The First Sparks and the Far Horizons Stirring up the Thinking on the Earliest Scandinavian Urbanization Processes – Again

BY MATS MOGREN

Mogren, Mats. 2013. The First Sparks and the Far Horizons. Stirring up the Thinking on Abstract the Earliest Scandinavian Urbanization Processes – Again. Lund Archaeological Review 18 (2012), pp. 73–88. The author stresses the need for new input into research on Early Medieval (6th–11th century) urbanization processes in . A multi-pronged approach is suggested, where the societal structure of the region and period must be re-evaluated using hitherto relatively untried theoretical tools, such as a heterarchical perspective, the heterogenetic- orthogenetic conceptual pair, and a deconstruction of the bewildering concept of central place. It also includes a necessary critique of the “power paradigm”, current for 30 years. It is further stressed that comparative material should be sought globally. The Swahili coast of East Africa and maritime Southeast Asia are presented as the most operative comparison regions, due to a number of similar traits shared by the three areas. Network building with the chosen areas aims at a shared future understanding of the causes of the first sparks of urbanization. Mats Mogren, The County Administrative Board of Scania, 205 15 Malmö. mats.mogren@ lansstyrelsen.se. Department of Archaeology and Ancient History, Lund University, Sandgatan 1, 223 50 Lund. [email protected]

The paradigmatic nature of research someti- questioned. Such a paradigm, or at least a line mes bears a certain resemblance to the labours of thought with certain paradigmatic features, of Sisyphus. An understanding of a field of is the idea that the societal elite was decisive inquiry is gradually built up through lots of for the emergence of the earliest urban centres individual and collaborative work until it ac- in Northern Europe. This understanding is quires the shape of a paradigm. After a while part of a wider one, seeing society as a whole the debates enter a lull, most participants ack- during the period in question as explicitly hie- nowledge the understanding as valid, and Si- rarchical. syphus himself, who has just reached the crest The line of thought was launched among of the hilltop, has a discreet smile on his face, historians in the 1970s and gathered conside- thinking that perhaps this time the task is ac- rable momentum within archaeology in the complished. That is the appropriate moment 1980s after the publication of a number of in- for Zeus, if he values the dynamics and fresh- fluential dissertations and articles by historical ness of research, to push the stone downhill archaeologists in Lund and by the members again. of the research group connected to the project The paradigms sometimes seem to have Fra stamme til stat (Mortensen & Rasmussen an astonishing longevity, without really being 1988, 1991) in , to mention just a

The First Sparks and the Far Horizons 73 few of the most important advocates. Still, sites like Hodde, Vorbasse, and others may many take the hierarchical society and the ac- have had a population of up to 200 inhabi- tive role of the societal apex more or less for tants at any one time, which is equivalent to granted. That situation has to be stirred up a many smaller towns with charters from the bit. Late Middle Ages and the Early Modern pe- So much qualitative research has been car- riod in Scandinavia. Certainly, just a few non- ried out on the societal structure and espe- agrarian households will not make an urban cially the Early Medieval1 aristocracy during setting, so scale has some importance, but ur- the last 30 years or so, that any criticism may banity cannot be quantified in this way.I t is a seem utterly quixotic, but there are several as- very qualitative concept. pects of the paradigm that need questioning. For reasons of the comparison methodo- Three main aspects of early urbanization logy chosen in our study we also want a con- are at issue here. First of all, there is the con- cept that is operative outside of our limited cept itself and its temporal parameters and cultural setting. Urbanization as a concept trajectories in the Scandinavian context. Se- may sometimes seem elusive if we strive for condly, we must once again discuss the rela- a globally valid definition, and many resear- tionship of urbanization to power and societal chers have pointed out that trying to find the structure. And finally, we must find a new way lowest common denominators for phenome- forward. na called urban in widely different parts of the world is futile, but I maintain that we must keep trying; we must finally rid ourselves of An elusive concept the spectre of Max Weber before we can leap forward. This is also the reason why I avoid This is not the place to repeat the well-known criteria lists, which more often than not are ut- history of definition attempts in urbanization terly Eurocentric. Towards the end of this text research. I trust that readers are well acquain- an operative level of comparison is suggested. ted with this more than century-old discus- For these reasons, my own definition is sion. Suffice it to say that the main distin- kept simple. A place or area is becoming ur- guishing features of the discussion have been ban when a substantial number of people move the varying hypotheses on the causes of the together to meet two or more non-agrarian emergence of urban sites and the criteria for needs, in a context of social plurality. The lat- urbanization. ter parameter is of special importance in the Urbanization is not seen here as a mo- Early Medieval period for distinguishing sites nolithic, unilinear concept. We must reckon that are in some respect urban from magnate’s with multiple urbanization processes, with residences, which will be apparent later on in differing characteristics, at work simultan- the text. eously or one after the other. The primary In downplaying morphology and structure parameters are function, with its connected you open up for the possibility of an urban concept of agency, and another factor of pa- site being seasonal and/or agglomerated/poly- ramount importance is communication (e.g. morphous. This is necessary if you are on the Sindbæk 2007a; 2007b; 2008). Parameters quest for the very first ignition of the process such as morphology and structure should be of urbanization in a certain region, the first regarded as secondary. sparks. Population size, on the other hand, cannot Another conceptual straitjacket can be av- have been decisive. Large Danish Iron Age oided by regarding urban sites as not neces-

74 MATS MOGREN sarily permanent features. The reason why we many other publications), and other sites with tend to think about permanence as natural evidence for ceremonial activities provided us and put all our efforts into explaining dis- with an embarrassment of riches never before continuities is of course the High Medieval seen in Scandinavian archaeology, and when stone- and brick-built town and the inertia it one magnate’s residence site from the Early provides, but a more fruitful approach to early Medieval period after the other enriched urbanization would be to turn the discussion our record, then the need to include this in upside-down and regard discontinuities as a a general discussion of the transformation of state of normality and instead try to explain North European society became evident, and relative permanence when it occurs. This imp- the term central place was coined. We have all lies that, while I more or less agree with Ma- used it, but do we know what it means? teusz Bogucki (2010) that the northern trade It may seem that my criticism of the con- and craft sites were not one leap in a long tra- cept is like preaching to the choir. Certainly, jectory from the urban centres of Antiquity a well-informed and creative discussion has to the High Medieval town (hence no “proto- been current for some time (e.g. Näsman towns”), but part of an independent urbaniza- 2011), and I make no claim to originality in tion process,2 I tend to disagree with his state- criticizing the concept, but it is important for ment that they were a cul-de-sac. Interrupted all of us to establish a benchmark by repea- development was the natural state of things. ting the critique. “Central place” is the sort What needs to be explained, again, is the of cautionary term that comes to mind when fact that the establishments of the 10th–12th you understand that you have to widen your centuries lingered on. All this being said, it is scope, but actually do not understand fully evident that, at the end of the day, urbaniza- what you are talking about. No harm in that, tion and urbanity are very abstract concepts. at the outset of the discussion, but when the Maybe that is why we sometimes mix them term continues to be in use one long decade up with other phenomena. after another it will bewilder your thinking. It is high time to deconstruct this concept and instead try to understand what it comprises A bewildering concept and define its parts, at least if it is urbaniza- tion that is our concern. Some decades ago, it became evident that Analytically, urban sites, magnates’ resi- medieval urbanization did not start at the ear- dences and ceremonial/religious centres are liest stratigraphical levels of the still existing three different phenomena. Sometimes, and towns. A plethora of sites popped up that po- indeed quite often, they combine. A magnate’s sed a problem. The long-known sites of , residence can develop into an urban site if pe- , and a few others could pre- ople in the trades and crafts, forming sepa- viously be treated as phenomena apart. Howe- rate households, move in and settle there. The ver when the excavations in Gross Strömken- same goes for ceremonial centres. An urban dorf, , Åhus, Löddeköpinge, Fröjel, centre can acquire a magnate’s residence if a Janów Pomorski and many more all around member of the societal apex moves in. A cere- the southern Baltic provided evidence for monial centre can also have its magnate’s re- long-distance trade preceding the well-known sidence. But to understand the incentives for emporia by at least half a century, when ex- urbanization the three must be kept apart. In cavations in Gudme-Lundeborg (e.g. Nielsen fact, all sites must be seen as unique; it is very et al. 1994), Uppåkra (Larsson 2004, among difficult to categorize 8th–9th-century sites.

The First Sparks and the Far Horizons 75 That may be the main obstacle to attaining The problem with the pyramid is its im- an understanding of the earliest urbanization plications. When society is presented as a py- processes. ramid it is easy to convey the implicit notion The pooling of trade and craft sites, with that it represents society as a whole, and that ceremonial centres and magnates’ residences, the pyramid “covers the map”, meaning that most probably has its implicit roots in the no- the apex can, at least indirectly, control every- tion of a pyramidal hierarchy with no room one beneath. The step from this notion to the for independent actors. idea of “confederations” and trans-regional sovereign kingdoms, maybe even early state formation, is a short one (e.g. Näsman 1991; The top(pling) of the pyramid 1998b). Most probably, everyone outside the royal In Scandinavian first-millennium studies, retinue and vassalage in the Middle Ages felt and especially in archaeology, the hierarchi- the political structure to be an exogenous im- cal, pyramidal understanding of society that plant in their local society. Apex-level politics, bloomed in the eighties and early nineties still and the polity, might perhaps be seen as a stands unchallenged, like dried flowers in an kind of “enterprise”, not as something organic old vase. In the nineties the central places were in society. starting to become differentiated, but only What this research needs is a heterarchi- scale-wise. They were ordered in a three-tiered cal perspective. As Carole Crumley has em- hierarchy; local, regional and supra-regional, phasized, hierarchy is a “pervasive structural but there were few attempts to understand metaphor for order itself”, hence the mista- their diversity as regards function and agency. ken assumption of the pyramid being equal The existence of hierarchies during the to society, but a hierarchy is just one of many first millennium is not an incorrect assump- patterns in which elements can be ordered tion, of course, as a multitude of archaeologi- (Crumley 1995, p. 2). cal finds indicate. A major leap in complexity Heterarchy is not the opposite of hierar- is evident at the end of the 2nd and beginning chy, as some tend to believe. Hierarchy is a of the 3rd century. Chiefly centres appeared, structure, while heterarchy is a perspective. some of them seemingly out of the blue, most With such a perspective it becomes evident conspicuously so Himlingøje on Zealand with that elements in a society can be unranked, its wealth of Roman import items. Major ce- or ranked in various ways, and that the hierar- remonial centres, as in Uppåkra, were created chies that do occur are always parts of a grea- by erecting structures that must be considered ter context. There is order without hierarchy, to be temples. and we must be wary of using hierarchy as a The structure with a societal base, a midd- reductionist metaphor for society at large. The le tier and an apex is very evident in many application of hierarchy as a concept in ar- regions of Scandinavia. It is also shown in chaeology has been totally out of proportion building traditions, where very large houses up until now. The opposite of order is chaos, are found from the last century BC and on. which nobody wants. We must therefore ack- They are generally larger in Scania than west nowledge the possibility of any societal unit of the Sound all through the first half of the having a self-organizing capacity. millennium (Artursson 2008; Carlie 2008), When the hierarchical thinking is applied which indicates a distinct socio-cultural bor- to urbanization studies, its reductionist cha- der during that period. racter becomes very evident. Most of the ar-

76 MATS MOGREN gumentation regarding the emergence of the under-age minors, peace was guaranteed by first trade/craft sites in southern Scandinavia the king, and the law was given to the sub- and elsewhere has built upon the assumption jects. The key concept was munt, meaning that the sites were founded by the societal apex protection. The king protected his subjects, groups, kings no less (e.g. Jensen 1991, p. 80). as parents protect their children, whether they This in turn seems to build on the astonishing want it or not. assumption that commoners are incapable of In Scandinavia there was ascending go- putting huts in straight rows. vernment, meaning that the king ruled with Until Claus Feveile published his very so- the sanction of his subjects. Peace was guaran- bering and clarifying summary of decades of teed by the people and the law was made by archaeological work in (2006), it was them (Ullman 1965, p. 56 f.). At least this was more or less taken for granted that the small the general nominal idea; in reality influential area of lots laid out in a regular pattern in the magnates could certainly manipulate the judi- area of present-day Sct. Nicolaigade, maybe cial system in their favour, but even then they 34−35 lots in two parallel rows on a sandbank had to play according to the rules. Anyhow, I by the marsh (Feveile 2006, fig. 10), was an suggest that this is one reason why we should initiative by some regal actor. This assump- not attempt political analogies with the conti- tion has been taken up wherever there is evi- nent for this period (cf. Näsman 1998b). dence of parcellation. Thus, for the last few The existence of a Scandinavian judicial decades in Scandinavian research, parcellation system, well established already in the 8th and has been seen as valid evidence of elite groups 9th centuries, is evident from the Forsa ring, at the supra-regional level, and an early uni- with its runic inscription telling about a set fied kingdom has been postulated. of legal sanctions against persons who neglect their duties in the maintenance of a vi, a sa- cred site. The ring’s inscription has been ling- The societal mastic uistically dated to probably the 9th century (Liestøl 1979), no later than the 10th,3 and it If everyone favours order, and if we question appears to refer to a well established societal the role of early kingship, what was it that structure, so we may surmise that the judicial created societal coherence and peaceful con- system was considerably older. ditions? We must not forget that Scandinavia The existence of such an “ascending” legal in the 8th and 9th centuries was not part of system is, I believe, also corroborated by Ste- Europe. Europe as a concept got its first de- fan Brink’s detailed analysis of the much later finition in those centuries, being equivalent Hälsinge Law (2010a), where he elucidates to Merovingian/Carolingian Latin Christen- early “folk” layers which are not to be found dom, which was very different in many re- in the Uppland Law, on which the Hälsinge spects. For one, Scandinavia did not share the Law otherwise was modelled. The judicial sys- European system of political thought. tem must be considered to be the mastic that The difference is most easily explained by gave coherence to Scandinavian society, not referring to the judicial sphere and it can be the odd aristocrats, as revealed by archaeology. illustrated by the concepts of descending and How should we comprehend early medie- ascending government. “Europe” i.e. Latin val kingship in Scandinavia? The term “king” Christendom, had descending government, has no other meaning than a societal leader meaning that the king ruled by the grace of who has no sovereign. All other meanings are God, his subjects were treated more or less as contextually dependent. That means that we

The First Sparks and the Far Horizons 77 must be ready to imagine kings without coun- Godfred. We can learn a lot from studying try, and the possibility of multiple kings in a what is known about him. given region, of which there in fact is textual evidence. Early Medieval kingship in Scan- dinavia depended totally on the retinue, the lith, and ultimately on plunder, tribute and What’s your racket? war − perpetual violence. Kingship was pre- datory. Hedeby had its roots in the 8th century, as the In 1988 Thomas Lindkvist launched the archaeological record tells us. Then it mainly distinction between external and internal consisted of what is today called the Südsied- appropriation. There are theoretical stances lung, south and outside of the 10th century in that text that have lost some of the follo- rampart (Hilberg 2009). It was a relatively wing they had then, but this part still has a small, open settlement, maybe a little bit like considerable explanatory value and could be Ribe or Gross Strömkendorf. Hedeby then developed further. The power of the elite be- got a coercive boost by the events of 808 CE, ing built on the appropriation of the primary as they are related in the Royal Frankish An- producer’s surplus seems as good a definition nals. of the emerging state as any. The That year a Danish military force attacked kingdoms lacked the capacity for internal app- and destroyed the relatively large trading sett- ropriation, and most of us see them as chief- lement of Reric in the land of the Slavic Abod- doms in the traditional anthropological sense. rites, and forced the traders there to resettle in In southern Scandinavia kings that were Hedeby. Reric is now generally identified with acknowledged at a supra-regional level are im- Gross Strömkendorf. This is more or less cor- possible to discern before the mid-10th cen- roborated by archaeological evidence which tury and the Jelling dynasty. Before that there gives the years around 810 as the start of the is reliable evidence for a short-lived (about decline (end of building activities) at Gross three generations or slightly more) realm pro- Strömkendorf and a simultaneous increase in bably comprising Funen and south Jutland building activities in the central settlement at during the earlier and middle parts of the 9th Hedeby. The end of Gross Strömkendorf is century (most probably also in the latter half archaeologically set at c. 835 (Hilberg 2009; of the 8th century, judging from the earliest Jöns 2009). datings from Danevirke). It probably also in- I propose that 808 is the iconic event for cluded overlordship over some areas (we know understanding Scandinavian kingship of the about a tributary relationship with Vestfold), period and indirectly for understanding the but it must be regarded as nominal and meant trajectories of the 8th-century trade and craft little apart from the paying of tribute; no real sites around the southern Baltic. If the per- control could be upheld and thus these areas petuation of kingship necessitated perpetual cannot be regarded as parts of the kingdom. violence, these events may be likened to a The Funen-South Jutland realm ultimately protection racket; perhaps it would be better dissolved and left the ground open to vari- to compare Godfred to Al Capone than to ous petty chieftains for probably more than , in order to better understand half a century (cf. an interesting view from a both him and the nature of kingship in gene- philological vantage point by Gazzoli 2011). ral. Godfred probably had heard of the politi- The polity was the stage for one of the classic cal ideas of Charlemagne’s empire, and about figures of early Viking Age Scandinavia, King munt, and thought it was a good idea, but

78 MATS MOGREN he had to interpret and implement what he ast Asian archipelago. The first half of 1857 heard within the confines of his own context.4 was spent in the Aru Islands, just west of New We must understand settlements like Guinea. He describes the seasonal trading site Ribe, Reric, Åhus, and others, as founded by of Dobbo as situated “on the small island of self-organizing groups of traders and crafts- Wamma, upon a spit of sand which projects men. This assumption is shared by other out to the north, and is just wide enough to researchers such as Callmer, who states that contain three rows of houses” (Wallace [1869] no aristocratic presence is discernible in the 1994, p. 432). archaeological record of several Scanian sites His description of Dobbo immediately (Callmer 1998b), and by Carlsson, who has brings to mind our earliest Scandinavian and investigated several sites around the coasts of West Slavic trade and craft sites of the first Gotland, all of which lack traces of apex struc- half of the 8th century, Åhus, Gross Ström- tures (Carlsson 2011). kendorf/Reric or Ribe. There was also seasonal These settlements could have continued occupation on sandbanks or in river estuaries. on a self-organizing basis for maybe two or th- But there is more to Wallace’s observations ree generations. Later apex groups, when they than just topography. became aware of the source of wealth possible He describes how at his arrival in January to tap at those sites, elbowed in, just like the only a handful of families occupied the simple Mafia, and created their own interpretation of houses, but with time passing more and more munt. people came, repaired their houses and started It has been postulated over and over again trading with each other and with the local po- in the literature that trade must be protected pulation. Some of them brought commodities by power. This has in fact been regarded as one mainly for the trading community itself. After of the cornerstones not only of the criteria for a few weeks the site had lively activity which urban sites but also for the factors behind the lasted until May when activity trickled off, the emergence of the sites (cf. Blomkvist 2001). traders started to leave and soon the site was Certainly, trading sites became “protected” by abandoned again. The traders who came and power after awhile, not in any benign, fatherly went were of different origin, from all over the sense of the word but as a racket. I propose Southeast Asian archipelago: Chinese, Bugis, that urban sites did not need the protection of Javanese, Ceramese, Timorese and others; power, but they could be struck by it. Dobbo was truly a poly-ethnic society. Now we have put the kings in place, but Perhaps the most astonishing fact about the question about what created order in the this site was that there was no power to keep earliest trade and craft sites remains to be order. No police, no military units, in spite answered. If we postulate that the presence of of the presence of pirates in the waters sur- some kind of power is not a necessity, are the- rounding the islands. The Dutch colonial go- re any sources indicating that we are justified vernment sent a commissioner from Amboina in doing so? Yes, in fact, there are. But this once a year, at best, to hear complaints, and requires us to look far beyond the horizon. the native chiefs lived far inland and showed no presence at Dobbo. The leading traders formed a court ad hoc to condemn thieves. Trade is the magic… The small heterogeneous and seasonal com- munity of Dobbo was self-organizing. Wallace Alfred Russel Wallace travelled and studied also finds this astonishing, and towards the for eight years (1854–1862) in the Southe- end of the account of his stay he provides us

The First Sparks and the Far Horizons 79 with his interpretation for this situation: by the anthropologists Redfield and Singer around six decades ago, and has in our time Trade is the magic that keeps all at peace, been taken up by John N. Miksic in his work and unites these discordant elements into on Southeast Asia (Miksic 2000). Orthogene- a well-behaved community. All are traders, tic sites are associated with stability and ritual. and all know that peace and order are es- They are placed in areas of surplus agricultural sential to successful trade, and thus a pu- production, are always in some respect cen- blic opinion is created which puts down all tral, are populated by a civil, religious and mi- lawlessness (Wallace [1869] 1994, p. 444). litary bureaucracy, might have an aristocratic/ royal presence, and are distinguished by some Wallace’s description has caught the imagina- form of monumentality, so they are usually tion of researchers like Miksic as being a role easy to discover. model for early urbanization in 10th–14th- Heterogenetic sites are associated with century Southeast Asia, and it certainly has change and entrepreneurship. They are often an interest for us in the Scandinavian 8th- liminally placed, between ecological zones, century context. It may suffice to remind the ethnic or political regions, or differing trans- reader what Callmer writes about the coastal port systems. They often lack monumentality, sites of first-millenniumS candinavia. He finds which can make them archaeologically invi- it reasonable to assume that trading sites also sible on the surface. Production and trade of in fairly close proximity to aristocratic centres commodities is the most distinguishing feature. were difficult to control entirely, that trading The two categories are actually opposite sites had their own life, their own meanings poles on a continuous scale; most sites have and rationality, and that they were not neces- traits of both. As is evident from the discu- sarily linked to power centres at all (Callmer ssion above, this conceptual pair has a very 1998b, p. 35). high explanatory value also outside of Miksic’s Southeast Asian context, and not least for Early Medieval Scandinavia. In trying to di- The far horizons sentangle the very unsatisfactory bundle of elements of the central place concept it comes This brings us to the widening of the hori- in very handy. zons, so necessary if we are going to find a In choosing comparative material we can- way forward from the treadmill of rehashing not pick it at random. Many attempts to our own limited data over and over again. We study urbanization comparatively on a global must break out of the confinement of a nar- scale have ended up in a mere catalogue of row North European sphere in making our case studies that really do not relate very well comparisons (cf. Andersson 2009). TheS can- to each other. But if societies that share se- dinavian discussion could surely benefit from veral traits are chosen for comparison, as in a better familiarity with urbanization research the case elaborated on below, advances can in other parts of the world. be made in the general understanding of the First of all, concepts used in one setting emergence of urban sites, and the widely dif- can have a very high explanatory value in fering geographical setting can be used to our another. Such is the case, for researchers on advantage. Medieval Scandinavia, with the conceptual We must look beyond the ancient river pair orthogenetic and heterogenetic. civilizations and their successor states, i.e. This analytical dichotomy was launched search for areas without earlier urban tradi-

80 MATS MOGREN tions (but exposed to influence from such are- present, but the search is on (Chami 1994; as). We must look for areas with similar socio- 1999). structural and technological conditions, areas During the 7th and 8th centuries new ur- that were part of long-distance trade networks ban sites emerged along the central parts of and preferably also subject to fundamental the coast that were oriented towards long-dis- ideological shifts, like Scandinavia. tance trade, sites which are pre-Islamic, lack- Two such attempts are being made at pre- ing masonry architecture and having disconti- sent. At the universities of York and Århus, a nuous trajectories. Good examples are Unguja group of researchers is building a network of Ukuu in Zanzibar (Juma 2004) and Tumbe research contacts in Europe, Africa and Asia, in Pemba (LaViolette & Fleisher 2009). Se- known as the Entrepot Project.5 At Lund Uni- veral other sites are still awaiting the archa- versity, another group of researchers is doing eologists. Later on, during the 10th and 11th exactly the same , and until recently centuries, the coast was gradually islamicized, these two groups worked without knowledge masonry architecture with sometimes monu- of each other. The important fact about this mental characteristics started emerging and parallelism is that it makes it evident that the during later centuries stone built cities with time is ripe for a general understanding of the mosques, palaces and city walls were created, necessity for widening horizons. from Mogadishu in Somalia to Mahilaka in The work in Lund has taken shape as se- Madagascar, representing the classic Swahili minars on the issues covered here, but also urban culture. as network building with colleagues in East The parallelism with Scandinavia is very Africa and Southeast Asia. A triangle is taking evident. The only major difference is that in shape with Lund, Dar-es-Salaam and Singa- Africa the Medieval urban centres became pore at its corners (and naturally with a link independent city-states, while in Scandinavia to Entrepot). Why these three? they became integrated into larger territo- It is not only a case of contemporary urba- rial kingdoms. Research regarded these sites nization processes (the Scandinavian and East for a long time as exclusively detached from African processes are almost entirely contem- the surrounding countryside, but modern re- porary, while the Southeast Asian ones seem search with extensive survey work has proved to be slightly later, but fully comparable). the opposite (Fleischer 2010; Wynne-Jones The three areas have early urban sites sharing 2007). many traits. Partly dependent on the hetero- In the maritime parts of Southeast Asia genetic nature of the early sites, they all dis- (the Austronesian cultural area), urban sites play marked discontinuities, all three areas are related to long-distance trade generally, and exposed to contacts with areas of earlier urban following Anthony Reid (1988–1993), have development, all three have experienced pro- been regarded as a very late phenomenon oc- cesses of religious change, and all three areas curring during the 15th to 17th centuries. are markedly maritime, with advanced ship However it has recently been claimed by se- technologies. veral researchers (Christie 1998; Miksic 2000; Along the East African coast from Soma- Wade 2009) that there was an earlier age of lia to Mozambique, an area which later expe- commerce in Southeast Asia, during the 10th rienced a process of ethnic genesis resulting in to 14th centuries, and, in fact, evidence of the Swahili culture, trading sites are known long-distance trade goes back to the centuries from Roman Antiquity. The earliest ones are around the beginning of the common era (Ar- archaeologically known only in Somalia at dika & Bellwood 1991).

The First Sparks and the Far Horizons 81 Several sites with heterogenetic traits have ourselves: why the 8th century? Would it not been presented by archaeology in the area, have been more natural for trade emporia to such as Tumasik (present-day Singapore), emerge during the first half of the millenni- Kota Cina (literally the stockade/fort of the um, when there was a major trading partner Chinese), Ladong and Barus (a craft and tra- in the south with whom to exchange goods de entrepôt with wide-ranging connections on a large scale? So far only Lundeborg is well all over the Indian Ocean and all the way to known from that period, a site that appears Egypt), all in Sumatra, and Banten Girang in anomalous in its context.6 Java. The post-Roman era was a period of re- There was also the trade “empire” of Srivi- lative peace, while the very troubled Roman jaya, notoriously difficult to study for reasons Iron Age was not, as Ulf Näsman has shown of topography as well as building technology, so eloquently in a simple diagram (Näsman but also because it differs very much from sta- 1998, p. 261). But the “war indicators” in te structures in India, China or Europe; it was South Scandinavia had been at a very low le- probably more of a network of trading sites, vel already from and including the 6th cen- acknowledging a nominal overlordship, rather tury, so peaceful conditions cannot be the than a territorial polity. Inscriptions indicate triggering cause, only a precondition. In the that its capital was at present-day Palembang same text Näsman also stresses the important in Sumatra. The sphere of influence of this fact that around the start of the 6th century thalassocracy or trade network comprised are- the patterns of exchange of luxury goods shif- as from southern Thailand to Java. ted from the Carpathian-Black Sea region to Working with this rich material within a the Rhineland-North Sea region, due to poli- comparative framework will certainly elucida- tical unrest and ethnic transformations in the te the urbanization processes of all three areas. former area. One implication of this is that a Logistics are made easier by the fact that the mainly overland and river-based communica- University of Dar-es-Salaam works as a hub in tion system had to be replaced by a seafaring an African network for archaeology, previous- system, which must have put emphasis on de- ly sponsored by the Swedish agency Sida, and veloping nautical technology. many East African archaeologists know each Knowledge about the development of other personally and have long-standing con- nautical technology has taken enormous tacts with Swedish researchers. Researchers strides forward with the ship finds especially in the Entrepot Project have also been active from Skuldelev and Roskilde but these ships along the Swahili coast. In Southeast Asia, the belong to a later age, from the 10th century National University of Singapore has worked, onwards. For the earlier periods very much more informally, as a similar hub for archaeo- of our understanding is still built on the logy from Burma to Indonesia. Prospects for three-stage development represented by the the future could be worse. Hjortspring boat (a paddled, sewn canoe) of the Pre-Roman Iron Age, the Nydam boat (a rowed, riveted boat) of the Late Roman Iron The first sparks Age, and the grand Norwegian ship burials at Oseberg and Gokstad (the iconic Viking ships All three of these areas are well suited for stu- par excellence, propelled by sail) of the 9th dying the causes of early urbanization, the and 10th centuries. Many ship finds on both ignition of the first sparks. Returning to the sides of the Baltic have been investigated in re- Scandinavian context, we are inclined to ask cent years (see a good overview by Indruszew-

82 MATS MOGREN ski 2005) and several replicas have been built demand. What had Scandinavians to offer in and evaluated at sea. Detailed knowledge of the greater European trade system? And what hull-building techniques has been obtained, did they want from abroad? It is well known but until now the most important question from historical anthropology that supply and related to our concern here, the introduction demand can alter society at large. The de- of sail in Scandinavia, has not been addressed mand for European commodities changed the at all (Indruszewski 2005), or answered only economy of the local populations in North within a very wide time frame (Crumlin-Pe- America and the supply of the goods chan- dersen 1999). It seems to have been sufficient ged political and military patterns. The same to frame the period of introduction by refer- happened in West Africa, in the South Paci- ring to the early and late Gotlandic picture fic and elsewhere, and the European demand stones; the former from the 5th century have for Chinese silk totally transformed not only vessels without sails, the later from the 8th the economy but also the landscape of South century have vessels with sails. China in the 16th century. Why is the introduction of sails so im- It has long been understood that Early portant to us? Simply because in a rowing Medieval trade communities in the Baltic boat of the Nydam type very little cargo can were multi-ethnic. The presence of Frisian tra- be carried; the crew and their provisions take ders has been suggested (cf. Callmer 1998a), up most of the space (see the discussion by but also the , and Scandinavians Englert 2009 for a later period). Urbanization themselves were active traders. If we have a built on trade requires a certain volume of notion of who was supplying, we must also commodities to get out of the starting blocks, give those demanding a bit of thought. Was so the sails are important as markers. it only the local aristocracy? Who wanted the In a forthcoming dissertation Ulla Isabel complete dress outfits produced at Åhus in Zagal-Mach Wolfe narrows the date of in- the 8th century? And furthermore, what did troduction considerably. She maintains that these demanding Scandinavians have to offer the introduction of sails had a tentative be- in return? Furs, as in North America? Or were ginning at the very outset of the 8th century, these sites actually turned into slave markets, and that the really large sails emerged during as in West Africa, when they had nothing else the 9th century together with a considerable to sell? By formulating relevant questions like development in sailing techniques (Wolfe these we may proceed towards a better under- pers. comm. & 2013). The beginning coinci- standing of the processes. des with the establishment of Ribe, Åhus and Gross Strömkendorf/Reric, so definitely there must be a connection. You do not move what is sacred The problem is that the peoples of the Bal- tic must have known about sailing ships long Finally, the question of permanence must be before that date, but without taking up the addressed. Uppåkra existed for more than a technology. Thus there must be another factor millennium, from the 1st century BCE to that caused them to apply the technology in the 11th CE. At the end of the 10th century their own home waters, one which also lay be- Lund was emerging as a competitor just two hind the emergence of the beach and estuary kilometres away, and now Lund has had per- trading sites of the early 8th century. manence for a millennium. During the 11th One starting point is of course to consi- century the situation in the area was charac- der the classic binary concept of supply and terized by five sites with central or even urban

The First Sparks and the Far Horizons 83 features, Uppåkra, Lund, Dalby, Lomma and relics of named saints, plus a number of na- Borgeby/Löddeköpinge, and it was in the ba- meless relics and 79 relics of other types. By lance which was to live on and which was to comparison it may be mentioned that Uppsala fade away. We know in retrospect that Uppå- cathedral had 64 named relics and a number kra after c. 1100 CE was a mere agrarian vil- of unspecified ones. The Lund collection was lage and that Lund from that date developed not quantitatively within the range of some into the Metropolis Daniae, a major royal and continental relic collections, like the ones in ecclesiastical centre, but avoiding teleologi- Halle, Braunschweig or Prague which num- cal reasoning we must try to understand the bered thousands, but in the Nordic context it causes of the development. This is in itself a was outstanding. The cathedral of Saint Law- formidable research task, which has only very rence claimed to possess various relics con- recently been tentatively addressed (Sjöbeck nected to Jesus Christ and Saint Mary, Saint 2012), and we may surmise that the causes John the Baptist, Saint John the Evangelist, were multiple, but one of the probably deci- assorted parts of the 11,000 virgins and even sive ones has been indicated by research el- a part of the tree of paradise, not counting sewhere. numerous saints that are less well-known to Frans Theuws (2004) points out the im- latter-day readers (Axel-Nilsson 1989). These portance of inalienable possessions. He stres- were inalienable possessions, indeed. ses the quality of markets as “total social events”, where three main spheres of exchange share the same space, the exchange of gifts, Countervailing the canon the exchange of commodities and the keeping of inalienable possessions, which are mostly This article has presented no new mate- of a sacred nature. In comparing rial. It is rather a crude sketch of a framework and Maastricht, the former having a relatively for future research which aims at showing al- short span of life, while the latter is still in ternatives to the prevalent canon. It has pro- existence, Theuws notices the lack of inaliena- posed a number of theoretical preconditions ble possessions in Dorestad. The possessions that oppose the orthodoxy of Scandinavian brought to Maastricht, relics of Saint Serva- early urbanization studies: tius, on the other hand, provided the “cosmo- logical authentication” necessary for creating • Function, agency and communication are permanence. the important elements in defining urban The material presented by Theuws can ser- sites. ve as an excellent model for explaining the ex- • There are many paths to urbanity; sites are treme longevity of Uppåkra as well as the site’s different, also within the broad analytical demise and the rise of Lund. In Uppåkra a categories used here. pagan temple which was rebuilt several times, • Discontinuity is the rule; permanence re- but representing permanence on the same site quires explanation. for at least 700 years, must have filled the re- • The concept of central place must be repla- quirement of “inalienable possessions”. When ced by a more exact set of definitions. the religion ultimately changed, Uppåkra had • Urban sites can emerge acephalously; kings lost its quality of inalienability. are not a requirement. Permanence was created almost instantly • Communities of traders and craftsmen are at the new site, where the cathedral of Saint perfectly capable of self-organization. Lawrence was provided with no less then 417 • Kings and other members of the aristo-

84 MATS MOGREN cracy played a role mainly as tappers of of the Mafia (camorra, cosa nostra, ’ndrang- the wealth, through “protection rackets”, heta, etc.), that are shared by Early Medieval which conforms well with the concept of Scandinavian kingship as we think we know it: concepts like honour, loyalty oaths, reli- external appropriation. Kingship was pre- gious connections, generosity/philanthropy, datory. protection, client relations, and violence. The • Heterarchical perspectives must be intro- Mafia is organized as “families”, it has a capo duced in urbanization studies; hierarchies (king), who can aspire to become capo di tutti are only part of the story. capi (overlordship), and who has a number of consiglieri (members of the lith) around him • New input can be had from comparative (some with their own aspirations, which gives global studies of similar cultural areas. unstable conditions), who in turn command soldati (warriors). Further studies of these ap- This must be seen as a working hypothesis. parent similarities may prove to give us a new We have still not identified the first sparks and interpretative model for the Early Medieval pe- what ignited them, but we have perhaps nar- riod. 5 The project runs in collaboration between York rowed down the range within which we have and Århus universities. It focuses on the ex- to search. So, let us hope that there is now pansion of maritime communication and net- a long push upslope before the next lull to work urbanism in the period c. 500–1200 AD anger Zeus. After all, isn’t it the duty of all re- through comparative studies of material flows searchers to try to imagine Sisyphus as happy? (http://www.york.ac.uk/archaeology/research/ current-projects/entrepot/). 6 a site, or rather agglomeration of sites, which may provide key evidence for understanding Notes not only the early chronology and functions of the trading sites, but also their relationship 1 in this text a continental periodization is used. to centres of various types, is the Ravlunda- It is the author’s conviction that there is no Haväng-Maletofta complex on the east coast need for a separate Scandinavian chronology of Skåne (Riddersporre 1998; Fabech 1999; (Mogren 2005, p. 21f). Thus Early Medieval Helgesson & Paulsson 2008). Artefact datings here means the period c. 550‒1100. give a chronological range from the 2nd cen- 2 Colleagues criticizing the reintroduction of tury into the High Middle Ages, but so far trade and crafts into the discussion someti- with a clear dominance for finds from the 6th mes take it for granted that the advocates of to the 11th centuries. There are indications of these two factors work teleologically from the a cult-judicial centre but the district itself is vantage point of the High Medieval town, liminal in many respects. The sites are still rela- which “makes us again fall into the trap of the tively unknown but research is underway and Frühe Stadt-concept” (Näsman 2011, p. 190). reports from recent excavations are pending. Nothing could be more mistaken; if there is anyone today who considers these early sites as predecessors of the High Medieval town, he or she deserves criticism, but there are no such References proponents that I know of. The 8th-century trading/craft sites, the 9th/10th-century em- Andersson, H. 2009. Historisk arkeologi och poria and the late 10th/12th-century towns globalisering. In Mogren, M., Roslund, M., represent three separate processes and trajecto- Sundnér, B. & Wienberg, J. (eds.), Triangu- ries. But they all represent urbanization. lering. Historisk arkeologi vidgar fälten. Lund 3 Carl Löfving’s recent “dating” of the ring to the Studies in Historical Archaeology 11. Lund. High Middle Ages (2010) has been effectively Ardika, I. W. & Bellwood, P. 1991. Sembiran. The refuted by Brink (2010b). beginnings of Indian contact with Bali. Anti- 4 Very tentatively, there are, in fact, several traits quity 65 (2). defining the (alternative) societal structure Artursson, M. 2008. Storgårdar och bebyggelse i

The First Sparks and the Far Horizons 85 Skåne och på Själland. Perioden 500 f.Kr–150 zania. First Incontrovertible Archaeological e.Kr. In Carlie, A. (ed.), Öresund – barriär eller Link with the Periplus. Current Anthropology bro? Kulturella kontakter och samhällsutveckling 40 (2). i Skåne och på Själland under järnåldern. Goth- Christie, J. W. 1998. Javanese markets and the enburg & Stockholm. Asian sea trade boom of the tenth to thirte- Axel-Nilsson, G. 1989. Thesaurus Cathedralis Lun- enth centuries A.D. Journal of the Social and densis. Lunds domkyrkas skattsamling. Acta Re- Economic History of the Orient 41 (3). giae Societatis Scientiarum et Litterarum Got- Crumley, C. L. 1995. Heterarchy and the Ana- hoburgiensis, Humaniora 30. Gothenburg. lysis of Complex Societies. In Ehrenreich, R., Blomkvist, N. 2001. The Concept of the Town and Crumley, C. L. & Levy, J. (eds.), Heterarchy the Dawn of Urban Life East and West of the and the Analysis of Complex Societies. Archaeo- Baltic. Lübeck Style? Novgorod Style? Baltic Rim logical Papers of the American Anthropologi- Central Places as Arenas for Cultural Encounters cal Association, No. 6. Washington, D.C. and Urbanisation 1100−1400 AD. Riga. Crumlin-Pedersen, O. 1999. Ships as Indicators of Bogucki, M. 2010. Viking Age Emporia Around Trade in Northern Europe 600–1200. In Bill, the – a Cul-de-sac of European J. & Clausen, B. L. (eds.), Maritime Topograp- Urbanization? In Buko, A. & McCarthy, M. hy and the Medieval Town. Papers from the 5th (eds.), Making a Medieval Town. Patterns of International Conference on Waterfront Archaeo- Early Medieval Urbanization. Warsaw. logy in Copenhagen, 14–16 May 1998. Publica- Brink, S. 2010a. Hälsingelagens ställning mellan tions from The National Museum, Studies in väst och syd, och mellan kung, kyrka och lo- Archaeology & History Vol. 4, Copenhagen. kala traditioner. Kungl. Vitterhets Historie och Englert, A. 2009. Cargo Ships as Indicators of Antikvitets Akademiens Årsbok 2010. Commerce and Urbanisation in Medieval Brink, S. 2010b. Är Forsaringen medeltida? Häl- Denmark 1000−1250. In Engberg, N., Nør- singerunor 2010. gaard Jørgensen, A., Kieffer-Olsen, J., Madsen, Callmer, J. 1998a. Archaeological Sources for P. K. & Radtke, C. (eds.), Archaeology of Med- the Presence of Frisian Agents of Trade in ieval Towns in the Baltic and North Sea Area. Northern Europe ca. AD 700−900. Studien PNM Studies in Archaeology & History 17. zur Archäologie des Ostseeraumes. Von Eisenzeit Copenhagen. zum Mittelalter. Festschrift für Michael Müller- Fabech, C. 1999. Centrality in Sites and Landsca- Wille. A. Wesse, hrsg. Neumünster. pes. In Fabech, C. & Ringtved, J. (eds.), Settle- – 1998b. Handelsplatser och kustplatser och de- ment and Landscape. Proceedings of a Conference ras förhållande till lokala politiska system. Ett in Århus, Denmark, May 4−7 1998. Århus. bidrag till strukturen i den yngre järnålders Feveile, C. 2006. Ribe på nordsiden af åen, 8.−12. samhälle. In Larsson, L. & Hårdh, B. (eds.), århundrede – oversigt og tolkning. In Feveile, Centrala platser centrala frågor. Samhällsstruk- C. (ed.), Ribe studier. Det ældste Ribe. Udgrav- turen under järnåldern. Uppåkrastudier 1. Acta ninger på nordsiden av Ribe Å 1984–2000. Archaeologica Lundensia, Ser. in 8º, No. 28. Bind 1.1 Jysk Arkæologisk Selskabs skrifter 51. Stockholm. Århus. Carlie, A. 2008. Storgårdar kring Öresund. Tra- Fleischer, J. B. 2010. Swahili Synoecism. Rural ditioner och förändring under perioden 150– Settlements and Town Formation on the Cen- 550 e.Kr. In Carlie, A. (ed.), Öresund – barriär tral East African Coast, A.D. 750−1500. Jour- eller bro? Kulturella kontakter och samhällsut- nal of Field Archaeology 35 (3). veckling i Skåne och på Själland under järnål- Gazzoli, P. 2011. Denemearc, Tanmaurk ala, and dern. Gothenburg & Stockholm. Confinia Nordmannorum. The Annales Regni Carlsson, D. 2011. Vikingatidens Västergarn – en Francorum and the Origins of Denmark. Vi- komplicerad historia. ArkeoDoks Skrifter 3. king and Medieval Scandinavia 7, Turnhout. Stockholm. Hårdh, B. 2010. Viking Age Uppåkra. In Hårdh, Chami, F. 1994. The Tanzanian Coast in the First B. (ed.), Från romartida skalpeller till senviking- Millennium AD. An Archaeology of the Ironwor- atida urnesspännen. Nya materialstudier från king Farming Communities. Studies in African Uppåkra. Uppåkrastudier 11. Acta Archaeolo- Archaeology 7. Uppsala. gica Lundensia, ser. in 8o, No. 61, Lund. – 1999. Roman Beads from the Rufiji Delta, Tan- Helgesson, B. & Paulsson, J. 2008. Maletofta Me-

86 MATS MOGREN talldetektering 2008. Arkivrapport 18/2008. Mortensen, P. & Rasmussen, B.M. (eds.), Fra Regionmuseet Kristianstad. Stamme til Stat i Danmark 2. Høvdingesamfund Hilberg, V. 2009. Hedeby in Wulfstan’s days. A og Kongemagt. Jysk Arkæologisk Selskabs Skrif- Danish Emporium of the Viking Age between ter XXII:2, Århus. East and West. In Englert, A. & Trakadas, A. Näsman, U. 1991. Det syvende århundrede – et (eds.), Wulfstan’s Voyage. The Baltic Sea region mørkt tidsrum i ny belysning. In Mortensen, P. in the early Viking Age as seen from shipboard. & Rasmussen, B. M. (eds.), Fra Stamme til Stat Maritime Culture of the North 2. Roskilde. i Danmark 2. Høvdingesamfund og Kongemagt. Indruszewski, G. 2005. From Iron Age to Middle Jysk Arkæologisk Selskabs Skrifter XXII:2, År- Age. TheA rchaeological Evidence of Ships and hus. Boats of the Baltic Sea. Questiones Medii Aevi – 1998a. The Justinian Era of South Scandinavia. Novae, vol. 10, 2005. An Archaeological View. In Hodges, R. & Jensen, S. 1991. Dankirke–Ribe. Fra handelsgård Bowden, W. (eds.), The Sixth Century. Produc- til handelsplads. In Mortensen, P. & Rasmus- tion, Distribution and Demand. Leiden. sen, B.M. (eds.), Fra Stamme til Stat i Dan- – 1998b. Sydskandinavisk samhällsstruktur i ljuset mark 2. Høvdingesamfund og Kongemagt. Jysk av merovingisk och anglosaxisk analogi eller i Arkæologisk Selskabs Skrifter XXII:2, Århus. vad är det som centralplatserna är centrala? In Jöns, H. 2009. Ports and Emporia of the Southern Larsson, L. & Hårdh, B. (eds.), Centrala platser Coast. From Hedeby to Usedom and . centrala frågor. Samhällsstrukturen under järn- In Englert, A. & Trakadas, A. (eds.), Wulfstan’s åldern. Uppåkrastudier 1. Acta Archaeologica Voyage. The Baltic Sea region in the early Viking Lundensia, Ser. in 8º, No. 28. Stockholm. Age as seen from shipboard. Maritime Culture of – 2011. Central Places in South Scandinavia – A the North 2. Roskilde. Transformation Twenty Years After. In Pan- Juma, A. 2004. Unguja Ukuu on Zanzibar. An Ar- huysen, T. A. S. M.. (ed.), Transformations in chaeological Study of Early Urbanism. Studies in North-Western Europe (AD 300−1000). Pro- Global Archaeology 3. Uppsala. ceedings of the 60th Sachsensymposion 19.−23. Larsson, L. (ed.) 2004. Continuity for Centuries, September 2009 Maastricht. Hannover. Uppåkrastudier 10, Stockholm/Lund. Nielsen, P. O., Randsborg, K. & Thrane,H . (eds.), LaViolette, A. & Fleisher, J. 2009. The UrbanH is- 1994. The Archaeology of Gudme and Lunde- tory of a Rural Place. Swahili Archaeology on borg, Copenhagen. Pemba Island, Tanzania, 700−1500 AD. Inter- Randsborg, K. 1980. The Viking Age in Denmark. national Journal of African Historical Studies 42 The Formation of a State. London. (3). Reid, A. 1988–1993. Southeast Asia in the Age of Liestøl, A. 1979. Runeringen i Forsa. Kva er han, Commerce, 2 vols. New Haven. og når vart han smidd? och sed 1979. Riddersporre, M. 1998. Ravlunda och Uppåkra. Lindkvist, T. 1988. Plundring, skatter och den feo- Två exempel på försvunna storgårdar? In Lars- dala statens framväxt. Organisatoriska tendenser son, L. & Hårdh, B. (eds.), Centrala platser, i Sverige under övergången från vikingatid till centrala frågor. Samhällsstrukturen under järn- tidigmedeltid. Opuscula Historica Upsaliensia åldern. Uppåkrastudier 1. Acta Archaeologica 1. Uppsala. Lundensia, Ser. in 8º, No. 28. Stockholm. Löfving, C. 2010. Forsaringen är medeltida. Forn- Sindbæk, S. M. 2007a. Networks and Nodal vännen 105. Points. The Emergence of Towns in Early Vi- Mogren, M. 2005. Den långa medeltiden. In king Age Scandinavia, Antiquity 81. Mogren, M. (ed.), Byarnas bönder. Medeltida – 2007b. The Small World of the . Net- samhällsförändring i Västskåne. Skånska spår – works in Early Medieval Communication and arkeologi längs Västkustbanan. Stockholm. Exchange, Norwegian Archaeological Review Miksic, J. N. 2000. Heterogenetic cities in premo- 40. dern Southeast Asia. World Archaeology 32 (1). – 2008. Local and Long Distance Exchange. In Mortensen, P. & Rasmussen, B. M. (eds.) 1988. Brink, S. & Price, N. (eds.), The Viking World, Fra Stamme til Stat i Danmark 1. Jernalderens New York & London. stammesamfund. Jysk Arkæologisk Selskabs Sjöbeck, A. 2012. Varför Lund? En analys av fem Skrifter XXII, Århus. urbana orter i tusentalets Skåne och jämförel- Mortensen, P. & Rasmussen, B. M. (eds.) 1991. In ser av förutsättningar för urbanisering. Bac-

The First Sparks and the Far Horizons 87 helor paper in Historical Archaeology, Lund University, unpubl. Theuws, F. 2004.E xchange, Religion, Identity and Central Places in the Early Middle Ages. Ar- chaeological Dialogues 10 (2). Ullman, W. 1965. A History of Political Thought. The Middle Ages.H armondsworth. Wade, G. 2003. An Early Age of Commerce in Southeast Asia, 900−1300 CE. Journal of Southeast Asian Studies 40(2). Wallace, A. R. [1869] 1994. The Malay Archipe- lago. The Land of the Orang-utan, and the Bird of Paradise. A Narrative of Travel with Studies of Man and Nature. Kuala Lumpur, third impres- sion. Wolfe, U.I. Zagal-Mach 2013. Grasping Tech- nology, Assessing Craft. Developing a Research Method for the Study of Craft Tradition. Diss., Archaeology, Lund University. Wynne-Jones, S. 2007. Creating Urban Com- munities at Kilwa Kisiwani, Tanzania, A.D. 800−1300. Antiquity 81.

88 MATS MOGREN