracle Book of Saint re- is with 25.5 hectares about as twice as large. During the ports that in the late 12th century, a wax trader from movement from one place to the other the peninsula Schleswig built, with the financial support of the Danish could have reached its capacity limits quickly, especially king, a navis magna, a very large ship. The book also tells since with graveyards and the royal us that the ship was so big that it wasn’t able to move palace with an additional open area for the entourage into the water even via the use of slide pads. Only under was also located there. Thus, it would be a plausible hy- the invocation of St. Thomas and a generous commit- pothesis that the dams were built for the purpose of land ment on the part of the merchant was the ship finally reclamation in the southern shallow waters. They could launched. (Englert 2000:1–2). be understood as a measure against the background of Finally, the special role of the wide plot and its a high population pressure in the 1080s. That there has corresponding dam in the western section of the excava- been a population increase is confirmed by studies of tion has to be mentioned. They differ not only in their

161 New Aspects on Viking-age Urbanism, c. 750-1100 dimension from everyone else, but also attract attention that they are to be understood as multi-functional de- by their different type of development. It turned out to vices. On the one hand, they served for settlement pur- be not only significantly lesser, but also, in contrast to poses, land reclamation and as flood protection, which the eastern part of the Plessenstraße, no clear remains of is also confirmed by the present course of the southern buildings and corduroy streets were determined. On the bank that has shifted 100 m further into the Schlei de- contrary the first dam built in 1087 was subdivided in spite the increase of the water level. On the other hand, seven to eight smaller rectangles of about 17 m² marked they could, at least in the later stages of construction by vertical timbers and wattle wood fences. So far the from the mid- onwards, act as jetties, which, de- purpose of these features is unclear, but finds of balances pending on their location, easily enabled even larger and weights provide a possible explanation. A big part of ships to berth while floating. What is beyond question all known weights ever found in Schleswig were located is that the features shown represent the medieval har- in the Plessenstraße and especially in the area of the wide bour of Schleswig. Even though in the past several plac- plot and the corresponding dam (Steuer 1997:91f, fig. es have been discussed as location for the harbour7, there 59). The weights are standardized copies, so-called poly- is no place on the Old Town peninsula that represents hedral and spherical weights which were widespread in the town’s importance in northern European long-dis- the entire Baltic up to the 13th century. They were closely tance trade in such a way as the area explored during connected with a trade network which used the weight the three excavations presented above. The structured of silver as currency for centuries (Steuer 1997:11–18). development already confirms the outstanding rank on Altogether the big size, the undeveloped open space, the its own. In addition there are many maritime findings rectangular subdivisions and the concentration of the like reused ship timbers and fish traps, which had been spherical zone weights give reason to suspect, that this thrown into one of the dams instead of brushwood. The part of the waterfront might have been a marketplace of the site into a marketplace as well as a - (fig. 4:3). But this hypothesis must be further strength- tential shipyard area should also be mentioned, because ened by more evidence, like the distribution of finds they are essential infrastructural aspects of trans-ship- that are connected with trade activities. Recent research ment centres. Even though the analysis is still in prog- on Hedeby harbour has shown that harbour facilities, ress and much was previously formulated as a hypothe- in this case jetties, can also serve as marketplaces (Kal- sis, one thing is certain: Right here, on the south bank mring 2010:443–450). In neither Hedeby nor in early of the Old Town peninsula, we find the portus and locus Schleswig such space is known from the inner settlement celeberrimus mentioned in the written sources. yet. This can be due to the fact that the marketplace To refer to the question of early urbanization hasn’t yet been found or, what’s way more likely, that it processes, a look on the first settlement phases of the is located elsewhere: at the waterfront, where the goods Plessenstraße should again be taken. Here it seems that arrived. If the presented hypothesis proves to be right, by the middle of the 11th century, a structured devel- it means that the waterfront marketplace in Schleswig opment was deliberately promoted. Behind this action is a model that has been adopted from Hedeby and there was most likely the intention to create an area that doesn’t follow the continental European trend of cen- was explicitly tailored to the needs of the economy: pro- tral marketplaces, which have been common since the duction8 and trade. This becomes even clearer during second half of the 10th century (Kalmring 2010:448). In the building boom in the 1080s and 90s where much Schleswig this is first the case in the 13th century, when of the revetments, dams, houses and corduroy streets it loses its important role as a transit place due to several have been built. On a field of at least 200 m in length structural changes. In this course the waterfront chang- a densely built-up area for different purposes arises. It es and today’s marketplace in the middle of the Old is obvious that, due to the major efforts that have been Town is set up (Harck 1995:151; Lüdtke 1997; Radtke made, the expansion the waterfront was of major inter- 2007b:331–332). est. So far there is no certain evidence as to who was responsible for this action. However, a royal ini- Conclusion tiative was likely. In the light of the parallel or short- As pointed out, the dams at the Schleswig Waterfront ly subsequent developments in Schleswig such as the cannot be reduced to a single function. Rather it seems establishment of the religious and power-political in-

162 The Schleswig waterfront frastructure, it seems unlikely that such a development took place without the influence of the king. Aspects like planned infrastructure in the form of plots, spe- cialized craftsmanship, power-political and religious central places are generally regarded as important fea- tures of early urbanization (Steuer 2005:450). But this is also true for so-called proto-urban settlements, like Hedeby for instance, which can be found at several lo- cations at the southwestern Baltic coast (Kleingärtner & Tummuscheit 2007). The case of Schleswig, however, presents a complex economic infrastructure which goes hand in hand with the establishment of religious and power-political structures that clearly stand out from the ones of the old centres. Thus already in the archaeo- logical features of the 11th century, the development to- wards a new type of medieval town becomes apparent, whose special legal status is later written down in the municipal law, which originates in the mid-12th century (Radtke 2007b:327).

Corresponding Author Felix Rösch Institut für Ur- und Frühgeschichte, Kiel university Kiel, Germany [email protected]

163 New Aspects on Viking-age Urbanism, c. 750-1100

References der Ostsee. Eine methodische Diskussion am Bei spiel der Ostseeförde Schlei, Schleswig-Holstein. In: U. Müller, S. Kleingärtner & F. Huber (eds.) Zwis Belasus, M. 2004. Das Möweninsel-Schiff. Ein mittelter chen Nord- und Ostsee 1997–2007. 10 Jahre Arbeits licher Schiffsfund aus der inneren Schlei. Unpublished gruppe für maritime und limnische Archäologie Diploma thesis. Kiel University. (AMLA) in Schleswig-Holstein. Universitätsforschun – 2009. Das hochmittelalterliche Schiff von der Mö gen zur prähistorischen Archäologie 165. Bonn. 177 weninsel bei Schleswig. In: U. Müller, S. Klein – 186. gärtner & F. Huber (eds.) Zwischen Nord- und Ost Englert, A. 2000. Large cargo vessels in Danish Waters, see 1997–2007. 10 Jahre Arbeitsgruppe für maritime AD 1000–1250. Unpublished Doctoral thesis. Kiel und limnische Archäologie (AMLA) in Schleswig-Hol University. stein. Universitätsforschungen zur prähistorischen van Es, W.A. & Verwers, W.J.H. 2009. Excavations at Archäologie 165. Bonn. 89 – 98. Dorestad 3, Hoogstraat 0, II-IV. Nederlandse Oudhe Bleile, R. 2008. Quetzin – eine spätslawische Burg auf den 16. Amersfoort. der Kohlinsel im Plauer See. Befunde Fansa, M. (ed.) 1995. der sassen speyghel. Sachsenspiegel und Funde zur Problematik slawischer Inselnutzungen – Recht – Alltag. Band I: Aus dem Leben gegriffen in -Vorpommern. Beiträge zur Ur- und – Ein Rechtsbuch spiegelt Zeit. Archäologische Frühgeschichte Mecklenburg-Vorpommerns 48. Mitteilungen aus Nordwestdeutschland, Beiheft 10. . Boshof. E. 2008. Die Salier. Stuttgart. Oldenburg. Brandt, K. & Kühn, H.-J. 2004. Der Prahm aus dem Feveile, C. 2010. Ribe. Emporium and town in the 8th Hafen von Haithabu. Beiträge zu antiken und mittel and 9th centuries. In: A. Willemsen & H. Kik (eds.) alterlichen Flachbodenschiffen. Kolloquium Schleswig Dorestad in an international framework. New research 2002. Schriften des Archäologischen Landesmuse on centres of trade and coinage in Carolingian times. ums, Ergänzungsreihe 2. Neumünster. Brink, S. & Congress Leiden 2009. Turnhout. 142 – 148. Price, N. (eds.) 2008. The Viking World. London & Grupe, G., von Carnap-Bornheim, C. & Becker, C. New York. 2013. Rise and Fall of a Medieval Trade Centre: Eco von Carnap-Bornheim, C. & Lüth, P. 2008. Neues aus nomic Change from Viking Haithabu to Medieval der Schleswiger Altstadt. Archäologische Nachrichten Schleswig revealed by Stable Isotope Analysis. Euro aus Schleswig-Holstein 2008 (15). 72 – 74. pean Journal of Archaeology 2013 (16):1. 137 – 166. Christophersen, A. 1999. The waterfront and beyond. Harck, O. 1995. Grundtypen der neuen dänischen und Commercial activity and the making of townscapes. schleswigischen Städte. In: E. Hoffmann & F. Lubo In: J. Bill & B.L. Clausen (eds.) Maritime Topogra witz (eds.) Die Stadt im westlichen Ostseeraum. Vorträ phy and the Medieval Town. International Confer ge zur Stadtgründung und Stadterweiterung im Hohen ence on Waterfront Archaeology 5, Copenhagen 14 Mittelalter. Teil 1. Tagung Kiel 1990. Kieler Werk 16.05.1998. Publications from The National Muse stücke. Reihe A. Beiträge zur schleswig-holsteini um. Studies in Archaeology & History 4. Copenha schen und skandinavischen Geschichte 14. Frank gen. 161 – 168. furt/Main, Berlin, Bern, New York, Paris & Wien. Crumlin-Pedersen, O. 1997. Viking-age ships and ship 145 – 154. building in Hedeby/Haithabu and Schleswig. Ships & Hilberg, V. 2007. Haithabu im 11. Jahrhundert. Auf boats of the North 2. Roskilde & Schleswig. der Suche nach dem Niedergang eines dänischen – 1999. Ships as indicators of trade in Northern Eu Emporiums der Wikingerzeit. In: C. Dobiat, M. rope 600-1200. In: J. Bill & B.L. Clausen (eds.), Posselt, & B. Zickgraf (eds.) Geophysik und Ausgra Maritime Topography and the Medieval Town. Inter bung. Einsatz und Auswertung zerstörungsfreier Pros national Conference on Waterfront Archaeology 5, pektion in der Archäologie. Internationale Archäolo Copenhagen 14-16.05.1998. Publications from The gie, Naturwissenschaft und Technologie 6. Rahden. National Museum. Studies in Archaeology & Histo 187 – 203. ry 4. Copenhagen. 11 – 20. – 2009. Hedeby in Wulfstan`s days: a Danish empo Dörfler, W., Jacobsen, O. & Klooß, S. 2009. Indika rium of the Viking Age between East and West. In: toren des nacheiszeitlichen Meeresspiegelanstiegs A. Englert & A. Trakadas (eds.) Wulfstan`s Voyage.

164 The Schleswig waterfront

The Baltic Sea region in the early Viking Age as seen dem Rathausmarkt von Schleswig. Ausgrabungen in from shipboard. Maritime Culture of the North 2. Schleswig. Berichte und Studien 12. Neumünster. 9 Roskilde. 79 – 113. – 84. Hilberg, V., Rösch, F. & Schimmer, M. 2012. Zwischen Mescoli, F. 2008. Von geplünderten Heiligtümern und Wikingern und Hanse. Der Übergang von Haitha geheiligten Plünderern. Die Christianisierung Skandi bu nach Schleswig im 11. Jahrhundert. Archäologi naviens im Mittelalter 814–1104. . sche Nachrichten aus Schleswig-Holstein 2012 (19). 64 Meyer, J. 1996. Die Keramik des 11. und 12. Jahrhun – 71. derts aus der Hafensiedlung Schleswig. Kiel University. Hübener, W. 1959. Die Keramik von Haithabu. Die Mühlbacher, E. 1908. Regesta imperii nach Johann Fried Ausgrabungen in Haithabu 2. Neumünster. rich Böhmer. Teil 1. Die Regesten des Kaiserreichs unter Jahnke, C. 2006. „… und er verwandelte die blühende den Karolingern 751–918. Innsbruck. Handelsstadt in ein unbedeutendes Dorf“. Die Rolle Müller, S. & Schneider, J. 2010. Deutsche Texte der Sa Schleswigs im internationalen Handel des 13. Jahr lierzeit. Neuanfänge und Kontinuitäten im 11. Jahr hunderts. In: G. Fouquet, M. Hansen, C. Jahnke & hundert . Tagung zur Canossa-Ausstellung Paderborn J. Schlürmann (eds.) Von Menschen, Ländern, Mee 2006. Mittelalter Studien 20. München. ren. Festschrift für Thomas Riis zum 65. Geburtstag. Radtke, Chr. 1977. Aula und . Überlegun Tönning. 251 – 268. gen zur Topographie und Struktur des Königshofes Janssen, W. 1987. Die Importkeramik von Haithabu. in Schleswig. Beiträge zur Schleswiger Stadtgeschichte Die Ausgrabungen in Haithabu 9. Neumünster. 1977 (22). 29 –47. Kalmring, S. 2010. Der Hafen von Haithabu. Die – 2007a. Schleswig und seine Geschichte. In: C. von Ausgrabungen in Haithabu 14. Neumünster. Carnap-Bornheim & M. Segschneider (eds.) Die Kleingärtner, S. & Tummuscheit, A. 2007. Zwischen Schleiregion. Land – Wasser – Geschichte. Führer zu ar Haithabu und Wolin - Die frühe Phase der Urbani chäologischen Denkmälern in Deutschland 49. sierung an der südwestlichen Ostseeküste. Quaestio Stuttgart. nes Medii Aevi Novae 2007 (12). 215 – 252. – 2007b. Schleswig ca. 1000–1250. Systemtheoretische Kühn, H. J. 2009. Frachtsegler und Fähre. Zwei bedeu Skizze eines Urbanisierungsprofils. In: S. Burmeister, tende Wracks des 12. Jahrhunderts aus der Schlei. In: H. Derks & J. v. Richthofen (eds.) Zweiundvierzig. U. Müller, S. Kleingärtner & F. Huber (eds.) Zwi Festschrift M. Gebühr. Internationale Archäologie. schen Nord- und Ostsee 1997–2007. 10 Jahre Arbeits Studia honoraria 25. Rahden, Westfalen. gruppe für maritime und limnische Archäologie – 2009. Haithabu. Perspektiven einer Stadtentwicklung (AMLA) in Schleswig-Holstein. Universitätsforschun in drei Stationen - 800, 900, 1000. Zeitschrift der Ar gen zur prähistorischen Archäologie 165. Bonn. 81 – chäologie des Mittelalters 2009 (37). 135 – 162. 88. Rohwer, B. 1937. Der friesische Handel im frühen Mit Labes, S. 2005. Der nacheiszeitliche Meeresspiegelan telalter. Kiel. stieg der Ostsee im Bereich der Schlei. Offa 2002/03 Schietzel, K. 1969. Die archäologischen Befunde der Aus (59/60). Festschrift für J. Reichstein. 223 – 227. grabung Haithabu 1963–64. Berichte über die Aus Legant, G. 2010. Zur Siedlungsgeschichte des ehemaligen grabungen in Haithabu 1. Neumünster. Lübecker Kaufleuteviertels im 12. und frühen 13. Jahr – 1981. Stand der siedlungsarchäologischen Forschung in hundert: nach den ältesten Befunden der Grabung Alf Haithabu - Ergebnisse und Probleme. Berichte über straße - Fischstraße - Schüsselbuden, 1985–1990. Lü die Ausgrabungen in Haithabu 16. Neumünster. becker Schriften zur Archäologie und Kulturge Schultze, J. 2008. Haithabu - Die Siedlungsgrabungen. I. schichte 27. Rahden, Westfalen. Methoden und Möglichkeiten der Auswertung. Die Lüdtke, H. 1985. Die mittelalterliche Keramik von Ausgrabungen in Haithabu 13. Neumünster. Schleswig. Ausgrabung Schild 1971–1975. Ausgra – 2012. Interessante Baubefunde aus dem Schleswiger bungen in Schleswig. Berichte und Studien 4. Neu Hafen. In: P. Gammeltoft & V. Hilberg (eds.), Beret münster. ning fra det niogtyvende tværfaglige vikingesymposium. – 1997. Die archäologischen Untersuchungen unter Højbjerg. 69 – 84. dem Schleswiger Rathausmarkt. In: V. Vogel (ed.), Siegloff, E. 2006. Die Bedeutung der Schleswiger Land Kirche und Gräberfeld des 11.-13. Jahrhunderts unter enge für den vorhansischen Schiffbau. Beiträge zur

165 New Aspects on Viking-age Urbanism, c. 750-1100

Schleswiger Stadtgeschichte 2006 (51). 7 – 22. 1200. Ergebnisse der Mittelalter-Archäologie. Kollo – 2012. Maritimes Kulturgut der Wikingerzeit und quium Köln 31. Januar bis 2. Februar 1984. Zeit des Mittelalters aus Hollingstedt. In: K. Brandt (ed.) schrift der Archäologie des Mittelalters Beiheft 4. Hollingstedt an der Treene. Ein Flusshafen der Wikin Köln. 257 – 262. gerzeit und des Mittelalters für den Transitverkehr zwi – 1989. Schleswig im Mittelalter. Archäologie einer Stadt. schen Nord- und Ostsee. Schriften des Archäologi Neumünster. schen Landesmuseums 10. Neumünster. 115 – 248. – 1991. Profaner Holzbau des 11. bis frühen 13. Jahr Stark, J. 1988. Haithabu-Schleswig-Danewerk. Aspekte hunderts in Schleswig. In: H.W. Böhme (ed.) Sied einer Forschungsgeschichte mittelalterlicher Anlagen in lungen und zur Salierzeit 1. In den Schleswig-Holstein. British Archaeology Reports. In nördlichen Landschaften des . Monographien ternational Series 432. Oxford. des Römisch Germanischen Zentralmuseums 27. Steuer, H. 1997. Wagen und Gewichte aus dem mittelal Sigmaringen. 263 – 276. terlichen Schleswig. Funde des 11. bis 13. Jahrhunderts – 1992. Der Schleswiger Hafen im Mittelalter. Verein aus Europa als Quellen zur Handels- und Währungs Schleswiger Briefmarkensammler e.V. (ed.) 900 Jah geschichte. Zeitschrift der Archäologie des Mittelal re Hafen Schleswig 1092–1992. Schleswig. 23 – 36. ters Beiheft 10. Köln. – 1997. Schleswig - Stadt und Hafen im 11./12. Jahr – 2005. Headword „Stadt“ - § B. Kulturgeschichtlich. hundert. In: G. De Boe & F. Verhaeghe (eds.), Travel In: H. Beck, D. Geuenich & H. Steuer (eds.) Reall Technology & Organisation in Medieval Europe. Pa exikon der germanischen Altertumskunde 24. Berlin & pers of the Medieval Europe Brugge 1997 Confer New York. 451 – 453. ence Vol. 8. I.A.P. Rapporten 8. Zellik 1997. Stoob, H. 1973. Deutscher Städteatlas I:9. . – 1999. Der Schleswiger Hafen im hohen und späten Tesch, S. 1996. Sigtuna - Royal Stronghold and Early Mittelalter. In: J. Bill & B. L. Clausen (eds.) Mari Town. In: L. Nilsson & S. Lilja (eds.) The emer time Topography and the Medieval Town. Internation gence of towns. Archaeology and early Urbanization in al Conference on Waterfront Archaeology 5, Copen Non-Roman, North-West Europe. The Swedish Insti hagen 14–16.05.1998. Publications from The Na tute of Urban History 75 Years. An Anniversary Sym tional Museum. Studies in archaeology & history 4. posium, September 3rd 1994. Stockholm. 113 – 126 Copenhagen. 187 – 196. Tesch, S. & Vincent, J. 2005. Vyer från medeltidens – 2002. Archäologische Belege für Fernkontakte der Sigtuna. Stockholm. Stadt Schleswig im 11. –13. Jahrhundert. In: K. Ulbricht, I. 1984. Die Verarbeitung von Knochen, Ge Brandt, M. Müller-Wille & Chr. Radtke (eds.) Ha weih und Horn im mittelalterlichen Schleswig. Ausgra ithabu und die frühe Stadtentwicklung im nördlichen bungen in Schleswig. Berichte und Studien 3. Neu Europa. Internationale Fachkonferenz „Haithabu münster. und die frühe Stadtentwicklung im nördlichen Eu Unverhau, D. 2002. „Möchte unser Hedeby vor ropa“ Schleswig 25. – 27.09.1998. Neumünster. 367 unseren Blicken entstehen“ (). In: – 378. K. Brandt, M. Müller-Wille & Chr. Radtke (eds.) Voss, F. 1967. Die morphologische Entwicklung der Haithabu und die frühe Stadtentwicklung im nördli Schleimündung. Hamburger Geographische Studien chen Europa. Internationale Fachkonferenz „Hait 20. Hamburg. habu und die frühe Stadtentwicklung im nördlichen van de Walle-van der Woude, T. & Groenman-van Europa“ Schleswig 25–27.09.1998. Neumünster. 11 Waatering, W. 2001. Die Lederfunde der Ausgrabung – 42. Plessenstraße in Schleswig. Ausgrabungen in Schleswig. Vogel, V. 1977. Die Anfänge des Schleswiger Hafens. Berichte und Studien 15. Neumünster. Beiträge zur Schleswiger Stadtgeschichte 1977 (22). 21 – 28. – 1983. Archäologische Stadtkernforschung in Schleswig (Endnotes) 1969-1982. Ausgrabungen in Schleswig. Berichte 1 For the history of research of the settlement excava- und Studien 1. Neumünster. 9 – 54. tions cf. Stark 1988:30f; Hilberg 2009:84f; for partic- – 1986. Zum Parzellengefüge in der Stadt um 1200. ular aspects like dwellings or the harbour cf. Schultze In: H. Steuer (ed.) Zur Lebensweise in der Stadt um 2008:123–132 resp. Kalmring 2010:15–19; for recent

166 The Schleswig waterfront research; Hilberg 2007; 2009; Hilberg et al. 2012; Hil- berg & Schultze (in this publication); for the oldest re- search cf. Unverhau 2002.

2 Cf. therefore especially the series „Lübecker Schriften zur Archäologie und Kulturgeschichte“ (1–27, 1978– 2010) as well as the „Lübecker Kolloquien zur Stadtar- chäologie im Hanseraum“ (1–7, 1997–2012) edited by M. Gläser.

3 Cf. therefore especially the series edited by V. Vogel: Ausgrabungen in Schleswig. Berichte und Studien (1– 17, 1983–2006).

4 The full project title is named: “Zwischen Wikingern und Hanse. Kontinuität und Wandel des zentralen Umschlagplatzes Haithabu/Schleswig im 11. Jahrhun- dert” (Between Vikings and Hanseatic league. Conti- nuity and change of the trans-shipment center Hedeby/ Schleswig in the 11th century). It has been set up by Ralf Bleile from the Archaeological State Museum of Schleswig-Holstein and Ulrich Müller from the Insti- tute of Pre- and Early history, Kiel University.

5 In terms of methodology cf. especially Schultze 2008, but also Kalmring 2010:194–207; 494–496.

6 These small-scale rescue excavations are generally poorly documented and haven`t been published or even mentioned in any articles yet. The only exception is an excavation consisting of two trenches in front of the Hafenstraße 13 and 16 (Vogel 1992).

7 Cf. Vogel 1997:194. The frequently estimated harbour in the Holmer Noor in front of the king’s palace/grey friary most likely goes back to a historicized map show- ing Schleswig in 1154 that has been drawn in 1641 by J. Mejer (cf. also Vogel 1983:com. 6). Due to swampy soil conditions archaeological or geological investiga- tions never took place in this area, so every consider- ation about the existence and the role of such a harbour is highly hypothetical.

8Even if the aspect of production hasn’t been highlighted in this paper yet, it has been proven at the waterfront. S. Ulbricht 1984; van de Walle-van der Woude & Groen- man-van Waatering 2001; Hilberg et al. 2012:68–71.

167 New Aspects on Viking-age Urbanism, c. 750-1100

Starup Østertoft – a landing place and settlement with a church from the 11th–12th Centuries

Anders Hartvig

Haderslev inlet cuts deep into the fertile moraine of right neat the water. The church is the parish church for south-east Jutland. The town of Haderslev is locat- the four medieval villages of Brorsbøl, Vandling, Lund- ed at the bottom of the inlet. Right down at the inlet, ing, and Lønt, all of which are located further inland. The and only a few kilometres east of Haderslev, is Starup present settlement of Starup, which is located some hun- Church - one of the most remarkable church buildings dred metres south of the church, only came into existence of southern Jutland. after 1800. Today, a rune stone with the inscription Eriks Kuml The present appearance of the church is the result of a stands in front of the church. The rune stone was found comprehensive restoration, which was carried out during during the restoration of the church (Danmarks Kirker the period 1908-1918. The building of the church was 1954:419). But who was this Erik and what was his re- initiated towards the end of the 11th century as a three- lation to the church? Since 1985, five archaeological ex- aisled Romanesque limestone with apses on the cavations covering a total of 4500 m2 have been carried eastern side of the choir and the two nave aisles, as well out around the church. as a strongly built western tower with its own entrance and an aristocratic gallery (Bertelsen 2015). The choir with apse was completed according to the original plan, Topography but the outer walls of the aisles were probably never con- Starup church is located on a sandy peninsula at the structed to their full height. The reason for this was that southern bank of Haderslev inlet. Towards the north the construction work was closed down shortly after it and the east, the peninsula is delimitated by the inlet; had begun, probably sometime in the beginning of the towards the west, by a marked valley. Originally, this 12th century. The walls of the nave aisles were torn down, valley was a shallow part of the inlet, but today, only a the arcades between the pillars in the nave were walled up small stream runs in the bottom of the valley. Towards in which way the church was transformed to an aisleless the south, the peninsula rises gently some 600 m to- church. At the same time, the western tower was given wards the modern east/west running main road. South up and the church was completed with a simple western of the road, the topography changes into the hilly mo- wall. During the Late Middle Ages (1400–1536) a second raine that dominates the area around Haderslev. phase of construction work took place at Starup Church. This time, a vestry, a porch and a western tower were add- ed to the church. The church now appeared like a com- Starup Church mon Danish aisleless village church with Late Medieval In many ways, Starup Church differs from the other extensions. This appearance was kept until the restoration churches in the area. It is located on an idolated spot in the beginning of the 20th century (Danmarks Kirker

168 Starup Østertoft

Fig. 1. Map showing the location of Starup Østertoft. Graphics: J. Andersen.

1954:397ff). west of the church, a north-south oriented ditch was The excavations excavated. It was approximately 2.5 m wide and 1.5 m The present head of the Museum Sønderjylland – deep. It was possible to identify the ditch through the Arkæologi Haderslev, Lennart S. Madsen, carried entire area of excavation and it is supposed to represent a out the first two excavation campaigns by the church division between the church and the secular settlement. in 1985 and 1987. The excavation areas were located The excavation south-west of the church revealed evi- immediately next to the church and covered a total of dence of four buildings: two square buildings and the 1500 m2. These excavations revealed evidence of dif- eastern ends of two aisleless houses of which one had a ferent craft acticvities such as iron forging and bronze fireplace at the southern wall (Madsen 1987b:210ff; cf. casting, but also fireplaces, wells and pits, which can be Madsen 1985; 1987a). related to the construction of the church. A few metres The construction of a new parish community cen-

169 New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 2. Reconstruction of the church. Top: the situation at the time when the construction work was stopped. Bottom: the reduced church. Graphics: J. Andersen & A. Hartvig. tre in 2001 immediately west of the area excavated in This area used to be a forest. Consequently, the level of 1987 offered another opportunity to excavate near the preservation of postholes and culture layers was not as church. In total, c. 650 m2 were excavated by Tenna R. high as had been the case at the preceding excavations. Kristensen, Museum Sønderjylland – Arkæologi Hader- The excavation showed – contrary to the previous ex- slev. Again, evidence of craft activities and at least two cavations – only few traces of craft activities. Instead, a longhouses, most likely used as dwelling houses, were number of buildings were indicated by many postholes revealed (Kristensen 2001: HAM 1801 Sb. 61). (Hartvig 2012: HAM 1801 Sb. 61). An enlargement of the churchyard and the extension The chapel, which was located just north of the area of the parish community centre gave way to a further excavated in 1987, was torn down in 2011. At this occa- excavation campaign in the autumn of 2010. This time, sion, the ditch found in 1987 could be identified once an area of c. 2200 m2 west of the old churchyard and the again. The ditch continued directly towards Haderslev parish community centre was excavated by the author. inlet and must thus have surrounded the church.

170 Starup Østertoft

Fig. 3. Plan of the excavations at Starup Østertoft. (Graphics: A. Hartvig & T. Hunniche).

and daub or rather plank or stave walling is uncertain. The Settlement Fireplaces were visible in only three of the houses which The settlement at Starup Østertoft appears to be fo- may be explained by the poor degree of preservation. cused on the excavated areas farthest away from the The lengths of the houses varied between 8 and 20 me- church. In total, the remains of at least thirty buildings tres, the widths between 3.6 and 5.6 metres. Based on have been excavated; all buildings have earth fast posts the fully excavated houses it appears that two types of and had been constructed as aisleless, east-west oriented constructions were preferred, i.e. houses with or with- buildings. The buildings date to the 11th –13th centuries. out central ridge posts. House 24, measuring 15.5 x 5 Due to the fact that the excavated areas were all rath- m is the largest of the fully excavated houses. Along the er small, only six buildings could be excavated in their central line of the house and 2.5 m and 3 m respec- full lengths; the remaining buildings lay partly outside tively from the outer ridge post were positioned two the excavated areas. Almost all the houses had straight strong ridge posts. The position of these two central side and gable walls. Whether the buildings had wattle ridge posts was slightly staggered compared to that of

171 New Aspects on Viking-age Urbanism, c. 750-1100 the wall posts in which way they would have supported The house is situated right down at the Haderslev in- the ridge (Sørensen 2011:105). Apart from having been let and existed in two phases. A culture layer above the roof-supporting posts, they may also have been used in house has been dated to the last part of the 12th – be- internal partitions, which divided the house into three ginning of the 13th centuries. The house underneath the sections: one large central room and two smaller rooms culture layer must thus have been built and abandoned on either side. A similar arrangement involving central prior to the beginning of the 13th century. In the oldest roof-supporting posts and partitions is known from phase the house measured at least 20 x 5.7 m. A central other houses at Starup Østertoft. As mentioned above, wall trench has divided the house into two rooms. The fireplaces had been preserved in only three buildings. remains of a paved fireplace were found in the west- However, based on the fact that the houses have been ern room. During the youngest phase, the house was separated into different rooms and that no larger build- moved further west and measured at least 8 x 5.4 m. ings have been excavated, all buildings are considered Also during this phase the house had been divided into to have been main houses used for dwelling purposes. two rooms. It is a little more difficult to date the con- Apart from the six fully excavated houses, traces from struction of phase 1. Two AMS 14C-analyses suggest a side or gable walls from another sixteen main houses probable date of the house in the second half of the 11th and seven four post-structures have been identified. century.1 Among the buildings that could not be fully excavated, Based on stratigraphy and the 14C-dates it is possible one house attracts particular attention, i.e. house 31. to suggest the following development of the settlement.

Fig. 4. Selected small finds from the excavations: Top left: Baltic Ware; top right: Ottonian-Carolingian disc brooch; bottom left: rivets; bottom right: buckle with animal head. (Photo: J. Andersen).

172 Starup Østertoft

Fig. 5. Map of the area of Starup and Vesterris.

The oldest houses date to the last part of the 11th through the beginning of the 12th century and consist of houses Craft activities 31, 24 and 1. This date is based on Baltic Ware, the The evidence for craft activities was primarily found in overlaying culture layer at house 31 as well as the fact the excavated area closest to the church. Forging and iron that house 24 is the oldest of a number of succeeding working predominates, but evidence of bronze working houses. Characteristics for the houses in the phase are and possibly of bone/antler working has also been found. their size and evidence of division into rooms. Large amounts of iron slag suggest that intensive forg- Hereafter, smaller houses of about 10-12 m length ing and iron working took place on the site. In total, supersede the initial three houses and continue until at least ten smithies have been excavated. Most of the the abandonment of the settlement in the 12th century. smithies were not within buildings and were situated A similar development is known from the settlements close to the church. The smithies consisted of a forge of Østergård and Stagebjerggård (Sørensen 2011:232; and an anvil placed roughly one meter apart. The forg- Hartvig & Pedersen 2013:202). Since there is no evi- es were constructed as simple hollows, often no more dence of fences or ditches, it has not been possible to than 15 cm deep; some had traces of stone paving in the confidently identify tofts or allotments. It is, however, bottom. Smaller, circular features with charcoal and a noticeable that the houses – often during different phases large number of hammer flakes indicate the anvils. Oc- - were located in groups. Whether this may be taken as casionally, there were traces of a central post or stone, evidence of a structure involving allotments or tofts is which could derive from the fundament of the anvil. In uncertain. medieval depictions, the smith is often seen kneeling

173 New Aspects on Viking-age Urbanism, c. 750-1100 and working on the ground (Thomsen 2008:299). This Date must also have been the case at Starup Østertoft. The settlement and craft activities at Starup Østertoft Three smithies were surrounded by a light fence or can be divided into four phases. These phases should – more likely – a smaller building. From the sites of not be seen as unconnected intervals but rather as flu- e.g. Viborg Søndersø and Klosterbakken such buildings id transitions in the development of the site. The small are known (Pedersen 2006:159ff; Thomsen 2008:271). amount of glazed pottery from the site is notable. Add- The material evidence from the smithies and the sur- ing to this fact the few finds which date to the High rounding area consists mainly of nails, rivets and other and Late Middle Ages it appears that the settlement was simple things. Several of the forges also contained small abandoned or very limited after 1200. bronze drops, pellets, waste and bronze slag suggesting Phase 1: The Invisible Phase. This phase is dated broad- that a versatile smith was at work here. ly to the Viking Period, 700-1000 AD. It has not been Evidence of six fireplaces and ovens was also found. possible to relate securely any of the structures or fea- The function of the individual fireplaces and ovens is tures to this phase but the presence of the rune stone, uncertain, but slag, iron objects, limestone and pottery metal detector finds as well as the place names Lund- suggest that they were contemporary with the smithies. ing and Vandling indicate activities in the area (Gam- A large ditch separated the settlement from the meltoft 2012:4). church. The ditch ran two metres west of the late me- Phase 2 comprises the craft activities and the old- dieval tower and could be followed over a total length est part of the settlement, including the longhouses. of 51 m. The ditch was up to 2.5 m wide and 80–120 During the last part of the phase, the ditch was dug out cm deep. In the upper part of the ditch, an up to 80 and the construction of the church began. This phase cm wide stone fundament of granite and limestone was can be dated to the 11th century – beginning of the 12th excavated. This fundament most likely represents an century (Hartvig 2015). earlier churchyard wall erected in connection with the Phase 3 consists of the smaller houses from the settle- construction of the limestone church. The Arnold’s wall ment as well as a small number of smithies. This set- at has also been dated to around tlement phase has been dated to the 12th century. At 1100 and has been erected on top of an older ditch some stage during this phase, the construction work at surrounding the churchyard (Ulriksen 2012:181). The the church was stopped and the church was reduced lack of building material waste underneath the wall sug- to a smaller church than originally planned (Bertelsen gests that the ditch had been already filled in before the 2015). construction work of the church had begun. The fact Phase 4 represents the abandonment of the settle- that the ditch predates the limestone church indicates ment. Judging from the pottery in the culture layers, that the ditch surrounded an earlier church, possibly a the upper layers of the wells and the few coins from the stave church. site, the abandonment or decline must have taken place during the 13th century. The settlement probably lost th Small finds influence in the area towards the end of the 12 cen- tury, but continued some years until it ceased and only The number and variety of finds from the excavations the church remained. is considerable. Pottery makes out the largest group of finds. The majority is locally produced grey ware, while about 1/7 is locally produced Baltic Ware (Sørensen Interpretation and discussion 2011:372-375). Sherds of Baltic Ware dominate the area As mentioned above, it has not been possible to relate of craft activities and the ditch whereas the local grey settlement evidence to Phase 1. It is, however, conceiv- ware is predominant within the settlement. Apart from able that the fertile mainland was preferred to the sandy pottery, everyday objects such as nails, rivets, knives, peninsula. A gold arm ring was found at Bygsgård near shears etc. have been found as well as more exclusive ob- Vandling and Lunding in 1742, some 2.5 km further jects such as bronze buckles and riding gear. inland from Starup Chuch (Parish nr.Sb 27 Sdr. Starup Parish)2. Could this be the place where the family with the rune stone lived? The natural harbour at the pen- insula may thus have been the harbour for the family’s

174 Starup Østertoft ship(s). In addition to the local topography, the place if this assumption is correct, the settlement could have name Starup also suggests some sort of harbour. The covered at least four hectares. first part of the name Starup comes from Stath, mean- Within the last few years several metal detector finds ing landing place (Madsen 1987b:210). During the have been made on the opposite side of the inlet, i.e. 11th century, probably in connection with the harbour at Vesterris. The objects are contemporary with the and the associated ships, craft activities, predominant- finds from Starup Østertoft. As was the case at Starup ly forging, developed on the site. The settlement grew Østertoft, iron and bronze slag have also been found at and large-scale construction works of the church were Vesterris. Vesterris and Starup Østertoft are only sepa- initiated; it is likely that the development of a prop- rated by the narrow inlet, but if or in what way the two er town was in intended. Soon, however, the church areas were related is not clear at present. Should they was reduced compared to the original project and the be regarded as two separate sites, or as one large? The settlement died out during the 13th century: only the parish name may indicate the answer. Starup Østertoft aisleless limestone church remained. The location right lies in Southern Starup Parish suggesting the presence down to Haderslev inlet allowed access to the Baltic Sea of a Northern Starup Parish, or perhaps rather an orig- towards the east and the shortest route across Jutland inal Starup Parish on both sides of the inlet. Perhaps to Ribe towards the west. This route across Jutland is Northen Starup Parish coincides with the present Ves- visible through a number of Viking Period sites appear- terris? Despite these puzzling questions it remains a fact ing like beads on a string from Haderslev inlet to Ribe that there was a significant settlement here during the (Eisenschmidt 2013:fig. 1). From medieval sources the 8th–13th centuries – perhaps even a local centre of the connection from Ribe to Haderslev and from here to area. the Baltic Sea is well known (Madsen 2008:144). The reasons for the decline – or rather, relocation – Apart from the location close to the inlet, the settlement of Starup may be numerous. The most obvious expla- of Starup Østertoft differs also in other ways from the nation should be sought a few kilometres farther west many medieval rural settlement sites. The considerable were the town of Haderslev developed during the 12th and varied spectrum of finds as well as the abundant century and soon took over the leading position in the evidence of craft activities suggest that agriculture did area (Kristensen 2015). not make out the primary subsistence economy. Starup Østertoft should not be compared with the many ear- ly medieval rural settlement sites. Rather, it should be Corresponding author compared to a small group of settlement sites based on Anders Hartvig, Museumsinspektør craft activities. This group includes the sites of Museum Sønderjylland – Arkæologi Haderslev and Kosel-West near Slesvig and Hedeby, both regarded Haderslev, Denmark as settlements where craft activities played an unusual- [email protected] ly important role (Meier 2007:124–125; Ethelberg et al. 2003:455). The similarities in terms of small finds, craft activities and house types between the three sites are significant. Starup Østertoft differs from the other two sites by the lack of Viking Period material evidence and by the presence of the church. The extent of the medieval settlement by the church is not entirely clear. Two trial trenches dug in 1988 and 2010 respectively on the field towards the north demonstrate settlement evidence in the form of post- holes, pits, ditches and culture layers spread all over the field. Prior to the construction of a modern house north of this field, postholes were also registered here (Witte 2010: HAM j.nr. 4560). It is uncertain how much farther north the medieval settlement reached, but it is assumed that the entire outer peninsula was settled;

175 New Aspects on Viking-age Urbanism, c. 750-1100

Unpublished excavation report. Museum Sønderjyl References land – Arkæologi Haderslev. Bertelsen, T. 2015. Storhed og fald. Den romanske – 1987b. Starup Fundet – En udgravning og dens re fråd stenskirke i Starup og Haderslevs første stenkir sultater Sønderjysk Månedsskrift 1987 (63). Hader ke. In: T.R. Kristensen (ed.) Haderslev – En Købstad slev. 206 – 211. bliver til. Udgravninger ved Starup og Møllestrømmen – 2008. Haderslev - de første to århundreder. In: H. = Museum Sønderjyllands årbog 2015. Haderslev. 49 Andersson, G. Hansen & I. Øye (eds.) De første 200 – 84. årene – nytt blikk på 27 skandinaviske middelalderby Danmarks Kirker 1954. Nationalmuseet (ed.) Dan er. Universitetet i Bergen Arkæologiske Skrifter 5. marks Kirker XX:1. Haderslev . Starup. Køben Bergen. 135 – 146. havn. 397– 438. Meier, U.-M. 2007. Die früh- und hochmittelalterliche Eisenschmidt, S. 2013. Häuser der Wikingerzeit in Siedlung bei Schuby, Kreis Schleswig-Flensburg. Of Nordschleswig. In: S. Kleingärtner, U. Müller & J. fa-Bücher 83. Neumünster. Scheschkewitz (eds.) Kulturwandel im Spannungsfeld Pedersen, S.L. 2006. En sønderjysk landsmedje fra von Tradition und Innovation. Festschrift für Micha 1200-tallet. Arkæologi i Slesvig/Archäologie in Schles el Müller-Wille. Neumünster. 195 – 213. wig 11. Haderslev & Kiel. 159 – 166. Ethelberg, P., Hardt, N., Poulsen, B. & Sørensen, A.B. Sørensen, A.B. 2011. Østergård – vikingetid og 2003. Det Sønderjyske Landbrugs Historie. Jernalder, middelalder. Skrifter fra Museum Sønderjylland 5. Vikingetid og Middelalder. Skrifter Historisk Sam Haderslev. fund for Sønderjylland 82. Haderslev. Thomsen, T. 2008. A workshop for different forms of Gammeltoft, P. 2012. Ældre sønderjyske stednavne. metal work from 1018-1025 A.D. at Viborg Sønderjysk Månedsskrift 2012 (88):1. Haderslev. 3 – Søndersø (Denmark) – an archaelolgical and strati 9. graphical survey. Probleme der Küstenforschung im Hartvig, A. 2012. Starup Østertoft HAM1801 Sb. 61. südlichen Nordseegebiet 2008 (32). 67– 96. Sdr. Starup Sogn. Unpublished excavation report. Ulriksen, J. 2012. Kulturhistorisk rapport: Kirkegård Museum Sønderjylland – Arkæologi Haderslev. og hustomter ved Konventhuset i Roskilde. ROMU. – 2015. Udgravninger ved Srd. Starup Kirke. In: T.R. Årsskrift fra Roskilde Museum 2012. Roskilde. 169 – Kristensen (ed.) Haderslev – En Købstad bliver til. 197. Udgravningerved Starup og Møllestrømmen = Muse Witte, F. 2010. Starup Kirkevej 2010 HAM j.nr. 4560. um Sønderjyllandsårbog 2015. Haderslev. 15 – 48. Sdr. Starup sogn. Unpublished excavation report. Hartvig, A. & Pedersen, S.L. 2013. Hvinderupgård og Museum Sønderjylland – Arkæologi Haderslev. Stage bjerggård – Bebyggelse fra vikingetid og Mid delalder omkring Kolding. Arkæologi i Slesvig/Ar (Endnotes) chäologie in Schleswig 14. Haderslev & Kiel. 189 – 210. 1 Three AMS 14C-analyses have been carried out Kristensen, T.R. 2001. Starup Østertoft HAM1801 Sb. (AAR16924-16926). One of the dates must be left out 61. Sdr. Starup Sogn. Unpublished excavation re of the calculation. The remaining two dates were statis- port. Museum Sønderjylland – Arkæologi Hader tically identical and allow an average to be calculated: slev. with one standard deviation the most likely date of the – 2015. Møllestrømmen. Et arkæologisk indblik i Ha house falls between 1090 and 1122. derslevs oprindel. In: T.R. Kristensen (ed.) Hader slev – En Købstad bliver til. Udgravninger ved Starup 2 The so-called SB numbers are a part of the Danish og Møllestrømmen = Museum Sønderjyllands årbog registration system of archaeological sites. Please refer 2015. Haderslev. 113 – 188. to http://www.kulturarv.dk/fundogfortidsminder/Lo- Madsen, L.S. 1985. Starup Østertoft HAM1801 Sb. 61. kalitet/82684/ Starup Sogn. Unpublished excavation report. Muse um Sønderjylland – Arkæologi Haderslev. – 1987a. Starup Østertoft HAM1801 Sb. 61. Sdr. Sta rup Sogn.

176 Aachen and Cologne in Carolingian and Ottonian times and the Viking raids

Aachen and Cologne in Carolingian and Ottonian times and the Viking raids

Sebastian Ristow

Fig. 1. Places that are mentioned in the text. (S. Ristow, graphics: H. Stelter, )

teries at Prüm and Stavelot. It took a in 892 to Historical sources at the time of the Viking end the activities of the invaders. invasions in the Rhine-Meuse region At St-Denis in 869, following the first raids of the It is recorded that towards the end of the 9th centu- Vikings, an elaborate system of defensive ditches was ry there were large incursions of the Vikings into the installed around the wealthy monastery at St-Denis northern Carolingian Empire (Simek & Engel 2004; (fig. 2), which held the important tombs of the Frank- Willemsen 2004; Mäkeler 2008; Jäschke 1975; Runde ish elite.1 At Stavelot current excavations have revealed 2003). After the first raids in the 860’s, in which the existence of destruction layers dating to the late 9th to the north of Cologne was attacked, great numbers century.2 This is also the case at , and of Viking warriors had reached the coast of Flanders by Duisburg, where we find ditches and graves contain- 879 (Runde 2003:256–259 with sources) (fig. 1). ing the dismembered bodies of those slain (Groothed- They moved inland, pillaging and destroying, to Paris de 2004:115–120; Bartels & Vermeulen 2009; Krause and St-Denis and also raided Cologne, Bonn, Aachen 2005:41 fig. 2; 43–47 with ref. therein). But what was and (all several times), as well as the great monas- happening in Cologne and Aachen?

177 New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 2. Monastery of St-Denis in Carolingian times. Narmer Architecture & ArchaeoPlanRistow (Zs. Vasáros, M.Z. Thót, G. Nagy & S. Ristow 2016).

Statements from the time follow a typical pattern. nales Xantenses et Annales Vedastini: 51). In the Fulda Annals, the report for the year 881 for The letter of Archbishop Hermann I of Cologne to the East Frankish Kingdom under King Ludwig III († Pope Steven V provides a good insight into the con- 20. 1. 882), states: “the (…) devastated very sequences of this incursion, stating that “the churches many places in the realm of our King, namely Cam- and all the buildings in the city of Cologne, and with brai, , the Hesbaye region and even Ribuaria, them the relics, after which they were named, (…) were also the chief monasteries there: Prüm, Inda, Stavelot, destroyed by fire» (Transl. Q. Hutchinson after Korth Malmedy and the Palatinate of Aachen, where the 1890:39; cf. Päffgen 2004:87). King’s chapel was located [i.e. the church of Charlem- The old city walls of Cologne, dating to between the agne, St. Mary’s?],” and there they made stables for their 1st and 4th centuries were certainly repaired in 883 (An- horses. They also burnt the churches and buildings of nales Fuldenses: 398), which probably spared the city Cologne and Bonn (Annales Fuldenses: 394; cf. Annales during the next invasion of 891. Mention is made of the Fuldenses sive Annales regni Francorum orientalis: 96– restoration of monasteries and churches as well as the 97; Transl. Q. Hutchinson after Kaemmerer 1980:93). reconstruction of the fortifications. The annals of St. Vaast in which are somewhat less The Chronicon of the Regino of Prüm records that detailed, report in 882 the destruction and incineration in the years 881/82 the Vikings of , among of Aachen and Cologne by the ‘Danes’ (Annales Vedas- other things, left the Aachen Palatinate in ashes having tini: 520; Scriptores rerum Sangallensium: 199; cf. An- already devastated Cologne and Bonn through murder,

178 Aachen and Cologne in Carolingian and Ottonian times and the Viking raids robbery and fire (Reginonis abbatis Pruiensis Chron- icon: 118). The “Normans” plundered the area of the Aachen Palatinate for a second time in 891 (Reginonis chron- icon: 602; Reginonis abbatis Pruiensis Chronicon: 136). There are further reports later of Viking move- ments. All these sources were recorded or composed close to the events, thus leaving no doubt as to their credibility. Obviously, places with accumulated riches were de- liberately targeted and plundered, the most preferable being the monasteries and centres of secular and epis- copal rule. The Viking invasions even produced a legendary Christian victim or ‘martyr’ in Abbot Egilhard of Inda (Kornelimünster). Allegedly, itinerant Vikings stabbed the abbot in . Leo Hugot has suggested that the findings of a of blue stone, sunken into the floor of the oldest building of the monastery church of Inda, is that of Abbot Egilhard dating to the late 9th cen- tury. The skeleton had a 4-inch hole in the skull (Hugot 1968:112). But the history of this abbot is first record- ed in a 17th century chronicle of the monastery at Kor- nelimünster (Kühn 1982:27)3. Victims of the Vikings in 9th century settlement areas can usually be archaeologically identified by way of their hasty burial in irregular and shallow graves, with the bod- ies lying in irregular positions. This can be seen at Duis- burg, Zutphen and also possibly at Aachen where one of Fig. 3. Tomb from Aachen. Prinzenhofstraße. Stadtarchäologie Aachen. these graves (NW 2011/0051) has AMS-dating between (Photo: A. Schaub). 779 and 864 (Analysis of CEZ Archäometrie gGmbH, Dr. Ronny Friedrich, 26th Dec. 2014; Krause 2005; Groothedde 2004; Schaub 2011/12:17 fig. 4) (fig. 3). of these two settlements at this time? What were their Numerous treasure hoards from the western Carolingian respective conditions for the emergence of conurbations Empire might also date back to the time of the Vikings in the 8th/9th century? Did the development of Aachen (Wamers 2005:117 fig. 41; cf. Haas 2011:153–154) but and Cologne continue during the 10th century (Trier it is not clear whether these hoards were a result of the 2002:307)? Viking threat, or connected to internal power struggles. From an archaeological perspective it is at first sur- During the 9th century several settlement agglomera- prising that, in Cologne and Aachen, layers resulting tions were encompassed by ditches, by either the inhabi- from fire and/or destruction, expected to date to the tants or the Vikings themselves, for example at Zutphen end of the 9th century, are completely absent. The same and Deventer. Furthermore, the Carolingian wells were is mainly true for Trier which was also a victim of Vi- filled in after only a short period of use, as at Duisburg, king destruction (Haas 2011:162–165). But are such Huy and Aachen.4 The interruption in the development layers necessary as evidence of a raid? There could be of these settlements, or their cessation in the late 9th cen- alternative and valid explanations for why this evidence tury, as at -Niederberg, can be taken as indica- is missing. Perhaps, after the dangers of raiding had tions of Viking activity (Rünger 2012:235). passed, the physical evidence of destruction was re- Has any archaeological evidence of these events been moved to allow the clearing up for re-settlement, and found in Aachen and Cologne? What was the nature other ‘finds’ got taken away.

179 New Aspects on Viking-age Urbanism, c. 750-1100

Roman colonia. This was a planned city with forti- fications, port and infrastructure, which is acknowl- edged in broad outline. Aachen developed as a spa resort in the 4th century with large baths and settle- ment area possibly with fortification in the area of the modern city centre. The Market Hill with today’s Town Hall was probably at the time a fortified late Roman settlement (). The site of the Roman road, to which the burgus might have been connect- ed, is as yet not precisely known in this part of the city. The latest recorded construction projects at Aachen date from the second third of the 5th century. Only a few pottery fragments can be dated to the 6th century and some very uncertain inscriptions possi- bly date from the 5th through to the 8th centuries. A clearly identifiable poverty in archaeological finds is characteristic of the Rhine-Meuse region, from the second half of the 5th through to the first half of the 6th century. For the 7th century, a fragment of a silver earring was found in a secondary level at the cathedral in Aachen. The discovery of an 8th century wooden coffin and an equal-armed brooch in the form of two conjoined cicadas (‘gleicharmige Fibel’) of the late 8th or early 9th century identify the use of the site and area west of the cathedral as a Late-Merovingian cemetery. This context also produced an extravagant limestone sar- cophagus intended for an infant of about six months from the highest echelons of Frankish society in the second half of the 8th century (Ristow 2012). Possibly belonging to the same context, several fragments of another sarcophagus made from expensive limestone Fig. 4. Features of the times before 800 at site. (S. from Lorraine (possibly Savonnière) survive among the Ristow & A. Kobe 2015). few finds from the numerous poorly documented older excavations at Aachen Cathedral. Some early medieval architectural remains recogniz- ably from the period before (fig. 4), differ Antique and Merovingian - temporal condi- from those of Roman times and have another orienta- tions of urban development tion, (NNE-SSW): a basilica in the north (‘Nordbasi- lika’) and a square or rectangular building (‘SW-Bau’), Aachen as well as fragments of an alignment west of the walls of Insofar as the modern cities of Cologne and Aachen the later ‘central building’ (‘Mittelbau’) from the second were both founded within the in the 1st half of the 9th century. Fragments of a rounded wall and century AD, the formal conditions for their emergence altar foundations (no. 281, 282) also date to pre-Char- are considered similar. The structure of the settlement lemagne times. Finally there is a ditch (no. 366) and of Aachen and its possible function as a regional centre a wall that might mark a boundary of the area in the for the Roman Empire cannot be accurately estimated northeast. Moreover, the fragment of a water pipe found due to gaps in the available sources. to the southwest of the buildings should perhaps also be On the other hand, Cologne was a fully developed included (Kyritz 2010)5 with the 8th century remains.

180 Aachen and Cologne in Carolingian and Ottonian times and the Viking raids

Most of these findings pre-date AD 800. They are thus likely to belong to the time of Pippin the Younger Aachen and Cologne in Carolingian and (also known as ; 714–768), the father of Ottonian times Charlemagne and so date to around the mid-8th century and indicate a functioning residential complex with its Aachen associated churches (Ristow in prep.). When Charlemagne decided to expand his residence in Aachen for his retirement, he improved upon the build- ings that belonged to the time of Pippin, his predeces- Cologne sor. First, he decided to replace the former church with a In Cologne the Roman praetorium was kept in use as a new cathedral, that of St. Mary, with its massive central centre of government from right up to edifice and – for that time – enormous dome. He left Carolingian or latest Ottonian times. Theaula regia only the northern basilica untouched, covering over its mentioned by Gregory of Tours is probably identifiable connection with a further church or memorial building with the Roman palace in the state of the 520’s (Gre- lying to the south, along with sections of an older cem- gorii Turonensis Opera: 231). However, the findings of etery (Ristow 2014). the excavations carried out since 1953 have not yet been According to dendrochronological data St. Mary’s published. A coin dating to the time of the dates from 793 and according to historical data it dates shows an emperor who ruled from 565–578 (Nuber from 796 and was completed for in 804. 1984:109 No. 76), thus as well as some early medie- A ring beam at a height of 25 m dates that part of the val ceramics and a bow fibula providing archaeological dome to 803 ± 10. proof of an expected Merovingian secular presence. As a counterpart to the older northern basilica still in Evidence exists of walls and pottery dating from the use, a new southern apsidal building was erected con- 5th to the 9th centuries, especially beneath and around nected by a passageway and with massive apse founda- the cathedral (Doppelfeld 1970; 1972; Trier 2002). tions, but of its interior design and function, the exca- The mapping of early medieval pottery finds, begun vation documentation from 1913/14 reveals no further by Walter Lung (1956) in the 1950’s,6 taken up by Mar- details. cus Trier and most recently continued by Thomas Hölt- The Atrium was built in its first phase of design with ken, indicates occupation from Antiquity in the east- two sets of three inner exedrae on both long sides prob- ern area of the former Roman city, between the “Hohe ably more or less contemporaneously with the church. Straße” (the former cardo maximus) and the banks of the The construction of the Auditorium and the Granus Rhine. It is only in the area of the former baths and of Tower (‘Granusturm’) possibly began slightly later. The the churches of St. Cecilia (Cäcilien) and St. Kolumba available dendrochronological data from the upper sec- that finds indicates the existence of small settlement tions of these structures would indicate a date between sites. We can expect a gradual resettlement of the entire 800 and 815. Thus, Charlemagne had died before the city during the late Merovingian period at the earliest. tower was completed. Evidence of extensive settlements was found within the The tower, aula and main corridor between the atri- framework of the archaeological investigations of 1992– um and the church were connected by a large passageway 94, 1996–98 and 2004 –09, carried out during the con- along the south side of the aula, leading in a southerly struction of the “Nord-Süd-U-Bahn” in the suburbs on direction. The first phase possibly ended at the former the Rhine at Heumarkt. Most importantly, the evidence south-west building. Later, the corridor was connected clearly shows continuity here from the Roman Empire to the atrium, enabling a direct approach to the upper through to the early Middle Ages. floors in St. Mary’s for the emperor and his entourage. A clear parallel development can be found in Mainz, The central building (‘Mittelbau’) is later and its pre- the second Roman capital of the province of . cise function is unknown. Maybe it served as a repre- An Arabic travelogue (Trier 2002:308) reported in sentative entrance area to the core Palatinate; a meeting 961/66 that only a fraction of the old Roman city area room, a basilica, or perhaps a tower-like building for a was further settled in the early Middle Ages. or library. It was possibly built before, or in consequence of, the Viking invasions at the end of the 9th century. The constructional features all date from the

181 New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 5. Phases of the churches of St. Pantaleon. (S. Ristow & H. Poesche).

th second half of the 9 century, to judge by the dendro- piece of opus sectile in green porphyry, and also some chronological and AMS data from plant residues in the th 7 9 century pottery. A second ditch contained pottery walls. of the 12th century. Evidence of the 10th century comes In Aachen only small insights into the core environ- from the dendrochronological data recovered from two ment of the Palatine are available from documentation. wells that date to the late 9th century. A horse skull Excavations in 2006 to the north, outside the Pa- which may have been thrown in to pollute the water latinate and beyond the ‘Barbarossamauer’ wall on the was found in one well. Only a summary preliminary ‘Templergraben 57’ plot produced evidence of Carolin- report of these findings has been published so far. gian date, especially a dividing ditch with an east-west Some of the settlement areas as identified from burial orientation (Schaub 2007). The small finds included a sites in the vicinity of the Palatinate were either pulled

182 Aachen and Cologne in Carolingian and Ottonian times and the Viking raids down or sacked around the time of the Viking inva- and gained a new profile by its architectural style and sions, perhaps in 891. content (building 2). This building was definitely sacred During excavations in 2008 at ‘Elisengarten’, to the from this time onwards, evidenced by the presence of the southeast of the centre of the Palatinate, finds from the enormously rich tombs. The date, ca AD 500, just de- 10th century were not safely identifiable and only a cades after the baptism of the Frankish King Clovis and few fragments of pottery might indicate a 10th century a great number of his leading warriors, could only signify date.8 a Christian church. Around the middle or late 6th cen- However, thus far, no recorded archaeological ma- tury, the church was expanded to accommodate a larger terial in the city of Aachen can be dated to the period design and was fitted with splendid interior furnishings between the late 9th and the late 11th centuries. In par- and an imposing baptistery (building 3a/b). It was fur- ticular, all products of the earlier characteristic ‘Vorge- ther extended as the bishop’s church in the second half of birgsware’ such as Badorf relief-band amphorae and ear- the 8th century (3 c/d) (Ristow 2002; 2013). Later, - ly Pingsdorf Ware are missing. On the other hand, finds bishop Hildebold built a 95-metre-long double-apsidal begin to appear from the late 1070’s, such as pottery building with a huge atrium to the west (the ‘Old Ca- from the Meuse region where settled potters, possibly thedral’) (Back et al. 2012). Otherwise, in the southern from the Rhineland, were producing new ceramic wares area, the little-known Palace of the Bishops was erected at Schinveld (Stoepker 2011). possibly in Ottonian times.9 This was perhaps also used Concerning the environment of the inner Palatinate in the 10th/11th centuries for visits by the king (Brühl at Aachen, there is still a considerable need for further 1990:37–38; Hillen & Trier 2012:26–41). archaeological research. It would currently appear from At the ‘Heumarkt’, in the heart of the late antique the archaeological evidence that the development of the and early medieval suburb lying to the east of Cologne rural landscape around the Palatine halted towards the on the banks of the Rhine, layers of black earth, up to end of the 9th century. 60 cm thick, have been found comprising the remnants Whether this interruption was due to the Viking threat, and waste from settlements and production activity or to the shifting interests of the royal presence, cannot from the 5th to 10th centuries (Höltken 2003; 2006). be determined. There is a complete absence of easily A 1.5-acre marketplace or mercatus coloniae is men- recognisable destruction layers. The next sequence of tioned for the first time in 992 (Trier 2011b: 185). finds in the Palatinate area dates to the late 11th century, Wooden timber-frame buildings up to 6.5 m wide allowing us to resume tracking its development after al- and 12 m-long, on low plinth walls, dominated the most 200 years of missing evidence. image of the urban settlement of the Carolingian /Ot- An access route had been made to the cathedral in tonian period (Höltken & Trier 2011:177). the zone of the former northern basilica. The northern These buildings differ significantly from those in apse of the basilica no longer remained, and a repre- the trading emporiums of the time, such as Dorestad sentative access was created from the northern square, and HedebyThomas Höltken and Marcus Trier postu- today’s ‘Katschhof’. Excavations in 2012 produced late late antique and Merovingian secular traditions for this Roman and early Carolingian finds in the buildings to architecture in Cologne (Höltken & Trier 2011:178). the southeast, next to the aula, today’s Town Hall. Large- At that time the city covered an area of approximately scale building projects were undertaken in the late 11th 40 ha (Hillen & Trier 2012:12). Occasionally the iden- century and one of the main streets of the high and late tification of special burials might suggest the presence medieval town (‘Krämerstraße’) was subsequently built of royal cemeteries (Höltken & Trier 2011:178–179)10 there. and provide an indication of local divisions across the settlements’ topography. Some findings from recent excavations shed light Cologne on our understanding of the urban agglomeration of The first substantial architecture on the site of Cologne Cologne during the 10th/11th century (Höltken & Tri- Cathedral, possibly even then a sacred building with an er 2011:182–183; Hillen & Trier 2012:13–16.). The internal apse and eastern atrium, was erected in the late development of the settlement appears ubroken and 4th or 5th century (building 1). It was enriched with the the town became dominated by the large mostly stone graves of the Merovingian kings of the mid- 6th century, church buildings of the high and late Middle Ages.

183 New Aspects on Viking-age Urbanism, c. 750-1100

The castle of Deutz remained part of the town in between the late 9th and the 11th century. the 8th/9th century and still functioned as a defensive This certainly was not the case in Cologne to judge fort, as shown by archaeological finds (Höltken & Tri- by the augmentation of the little-known Bishops Palat- er 2011:176). The fort area probably also contained a inate and of its churches. church, which would explain the election of Archbish- Only the external walls of the Church of St. Panta- op Paul, in 870, as taking place in ‘diuza castrum’11. leon show any evidence of possible damage from Viking attacks (fig. 5). A clear break from the representative church foundations in the first half of the th9 century Conclusion (Pantaleon II), which is prominently mentioned in the Cologne and Aachen, following on their existence as 866 description of the goods of Bishop Gunthar, led to Roman towns and settlements, had different historical the new foundation of an extensive Benedictine monas- approaches to settlement patterns. tery by Archbishop Brun in 955–957 (Pantaleon V) (Ri- However, in both places there appears to be a basic stow 2009:108). This construction phase might be par- continuity of settlement from antiquity to the Middle alleled in building phases D and E of St. Severin, outside Ages. Only archaeological material evidence and dated and to the south of Cologne. But both these sites lack contemporary written sources can be used as arguments any definite archaeological layers showing destruction in for continuity. Analyses adopting models of social, eth- the 9th century. nical or religious-historical phenomena should not be Possibly the establishment of the collegiate church of used, because the risk of missing probabilities is too St. Cecilia (Cäcilien) within the city walls in the year high. The growth of Christianity shouldnot serve as 888, if we believe the tradition of its founding, could be evidence of continuity, because in the Rhineland, espe- a reflection on these events (Oediger 1954–61:91). cially, we find discontinuity and change in the advance- In Aachen we have to await the interpretation of the ment of Christianity during late antiquity and Carolin- excavations of the settlement areas outside the inner gian times. Palatinate ditches, especially at Templergraben. Mean- In Aachen the evidence for early Christianity in late while the destruction layers in Duisburg and other plac- antiquity is low, while in Cologne a pronounced Chris- es on the will possibly show that these are tian element seems to have existed by the 5th century. the remains of Viking attacks. Inside the Palatinate of This is indicated by the traditions associated with the Aachen it seems that there is no visible continuity be- earliest bishops and the mention of Christian architec- tween the late 9th century and the emergence of finds in ture of the 4th century. Following an interruption, the the late 11th century. process of Christianisation resumes in the second half of the 6th century, as can be seen throughout the Rhine- land and parts of the Meuse region, such as at Tongeren. Acknowledgements While Cologne became a residential centre for secu- For editing the English translation I would like to thank lar rulers and bishops under the Merovingians and Car- Quentin Hutchinson, Seahouses. For discussion and olingians, Aachen and the Palatinate rose to outstand- advice I would like to thank Wolfram Giertz and An- ing fame only under Pippin, Charlemagne and his sons. dreas Schaub,both Aachen, as well as Thomas Höltken Nothing is known about any associated contemporary and Marcus Trier, both Cologne. settlement around the Palatinate. Perhaps the devel- opment of the ‘secular’ Palatinate, Ingelheim and the bishopric of Mainz reflected the relationship between Corresponding Author Aachen and Cologne. Sebastian Ristow It would certainly be a worthwhile research enter- Archäologisches Institut, Cologne university prise to study the Palatinate of Carolingian times, and Cologne, Germany other central places, more representative of and particu- [email protected] lar to the churches’ background. In Aachen, the Viking invasions may have signifi- cantly disrupted or stopped residential development as indicated by the absence of archaeological finds dating

184 Aachen and Cologne in Carolingian and Ottonian times and the Viking raids

Berthold J. & Kempken, F. 2012. Wohnen und Arbei References ten am Rhein – Der Kurt-Hackenberg-Platz im frü hen Mittelalter. In: M. Trier & F. Naumann-Steck Primary Sources ner (eds.) Zeittunnel. 2000 Jahre Köln im Spiegel der . Scriptores rerum Germanicarum in U-Bahn-Archäologie. Köln. 108 – 111. usum scholarum separatim editi 5. Monumenta Brühl, C. 1990. Palatium und Civitas. Studien zur Pro Germaniae Historica. Ed. G. Waitz (1883). Hanno fantopographie spätantiker Civitates vom 3. bis zum ver. 13. Jahrhun dert 2: Belgica I, beide Germanien und Annales Fuldenses. Scriptores (in Folio) 1. Monumenta Raetia II. Köln. Germaniae Historica. Ed. G.H. Pertz (1826). Han Carruba B. & Wirtz, R. 2012. Der Alter Markt im nover. Früh- und Hochmittelalter. In: M. Trier & F. Nau Annales Fuldenses sive Annales regni Francorum orient mann-Steckner (eds.) Zeittunnel. 2000 Jahre Köln im alis. Scriptores rerum Germanicarum in usum scho Spiegel der U-Bahn-Archäologie. Köln. 146 – 149. larum se paratim editi 7. Monumenta Germaniae Doppelfeld, O. 1972. Köln von der Spätantike bis zur Historica. Ed. F. Kurze (1891). Hannover. Karolingerzeit. In: H. Jankuhn (ed.) Vor- und Früh Annales Vedastini. Scriptores (in Folio) 1. Monumenta formen der europäischen Stadt im Mittelalter 1. Be Germaniae Historica. Ed. G.H. Pertz (1826). Han richt über ein Symposium in Rheinhausen bei Göt nover. tingen vom 18. bis 24. April 1972. Abhandlungen Annales Xantenses et Annales Vedastini. Scriptores re der Akademie der Wissenschaften in Göttingen. Phi rum Germanicarum in usum scholarum separatim lologisch-Historische Klasse. 3. Folge Nr. 83: Göt editi 12. Monumenta Germaniae Historica. Ed. B. tingen. 111 – 129. de Simon (1909). Hannover. – 1970. Das Fortleben der Stadt Köln vom 5. bis 8. Gregorii Turonensis Opera. Scriptores rerum Merovin Jahrhundert nach Chr. Antikvariskt Arkiv 1970 (38). gicarum 1:2. Monumenta Germaniae Historica. Ed. 35 – 42. B. Krusch (1885). Hannover. Fremersdorf, F. 1925. Köln. In: K. Schumacher (ed.) Reginonis abbatis Pruiensis Chronicon cum continua Die merowingische und karolingische Zeit. Siedlungs- tion Treverensi. Scriptores rerum Germanicarum in und Kulturgeschichte der Rheinlande von der Urzeit usum scholarum separatim editi 50. Monumenta bis in das Mittelalter 3. Mainz. 156 – 165. Germaniae Historica. Ed. F. Kurze (1890). Hanno Giertz, W. 2005/06. Zur Archäologie von Pfalz, vicus ver. und Töpferbezirk Franzstraße in Aachen. Notber Reginonis chronicon. Scriptores (in Folio) 1 Monu gungen und Untersuchungen der Jahre 2003 bis menta Germaniae Historica. Ed. G.H. Pertz (1826). 2005. Zeitschrift des Aachener Geschichtsvereins Hannover. 2005/06 (107/108). 7 – 89. Scriptores rerum Sangallensium. Scriptores (in Folio) 2 – 1996. Middle Meuse valley ceramics of Huy-type: a Monumenta Germaniae Historica. Ed. G.H. Pertz preliminary analysis. Medieval Ceramics 1996 (20). (1829). Hannover. 33 – 64. Groothedde, M. 2004. The Vikings in Zutphen (Neth Secondary Sources erlands). Military organisation and early town devel opment after the Viking raid in 882. In: R. Simek & Back U., Höltken, Th. & Hochkirchen, D. 2012. Der U. Engel (eds.) Vikings on the Rhine. Recent Research Alte Dom zu Köln. Studien zum Kölner Dom 12. on Early Medieval Relations between the Rhineland Köln. and Scandinavia. Studia Medievalia Septentrionalia Bartels M.H. & Vermeulen, B. 2009. Die frühmittel 11. Wien. 111 – 132. alterlich umwallten Siedlungen Deventer und Zut Haas, J. 2011. Wikinger in Trier. Zur Frage historischer phen in den östlichen Niederlanden. In: M. Seg oder typologischer Schilderung der Vernichtung der schneider (ed.) Ringwälle und verwandte Strukturen Stadt im Jahr 882. Kurtrierisches Jahrbuch 2011 (51). des ersten Jahrtausends n.Chr. an Nord- und Ostsee. 151 – 179. Schriften des Archäologischen Landesmuseums. Hillen, Ch. & Trier, M. 2012. Zur Geschichte der Köl Ergänzungsreihe 5. Neumünster. 219 –248. ner Königspfalz. Geschichte in Köln 2012 (59). 5 –42.

185 New Aspects on Viking-age Urbanism, c. 750-1100

Hugot, L. 1968. Kornelimünster. Untersuchung über die Mäkeler, H. 2008. Die Wikinger im Frankenreich. In: baugeschichtliche Entwicklung der ehemaligen Bene A. Koch (ed.) Die Wikinger. : 227 – 233. diktinerklosterkirche. Rheinische Ausgrabungen 2. Nuber, E. 1984. Stadt Köln 1:1. Die Fundmünzen der Köln. römischen Zeit in Deutschland. Abteilung VI. Nord Höltken, Th. in prep. Early medieval crafts in Cologne. rhein-Westfalen. Berlin. In: F. Verhaege (ed.) Early towns and artisan produc Oediger, F.W. 1954–61. Die Regesten der Erzbischöfe von tion in the Low Countries an archaeological perspective Köln im Mittelalter 1. Publikationen der Gesellschaft 500–1050. Congress 8./9.12.2011 Amersfoort. Me für Rheinische Geschichtskunde 21. Bonn. 313 – dieval and Modern Matters. 1099. – 2006. Heumarkt V. Karolingisch-ottonische Hausbe Päffgen, B. 2004. Urban settlements and sacral topog funde vom Heumarkt in Köln. Kölner Jahrbuch 2006 raphy in the Rhineland at the time of the Viking (39). 457 – 520. raids. In: R. Simek & U. Engel (eds.) Vikings on the – 2003. Keramikfunde des 8.–10. Jahrhunderts vom Rhine. Recent Research on Early Medieval Relations be Heumarkt in Köln. Kölner Jahrbuch 2003 (36). 511 tween the and Scandinavia. Studia Medie – 566. valia Septentrionalia 11. Wien. 83 – 110. Höltken, Th. & Trier, M. 2011. Köln – Stadtentwick Ristow, S. in prep. Evaluation of excavations since 1911 lung zwischen Kirchen und Kaufleuten. In: G.K. at the Palatinate of Aachen. Stasch & F. Verse (eds.) König Konrad I. Herrschaft – 2014. Archäologie des Aachener Domes zwischen und Alltag. Begleitband zur Ausstellung 911 - Kö spätantiker und ottonischer Zeit (400–1000). In: H. nigswahl zwischen Karolingern und Ottonen. König Müller, C. Bayer & M. Kerner (eds.) Die Aachener Konrad der Erste - Herrschaft und Alltag. Vonderau Marienkirche. Aspekte ihrer Archäologie und frühen Museum Fulda 28. Fulda. 171 – 184. Geschichte. Der Aachener Dom in seiner Geschichte. Jäschke, K.-U. 1975. Burgenbau und Landesverteidigung Quellen und Forschungen 1. . 43 – 79. um 900. Überlegungen zu Beispielen aus Deutschland, – 2013. Forschungsstand und Forschungsstandpunkte Frankreich und England. Vorträge und Forschungen. zu den Anfängen der christlichen Religion im Rhein Sonderband 16. Sigmaringen. land. Rheinische Vierteljahrsblatt 2013 (77). 1 – 24. Kaemmerer, W. 1980. Aachener Quellentexte. Veröffent – 2012. Sarkophag für einen Säugling aus der Dom lichungen des Stadtarchivs Aachen 1 Aachen. grabung in Aachen. Bonner Jahrbücher 2012 (212). Kluge-Pinsker, A. 2001. Produktion und Verbrauch von 290 – 292. Keramik im mittelalterlichen Duisburg des 9.–10. – 2002. Die frühen Kirchen unter dem Kölner Dom. Be Jahrhunderts. Archäologie und Denkmalpflege in funde und Funde vom 4. Jahrhundert bis zur Bauzeit Duisburg 5. Duisburg. des Alten Domes. Studien zum Kölner Dom 9. Köln. Korth, L. 1980. Köln im Mittelalter. Annalen des His – 2009. Ausgrabungen von St. Pantaleon in Köln. Ar torischen Vereins für den Niederrhein 1980 (50). 1 – chäologie und Geschichte von römischer bis in karo 91. lingisch-ottonische Zeit. Zeitschrift für Archäologie Krause, G. 2005. Historisches Ereignis und archäolo des Mittelalters. Beiheft 21. Bonn. gischer Befund – Drei Fallbeispiele aus der Duisbur Runde, I. 2003. Xanten im frühen und hohen Mittel ger Stadtarchäologie. Mitteilungen der Deutschen Ge alter. Sagentradition, Stiftsgeschichte, Stadtwerdung. sellschaft für Archäologie des Mittelalters und der Neu Rheinisches Archiv 2003 (147). 266 – 273. zeit 2005 (16). 39 – 48. Rünger, T. 2012. Die Keramik der karolingerzeitlichen Kyritz, D. 2010. Der karolingische Paukanal in Aachen. Wassermühlen bei Erfstadt-Niederberg (Rhein-- Archäologie im Rheinland. 143 – 144. Kreis, Nordrhein-Westfalen). In: L. Grunwald, H. Kühn, N. 1982. Die Reichsabtei Kornelimünster im Mit Pantermehl & R. Schreg (eds.) Hochmittelalterliche telalter. Geschichtliche Entwicklung, Verfassung, Kon Keramik am Rhein. Eine Quelle für Produktion und vent, Besitz. Veröffentlichungen des Stadtarchivs Aa Alltag des 9. bis 12. Jahrhunderts. RGZM Tagungen chen 3. Aachen. 13. Mainz. 231 – 235. Lung, W. 1956. Zur Topographie der frühmittelalter Schaub, A. 2011/12. Von römischen Wehrgräben un go lichen. Kölner Altstadt. Kölner Jahrbuch für Vor- u. tischen Heiligenbildern. Archäologie in Aachen Frühgeschichte 1956 (2). 54 – 70. 2010/11. Zeitschrift des Aachener Geschichtsvereins

186 Aachen and Cologne in Carolingian and Ottonian times and the Viking raids

2010/11 (113/114). 7 – 29. mentioned in the Annales Bertiniani (1883:98) for the – 2007. Das archäologische Jahr 2006 in Aachen. Zeit year 869. For evidence for the ditch surrounding the schrift des Aachener Geschichtsvereins 2007 (109). 1 – monastery see Wyss (2001:180). 17. Simek R. & Engel U. 2004. Vikings on the Rhine. Recent 2 Many thanks to Brigitte Neuray, Stavelot, for infor- Research on Early Medieval Relations between the mation. Rhinelands and Scandinavia. Studia Medievalia Septentrionalia 11. Wien. 3 Many thanks to Clemens Bayer, Bonn, for discussion Stoepker, H. 2011. Het begin van de aardewerkproductie about Kornelimünster. in Brunssum en Schinveld in het licht van de regionale neder zettingsgeschiedenis: een archeologisch bu 4 Carolingian wells: Duisburgfilled-in at the end of reauonderzoek en een onderzoeksprogramma. Archeo 9th century (Kluge-Pinsker 2001:98-99 pl. 3-4); Huy: coach studies 4. Wijlre. Giertz 1996: 57 note 5 (and finds from Period IIIb); Trier, M. 2011a. Agrippina Colonia und das Militärla Aachen: Goldschmidt Archäologie & Denkmalpflege ger Divitia am Übergang von der Antike zum Mittel Excavation at ‘Templergraben 57’ (NW2006/1173) alter (400–700 n.Chr.). In: M. Konrad & Ch. Wit with two Carolongian wells, one of them dating defi- schel (eds.) Römische Legionslager in den Rhein- und nitely after 856 and possibly 880 ± 10, the other after Donauprovinzen – Nuclei spätantik-frühmittelalterli 887 andpossibly 910 ± 10 (report B. Schmidt, Univ. chen Lebens? Abhandlungen der Bayerischen Akade of Cologne, Labor für Dendrochronologie, report no. mie der Wissenschaften. Philosophisch-historische 2007-16), with warm thanks to Frank Goldschmidt. Klasse 138. München. 175 –195. – 2011b. Köln im Mittelalter: Erzbischöfliche Kapitale 5 The report of the Leibniz Labor für Altersbestimmung und europäisches Handelszentrum. In: L. Heeg Kiel (M. Hüls) revealed the date-range 661–772. (red.) Die Salier. Macht im Wandel. Speyer. 184 – 186. 6 First approaches to the interpretation of continuity – 2002. Köln im frühen Mittelalter: Zur Stadt des 5. bis in the early medieval settlement of Trier: Fremersdorf 10. Jahrhunderts aufgrund archäologischer Quellen. 1925:162–163; Trier 2011a:180–181 with fig. 2; Hölt- In: J. Henning (ed.) Europa im 10. Jahrhundert. Ar ken in prep. chäologie einer Aufbruchszeit. Internationale Tagung in Vorbereitung der Ausstellung „Otto der Große, 7 With possible interpretation as a kind of proto-donjon Magdeburg und Europa“. Mainz. 301 – 310. in response to the Viking threat: Giertz 2005/06:73. Wamers, E. 2005. Die Spur des Silbers – fränkisches Beutegut im Norden. In: E. Wamers & M. Brandt 8 Information from Wolfram Giertz, Aachen. – The ex- (eds.) Die Macht des Silbers. Karolingische Schätze im cavations have not yet been evaluated and published; Norden. Regensburg. 112 – 118. archaeologist: Gary White. Willemsen A. 2004. Wikinger am Rhein. With contri butions of R. Simek & R. de Bruin. Utrecht. 9 In detail, with all finds, including Old Finds: Hillen & Wyss, M. 2001. Die Kosterpfalz Saint-Denis im Licht Trier 2012:20–25. der neuen Ausgrabungen. In: L. Fenske, J. Jarnut & M. Wemhoff (eds.) Splendor palatii. Neue Forschun 10 Other tombs of children had been interred between gen zu Paderborn und anderen Pfalzen der Karolinger 890 and 895 at special areas of the settlement (Carruba zeit. Deutsche Königspfalzen. Beiträge zu ihrer histo & Wirtz 2012:146); a single tomb of a 6/7-year old rischen und archäologischen Erforschung 5. Veröf child of the Carolingian or Ottonian period was found fentlichungen des Max-Planck-Instituts für Ge some 100 meters away at ‘Kurt-Hackenberg-Platz’ schichte 11:5. Göttingen. 175 – 192. (Berthold & Kempken 2012:110).

11 (Endnotes) For example in the Chronicon of Regino of Prüm for the year 869 (Reginonis abbatis Pruiensis Chronicon: 1 The ‚castellum in giro … et lingo et lapide‘ at St-Denis is 99).

187 New Aspects on Viking-age Urbanism, c. 750-1100

Antwerp (Belgium) – a ninth century Viking town? From Early Medieval trading place to Ottonian power centre

Dries Tys & Barbora Wouters

Fig. 1. Present situation (left) and situation before straightening and re-development of the quay between 1870 and 1880 (right, figure adapted from Van de Walle 1961, fig. 34). The letter A represents the 1955–1957 excavation of the Mattenstraat by Van de Walle, the letter B shows the site excavated in 2008–2009.

town centre of , where archaeologists excavated Introduction and studied the 9th- and 10th-century fortress and comi- The origin of the large medieval crafts and trade towns tal palace that are still regarded as the origins of this in Flanders has more than once been the centre of famous town (De Witte 1991). Since then, especially study and debate (Pirenne 1925; Verhulst 1999). This in eastern Flanders, interesting excavations and research debate was predominantly led by historians and their have taken place at the 10th- and 11th century Pfalz with interpretation of topographical information based on a urban ambitions Ename and in the 11th-century town wide range of mainly written sources (Verhulst 1999). of Aalst (De Groote 2010). The problem with town ar- Town Archaeology has only modestly participated in chaeology in Flanders was, and to a certain extent still the debate, since its data on the origins of these towns is, that it is hampered in its working by the constraints were rather scarce until the late 1980s and often not of preventive archaeology. Limited time to engage with up to international standard, as is painstakingly clear proper fieldwork, the use of small test trenches instead from the conference proceedings of Spa (Handelingen of full scale excavations and insufficient means for en- 1990). One of the exceptions was the research of the vironmental and other sampling and analyses are the

188 Antwerp (Belgium) - a ninth century Viking town?

Fig. 2. Digital Elevation Model of Antwerp showing the position of the town on the meeting point of the lower lying wetland areas to the north and the slightly elevated sandy soils to the south. The Fig. 3. underlying part of the earthen rampart construc- elevated D-shape, situated on the eastern bank of the Schelde, is tion. The thickness of the dark earth is ca. 30 cm. indicated with the letter D. In the top left corner of the figure, north of the D-shape, the docks of the current harbour of Antwerp can be discerned. result of a legislative framework that only allows for res- the early medieval stratigraphy on site (Wouters 2011; cue work. Devos et al. 2013), and further research is still taking In Antwerp, the Department of Archaeology of the place on various aspects of the excavation (Bellens et al. City of Antwerp, supported by the developer of the 2012). The well-dated sequence, structures and finds of site, Immpact, as well as the Vrije Universiteit Brussel the site have allowed to re-evaluate the origins of Ant- and the Flemish Heritage Institute were able to develop werp thoroughly and to reconsider the historic origins large-scale excavations in the oldest centre of the town of the town, relating it to 9th- and 10th-century trade in 2008–2009. This oldest centre is the so-called “Burg” and struggles for social, economic and political power of Antwerp, a D-shaped fortress that goes back to at in the Schelde estuary. least the Ottonian period (fig. 1). Already in the 1950s and 1960s, the Antwerp burg was subject to archaeo- Location logical research, notably with excavations by A. L. J. Antwerp is situated on the banks of the river Schelde. It Van de Walle (1961). is located on the spot where the most land-inward Ho- The 2008–2009 campaign spanned nearly two years locene maritime peat formation reaches into the Pleis- and was combined with several interdisciplinary proj- tocene sandy soils (fig. 2). ects, amongst which a micromorphological analysis of

189 New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 4. The wattle wall or fence (4a) and wooden trackway (4c) uncovered during the 2008–2009 excavations, compared with 190 the wooden trackway and houses (4b, 4d) found at the 1955–1957 excavation of the Mattenstraat led by Van de Walle. (Figure 4b and 4d from Van de Walle 1961, respectively plate XVII; figures 35, 38, 39). Antwerp (Belgium) - a ninth century Viking town?

of Antwerp (Annaert 2010).

Non-urban origins In this setting and context, Antwerp seems to have been a central place of certain importance related to trade and exchange during the Early Middle Ages. Antwerp first appeared as regional central place in the 7th century, although we cannot determine its settlement character exactly. There was a small 2nd–4th century Roman river settlement in the surroundings of the later D-shaped core. Furthermore, Antwerp was also a place where the monetarius Chrodigisilus produced golden tremisses around the middle of the 7th century (Tourneur 1942). Antwerp may even have pos- sessed a church, endowed by a regional landowner to the abbey of in 726, although the text in which the abbey tries to prove its ownership rights is to a certain extent a falsum (Rombaut 1989). Finds from this period are scarce in Antwerp. Around 1870 some “Frankish weapons”, including the umbo of a shield were discovered during the construction of the quayside near the core of the town. Other finds include spindle whorls, comparable to the 8th-century examples from Dorestad and some glass and amber finds (Warmenbol 1987). Moreover, over the past few years there have been several dispersed finds of bicoic Merovingian pottery, chaff tempered and burnished wares from the 6th, 7th and 8th century.1 However problematic, the scarce written data suggest that 7th- to early 9th-century Antwerp could have been Fig. 5. Adelbert Van de Walle overlooking the complex stratigraphy a rural estate center of regional importance, involved in of the Burg zone, site Mattestraat, 1958. the trade network between east and west. Geoarchaeo- logical and micromorphological research of the deposits underlying the 9th-century wooden remains excavated North and northwest of Antwerp, we find a more mar- in 2008–2009 (see next section), confirm the possibility itime landscape, while southwards and eastwards, the of such activities at the site. A dark earth2 with a thick- town is connected with the region of Maastricht (east) ness of ca. 30 cm, characterized by its thick, dark, ho- and Brussels (south). In the immediate region around mogeneous, humus-rich, non-peaty appearance (sensu Antwerp, several smaller tributaries connect with the Devos et al. 2009:270) (fig. 3) in this case proved to be river Schelde, providing the immediate surroundings of of rural origin. This is suggested by micromorphological the actual town with a low-lying landscape of wetlands evidence for crop growing and ground raising preced- th th and meadows. From the 5 or 6 century on, the region ed by, or possibly in combination with, use of the land was relatively densely inhabited, with numerous farms as pasture. The precise timespan over which the dark belonging to groups and settlers located in the vicini- earth has formed is difficult to date, but it appears that ty of the contemporary town. These must have had a these activities were long lasting and continuous, possi- certain access to international luxuries, which is illus- bly spanning from the Roman period until the early 9th trated by the prestigious grave goods in the extensive century. The formation of the dark earth stopped when Merovingian cemetery at Broechem, 10 kilometers east it was covered by the depositing of a sandy layer and

191 New Aspects on Viking-age Urbanism, c. 750-1100 became sealed by the 9th- and 10th-century settlement time, it was possible to take 14C samples, which showed (Devos et al. 2013). that the wooden trackway had a possible chronologi- cal range of AD 780–905, while a wattle wall from the

th same campaign dated to AD 680–880 (Devos et al. A wooden port town in the 9 century 2013:169)4. This implies that, already in the 9th centu- Probably the best evidence that Antwerp was indeed an ry, the wooden port in and around the Burg area was a early medieval central place of certain significance is the densely inhabited town with craft production (leather, fact that it was the focus of a Viking action in 836. Addi- antler and combs, textiles and , metal artefacts) tionally, this apparent act of looting was explicitly men- and trade of the same types of pottery as found in Dor- tioned in several written sources, such as the Annales estad (Badorf wares, Reliëfband amphora, red painted Fuldenses (Verhulst 1978, Henderikx 1995). Remark- wares including Hunneschans pottery, and more) with ably, the texts remain silent about Antwerp after 836, up the exception of Tating pitchers (Oakley 2011). to the middle of the 10th century. The archaeological ev- The resemblance with the structures and urban lay- idence, however, shows that Antwerp did not disappear out of ports like Birka, Haithabu, and Dublin, to name during the 9th century. Inside and around the area of only three of them, is striking (Clarke & Ambrosiani the D-shaped core of the present town, impressive traces 1991). Already in his time, Van de Walle suggested a link were found of a 9th-century port or trading settlement with the Scandinavian world (Van de Walle 1961:127), with a distinctly urban character. Both the excavations but without carbon-dated sequences he was not able to of 2008–2009 and in particular the excavations between prove this. Nevertheless, more arguments exist to sug- 1952 and 1961 have shown the presence of wood- gest a Scandinavian influence or presence. The site was en houses on small urban plots, surrounded by wattle not mentioned in written sources between 837 and the walls or fences, oriented onto wooden trackways (fig. 4) middle of the 10th century, and had no official position, (Van de Walle 1961; Oakley 2011; Bellens et al. 2012). nor was there any claim from a public Frankish author- The size of the plots and houses varies between 5–6.5 ity. At the same time, there are clear indications that m (width) and 12–13 m (length). The best-preserved and the Schelde estuary were controlled by Dan- houses, platforms and trackways were found during the ish rulers, like Harald and Rorik, and their garrisons, excavations by Van de Walle. He investigated a number between ca. AD 840 and 870 (Tys 2010). There are suf- of clearly non-rural tenements on the Mattenstraat, near ficient arguments strongly suggesting that th9 -century the waterfront in a complex and long lasting occupation Antwerp could well have been a port, town and crafts sequence, which was never fully excavated or published centre under influence of Scandinavian traders. This (fig. 5). The earliest excavated buildings were construct- needs confirmation, but data from the recent excava- ed in wood and probably dated to the ninth century, but tions appear to support this perspective. there were also significant unexcavated structures and deposits below these buildings. Additionally, a range of th th 9 –10 century imports suggested a trading settlement Antwerp between the late 9th and the late 10th (Van de Walle 1961:123–136; Loveluck & Tys 2006; century - aspects of crisis and the relation be- Oakley 2011). tween politics, power and urbanisation In 1988, remains of wooden settlement structures probably dating from the 9th century were found im- The actual Burg site is clearly recognisable by its mediately north of the so-called Burg-area, on the Gul- D-shaped structure (cf. fig. 1). The D-shape consists of denberg-site (Tys 1998)3. In 1266, this area carried the a chalkstone wall and, inside the wall, an older earthen name of Craywick and Leerwich, two apparent wic top- rampart that is remarkably well preserved underneath onyms of which the significance remains unclear (Prims modern houses and street levels. The archaeological & Verbeeck 1938:181). East of the Burg-area, Johan campaign of 2008–2009 has uncovered significant in- Veeckman excavated the remains of an organised grave- formation about the structure and chronology of the yard, dating from the 9th century (Veeckman 1998). D-shaped earthen rampart and the wall. The construc- The excavations of 2008–2009 confirmed the inter- tion of earthen rampart seems to originate from the end th pretation of previous results and showed that the actual of the 9 century (Bellens et al. 2012), with significant town centre originated in the 9th century at latest. This implications for the site. One of the results was that the

192 Antwerp (Belgium) - a ninth century Viking town? rampart formed a clear delineation in the urban space market, a fish market and a corn market. In the course that structures the core of the town up till today. How- of the 10th century, the urban settlement expanded well ever, the presence of 9th-century occupational structures outside the D-shape. These developments fit into what is outside the D-shape suggests that the older settlement known elsewhere as the Ottonian frontier policy, for ex- was initially larger than the late 9th-century D-shape and ample the erection of the fortress on the river Schelde in not necessarily shaped in this way. The rampart’s main Ename in AD 974 (Callebaut 2010). At the same time, significance is that the town clearly entered a phase of in the light of the new assessment of the 9th century Dan- crisis and stress towards the end of the 9th century. This ish influence in the development of Antwerp as a trade is not unexpected, considering the general crisis in the town, it is not unreasonable to interpret this as an official Low Countries around this period (Hendrikx 1995; Tys ‘recuperation’ of the urban space by the Frankish powers 2010). For instance in Gent, Mechelen, Namur, Aalst that had no control over this town earlier. and other centres, the development of the first defensive This transformation process from portus under Danish structures dates between the end of the 9th and the mid- influence to Ottonian burg illustrates the importance of dle of the 10th century, indicating the unstable character ideological space in shaping people’s experience of public of politics and society at the time. Unfortunately, the na- power. ture of the political events that caused the construction of the earthen rampart remains entirely unclear. Acknowledgements Data from written sources strongly suggest that An- We would like to thank Tim Bellens and Anne Schryvers twerp had become an Ottonian centre by the middle at the Department of Archaeology, City of Antwerp, of the 10th century. In 941, an inhabitant of the vicus as well as Yannick Devos, Centre de Recherches en Anuuerpis is mentioned in the Liber Traditionum of the Archéologie et Patrimoine, Université Libre de Brux- St-Peters abbey of Gent (Gijsseling & Koch 1950:138). elles. More important is the fact that Antwerp got hold of tolls on traffic and trade on the river, as well as tolls lev- ied on the market of Antwerp, of which traders from Corresponding Authors Ottonian ports and towns such as and Aachen seem Dries Tys to have been exempted (Doehaerd 1947:39-40; Verhulst Department of Arts and Archaeology, Vrije 1978:30–31). Notwithstanding the fact that the toll is Universiteit Brussel only described in later sources, the toll itself is an Ot- Brussels, Belgium tonian creation, which fits in the Ottonian network of [email protected] places of power and trade. In AD 980, a letter by bishop Notger of Liège describes Antwerp as a castrum (Gijssel- Barbora Wouters ing & Koch 1950:242)5. Department of Arts and Archaeology, Vrije Recent research on sections of the chalkstone wall Universiteit Brussel around the D-shape that were discovered during the Brussels, Belgium 2008-2009 campaign date its construction between the Department of Archaeology, University of Aberdeen end of the 10th and beginning of the 11th century (Bel- Aberdeen, lens et al. 2012:13). Additionally, the construction of a [email protected] Saint Walburga chapel in this fortress (stratigraphically on top of 9th century houses) fits this transformation of the D-shaped core of the town into an impressive and high-status Ottonian castral centre. This change proba- bly took place during the last quarter of the 10th century. Inside the fortress castral elements are present, such as the house of the villicus and the schepenbank, likely the oldest juridical court (Verhulst 1978). It is apparent that the Ottonians transformed the for- mer trade town into a high-status fortress or castrum around which new markets developed, such as a cattle

193 New Aspects on Viking-age Urbanism, c. 750-1100

Henderikx, P.A. 1995. De ringwalburgen in het mond References ingsgebied van de Schelde in historisch perspectief Annaert, R. 2010. The Merovingian cemetery of Bro In: R.M. Van Heeringen, P.A. Henderikx & A. Mars echem (B, province of Antwerp) in the North-Aus (eds.) Vroeg-Middeleeuwse ringwalburchten in Zee trasian Renensis. In: K. De Groote, D. Tys & land. Goes. 70 – 112. M. Pieters (eds.) Exchanging Medieval Material Cul Loveluck, C. & Tys, D. 2006. Coastal societies and ture. Studies on archaeology and history presented to identity, AD 600–1200: rethinking the social dis Frans Verhaeghe. Relicta monografieën 4. Brussel. 205 tribution of material culture along the Channel and – 215. North Sea coasts. International Journal for Maritime Bellens, T., Schryvers, A., Tys, D., Termote, D. & Na Archaeology 2006 (1):2. 140 – 169. kken, H. 2012. Archeologisch onderzoek van de An Nicosia, C., Devos, Y. & Borderie, Q. 2013. The contri twerpse burcht. Monumenten en Landschappen 2012 bution of geosciences to the study of European Dark (31). 4 – 21. Earths: a review. European Journal of Post-Classical Callebaut, D. 2010. Ename and the Ottonian west Archaeologies 2013 (3). 145 – 170. border policy in the middle region. In: K. De Oakley, E. 2011. Settlement and economy in Early Groote, D. Tys & M. Pieters (eds.) Exchanging Medie Medieval Antwerp: an assessment of pottery from val Material Culture. Studies on archaeology and his St.-Walburgis excavations 1957-1961. Medieval and tory presented to Frans Verhaeghe. Relicta monogra Modern Matters 2011 (2). 153 – 166. fieën 4. Brussel. 217 – 248. Pirenne, H. 1925. Medieval cities: Their Origins and the Clarke, H. & Ambrosiani, B. 1991. Towns in the Viking Revival of Trade. Princeton. Age. Leicester – London. Prims, F. & Verbeeck, M. 1938. Antwerpsch Straatnam De Groote, K. 2010. The contribution of archaeologi enboek. Antwerpen. cal sources to the research in the formation of towns. Rombaut, H. 1989. De Echternachseteksten betref The exemple of Aalst, a border town in the county fende Antwerpen: nieuwe argumenten Antwerpen. of Flanders. In: K. De Groote, D. Tys & M. Pieters Bijdragen tot de Geschiedenis 1989 (72). 3 – 26. (eds.) Exchanging Medieval Material Culture. Studies Tourneur, V. 1942. L’atelier monétaire d’Anvers des temps on archaeology and history presented to Frans Verhae mérovingiens au 12 siècle. Centennial volume of the ghe. Relicta monografieën 4. Brussel. 249 – 266. American Numismatic Society. New York. Devos, Y., Vrydaghs, L., Degraeve, A. & Fechner, K. Tys, D. 1998. De rijkdom van grijs. Het middeleeuws 2009. An archaeopedological and phytolitarian study reducerend gebakken aardewerk uit enkele Ant of the “Dark Earth” on the site of Rue de Dinant werpse sites. In: J. Veeckman (ed.) Berichten en Rap (Brussels, Belgium). Catena 2009 (78). 270 – 284. porten over het Bodemonderzoek en Monu Devos, Y., Wouters, B., Vrydaghs, L., Tys, D., Bellens, mentenzorg 2. Antwerp. 161 – 199. T. & Schryvers, A. 2013. A soil micromorphological – 2010. The Scheldt Estuary as a Framework for Early study on the origins of the early medieval trading Medieval Settlement Development. In: A. Wil centre of Antwerp (Belgium), Interna lemsen & H. Kik (eds.) Dorestad in an international tional 2013 (315). 167 – 183. framework. New research on centres of trade and coin De Witte, H. (ed.) 1991. De Brugse Burg. Van grafeli age in Carolingian Times. Turnhout. 168 – 176. jke versterking tot modern stadskern. Archeo-Brugge 2. Van de Walle, A. 1961. Excavations in the Ancient Brugge. Centre of Antwerp. Medieval Archaeology 1961 (5). Doehaerd, R. 1947. Comptes du tonlieu d’Anvers 1365- 123 – 136. 1404. Brussel. Veeckman, J. 1998. De oudste Antwerpenaar op het Gijsseling, M. & Koch, A.C.F. 1950. Diplomata Belgica spoor. Archaeologia Medievalis 1998 (21). 65 – 66. ante annum millesimum centesimum scripta. Brussel. Verhulst, A. 1978. Het ontstaan en de vroege topografie Handelingen 1990. Ontstaan en vroegste geschiedenis van van Antwerpen van de Romeinse Tijd tot het begin de middeleeuwse steden in de zuidelijke Nederlanden. van de 12de eeuw. In: L. Voet (ed.) De stad Antwer Een archeologisch en historisch probleem. Handelingen pen van de Romeinse Tijd tot de 17de eeuw. Topograf van het 14de Internationaal Colloquium Spa, 6–8 ische studie rond het plan van Virgilius Bononiensis september 1988. Brussel. 1565. Brussel. 13 – 40.

194 Antwerp (Belgium) - a ninth century Viking town?

– 1999. The Rise of Cities in North-West Europe. Cam bridge. Warmenbol, E. 1987. Hoe Frankisch zijn de oudere Antwerpse vondsten wel? In: E. Warmenbol (ed.) Het Ontstaan van Antwerpen, Feiten & Fabels. Ant werpen. 151 – 156. Wouters, B. 2011. A micromorphological study of the dark earth of Antwerp (burchtsite). Unpublished MA thesis. Vrije Universiteit Brussel.

(Endnotes) 1 These were found in a medieval ditch near the An- twerp Academy, on the south bank of a rivulet that flowed into the Schelde just north of the actual town centre.

2 These are often, but not at all exclusively found at Late Roman and Early Medieval town sites (Nicosia et al. 2013); the general term “dark earths” refers to their similarity in appearance, taking into account that the activities and processes of which they are the result may differ considerably.

3 Excavations Antwerpse Vereniging voor Bodem en Grotonderzoek.

4 The trackway was dated between the end of the 8th and the 10th century (For a probability of 68.2%: AD 780e905 (23.1%) and AD 915e970 (45.1%), and a probability of 95.4%: AD 780e790 (2.2%) and AD 800e890 (93.2%)).

5 The letter recounts a miracle that would have hap- pened in Antwerp around AD 980 and fits well with the rising attention of Notger for the Schelde-region, where he is known to have invested in prestigious abbeys and churches.

195 New Aspects on Viking-age Urbanism, c. 750-1100

What Scandinavians may have seen and experienced in Miklagarðr: the urban fabric of Constantinople during the Middle Ages

Jesper Blid

This brief article investigates some of the possible ur- the Roman Empire, a status it maintained throughout ban features of Constantinople during the 9th century, the Byzantine era (until 1453). Historians have estimat- at the time when Scandinavians began to journey there ed that its population during the centuries immediately via the Russian rivers.1 As the general aim of this volume following Constantine’s refoundation numbered several is to shed light on possible urbanistic concepts in Scan- hundred thousands, maybe as many as half a million. In dinavia during the Viking Age, my contribution offers a terms of urban outlines, Constantine’s city had already hypothesis of Constantinople’s urban development and undergone change in Late Antiquity, and archaeological infrastructure at this time and also discusses the archi- excavations have exposed evidence of destruction from, tectural reuse of the earlier Roman monuments in the for example, fires and earthquakes, and also consider- city during the Middle Ages. As a classical archaeolo- able rebuilding. Even municipal buildings and the in- gist, I am not in a position to judge whether or not the frastructure of the city had begun to reform at this time, Northmen’s view of the city influenced Viking-period particularly as a result of increased privatization. During urbanism in Scandinavia; however, I can highlight cer- the Middle Ages/the Viking Age, when contacts with tain urbanistic tendencies at Constantinople and leave Scandinavia can be proved with certainty, it seems that it to my colleagues in Nordic archaeology to draw con- the Roman orthogonal city grid had largely transformed clusions. I should, however, point out that the antique into a more organic formula. The main structures re- appearance of the city – the traditional Graeco-Roman maining from Constantine’s city were isolated monu- urban layout with an orthogonal city grid (and presum- ments, as reproduced in late medieval and ably an egalitarian system of quarter distribution) – was vedute and maps. probably no more than a distant memory in Constan- The reuse of earlier structures did not end in the Mid- tinople by the Middle Ages. Yet, a study of the form, dle Ages: Byzantine buildings were frequently rebuilt/ function, and spatial relations of individual buildings reused during the Ottoman period (i.e. after 1453). in Constantinople and Viking Age Scandinavia might These include Byzantine monasteries and parish church- reveal relevant points of comparison. es, many of which survive and are still partly in use, and more of which remain scattered throughout the city as abandoned ruins. Only one of the Byzantine churches, A city of constant renewal namely Panaghia Mouchliótissa, is still used today by It is evident that Byzantium/Constantinople/Miklagarðr/ the Greek Orthodox Patriarchate. The others, with the Istanbul had a dynamic, ever-changing urban fabric from exception of two that serve as workshops for carpen- antiquity onwards. The refounded city, which Constan- try and for cars, have all been converted into mosques. tine inaugurated in AD 330, was in part built on the Thus, sacred architecture maintained its original pur- ruins of the older Greek of Byzantium.2 With the pose, although in a somewhat different form. It is inter- coming of Emperor Constantine, the city, which would esting to note that these older medieval churches have in time bear his name, underwent rapid urban expansion served continuously as houses of prayer, some for as long and eventually became the most significant metropolis of as 1,500 years.

196 What Scandinavians may have seen and experienced in Miklagarðr

Fig. 1. Runic graffito with a Scandinavian name located on the balustrade of the southern gallery of Hagia Sophia.

in the disposition of public spaces, a situation General urban development from Con- which, in my opinion, deserves more attention. stantine to the Byzantine Middle Ages The heart of the traditional Roman city was the As early as Late Antiquity, Constantinople was forum (agora in Greek). This was the location of strongly marked by Christianity. Hundreds, if not a number of institutions that were symbolic of thousands, of chapels, churches and monasteries the city and Roman culture in general. Council were erected in the city alongside traditional Ro- quarters, administrative buildings, and temples man buildings such as temples, porticoes, theatres, flanked the portico-lined square. Both secular and baths. Before we look more closely at individ- and sacred buildings lay side by side here, often ual examples of the city’s medieval development, without clear separation. Temples in honour of we must direct our attention to the general devel- the emperor’s genius, as well as other divinities, opment of Roman cities during the early medieval created a natural focus along one of the shorter period. sides of the square. This ‘temple area’, called the Traditionally, the urbanistic concept of the ‘Capitolium’, would have contained between one Graeco-Roman city is viewed as the direct opposite and three temples raised to the ‘Capitolian triad’: of its medieval counterpart. The antique settlement Jupiter, Juno, and Minerva. The twofold social is characterised by an orthogonal city grid, while stratification of Roman society was evident at the the medieval is distinguished by a more organically forum. On the one hand, the emperor and other developed city plan, with meandering streets and influential citizens could demonstrate their posi- alleys. Be that as it may, there is a great difference tion by raising self-aggrandizing monuments. On

197 New Aspects on Viking-age Urbanism, c. 750-1100 tion by raising self-aggrandizing monuments. On the into a gambling hall and used especially for games of other hand, council buildings and institutions embod- dice. Constantinople’s Capitolium, which was presum- ied the wishes and rights of the Roman citizens. Even ably erected by Constantine at the beginning of the 4th though the Roman system was never a in century, became an academy of higher education by the classical Athenian sense of the word, a similar, but imperial order in 425. However, by that time the build- hybrid, system was adopted. An example of this is the ing had been abandoned for long. In the second half of circular forum that Emperor Constantine erected in the 5th century, there is a noticeable move throughout Constantinople. This contained a great commemora- the Roman Empire to convert old temples into church- tive column at the centre, which was crowned with a es. This reuse intensified even more during theth 6 and colossal statue of the emperor himself. The forum of 7th centuries. To the best of my knowledge, this never Constantine also contained the senate building, the happened in Constantinople. Some of the more well- curia. This provided a symbolic image of the emperor known examples from elsewhere are the 6th-century and senate (i.e. the people) collectively ruling the land, cathedral in the Parthenon on the Acropolis of Athens which was quite far from the reality. and the conversion of the Pantheon in Rome at the be- Thus, the forum naturally required a very large space, ginning of the 7th century. and, as shown above, it was a symbolic centre for the city, the state, and traditional Roman values. Trade and lively social interaction among the citizens also took Continuous use of Roman public architecture place here midst all the customary buildings. However, Perhaps the most appreciated buildings among the gen- the increasing influence of the Church on the Roman eral populous in Roman times were the public baths, Empire began to affect the old system. One can speak the thermae. While the Church maintained a dualistic of a social reform, which resulted in much of the previ- view on traditional bathing culture, several prominent ous public spaces, such as the forum, being left without theologians were openly critical. , any clear function. This initiated a large reassignment of who was patriach of Constantinople at the beginning of older public and religious monuments in Late Antiquity the 5th century, considered the public baths to be a seed- and the early Middle Ages. It should be mentioned that bed for the root of all evil: female vanity. The Church several traditional institutions survived, but often only Sokrates, who wrote his Historia Ecclesiastica as symbols of former glory days. in the 5th century, welcomed a rise in morals among the The pagan cults were officially banned in the 390s by men of the Church. Previously, according to Sokrates, Emperor Theodosius I. However, polytheism still sur- if a bishop was asked how a man of his dignity could vived in many places well into the 6th century. Yet, as visit the baths up to twice a day, he was likely to answer: recent archaeological and historical research has shown, ‘because three times would be overdoing it’. It should many temples were already abandoned by the time of be clarified, however, that it was the custom of men and Theodosius’s . The sacred places stood empty, and women bathing together that was especially worrying to this created a problem – what were the citizens to do certain prelates and . Mixed bathing was with the pagan temples? Both textual sources and ar- banned at the Council of Laodikeia in the 4th century chaeological evidence have, however, confirmed that the and yet again at the Council of Trullo in 692 (possibly a abandoned buildings were quickly reused for activities sign that the first prohibition did not work as it should). other than polytheistic cults. This was a natural reac- It must, however, be pointed out that even the Romans tion, as the large and richly decorated temples were the themselves could occasionally consider shared bathing a most monumental buildings of the ancient cities. Yet, moral deviation. A preserved imperial decree from the how were these temples reused? The Byzantine historian time of Hadrian (2nd century AD) bans men and wom- John Malalas, who lived in the 6th century, provides us en from bathing in the same place at the same time. with some clues. Beside the Theatre of Constantinople, Yet, what really happened to the Roman baths of Con- which is now the site of the Ottoman Topkapı Palace, lay stantinople during the Middle Ages? The majority a temple dedicated to the goddess Aphrodite. Accord- seem to have been annexed by the Church in order to ing to Malalas, this building was used as a coach house be used for commercial purposes, baptisms, and other by the prefect of the Praetorian Guard. Malalas further sacred ceremonies. However, the population of Con- states that the city’s Temple of Artemis was converted stantinople does not seem to have bathed any less than

198 What Scandinavians may have seen and experienced in Miklagarðr

Fig. 2. Map of Constantinople by Sebastian Münster, c. 1550. before (the Turkish Hamam tradition in Istanbul is The monumentality of the building contributed to it nothing new). The difference was that bathing now being used for prestigious meetings. The Seventh Ecu- took place in Church-owned facilities, where men menical Council is said to have been held in the baths and women were kept apart in order to avoid the in 681 (although surely not in the bathing halls). temptations of the flesh. An archaeological exam- Two other buildings which were preserved in the ple that illustrates this is a minor bathing complex medieval period and partly retained their original func- built c. AD 400 close to the Aqueduct of Valens in tions were the Hippodrome and Kynegion (probably a central Constantinople. In the 6th century, a church Roman amphitheatre, which to date has not been iden- was built partly within the baths (today, the origi- tified with certainty). Chariot racing inside the hip- nal Byzantine church on the site serves as a mosque podrome was a major source of entertainment during called Kalenderhane Camii). Constantinople’s largest Roman times and also throughout the Middle Ages and thermae were the Baths of Zeuxippos. These lay right even into the Ottoman period. At the beginning of the beside the imperial palace and had been built during 17th century, most of the complex was demolished and the time of Constantine or, as some scholars believe, the stone blocks were reused to build the nearby Blue as early as the 3rd century AD. Baths during classical Mosque. Kynegion is said to have been reused for exe- antiquity were very costly complexes, containing not cutions during the Middle Ages. This function does not only swimming pools and sports halls but also art- differ much, however, from the activities which were work, libraries, cultic areas, etc. The Baths of Zeux- carried out in Roman amphitheatres during antiquity. ippos undoubtedly formed such a complex. The last recorded use of this complex as a bath house was at the beginning of the 8th century, but there is reason The naval infrastructure of the city to believe that the building was primarily used for As with most pre-modern centres of power, Constan- other purposes as early as the 6th and 7th centuries. tinople’s contacts with the rest of the world were con-

199 New Aspects on Viking-age Urbanism, c. 750-1100 ducted mainly along waterways. The Greek city of Byz- sometime around the turn of the 10th–11th century. Ar- antium was founded on a promontory on the western chaeologists believe that the destruction was caused by side of the Bosporus inlet, a strategic site that com- a heavy storm. Owing to the suddenness of this event, manded the waterways between the Sea of Marmara and many ships still retain their cargo in its original position the . As Constantinople’s importance grew in inside their hulls. The cargo consists mainly of contain- the 4th century AD, the city was provided with several ers (amphorae) filled with wine and oil. The early finds, harbours stretching out towards the Golden Horn and those dated to before the mid-7th century, come mainly the Sea of Marmara. Over the centuries, the majority of from Egypt and Palestine, but following the Arab con- these were filled in with the city’s refuse and now lie a quest of these areas, the find material is dominated by little inland from today’s coastline. The most extensive imports from the Aegean Islands and the Black Sea. excavations conducted as part of the creation of Istan- During work on Istanbul’s new metro system, ar- bul’s new metro system concern one of these antique chaeologists excavated buildings from the Byzantine harbours. Emperor Theodosius I (AD 379–395) built a and Ottoman periods at different sites within the city. new harbour at the site known today as Yenikapı (New Reports of several private houses, palaces, and churches Gate). Large breakwaters surrounded the harbour. Por- now await publication. This means that we will soon tus Theodosiacus, as the harbour came to be called, was have more information about old Constantinople, a reportedly used for annona, the annual transportation city whose archaeological remains are currently little of grain, which reached the city from Alexandria. Like known. most major cities in antiquity, Constantinople was not self-sufficient, but rather de pended on the importation of grain from the provinces. The archaeological evidence at Portus Theodosiacus reveals that imports from Egypt The Northmen and Miklagarðr stopped in 641, the time of the Arab conquest of the We have seen how Constantinople underwent an ur- Nile Valley. ban metamorphosis from Late Antiquity onwards. As The excavated finds from the harbour indicate that early as the 5th and 6th centuries, some polytheistic tem- it continued in use into the first half of the 11th centu- ples were being transformed into, for example, garag- ry. The archaeological finds from Portus Theodosiacus es, universities, and casinos. The city’s imperial baths are extensive. So far, several ships of Late Antique and became an auditorium for a church council. There are medieval date have been uncovered. Many of the ships naturally many more examples of Roman buildings that have been transferred for conservation to various sites in were reused in one way or another during the Middle western , and there is a plan to build a museum Ages, whether for a new function or in a remodelling of in Istanbul for the permanent display of all the ships. their original one. It should, however, be pointed out A synthesis of the finds from Portus Theodosiacus, or that much of the city of Constantinople/Miklagarðr, Yenikapı, as the excavation site is usually called, bears with which the Northmen came into constant contact witness to several dramatic incidents in Constantino- during the Viking period, was most likely not the same ple’s history. Geologists have studied the harbour’s vari- city that Constantine inaugurated in the 4th century. ous soil layers in collaboration with archaeologists, with Much of the orthogonal street plan had transformed many important results. According to surviving writ- into meandering streets and blind alleys. Furthermore, ten sources, a major earthquake hit Constantinople in the city was dominated by Christian monumental ar- 553. Geologists theorise that this earthquake triggered chitecture that had replaced the functions and ideolo- a major tsunami which engulfed the whole of Yenikapı. gies of the ancient polytheistic cults and Roman public The soil layers contain wrecked ships and even the com- places. The social role of the Church and its place in plete skeletons of what were probably working animals the cityscape might have been a source of influence for such as camels and horses. Osteological evidence shows the Christian Scandinavians who returned home from that these pack animals suffered serious injuries from Constantinople. The Byzantine influence of some me- bearing heavy burdens. When the tsunami swept over dieval murals at the churches of Garde and Dalhem on the city, these animals were apparently in the harbour Gotland, for example, has already been well researched, area. The majority of the excavated ships are medieval, but, in my opinion, the ideologies of the Byzantine and these also suffered violent destruction, possibly Church as a spiritual and social institution need to be

200 What Scandinavians may have seen and experienced in Miklagarðr examined more closely as a possible influence on the development of the Church in Scandinavia (not least on Gotland). Other less tangible points of reference between Constantinople and the Northmen might be more difficult to trace, such as commercial contacts, practical conditions of trade, and administration. A further very important link between Byzantium and Scandinavia is linguistic exchange. What do we real- ly know about Greek-speaking people in Viking Age Scandinavia? The famous late-13th-century Gotlandic seal inscription [GUTENSES SIGNO XPISTVS SIG- NATVR IN AGNO] demonstrates remarkable Greek influence in the spelling of Kristus [XPISTVS], which uses the Greek letters X and P. Furthermore, it is possi- ble that some citizens of Constantinople understood the enigmatic runes which Halvdan cut into a gallery balus- trade in Hagia Sophia. The new excavations at Yenikapı have clarified Constantinople’s position, its enormous commercial potential, and its wide trading connections during Late Antiquity and the Middle Ages. At the time that the Scandinavians arrived, Portus Theodosiacus was one of several functioning harbour sites most certainly connected to a complicated administrative apparatus; when the Scandinavians anchored in Constantinople (on peaceful expeditions), they could not have avoided Byzantine bureaucracy. It seems, in my opinion, more likely that travelling foreigners would have adopted rel- evant administrative customs and activities rather than have focused on fragmentary remnants of a long-altered Roman urbanism. Thus, to conclude, I consider it more probable that the traditions and institutions that were alive in the city of plenty (Miklagarðr) during the Vi- king Age exercised greater influence than any legacies from a fading classical Roman world.

Corresponding Author Jesper Bild Institutionen för arkeologi och antikens kultur, Stock- holm university Stockholm, Sweden [email protected]

201 New Aspects on Viking-age Urbanism, c. 750-1100

dle Byzantine Period in the Byzantine Empire, and the References (further reading) Viking Age in Scandinavia. I will use the term ‘medie- Androshchuk, F. 2012. Byzantium and the Scandina val period’ for the 8th century onwards. For the period vian world in the 9th-10th century: material evi spanning from the 3rd to the 7th century, I will use the dence of contacts. In: L. Bjerg, J. Lind & S.M. Sind commonly accepted term ‘Late Antiquity’. bæk (eds.) From to Varangians: Communication and Cultural Exchange Between the Baltic and the 2 The Greek historian Herodotus mentions a certain Black Sea. Black Sea Studies. Aarhus. 149 – 193. Byzas from Megara as the founder of the city of Byz- Balicka-Witakowska, E. 2010. Constantinople in the antium; the foundation is supposed to have occurred transition from Late Antiquity to the Middle Ages. in 668 BC. In the 1st century BC, Byzantium finally In P.J.J. Sinclair, G. Nordquist, F. Herschend & Ch. came under the control of the Romans. In AD 196, the Isendahl (eds.) The Urban Mind: Cultural and Envi city had the bad luck to support the wrong side in the ronmental Dynamics. Studies in Global Archaeology power struggle between Septimius Severus and his rival 15. Uppsala. 329 – 365. Pescennius Niger. When Severus finally triumphed, Byz- Bassett, S. 2004. The Urban Image of Constantinople. antium was burned to the ground. Septimius Severus, Cambridge. however, soon began rebuilding Byzantium. We know Blid Kullberg, J. 2106. When bath became church: spa that the city was provided with new temples and that tial fusion in Late Antique Constantinople and be the oldest phase of the Hippodrome, the city’s chariot yond. In: B. Shilling & P. Stephenson (eds.) Foun racing track, dates to that period. Byzantium continued tains and Water Culture in Byzantium. Cambridge. to be a medium-sized Roman city until Emperor Con- Kırımtayıf, S. 2001. Converted Byzantine churches in stantine (AD 272–337) moved his seat of power there. Istanbul: their transformation into mosques and mas jids. Istanbul. Kocabaş, U. 2008. The “old ships” of the “new gate” = Ye nikapı’nın eski gemileri. Istanbul. Kostenec J. & Örner A.T. (eds.) 2007. Walking thru Byzantium: the Great Palace region. Istanbul. Magdalino, P. 1986. Church, bath and diakonia in me dieval Constantinople. In: R. (ed.) Church people in Byzantium. Society for the promotion of . Twentieth spring symposium of Byzantine studies. Manchester. 165 – 188. Müller-Wiener, W. 1977. Bildlexikon zur Topographie Istanbuls. Tübingen. Saradi, H.G. 2006. The Byzantine city of the sixth centu- ry: literary images and historical reality. Athens. Striker, C.L. & Doğan Kuban, Y. (eds.) 1997. Kalen derhane in Istanbul. The buildings, their history, archi tecture and decoration. With contributions by A. Berger, N. Göyünç, E.J.W. Hawkins (†), Y. Doğan Kuban, U. Peschlow, C.L. Striker, & G. Traina. Mainz. The Chronicle of John Malalas. Transl. E. Jeffreys, M. Jeffreys, & R. Scott (1986). Melbourne.

(Endnotes) 1 According to the traditional , the 9th cen- tury equals the early Middle Ages in Europe, the Mid-

202 Baghdad and Samarra

Baghdad and Samarra. Imperial capitals of the Abbasid Empire

Andrew D. Petersen

Fig. 1. Map of Abbasid Empire.

The arrival of in the Arabian Peninsula in the of jihad which allowed for the conquest of new terri- first quarter of the seventh century and its rapid tories and also the enslavement of non-believers (see growth and expansion to become both a major reli- 2010). gion and a world empire are unique in world history. There are many explanations for this phenomenon Umayyads and Abbasids however a few factors which certainly played an im- Within the period of a century prior to the rise of the portant part. The first of these is that the were Abbasid dynasty the early Islamic state went through able to defeat two major empires, the Byzantines and a number of changes both in organization and ori- the Persians and in so doing were able to draw on entation which had a profound impact on the sub- the expertise and traditions of both civilizations. Sec- sequent development of Islamic civilization. In the ondly as a proselytising religion the new Islamic state early years during the life of the prophet Muhammad was able to incorporate the diverse peoples of these and his immediate successors , the rightly guided empires to create a cosmopolitan and vibrant society caliphs Abu Bakr (632–634 AD), ‘Umar (634–644 which stretched from the to the In- AD), ‘Uthman (644–656 AD), and ‘Ali (656–661 dus Valley and beyond. A third and often misunder- AD) the Islamic state had remained very much domi- stood aspect of the early Islamic state was the concept nated by the Arab tribes. During the rule of the Uma 203 New Aspects on Viking-age Urbanism, c. 750-1100 yyad dynasty (661-750 AD) the focus of the empire ly moved northwards to Samarra during the ninth shifted northwards to and the Mediterranean. Al- century later returning to Baghdad at the end of the though Arabic was firmly established as the language of century. the religion Greek, as the language of the former Byz- antine rulers, was widely used in administration and Origins of Islamic Urbanism culture. There is also a growing body of archaeological Because of the hybrid nature of the early Islamic evidence pointing to the continuity of Christian wor- state the first new towns built by the Arabs had a ship throughout the former Byzantine territories (see variety of cultural and historical roots (Petersen in for example Najjar & Sa‘id 1994). The establishment press). In the past there was a general idea that the of Damascus as capital of the Umayyad state demon- Arabs had no concept of either architecture or town strated the cultural and geographical direction of the planning and that any town built during the early empire. Whilst this arrangement worked well for the Islamic period were purely derivative of either Ro- western portion of the Islamic state it increasingly man, Persian or other local traditions. However in seemed to be out of step with the eastern half of the recent years there has been a reassessment of these Islamic state which was not only continually expanding view partly in the light of more extensive fieldwork but also producing much of the wealth upon which the in the Arabian Peninsula and partly because of a new Umayyads relied for their extravagant lifestyle. There willingness to accept that the Arabs made a contri- were also questions of legitimacy as well as religious or bution to the early Islamic state beyond the funda- doctrinal issues which became increasingly problematic. mentals of the religion (see for example King 1990). The problems came to a head in the mid-eighth century In considering the origins of Islamic urbanism au- when followers of the murdered caliph ‘Ali began a re- are now willing to include examples of pre-Is- bellion which challenged and finally crushed the Uma- lamic Arab towns such as Umm al-Jimal in Jordan, yyad rulers purging them from the entire empire except Dumat al-Jandal in , Shibam and other from where a fugitive branch of the family was early towns in Yemen as well as early urban centres able to establish itself. such as Malaiha in the (for In place of the Umayyads the Islamic state was now a discussion see Whitcomb 1996; Petersen 2009). ruled by a new dynasty- the Abbasids. Whilst in terms The archaeological evidence from these sites indi- of religious observance the Abbasids were no particu- cate that although the towns were related to other lar improvement on the Umayyads they were at least traditions outside Arabia they did have certain com- seen as having more legitimacy and significantly had mon ‘Arabian’ features. These may be characterised more of an eastern orientation. Whereas the Umayyads as a spatial arrangement based on tribal affiliations, had largely inherited Byzantine/Roman culture the Ab- the absence of town walls (although boundaries in basids took on and developed Persian culture from the the form of ditches may have existed), a generally elaborate court ceremonials to architecture and other non-orthagonal plan and the provision of large scale branches of culture.1 However the most obvious sign communal water sources. The first Arab/Islamic of the new orientation was moving the location of the new towns certainly conformed to this pattern. The imperial capital from the western side of the Syrian des- earliest and most famous Islamic new towns were sert to Mesopotamia () in the east. The first Abbasid the cities of Basra and Kufa both in southern Iraq capital was Raqqa which although still in Syria was on established in the seventh century as garrison cities the river and therefore in direct communica- for Arab tribes engaged in campaigns further east tion with Mesopotamia. Finally a new capital was built in Iran and Khurassan. Although very little archae- at Baghdad, in the centre of Mesopotamia by the 2nd ological work has been carried out in either city the Abbasid caliph al-Mansur in 768 AD. This new loca- literary accounts provide considerable information tion had a decisive effect on the Islamic state not only on their location and layout. Neither city was pro- by establishing the wealthiest capital of the early medi- vided with a wall and instead a ditch was used to eval world but also because it stood at the edge of the mark the borders of the settlement. At the centre Central Asian world and also the vast possibilities of of each city was the governor’s palace, the central the Indian Ocean cultures. For reasons which will be mosque and the market place. The housing space discussed in greater detail below the capital was brief- was allocated on a tribal basis and it is probable that

204 Baghdad and Samarra this reflected the hierarchy of family groupings. With the rise of the Umayyads in the late seventh cen- tury Roman concepts of urban planning became the standard for Islamic new towns. A Roman plan was used as early as the 670’s when the Islamic new town of Ayla was developed alongside its Roman predecessor.2 However under the Umayyads Roman planning was used in many categories of construction from palaces to cities. The derivation of plan of the so called ‘Des- ert Palaces’ from Roman frontier forts is well known (Genequand 2006) whilst the architectural origins of Syrian mosques from Byzantine churches and ulti- mately the Roman also testifies to the strong classical influences on Umayyad towns and cities. Probably the most famous example is the Umayyad city of ‘Anjar in Lebanon which is based on the plan of a Roman legionary fortress. Despite intensive research little is known of the original plan of the Umayyad city of in Palestine although the consensus is that it closely followed classical precedents. Certainly the fact that the city employed much re-used material from the nearby Roman/Byzantine city of Lydda will have ensured that it had a classical appearance.3 The advent of Abbasid rule, or the ‘Abbasid revo- lution as it is known, marked a fundamental change in the character of the early Islamic state (fig. 1). Amongst archaeologist working in the Levant (Syria, Palestine, Jordan and Lebanon) the Abbasid period is generally regarded as one of decline – where the fo- Fig. 2. Iraq. Principal Cities during the ninth century AD. cus of attention shifted irrevocably eastwards leading to economic and urban decline in the western region. Certainly the apparatus of government was no longer from the Round City and the only monuments to sur- focussed on the region and there are few major building vive from Baghdad before the Mongol invasion are some 4 of the buildings on the east side of the river and date to projects that can be dated to this period. In Mesopo- th tamia on the other hand there are numerous examples the early 13 century – examples include the Bab al- of palatial construction the most spectacular of which is Wastani, the Mustansariyya Madrassa and the so called the desert palace of al-Ukhaidhir (Bell 1914). Abbasid Palace. The only part of early Abbasid Bagh- The new Abbasid rulers were aware that they needed dad to have survived is a marble mihrab from the Great a capital in Iraq so that they could be in direct contact Mosque which is now in the Iraq Museum. There have with their original support base in Khurassan and the also been some excavations in the area of West Baghdad garrison cities (amsar) of Basra and Kufa (fig. 2). which have revealed some ninth century housing and ceramics, probably from the area of al-Karkh, but were on too small a scale to greatly add to our knowledge of Baghdad the Round City. In the absence of archaeological information we are forced to rely on the descriptions of near contemporary Archaeology and Historical Sources Arabic writers. Important descriptions of the Round One of the problems discussing the plan and appearance City come from the famous tenth century Persian his- of Abbasid Baghdad is that nothing remains standing torian al-Tabari whilst detailed descriptions of the to-

205 New Aspects on Viking-age Urbanism, c. 750-1100

Fig. 3 Baghdad. Location of the Round City) . pography of early medieval Baghdad are provided by the focus shifted westwards to Syria although in Iraq the ninth century historian al- Ya’qubi († 897AD) and the town of Wasit was built between Basra and Kufa as Ibn Serapion (writing c. 900 AD). Additional material a regional capital and as a means of retaining Umayyad historical information is supplied al-Baladhuri († 892) control of the region. In any case the first Abbasid ruler a famous historian and long-time resident in Baghdad, al-Saffah clearly saw the need to be based in Iraq and al-Muqaddasi a tenth century Palestinian geographer established a base (al-Hashimiyya) near the ancient city and al-Khatib another resident of Baghdad writing in of Anbar on the Euphrates. However during his short the mid-eleventh century (Le Strange 1899; Lassner reign of four years al-Saffah was too pre-occupied with 1970) defeating the remnants of the Umayyads and other op- position to establish a new capital. However al-Mansur, Foundation and Architectural Origins the second Abbasid caliph, had both more incentive, Although the first Islamic capital outside Arabia had more wealth and more political stability to establish been located in Iraq – at Kufa near the present city on a new capital. On first assuming the title of caliph in the Euphrates – this had proved problematic because 754 AD he had based his court near Kufa but found of rebellions amongst the troops and also because the this problematic both because of difficulties with the city had been built as a but rather than local Arab population and because of religious fanatics as a an imperial capital. During the Umayyad period in the region. Al-Mansur’s search for a suitable location

206 Baghdad and Samarra for a new palace and capital took him along the banks shape of the new city were decided with the help of the of the Tigris from al-Jarjaraya in central Iraq 100 miles Persian astronomers Naubakht and Mashallah south of modern Baghdad to Mosul in the north. The (Allawi 1988). The layout was based on a concentric old Persian village of Baghdad was eventually chosen circle with four main gates leading to the centre where primarily because of its location between the Tigris and the ruler’s palace and the main mosque were located (fig. Euphrates rivers which at this point are at their closest 4). The gates aligned to North-West, North-East, South- until they unite in southern Iraq. The same had West and South-East and were known as the al-Sham or been used by the Sasanians who established Ctesiphon Syria Gate (NW), the Khurassan Gate (NE), the Kufa (Ar. al-Mada‘in) as the capital of their Persian Empire Gate (SW) and the Basra Gate (SE). According to the next to ancient Seleucia. Significantly in late Sasanian Arabic sources this city plan was the first ever example and early Islamic times Baghdad was known as souq of a round city and was a testament to the originali- Baghdad and was the location of a famous market. By ty and intelligence of the founder. However as K.A.C. the mid seventh century much of the area was owned by Creswell (1989) and other authors have pointed out Nestorian monasteries and it was presumably relatively there are several Persian and Mesopotamian examples of easy to acquire the land. The precise location of the site cities with a round plan – the most important example was also well chosen as it lay within a sharp bend in the of which is Firuzabad in Iran (Huff 1999). This Royal river which had obvious defensive advantages as well as Sasanian city founded by the emperor Ardashir I in the access to the river as a supra-regional trade route from early third century AD was completely circular and di- North to South (fig. 3). The site was also connected to vided into twenty segments by radial streets which con- the Euphrates river by the Sarat enabling commu- verged on the imperial palace at the centre. Although nication with Syria and the West. the town had dwindled into insignificance by the early Once the site had been chosen work on the new Islamic period its similarity to Baghdad is striking and capital began in 762 AD. The precise dimensions and is the most likely source of Baghdad’s Round Plan. The

Fig. 4. Baghdad. Plan of the Round City 207 New Aspects on Viking-age Urbanism, c. 750-1100 round shape also has considerable cosmological signif- icance representing itself as the centre of the earth. In addition to its symbolic significance a round plan had several advantages over the rectangular plans favoured by Roman, Byzantine and Umayyad city planners. Firstly a circle is able to enclose a much greater area for the length of wall required than a typi- cal rectangular plan and therefore in terms of materials was considerably cheaper to build. The round shape also had defensive advantages as there were no weak corners to defend and each point of the fortifications was equidistant from the centre and so could easily be reached.

Growth and Development The Round City was formally known as Madinat al- Salam (Ar. City of Peace) to distinguish it from the pre-Abbasid Persian village of Baghdad which had pre- viously occupied part of the site. In addition to the caliph’s palace and the central mosque the Round City originally included housing and markets for merchants indicating that it was intended to be a real city rather than just an enormous fortified palatial compound. The markets were located in the arcades of the radial streets and were arranged according to their different crafts. However soon after completion of the Round City the markets within the city were moved outside to the area of al-Karkh as they were regarded as a security threat to the caliph and his advisors. The al-Karkh area was lo- cated immediately outside the Basra gate was laid out as a suburb on the orders of the caliph al-Mansur although Fig. 5. Samarra. Plan from aerial photographs courtesy of A. it had not been included in the original city plan. This Northedge immense area had formally been mostly arable land and became the main residential and commercial area during of Science). During the rule of the later Abbasid caliphs th th the reign of al-Mansur. During the reign of al-Mahdi the (12 –13 centuries) the city on the east side was protect- east bank was developed and became known as Mu‘ask- ed by a large ditch and a wall with four gates one of which er al-Mahdi. During the ninth century the city contin- survives (Bab al-Wastani). Although Baghdad ultimately ued to flourish and grow even though between the years re-emerged after the Mongol conquest of 1258 it never 836–892 AD the caliph and his entourage moved north recovered its former glory. At its height in the ninth cen- to a new capital at Samarra (see below). The significant tury it is estimated that Baghdad was one of the largest change was that when the caliph al-Mu‘tamid returned cities in the world with a population of 900,000 surpass- from Samarra he established himself on the east bank of ing that of Constantinople. the Tigris which has remained the centre of the city till Samarra the present day. Although during the tenth and eleventh centuries the city lost its political significance it was still Samarra is in many ways the opposite of Baghdad. Bagh- a major cultural and scientific centre for the Arab world dad has continued to flourish as a thriving prominent examples of which were the famous hospi- despite numerous assaults and including attacks th tal Bimaristan al-Aduli and the Dar al-‘Alim (Ar. House by both the Mongols in the 13 century and the Amer-

208 Baghdad and Samarra

Fig. 6. Samarra. Great Mosque with spiral minaret. icans in the 21st century – Samarra on the other hand on other occasions. Other historical sources include the was established fairly quickly but only functioned as a late ninth early tenth century author al-Hamdani and capital for less than sixty years and ultimately shrunk various sections in the geographical works of the 13th to a fraction of its former size today resembling a small century Syrian author Yaqut. rather than an imperial capital. However Whilst the historical sources are plentiful the archae- Samarra’s lack of success as a major city has been of great ological evidence for Samarra is overwhelming. The first benefit to archaeologists who have been able to explore systematic excavations at Samarra were carried out over its vast ruin field extending for over 45 kilometres from a century ago by the German expedition to Samarra but North to South and a width of up to 10 km: Samarra were interrupted by the First .5 Since then is recognized as the largest single archaeological site in there have been a number of smaller scale excavations the world. mostly during the 1970’s and 1980’s. Probably the most significant contribution to the study of Samarra as a city is the analysis of RAF aerial photographs and mapping Archaeology and Historical Sources carried out by A. Northedge since 1983 (fig. 5). This As with Baghdad there are plentiful historical sources work has been able to reconcile the surviving remains the most complete of which is the account of al-Ya‘qu- with historical accounts to provide a detailed descrip- bi. In fact Ya‘qubi’s description of Samarra seems so tion of the historical topography of the Abbasid city. detailed and vivid that it seems likely that he actually lived in the city for a period of time in the ninth cen- tury (Northedge 2007: 30). Other authors who pro- Foundation and Architectural Origins vide useful descriptions include the famous historian The reasons for the move from Baghdad and the foun- al-Tabari who visited Samarra in 863 AD and possibly dation of Samarra as a new imperial capital in 836 AD

209 New Aspects on Viking-age Urbanism, c. 750-1100 are well known and recorded by many of the early Arab as the congregational mosque markets and main streets. writers. For example al-Yaqubi records that there were One of the principals in the planning of the city was continual skirmishes between the caliph al-Mu‘tasim’s the separation of the Arab population from the Turkish troops and the local population of Baghdad during the troops. This was enforced spatially and also by regula- early eighth century. tions which made it illegal for Turkish slave troops to “When these barbarian would gallop around on marry from amongst the local Arab population. horseback, they would crash into people left and right, so the rabble would pounce on them and kill some of them and beat up other. [The Turks’] blood could be shed with Growth and Development impunity; nothing was done against those committing these Although the construction of Samarra began in 836 AD deeds. This weighed heavily on al-Mu‘tasim and he decid- the government consisting of officers and notables did ed to leave Baghdad” (al-Yaqubi 1892:255–256 cited in not move there until 838 AD. According to al-Ya‘qubi Gordon 2010:72). when al-Mu‘tasim died it had been assumed that the The troops were part of a Turkic slave army which city would be dismantled. The fact that the new caliph the caliphs had begun accumulating as personal body al-Wathiq (842–847 AD) chose to remain in Samar- guards and as internal security to counter the problem ra encouraged further investment an expansion of the of Arab tribal conflicts and rebellions. The site for the city to the south and the north. However the biggest new capital was also on the Tigris but this time it was expansion of the city took place under al-Mutawakkil located on the east bank in the vicinity of a Sasanian ir- (847–861 AD) who not only expanded the existing city rigation canal known as Qatul al-Khisrawi dug between but also chose to add a completely new area to the north 531 and 578 AD. Although the site was on the east known as al-Mutawakiliya. Probably the most famous bank the Qatul al-Khisrawi provided some level of pro- building in Samarra is the great Mosque built by al-Mu- tection from the east and marked a boundary to almost tawakil to the east of the original city. The mosque is of all of the urban development. On the other hand the immense size (239 x 156 metres) with 17 aisles and at Tigris did not form a barrier to the development and the north end a spiral minaret 52 metres high (fig. 6). several palaces are located on the west bank overlooking This was an immense building project which employed the main city. elephants for moving stone to build the fountain in the Unlike Baghdad the site chosen for the new imperial central courtyard. The walls of the mosque were lined capital was not generally regarded by contemporaries as with sheets of blue glass whilst the roof was supported a good choice and it has been suggested that the main on fired brick piers with marble columns at each corner. reason for choosing the site was that the caliph liked the In Mutawakkil’s new suburb in the north (Mutawak- area for hunting (see Northedge 2007:98). The area had kiliya) he built a slightly scaled down replica (214 x 135 been famous for hunting during the Sasanian period metres) of the Great Mosque known as the mosque of and there are the remains of a large pre-Islamic hunting Abu Dulaf complete with a spiral minaret 34 metres enclosure to the north of the Abbasid city. high. The model for both minarets may have been the Unlike Baghdad which was based around the con- stair tower in the centre of Firuzabad.6 cept of a Round City there was no central geometric or One of the notable features of Samarra is the pres- symbolic principal to the layout of Samarra. The layout ence of a number of enclosures and race tracks. The of the city was articulated by the course of the Tigris enclosures were probably mostly used for hunting al- and a central street or avenue Shari‘ al- A‘zam which though some may also have functioned as recreational ran North to South roughly parallel to the river. The parks. The race tracks have been studied in detail by main components of the city were palatial complexes Northedge who distinguished three distinct race cours- and numerous military camps providing accommoda- es all located to the east of the city opposite the caliphal tion for the Turkic troops and their families. Whilst the palace (Northedge 2007:152-162). caliph himself took of land allocations he left it to his generals and followers to build their own palaces possibly with a grant to finance the construction. It is Conclusion probable that the caliph was responsible for the con- In many respects Baghdad and Samarra appear to be struction of his own palace, the Dar al- ‘Amma as well two very different forms of city- one centralised and

210 Baghdad and Samarra commercially active (Baghdad) and the other (Samar- ra) a sprawling complex of palaces with little to sustain it apart from the presence of the caliph and his court. However there are some strong similarities not the least of which is the immense scale of construction in both cities which were purposefully built as imperial centres. The other feature shared by both cities is the concern with security and the increasing need to isolate the ca- liph from the general population. At Samarra the streets were purposefully designed as wide avenues to prevent the encounters between the mounted Turkish troops and the local population. It is also probable that the spiral minarets at Mutawakkil’s two mosques, the Great Mosque and the Abu Dulaf Mosque, also had a security function and could be used as over the city – a similar function is suggested for the square tower minaret at Ramla (see Cytryn-Silverman 2008:385– 386). The murder of Mutawakkil and the return of the capital to Baghdad in 892 AD mark the beginnings of the collapse of the Abbasids as an imperial power. This decline also marks the end of the unified Islamic state which had existed since the time of Muhammad in the seventh century. From the tenth century onwards the Islamic world would increasingly be subdivided into competing emirates and sultanates sometimes coalesc- ing into larger units but never regaining the unity and power of the early Islamic state.

Corresponding Author Andrew Petersen School of Archaeology, History and , University of Trinity Saint , United Kingdom [email protected]

211 New Aspects on Viking-age Urbanism, c. 750-1100

Starkey (ed.) Proceedings of the Seminar for Arabian References Studies 39. Oxford. 307–320. Allawi, I. 1988. Some Evolutionary and Cosmological – in press. Creole Identity and Syncretism in the Archae Aspects to Early Islamic Town Planning. In: M.B ology of Islam. In: T.A.R. Clack (ed.) Archaeology, Sevcenko (ed.) Theories and Principles of Design in the Syncretism, Creolisation. Oxford. Architecture of Islamic Societies. Massachusetts, 1988. Whitcomb, D. 1996. Urbanism in Arabia. Arabian Aga Program for . 57 – 72. Archaeology and Epigraphy 1996 (7). 38 – 51. Bell, G.L. 1914. Palace and Mosque at Ukhaidhir. A – 2006: The walls of early Islamic Ayla defence or sym Study in Early Mohammadan Architecture. Oxford. bol. In: H. Kennedy (ed.) Muslim Military Architec Creswell, K.A.C. 1989. Early Islamic Architecture. Re ture in Greater Syria from the Coming of Islam to the vised and updated by J.W. Allan. Aldershot. Ottoman Period. Leiden. 61–74. Cytryn-Silverman, K. 2008. Three Minarets in Ramla. Studies in Arabic and Islam 2008 (Endnotes) (35). 379 – 432. Genequand, D. 2006. Umayyad Castles: The shift from 1 For a full discussion of the Abbasid regime see Kenne- the Late Antique Military Architecture to early Is dy 2004. lamic palatial building. In: H. Kennedy (ed.) Muslim Military Architecture in Greater Syria from the Com 2 For a discussion of Roman and Islamic Ayla see Whit- ing of Islam to the Ottoman Period. Leiden. 3–25. comb 2006. Gordon, M.S. 2010. Preliminary Remarks on Slaves and Slave Labour in the Third/Ninth century ‘Abbasid 3 For a reconstruction of the Umayyad city of Ramla see Empire In: L. Culbertson (ed.) Slaves and Households Luz 1997. in the . Oriental Institute Seminars No.7. Chicago. 71 – 84. 4 The al-‘Anaziyya water at Ramla are an excep- Huff, D. 1999. headword “Firuzabad”. In: E. Yarshater tion to this general rule see Creswell 1989. (ed.), Encyclopaedia Iranica IX. Winona Lake. 633 –636. 5 The final report has only recently been published as Kennedy, H. 2004. When Baghdad Ruled the Muslim Leisten 2003. World: The Rise and Fall of Islam’s Greatest Dynasty. Cambridge. 6 For a description of the tower see Huff 1999. King, G.R.D. 1990. Creswell’s appreciation of Arabian architecture. Muqarnas 1990 (8):1. Leiden. 94 – 102. Lassner, J. 1970. The Topography of Baghdad in the Early Middle Ages. Detroit. Leisten, T. 2003. Architecture. Final Report of the first campaign 1910-12. Excavation of Samarra 1. Bagh dader Forschungen 20. Darmstadt. Le Strange, G. 1899. Baghdad During the Abbasid Ca liphate. Oxford. Luz, N. 1997.The construction of an Islamic city in Palestine. The case of Umayyad al-Ramla. Journal of the Royal Asiatic Society 1997 (7):1. 27 – 54. Najjar, M. & Sa‘id, F. 1994. A new Umayyad Church at Khild-Amman. Liber Annuus 1994 (44). 547 – 560. Northedge, A. 2007. The Historical Topography of Sa marra. Samarra Studies I. London. Petersen, A.D. 2009. Islamic Urbanism in eastern Ara bia: the case of the al-‘Ayn-al-Buraymi oasis. In: J.

212