Borders, Barriers, and Ethnogenesis Frontiers in Late Antiquity and the Middle Ages Edited by Florin Curta BREPOLS STUDIES IN THE EARLY MIDDLE AGES EDITORIAL BOARD UNDER THE AUSPICES OF THE CENTRE FOR MEDIEVAL STUDIES UNIVERSITY OF YORK Elizabeth M. Tyler (University of York) Julian D. Richards (University of York) Ross Balzaretti (University of Nottingham) Volume 12 British Library Cataloguing in Publication Data Borders, barriers, and ethnogenesis. - (Studies in the early Middle Ages ; v. 12) 1 .Middle Ages - History 2.Ethnology - Europe - History - To 1500 3.Ethnology - Rome 4.Ethnicity - Europe - History - To 1500 5.Rome - History - Empire, 284-476 6.Europe - Boundaries 7.Rome - Boundaries I.Curta, Florin 305.8'0094'0902 ISBN-10: 2503515290 © 2005, Brepols Publishers n.v., Turnhout, Belgium All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of the publisher. D/2005/0095/135 ISBN: 2-503-51529-0 Printed in the E.U. on acid-free paper. Contents Introduction FLORIN CURTA Part 1. Frontiers, Forts, and Fortifications The Byzantine-Arab Borderland from the Seventh to the Ninth Century 13 RALPH-JOHANNES LILIE Civilization versus Barbarians? Fortification Techniques 23 and Politics in Carolingian and Ottonian Borderlands JOACHIM HENNING The Limes Saxoniae as Part of the Eastern Borderlands 35 of the Frankish and Ottonian-Salian Empire MATTHIAS HARDT Remarks on the Archaeological Evidence of Forts 51 and Fortified Settlements in Tenth-Century Bulgaria RASHO RASHEV Moving Earth and Making Difference: Dikes and Frontiers in Early Medieval Bulgaria PAOLO SQUATRITI 59 Reconceptualizing the Seljuk-Cilician Frontier: Armenians. Latins, and Turks in Conflict and Alliance during the Early Thirteenth Century SARA NUR Y1LDIZ 91 Part 2. Frontiers, Ethnicity, and Material Culture Ethnic and Territorial Boundaries in Late Antique 123 and Early Medieval Persia (Third to Tenth Century) TOURAJ DARYAEH Acculturation and Ethnogenesis along the Frontier: 139 Rome and the Ancient Germans in an Archaeological Perspective SEBASTIAN BRATHER Frontier Ethnogenesis in Late Antiquity: 173 The Danube, the Tervingi, and the Slavs FLORIN CURT A The Shadow of a Frontier: The Walachian Plain during the Justinianic Age 205 EUGL'NS. TEODOR Ethnicity. Rulership, and Early Medieval Frontiers 247 MICHAEL KULIK.OWSKI Frontiers and Ethnic Identities: Some Final Considerations WALTER POHL 255 Civilization versus Barbarians? Fortification Techniques and Politics in the Carolingian and Ottonian Borderlands JOACHIM HENNING The north-eastern frontier between the Christian Prankish Empire and the pagan Saxons existing in the seventh and eighth centuries until Charle- magne^ conquest of Saxony is commonly regarded as a boundary separating the Frankish heirs of the late antique civilized world from the Saxon tribes, whose primitive ways of life epitomize paganism.1 A similar situation has been postulated for the later frontier at a distance of one hundred years, at the time the Ottonians began pushing the borders of their empire to the east.2 By then, the already converted Saxons had become 'civilized', and had begun styling themselves as defenders, if not heirs, of the antique traditions, as they watched over a line of political demar- cation, but also of sharp cultural contrast with the Slavs on the other side, whose backwardness and primitive ways of life had meanwhile become the new epitome of stubborn paganism.3 1 For the seventh- to eighth-century line of fortifications on the Saxon frontier, see Norbert Wand, "Die Biiraburg bci Fritzlar - eine frankische Reichsburg mil Bischofssitz in Hessen'. in Friihmittelalterlicher Burgenbau in Mittel- and Osteuropa: Tctgung. Nitra vom 7. bis 10. Ok- tober 1996. ed. by Joachim Henningand Alexander T. Ruttkay (Bonn, 1998). p. 175 with fig. 1. 2 For the Elbe-Saale frontier during the ninth and tenth centuries, as well as a map of border fortifications mentioned in the written sources, see Joachim Henning, Der slawische Siedlungsraum und die ottonische Expansion ostlich der Elbe: Ereignisgeschichte, Archao- logie, Dendrochronologie'. in Ewopa im 10. Jahrhunderl: Archaologie einer Aujbritchszeit, ed. by Joachim Henning (Mainz. 2002). p. 134 with fig. 3. J For the historical background, see Matthias Hardt, 'Hesse, Elbe, Saale and the Frontiers of the Carolingian Empire', in The Transformation of Frontiers: From Late Antiquity' to the Carolingians, ed. by Walter Pohl. Ian Wood, and Helmut Reimitz (Leiden, 2001). pp. 219-32; Linien und Saume, Zonen und Raume an der Ostgrcnze des Reiches im fruhen und hohen 24 JOACHIM HENNING However, this picture is nothing else than a stereotype rooted in the militant char- acter of available written sources, all of which were concerned primarily with the struggle against pagans. Recent archaeological studies of Carolingian and Ottonian fortifications from two border areas suggest that the presumably sharp cultural dif- ference marked by frontier lines was not so evident as previously thought. In fact, it became clear that in many ways the situation on the north-eastern and eastern fron- tier of the Carolingian and Ottonian empires is an early medieval replica of phenom- ena associated with the frontiers of the Late Roman Empire.4 Cultural differences between the core areas of the 'civilized" world, such as the Paris basin or the Rhine lands during the early Middle Ages, and the peripheral regions of Hessia and Bavaria to the east, as well as Brittany to the west, were much more important than contrasts supposedly created by the implementation of frontier lines. The economic and social resources available at that time were not sufficient for supporting cultural uniformity across the entire area under the direct control of the Carolingians or of the Ottonians. On the other hand, and despite the bloody military confrontations taking place in borderlands, political frontiers were not walls separating groups of people, but areas of cultural exchange, a melting pot of cultures, economies, and societies.3 All known cases of frontiers, from Late Antiquity to the tenth century, imply the existence of buffer zones, especially in areas where tribal groups existed in close contact with neighbouring empires. In that sense, "barbarians' existed on both sides of the political frontier,6 either as buffer groups on one side (e.g. the so-called Sorabian March on the eastern border of the Carolingian Empire in Thuringia, also known as the limes Sorabicus)1 or as 'federates' on the other side, watching over frontier fortifications8 or just settled as peasants in frontier hinterlands. In both cases, Mittelalter', in Gretize und Different im friihen Mittelalter, ed. by Walter Pohl and Helmut Reimitz (Vienna, 2000), pp. 39-56; and "The Limes Saxoniae as Part of the Eastern Borderlands of the Frankish and Ottonian-Salian Empire' (in this volume). 4 David Harry Miller, 'Frontier Societies and the Transition between Late Antiquity and the Early Middle Ages', in Shifting Frontiers in Late Antiquity, ed. by Ralph W. Mathisen and Hagith S. Sivan (Aldershot. 1996), pp. 158 -71. 5 Hugh Elton, 'Defining Romans, Barbarians, and the Roman Frontier', in Shifting Fron- tiers, ed. by Mathisen and Sivan, pp. 126-35. 6 See the studies collected in Germanen beiderseits des spatantiken Limes, ed. by Thomas Fischer, Gundolf Precht, and Jaroslav Tejral (Cologne, 1999). 7 For the archaeology of this frontier, see Hansjiirgen Brachmann, 'Der Limes Sorabicus: Geschiclite und Wirkung', Zeitschrift fur Archdologie, 25 (1991), 177-207; and the written record; Matthias Hardt. 'Limes Sorabicus", in Reallexikon der germanischen Altertumskunde. ed. by Hcinrich Beck, Dieter Geuenich, and Heiko Steuer, vol. xvui (Berlin. 2001), pp. 446-48. 8 Peter J. Heather, Foedera ani foederati of the Fourth Century', in Kingdoms of the Empire: The Integration of Barbarians in Late Antiquity, ed. by Walter Pohl (Leiden, 1997), pp. 57-74; Maria Cesa. Romisches Heer und barbarische Foderaten: Bemerkungen zur westromischen Politik in den Jahren 402-412\ in L arinee rornaine et les barbares du llf au Vlf siecle, ed. by Francoise Vallet and Michel Kazanski (St Germain-des-Pres, 1993), pp. 21-29. I Civilization versus Barbarians? 25 there was a strong tendency for 'barbarians' to identify with the policies and culture of the Empire whose interests they served, a process which in turn triggered ethnic changes and assimilation. Immediately after the Frankish military conquest of the border territories between the Lower Elbe and the Wismar Bay, until then controlled by Saxons and Danes, Charlemagne ordered the Slavic Obodrites to move into the region as Frankish 'federates'. Much like their late antique counterparts, the late eighth-century 'federates' were spread over a number of civitates, castella, and oppida in order lo defend the frontier of the Empire against Danish raids.9 In 789 Charlemagne built castrae on the left bank of the Elbe River against the Ligones, most likely allies of the Wilzi. According lo the Royal Frankish Annals, castella were also built on the opposite bank, as bridgeheads in foreign territory. One such fort is described in more detail as built of timber and earth ('ex lingo et terra aedi- ficavit'). The same is true about the forts erected in 806, after King Charles, Charle- magne's son, crushed the revolt of the Sorabian federates east of the Saale River. According to the Chronicle of Moissac, a monastery in the diocese of Cahors in southern France, Charles gave orders to the Slavic reges to build fortifications on the eastern bank of the Elbe and Saale rivers. However, the Royal Frankish Annals indicate that the fortifications were erected by the army ('castella ab exercitu aedi- ficata'). This seems to suggest that the Sorabian federates had been once again incor- porated into the Carolingian exercitus of the borderlands. Only few such Carolingian frontier forts have been identified by archaeological means. The all-mysterious fort built in 808 'ad aquilonem partem Albiae contra Magadaburg' and iuxta fluvium Albim' may well be the fortification identified in 2003 during the geophysical survey carried out in the Magdeburg region by a team from the Johann Wolfgang Goethe University in Frankfurt. Generations of historians and archaeologists have tried in vain to find the fort underneath a number of stone castles in the hinterland of Magdeburg, but just as described in contemporary sources, the fort was made of timber and earth. According to those accounts, it was visible from Magdeburg, as it watched over the Elbe valley north of that city on the opposite bank of that river. Furthermore, it is said to have been manned by Slavic contingents, and the survey revealed two sunken-featured buildings, each with a stone oven in the corner. Both houses produced ceramic assemblages marked by the presence of handmade pottery similar to the so-called Prague type. It is possible that the territories east of the limes Saxoniae (east of present-day Schleswig) and of the limes Sorabicus along the Saale were the first to experience the transfer of cultural elements of Frankish power represenlation and administrative structures onto the Slavic world. Indeed, it is in these areas that written sources place two rival ethnic groups, the one allied with the Franks (the Obodrites), the other 9 Royal Frankish Annals a. 808, ed. by Friedrich Kurze (Hannover, 1895; repr., 1930), MGH SS rer. Germ. 6:125; Chronicle of Moissac a. 808, ed. by Georg Heinrich Pertz (Hannover, 1826. repr.. 1925), MGH SS 1:308. 26 JOACHIM HENNING staunchly opposed to them (the Wilzi).10 Even groups on the other side of the political frontier that resisted Frankish or Ottonian encroachment, such as the Saxon tribal federation or the burgeoning states of the Western Slavs, were in the end quite willing to imitate the military and defensive techniques, if not also elements of the political structure, of the Empire." In the light of recent archaeological studies of early medieval fortifications on the Carolingian and Ottonian borderlands in the north-east, the question of cultural dif- ferences across the frontier between Christians and pagans can now be rephrased. Previous studies have been based on the assumption that 'civilization' was to be identified by means of castles built in stone similar to Roman frontier forts, whereas 'barbarians' were associated with ringforts of prehistoric tradition, all built in timber and earth. Thus, scholars studying the Biiraburg near Fritzlar in northern Hessia have long imagined the life of the garrison manning the Roman-like stone castle built in the early eighth century in the newly conquered territories east of the Rhine as iden- tical to that of the tenth-century soldiers placed by King Henry I on the German- Slavic borderline on the river Saale (Fig. I).12 By contrast, outside the castle and on the other side of the frontier, we are invited to imagine a network of ringworks erected by pagan Saxons. Such forts, widely spread in northern Germany, were first called 'Saxon ringforts, by the German archaeologist Carl Schuchhardt, who conse- quently dated their origins back to the pre-Carolingian period.13 Biiraburg is well known from written sources as a site of strategic importance on the Saxon frontier of the Carolingian Empire. St Boniface had called the site an oppidum, and Biiraburg was one of three bishoprics established as early as 742 in the lands east of the Rhine. It was, however, abandoned a few years later. Small-scale excava- tions outside the stone rampart on the eastern side of the fortified hill produced a number of pits, hastily interpreted as the remains of a suburbiwn including handi- craft workshops and houses of merchants engaged in long-distance trade. Biiraburg 10 Given the close contact between Franks and Slavs in thai particular area, it is not impos- sible to see the all-Sla\ic word for 'king' {korol Vkral) deriving from Charlemagne's name as originating from that same region. In any case, the archaeological evidence clearly demon- strates the adoption of Carolingian architectural elements of power representation by local Slavs. See Ingo Gabriel, '"lmitatio imperii'' am slawischen Fiirstenhof zu Starigard/Oldenburg (Holstein): Zur Bedeutung karolingischer Konigspfalzen fur die Herausbildung friiher Herr- schaftszentren und Burgstadte bei den nordwestlichen Slawen'. in Trudy V Mezhdunarodnogo Kongressa arkheologox-slavistov, Kiev 18-25 sentiabria 1985 g., ed. by V. V. Sedov, vol. I (Moscow, 1987). pp. 50-66. 11 Peter G. Heather, 'Frankish Imperialism and Slavic Society', in Origins of Central Europe, ed. by Przemyslaw Urbahczyk (Warsaw, 1997), pp. 171-90. Ralf Gcbuhr, 'Burg und Landschaft: Kulturhistorischc Untersuchung zur Archaologic frtihgeschichtlicher Wehrbauten an Elbe und Elster am Beispiel der Burg auf dem "Griinen Berg" bei Gehren' (unpublished M.A. thesis, Humboldt University, Berlin, 1996). 13 Carl Schuchhardt. Die Burg im Wandel der Weltgcschichte (Potsdam, 1931). pp. 228-29. Civilization versus Barbarians? 27 Figure 1. A Roman-like stone castle of the early tenth century on the German-Slavic frontier on the river Saale. Romanticized German textbook illustration of the 1930s (after R. Gcbuhr). thus appeared as the first post-Roman early town founded in the region east of the Rhine River, an almost urban settlement without any Roman roots, but with a sig- nificant role in spreading the Roman influence to the east. As such, Biiraburg would have indeed been an exceptional case for the early 700s, and it is therefore no surprise that the site received much attention from German historians. However, archaeological excavations, geophysical surveys, aerial photography studies and micromorphological soil investigations carried out since 1997 by the Johann Wolf- gang Goethe University in Frankfurt and sponsored by the Commission of Archaeo- logical Research in Hessia have radically changed that picture (Fig. 2). It became clear that the assumed suburbium of craftsmen and traders on the Biiraburg hill was nothing but a figment of scholarly imagination, a result of delib- erate misinterpretation of archaeological structures. The relatively numerous features identified outside the ramparts, initially thought to be the remains of the Frankish suburbium. are in fact of Neolithic origin. The idea of the earliest Frankish town east of the river Rhine must now be abandoned.14 14 For details, see Joachim Henning and Richard I. Macphail, Das karolingerzeitliche Oppidum Biiraburg: Archaologischc und mikromorphologische Studien zur Funktion ciner friihmittelalterlichen Bergbefestigung in Nordhcssen'. in Parerga Praehistorica: Jubilaums- schrift zur Prahistorischen Archaologie - 15 Jahre VP A, ed. by Bernhard Hiinsel (Bonn, 2004), pp. 221-51. 28 JOACHIM HENNING ---- 1 . fcfer.. 1 2 3 • Figure 2. The early medieval fortification on the Biiraburg hill. Field research of the Johann Wolfgang Goethe University in Frankfurt am Main (1997-2004): 1. stone rampart. 2. area of geophysical sensing, 3. archaeological trenches, 4. area of micromorphological investigation, 5. palisade rows, 6. important dis- coveries: palisade (P) and Neolithic structures (W). 7. pointed ditches (excava- tions until 1973), 8. formerly assumed ditch (revised by excavations of 2000). 9. recently reconstructed ditches (excavations of 2000). Moreover, investigations in the forcfield of the stone rampart produced evidence of multiple rows of timber palisades dating to the early Middle Ages. The chronol- ogy of these palisades was determined by means of seven radiocarbon datings and is roughly the same as that of multiple rows of pointed ditches that, together with the palisades, form an effective and independent fortification system. It is therefore very probable that the Carolingian occupation of the Biiraburg hill consisted of a simple earth-and-timber fortification. Excavations of the dark-earth deposits behind the stone rampart, which has long been viewed as Carolingian. did not produce clearly Civilization versus Barbarians? 29 stratified occupation layers but an irregular mixture of Carolingian, Ottoman, and prehistoric finds. The micromorphological analysis of these dark-earth deposits made in the London University College lab15 has demonstrated their primarily collu- vial nature. The depositing of this earth matrix was not the result of occupation activities behind the stone rampart, but of the erosion process that altered the hill slope and brought layers of the dark earth behind the stone rampart. The formation of the colluvium behind the stone rampart, together with a few remains of real, albeit temporary, occupation of the site cannot be dated earlier than the late ninth to early eleventh century, and as such the stone rampart could not have come into being at the time St Boniface established his bishopric in Biiraburg on the Frankish-Saxon frontier. The Carolingian fortification of these early times was no more than a simple palisade with a pointed ditch, which was in fact not at all different from Saxon strongholds on the other side of the frontier, but similar in many respects to other contemporary fortifications of Carolingian age in the eastern borderlands (Magde- burg, Halberstadt, Esesfelt).16 The Biiraburg/Fritzlar settlement and fortification complex must have been of some importance until late into the ninth or the early tenth century, perhaps even later. The first king of the Ottoman dynasty, Henry I, was elected in Fritzlar, at the foot of the Buraburg hill. The site was located in the heartland of the Conradine principality, initially the most powerful opponent of the Liudolfings, later their most important ally. The choice of Fritzlar, and of Conradine instead of Liudolfingian territory, for the investiture of Henry I must be seen as an act of great significance for the relation between the two powerful families. It is therefore very unlikely that the Conradines had no control over a fort of such importance throughout the ninth century. Judging from the existing evidence, the Conradine fort was the Buraburg fortification with ramparts of stone bonded with mortar built on top of older structures of Carolingian age. A similar picture emerges from the evaluation of other fortifications in the fron- tier region between Franks and Saxons. The communis opinio influenced by the older Buraburg interpretation is that on some such sites the stone-mortar ramparts must date back to early Carolingian times. Excavations at Buraburg proved this in- terpretation to be wrong. At the Christenberg hill near Mtinchhausen (Middle Hessia),17 a recent redating to the late ninth or tenth century of three spurs found 15 Analysis by Richard I. Macphail (report in the archive of the Johann Wolfgang Goethe University in Frankfurt am Main). 16 Hansjiirgen Brachmann, Der friihmittelalterliche Befestigungsbau in Mitteleuropa: Untersuchungen zu seiner Entwicklung und Fwikrion im germanisch-deutschen Bereich (Berlin. 1993). 17 Rolf Gensen, 'Christenberg, Bnrgwald und Amoneburger Becken in der Merowinger- und Karolingcrzeif, in Ahhessen im Frankenreich, ed. by Walter Schlesinger (Sigmaringen, 1975). pp. 121-72. 30 JOACHIM 11ENNING behind the stone-mortar rampart18 consequently triggered a reinterpretation of the fort. The spurs are very similar to isolated finds from the Buraburg fortification.19 At Christenberg they were clearly associated with features adjacent to the stone rampart. With the redating of the stone rampart to the late ninth or tenth century, the idea of a Carolingian stone fort in Christenberg becomes as dubious as in Buraburg. The evidence discussed so far suggests that Frankish hillforts built east of the river Rhine following the Carolingian conquest were not very different from those erected by Saxons on the other side of the frontier. Very similar to this archaic hill fort type were also contemporary Slavic strongholds of the so-called Feldberg type in western Pomerania and in Mecklenburg,20 which in AD 789 had been the target of a large military campaign of Charlemagne. Close to the end of the Saxon wars and following the conversion of most parts of Saxony to Christianity, the Frankish army now moving against the Slavic tribal federation of the Wilzi received reinforcements of Saxon contingents.'1 The chronology of Schuchhardt's 'Saxon ringforts'. which differ considerably from older hillforts in terms of their almost circular plan, limited size, as well as location in wetlands, have also been recently revisited. Results of dendrochronologi- cal analysis of timber remains from some of those lowland forts indicate that none was in existence during Charlemagne's lifetime. Consequently, these forts could not have played any role in the Saxon wars of the eighth century. Ringforts of the For the first mention of a need to redate the Christenberg finds, see Norbert Gossler. Mitlelaltcrliches Reitzubehor von hessischen Burgen', in Burgenforschung in Hessen: Begleitband zur Ausstellung im Mai burger Landgrafenschloss vom J. November J 996 bis 2. Februar 1997, ed. by Bernhard Schroth (Marburg. 1996), pp. 161-76. For a comprehensive analysis of the Christenberg spurs and stirrup, see Thomas Kind, 'Archaologische Funde von Teilen der Reiterausriistung aus Europa und ihr Beitrag zur Kultur- und Sozialgeschichte der Ottonenzeit", in Europa im JO. Jahrhundert', ed. by Henning, pp. 283-99. The presence of a clear Ottoman phase of occupation at Christenberg lias meanwhile been accepted by the excavator of the site; see Rolf Gensen. 'Ein Keramikkomplex mit dem Schlussdalum 753 vom Christenberg, Gde. Munchhausen am Christenberg, Kr. Marburg-Biedenkopf, in Archaolo- gische Beilrdge zur Geschichte Westfalens: Festschrift fiir Klaus Gunther, ed. by Daniel Berenger (Rahden, 1997), pp. 219-28. A new analysis of the materials from the Christenberg excavation drew the same conclusion: Andreas Thiedmann, 'Neue Forschungen zum Christen- berg bei Munchhausen'. Hessen Archdologie, 2001 (2002). 126-28. 1; Norbert Wand. Die Buraburg bei Fritzlar: Burg - oppidum ' - Bischofssitz in karolin- gischer Zeit (Marburg. 1974). 20 Sebastian Brather. "Karolingerzeitlicher Befestigungsbau im wilzisch-abodritischen Raum: Die sogenannten Feldberger Hohenburgen", mFruhmittelalterlicher Burgenbait, ed. by Henning and Ruttkay. pp. 115 26. Royal Frankish Annals a. 789. ed. by F. Kiuze (Hannover. 1895), MGH SS rer. Germ. 6:84; Einhard. Vita Caroli Magni, 12. ed. by Georg Fleinrich Pertz (Hannover. 1911; repr.. 1927), MGH SS 2:449. Civilization versus Barbarians? 31 'developed', small circular or slightly rectangular, type in Saxon territories only appear around the mid-800s, and their number rapidly increased shortly after AD 900. As such, they regularly appear on both sides of the new frontier between the Ottoman Saxons and their Slavic neighbours.22 The ninth- and tenth-century ringforts of the North German and Polish lowlands, especially those of Niederlausitz, just south of Berlin, and of Mazovia, have been the subject of a long-term research project of the Johann Wolfgang Goethe University. The careful study of building and fortification techniques already suggests that no stone-cum-morlar ramparts were erected in these eastern borderlands during the Ottonian period. The dendrochronological analysis of circular strongholds in the re- gion of Magdeburg indicates that Ottonian fortifications erected against Slavic raids were earth-and-timber structures" similar to those built by the Slavs on the other side of the frontier. Ringforts in Slavic territory have been traditionally dated to a very early date, namely to the late sixth or early seventh century. Some archaeol- ogists viewed such forts as the result of social and economic changes taking place in Slavic society, especially the rise of a class of noble landlords, of feudal-like structures not very different from that of the Merovingian West. Recent excavation of more than twenty-five ringforts in the Niederlausitz region with dates firmly established by means of dendrochronological analysis has shown this interpretation to be wrong 24 No strongholds existed during the entire period between the assumed " Henning, 'Der slawische Siedkingsraum'. pp. 134-35. 23 Wolfgang Schwarz, Burgwallgrabung in Osterburg, Ldkr. Slendal', in Archciologische Berichte Sachsen-Anhalt, 1994 (Halle, 1996), pp. 163-72. 24 For the first survey of results of dendrochronological analysis of Slavic. Tornow-type of fortifications in eastern Germany and Poland, see Joachim Henning, "Der Burg-Siedlungs- Komplex von Presenchen: Probleme und Perspektiven slawischer Archaologie im Braunkoh- lengebiet der Niederlausitz', in Archciologische Erkundung und Rettungsarbeit in Tagebauge- bieten Mitteleuropas: Internationale Arbeitstagung Sallgast, Kr. Finsterwalde. 10.-14. April 1989. ed. by Bernhard Gramsch and Giinter Wetzel (Berlin. 1991), pp. 141-46: Germanen. Slawen, Deutsche: Neue Untersuchungen zum friihgeschichtlichen Siedlungswesen ostlich der Elbe', Prdhistorische Zeitschrift, 66 (1991), 119-33; Joachim Henning and Karl-Uwe HeuBner. 'Zur Burgengeschichte im 10. Jahrhundert: Neue archaologische und dendrochrono- logische Daten zu Anlagen vom Typ Tornow", Ausgrabungen und Fimde. 37 (1992), 314-24. For the results of the project since 1991, see Joachim Henning, "Archaologische Forschungen an Ringwallen in Niederungslage: Die Niederlausitz als Burgenlandschaft des tistlichen Mit- teleuropa im friihen Mittelalter', in Friihmittelalterlicher Burgenbau, ed. by Henning and Rultkay, pp. 8-21; "Neues zum Tomower Typ. Keramische Formen und Formenspektren des Fruhmittelalters im Licht dendrochronologischer Daten zum vvcstslawischen Siedlungsraum', in Kraje slowianskie w wiekach srednich: Profanum i sacrum, ed. by Wladyslaw Losinski and Hanna Kocka-Krenz (Poznah, 1998), pp. 392-408; Joachim Henning and Thorsten Westphal. 'Forschungen zur archaologischcn Chronologie des Fruhmittelalters und das dendrochronolo- gische Labor an der Universitat Frankfurt am Main', in Probleme der mitteleuropciischen Dendrochronologie. ed. by Jitka Dvorska and Lumir Polacek (Brno, 1999), pp. 28-34; and 'Der slawische Siedlungsraum'. 32 JOACHIM HENNING Slavic colonization of this area and the beginnings of the east Frankish and Ottonian encroachment in the late 800s. The plotting of the dendro-dates from Niederlausitz, combined with that of dendrochronological results of excavations in the region be- tween the Elbe and the Oder/Weifie rivers, illustrates an increase in building activity, in sharp contrast to more peaceful periods, whatever interpretation we may choose for the lack of fortification building activities during such times (Fig. 3). Peaks of building activity can easily be associated with preparations before military cam- paigns launched against the Slavs by Saxon kings and emperors. For example, the building activity in the Niederlausitz region reached a peak in c. 919, the year in which Henry the Fowler was crowned King of East Francia. Henry had already made a name for himself during the expeditions he led into Slavic territory at the orders of his father, Otto, the Prince of Saxony.25 The Slavic tribes on the river Elbe could not expect anything good from his coronation in Fritzlar. Indeed, in 928 at the head of a well-equipped army, Henry began a sustained and ultimately successful campaign against the Slavic centres in the regions of Brandenburg, Gana (near Meissen), and Prague." There is no mention in literary sources of the conquest of small strong- holds, of which there were hundreds in each of the three areas. Nevertheless, there can be little doubt that they could not have escaped unscathed. The Annals of Hildes- heim and the Annuls of Wissembourg only mention that, in 932, Henry was in 'Lonsicin' (Lausitz),27 perhaps an indication that the King was busy imposing the pax Saxonica on his Slavic neighbours. This may indeed be the reason for the sud- den halt in building activities in Lausitz. The same is true about forts in the northern Slavic territories. A significant peak coincides here with the independence move- ment ending in 955, the year in which Otto I inflicted a crushing defeat on the Slavs at the Recknitz River. Moreover, building activities revealed by means of dendro- chronology increased considerably by the time of the major Slavic revolt of 983. The results of the joint studies of Slavic strongholds east of the river Vistula by a team from the Johann Wolfgang Goethe University in Frankfurt and the Polish 25 Widukind, Rerum gestamm Saxonicarum, I 17, ed. by Paul Hirsch and Hans-Eberhard Lohmann (Hannover, 1935), MGH SS rer. Germ. 60:27. 26 Widukind, Rerum gestarum Saxonicarum. I 35, MGH SS rer. Germ. 60:50. For the archaeological evidence, see Karl-Uwe Heuliner and Thorsten Westphal, ' Dendrochronolo- gische Untersuchungen an Holzfunden aus liiihmittelalterlichen Burgwallen zwischen Elbe und Oder', in Friihmittelalterlicher Burgenbau, ed. by Hcnning and Ruttkay. pp. 223-34 (Brandenburg); Tomasz Herbich, Roman Kfivanek, Krzyszlof Misiewicz, and Judith Oexle, 'Magnetic Surveys of the Site Burg Gana (Hof/Stauchitz) in Saxony", Archaeologia Polona, 41 (2003). 197-200 with figs 1-3 (Gana); Ji'tka Dvorska and Ivana Bohacova, 'Das historische Holz im Kontext der archaologischen Untersuchungen der Prager Burg', in Probleme der mitteleuropdischen Deudrochronologie, ed. by Dvorska and Polacek, pp. 55-67 (Prague). 27 Annals of Hildesheim a. 932, ed. by Georg Heinrich Pertz (Hannover 1839; repr., 1987), MGH SS 3:54; Annals of Weissembwg a. 932, ed. by Georg Heinrich Pertz (Hannover 1839; repr.. 1987), MGH SS 3:55. Civilization versus Barbarians? 33 Number of dendro data (waney edge/sapwood dating) 600 700 800 900 1000 1100 A.D. moorland way of Sukow: © ringforts in Mecklenburg: (i) around 800 (c) around 950 @ around 980 A.D. ringforts in Brandenburg/Nlederlausitz: (e) around 890 0 around 920 @ around 960 A.D. Figure 3. Construction activities in early medieval ringforts between the Elbe and the Odcr/Neifle rivers: plotting of dendro-dates from Niederlausitz, Brandenburg, and Mecklenburg-Vorpommern. Academy of Sciences in Warsaw clearly confirm this emerging picture.28 In Poland, strongholds of quasi-circular plan seem to have been built even later than those on the frontiers with the Prankish and Ottonian empires. In conclusion, the association between strongholds, on the one hand, and military or political activity, on the other, was much stronger than previously thought. By contrast, social and economic Friihe Burgen in Masowien and die Entstelnmg des Piastenstaates: Archaologie und Dendrochronologie zur Geschichte Europas im 9. und 10. Jahrhundert, ed. by Marek Dulinicz and Joachim Henning (Bonn, 2006), in preparation. 34 JOACHIM HENN1NG developments, while of fundamental significance, do not seem to have had a direct influence on the decision-making process leading to the construction of strongholds. The circular ringforts emerged as a Central European phenomenon of the early Middle Ages, first tested against Viking raids into coastal areas (Netherlands and Frisia),29 then against Slavic and Magyar attacks across the Elbe. This type of stronghold was later adopted by Vikings (e.g. the so-called Trelleborg forts in Den- mark^) and Slavs. By the late 800s. the geographic distribution of circular strong- holds had reached the river Oder. Shortly before or after 900, it crossed the Vistula. From the point of view of fortification methods, the eastern frontiers first of the Carolingian and later of the Ottoman Empire were much more areas of cultural equalizing, compensation, and exchange than a shield of civilization against barbarians.11 29 For the Netherlands, see R. Van Heeringen, P. A. Hendrikx, and A. Mars, Vroege- Middeleeuwse ringwulbiirgen in Zeeland (Amersfoort, 1995). For Holstein, see Dirk Laggin, "Die Stcllerburg in Ditmarschen", Hammaburg, 9 (1989), 191-98. For Brittany, see Jean- Pierre Nicolardot, "Elements de datation du champ de Peran, Pledran (C6tes-du-Nord)\ in Brelagne, pays de Loire. Touraine. Poilou d I epoque merovingienne: Actes de la vf Journee Nationale de I Association Francaise d'Archeologie Merovingienne, Rennes, Jtiin 1984 (Paris, 1988). pp. 73-77. 30 Else Roesdahl. Dendrochronology and the Viking Studies in Denmark, with a Note on the Beginning of the Viking Age", in Developments around the Baltic Sea in the Viking Age: The Twelfth Viking Congress, cd. by Bjom Ambrosiani and Helen Clarke (Stockholm, 1994), pp.106-16. 11 I wish to express my gratitude to Angela Ehrlich (Frankfurt am Main) for the computer- graphics drawing of Figures 2 and 3.