Landscapes of Power, Landscapes of Conflict - [PDF Document] (2024)

Landscapes of Power, Landscapes of Conflict - [PDF Document] (1)

Landscapes of Power, Landscapes of Conflict - [PDF Document] (2)

Landscapes of Power, Landscapes of Conflict

State Formation in the South Scandinavian Iron Age

Landscapes of Power, Landscapes of Conflict - [PDF Document] (3)


Series Editors: Gary M. Feinman The Field Museum, Chicago, IllinoisT. Douglas Price University of Wisconsin – Madison

Ofer Bar-Yosef, Harvard University • Christine Hastorf,University of California - Berkeley • Jeffrey Hantman, University of Virginia • Patty Jo Watson, WashingtonUniversity • Linda Manzanilla, Universidad Nacional Autónoma de México • John Parkington, University ofCapetown • Klavs Randsborg, University of Copenhagen • Olga Soffer • University of Illinois • Matthew Spriggs, Australian National University • John Yellen, NationalScience Foundation

Editorial Board:


FOUNDATIONS OF SOCIAL INEQUALITY Edited by T. Douglas Price and Gary M. Feinman

LANDSCAPES OF POWER, LANDSCAPES OF CONFLICT State Formation in the South Scandinavian Iron Age Tina L. Thurston


LIFE IN NEOLITHIC FARMING COMMUNITIES Social Organization, Identity, and Differentiation Edited by Ian Kuijt

A Continuation Order Plan is available for this series. A continuation order will bring delivery of each new volume immediately upon publication. Volumes are billed only upon actual shipment. For further information please contact the publisher.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (4)

Landscapes of Power, Landscapes of Conflict

State Formation in the South Scandinavian Iron Age

TINA L. THURSTON Baylor University

Waco, Texas

KLUWER ACADEMIC PUBLISHERS New York Boston Dordrecht London Moscow

Landscapes of Power, Landscapes of Conflict - [PDF Document] (5)

eBook ISBN:Print ISBN:

©2002 Kluwer Academic PublishersNew York, Boston, Dordrecht, London, Moscow

All rights reserved

No part of this eBook may be reproduced or transmitted in any form or by any means, electronic,mechanical, recording, or otherwise, without written consent from the Publisher

Created in the United States of America

Visit Kluwer Online at: http://www.kluweronline.comand Kluwer's eBookstore at:


Landscapes of Power, Landscapes of Conflict - [PDF Document] (6)


Tina Thurston’s Landscapes of Power; Landscapes of Conflict is a third-generation processual analysis of sociopolitical evolution during the Iron Age in southern Scandinavia. Several red flags seem to be raised at once. Are not archaeologists now postprocessual, using new interpretive approaches to un-derstand human history? Is not evolution a discredited concept in which so-cieties are arbitrarily arranged along a unilinear scheme? Should not modern approaches be profoundly historical and agent-centered? In any event, were not Scandinavians the ultimate barbarian Vikings parasitizing the complex civilized world of southern and central Europe? Tina Thurston’s book focuses our attention on the significant innovations of anthropological archaeology at the end of the twentieth century.

A brief overview of processual archaeology can set the context for ap-preciating Landscapes of Power; Landscapes of Conflict. During the 1960sthe emergent processual archaeology (a. k.a. the New Archaeology) crystal-lized an evolutionary paradigm that framed research with the comparative ethnography of Service and Fried. It was thought that human societies pro-gressed through stages of social development and that the goal was to dis-cover the evolutionary prime movers (such as irrigation, warfare, trade, and population) that drove social and cultural change. By the 1970s prime movers had fallen from favor and social evolution was conceived as complicated flows of causation involving many variables. Internal dynamics of social strat-ification, status rivalry, and ritual identification became important especially to the Michigan group, heavily influenced by Flannery, and to the British so-cial archaeologists following Redrew’s lead. In the 1980s self-questioning byyoung processualists was remade into Hodder’s radical critique. He pointed to fundamental weaknesses of the original processualism—their dehumaniz-ing of the evolutionary process, the historical uniqueness of each region, the


Landscapes of Power, Landscapes of Conflict - [PDF Document] (7)


enthocentrism of evolutionary approaches, and so forth. But where from there? Processualism lives, although much transformed. During the 1990s American archaeology has become dominated by the progressive Young Turks of the 1970s, and their students have internalized the critique of the postprocessualists to fashion a sophisticated revival in theoretical and methodological approaches to social evolution. Tina Thurston (Wisconsin, 1996), student of rebel processualists Douglas Price (Michigan, 1975) and Gary Feinman (CUNY, 1980), exemplifies the revival.

To read Tina Thurston’s book is to experience how the new processual archaeology is transformed. She focuses on a single dramatic case: the devel-opment of the state in southern Scandinavia during the Migration and Viking Periods. Central is the historical specificity of place and time.Working withinthe broader school of Scandinavian scholarship, the specifics of the historic moment are important. The anthropologist’s (or historian’s) challenge, met by Thurston’s work, is to transport the reader to the unique circ*mstances of another world and, at the same time, to highlight the general processes re-sponsible for specific changes. Change in Iron Age Scandinavia was forged in the landscape by a radical restructuring of the economy, needed to finance new state institutions, through a fundamental change in the settlement of people on the landscape. The world Thurston reconstructs is alive with peo-ple—farming families trying to make a living and state-building chieftains try-ing to mold autonomous regions into larger polities. The state-buildingprocess that she describes is extraordinarily “volunteeristic.” The fierce Vikingchieftains were an unruly lot, and their unification required leaders to subtly mold cultures of collaboration.

Thurston accomplishes her goals by presenting an impressive amount of evidence. She makes full use of the available medieval documents that in-clude annals, chronicles, and narratives both from contemporary written sources of southern writers and from Scandinavian oral histories later tran-scribed. These data sources allow for a broad view of Scandinavia through the later Iron Age. Thurston then focuses on a specific region in southern Swe-den to combine archaeological settlement survey using phosphate analyses,test excavations, and a historical analysis of place names. This is what “his-torical” archaeology is all about, the systematic use of all available sources tocross-check validity, fill in gaps, and provide views of the changes at the dif-ferent scales of human experience. Beyond being a model for European ar-chaeology, her work should be taken seriously by medieval historians as showing how a reasonable understanding of the period must involve large-scale and systematic archaeology.

Beyond the historical specifics of southern Scandinavia, Thurston hassubstantially built a useful case for comparison in our investigation of socialevolution. The critiques of social evolution that came out of the 1970s and

Landscapes of Power, Landscapes of Conflict - [PDF Document] (8)


1980s showed conclusively that unilinear schemes are inadequate. The sim-ple response is to return to area-specific studies, assuming that any general process of social evolution is inappropriate or unknowable—a mistake not made in this book. We require many well-documented archaeological cases, analogous to the earlier ethnographies, to understand the divergent pathways to complexity.

During the first half of the twentieth century, ethnographers established a corpus of descriptions that permitted comparative analysis of human soci-eties. Is the same enterprise possible for archaeology? I would argue that it is. As our simplistic theories of social evolution have crumbled, we require stud-ies of variation in social evolution; Tina Thurston’s book provides an excel-lent example of what these studies should look like. First, they should be diachronic. The problems with the classic ethnographies were their timeless character, societies frozen in time. Cultures, however, are works in progress, and it takes detailed historical and archaeological work to capture them as changing dynamically within historically constrained circ*mstances. I advo-cate a comparative history through archaeology, looking at process of change rather than societies as finished works. Second, systematic comparison re-quires cannons of adequate evidence. Comparison across cases is made ex-traordinarily difficult because rarely do we have comparable descriptions. Her use of settlement analyses that have become routine across many regions of study is just what will make comparisons feasible.

Thurston has accomplished a lot. She gives anthropological archaeology a rich case for secondary state development, new to many of us interested in comparative studies of social evolution. Her book provides a substantial re-source that begs for rigorous comparisons of state building.

Timothy K. Earle Northwestern University Evanston, Illinois

Landscapes of Power, Landscapes of Conflict - [PDF Document] (9)

This page intentionally left blank.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (10)

PrefaceReading Landscapes of the Past

Human beings experience and transform the natural world to a human worldthrough their direct engagement as reflective beings within its sensuous, material reality. . . . The symbolic appropriation of the world produces distinctive lifestyles and distinctive landscapes which are historically and geographically specific. Seiz-ing and understanding this dimension of human interaction with nature and its role in the ordering of space is the task of cultural geography. (Cosgrove 1983:15)

If this book had been written ten years ago, it likely would have been called Regions of Power, Regions of Conflict. During the twenty year period be-tween 1970 and 1990, many Euro-American archaeologists came to under-stand that the study of large geographic areas (or regions) built on a firm. foundation of site-specific studies revealed a great deal more than perspectives that made vast assumptions based only on small-scale investigations. By the time I began graduate studies, this approach had matured into an established theoretical orientation, and methodologically—through systematic survey and empirical observation of spatial changes through time—it had vastly improved our understanding of single sites, subregions, and whole culture areas.

I was trained in this tradition, and it was one of the defining foundations of my thinking on archaeological problems. Yet almost as soon as I felt at home in my empirical, positivist skin, the irritating and uncomfortable ideas of critical theory, context, phenomenology, and agency were thrust on me. There is no need here to reiterate the so-called processualist versus post-processualist debate that raged in most academic departments during this time. Following the lead of just about everyone in American archaeology, for several years I scorned such notions-how can one recover anything butartifacts, features, and distributions? Recovering intentionality was laughable,


Landscapes of Power, Landscapes of Conflict - [PDF Document] (11)


the role of individuals impossible to discern. The foolishness of such ap-proaches was so obvious!

Yet as I immersed myself in my chosen time and region—late prehistoric and protohistoric southern Scandinavia, I began to realize that there are a number of ways to understand more than artifacts and regions—by studying the changing patterns of landscapes and through them, people, places and the relationships between them. If you add other humanistic cultural pro-ductions, which in my time-area is indeed possible, even more can be dis-cerned. I did not need to run away with a postmodern circus or even reject my background in empirical regional studies to do this. I merely had to add more dimensions to what I already knew how to do. T. Douglas Price, whose influence on my work has been profound, sometimes calls himself an “en-lightened processualist.” This volume might best be viewed similarly.

Thus, the study that follows is, I hope, a hybrid of an empirical, objec-tive archaeological survey and a cultural study—more subjective in nature— of an inhabited, or peopled, time and place. The difference between what you read here and the “regional analysis” I planned before my “enlighten-ment” is mainly in the incorporation of two elements. The first, cultural con-text, consists of informing archaeology with the scraps of indigenous symbolism, myth, ethnohistory, and what we can infer about social organiza-tion. Of course, not every archaeologist has such material to work with, but a good many do. I do not believe that this type of evidence, which is a far deeper well than is apparent on the surface, is exploited nearly as much as it could be. The second element consists of landscape analysis, because in many ways it is easier to see the linkages between the socioeconomic, polit-ical, and ideological realms of culture while looking at landscapes, because landscapes are a type of cultural production, unlike regions, which are merely geographic areas. Understanding this difference was the turning point in my conceptualization of how to understand the past.

While “landscape studies” originated as an energetic and sometimes radical subdiscipline within cultural geography, the perspective long ago overflowed these confines and has entered into many disciplines, from an-thropology to social history to environmental studies. Landscapes, as they are currently conceptualized, are orthogonal, interpenetrating combinations of natural and built features, produced, altered, used, or conceptualized by hu-man beings. Embedded into them are human purposes, ideals, plots, and schemes: “The reading of the landscape is an immensely complex task. . . . It involves the rigorous testing of evidence of every kind, together with a close examination, not only of the forms and functions of buildings and boundaries but also of the ideas, assumptions, and priorities of those who made and used them” (Reed 1984a:7). The chapters of this book contain many more specific discussions of the landscape concept and how it can be used, judiciously, in

Landscapes of Power, Landscapes of Conflict - [PDF Document] (12)


archaeology, so I will merely say that it attempts to integrate many aspects of culture through the study of the natural and ideational worlds, the empirical and the subjective.

In the following study, I describe both a traditional archaeological survey and a landscape analysis, the case in point of course being my own archaeo-logical research area—the parts of southern Scandinavia that constituted the late Iron Age through early medieval polity of Denmark. This time period— just a tad longer than the first millennium A.D., from the first century A.D. until about A.D. 1200—was largely prehistoric, yet encompasses the region’s trans-formation from many autonomous, complex, middle-range societies into a unified and centralized state. This transition has long puzzled a variety of scholars: archaeologists grappled with interpreting single sites or microre-gions, and historians—ill equipped to analyze prehistoric activities—found that early texts contained conflicting information on the history of this area. It was not until 1980, when Klavs Randsborg published his landmark work Viking Age Denmark: The Formation of a State, that archaeologists consid-ered approaching these issues by examining Denmark on a larger, regional scale. My work, in effect, attempts not to overturn the conclusions of this original and later regional analyses, but to augment them in a number of ways and to explain the sequence more fully and more comprehensively,

In doing so, I also hope to present a case study that might be useful to others. I hope to encourage people to try methodologies they may not have used before, such as landscape analysis and soil chemical characterization. I also hope that by presenting a well-grounded integration of cultural context with archaeological evidence, more people will try contextualization before dismissing it as storytelling. On a broader level, I believe that locational and organizational changes in southern Scandinavia, as reflections of the so-ciopolitical and economic processes that underlie them, are not unique and have a great deal of comparative value for those who are interested in state formation and its many possible trajectories.


The study presented in this book was conceived in 1989, and in one form or another has been ongoing since that time. Over the past years, the assistance, guidance, advice, and cooperation of many people has been crucial.

I would like to especially thank T. Douglas Price for sharing his knowl-edge and insight and extending both day-to-day and long-term support. I would also like to thank Gary M. Feinman for his willingness to spend long hours discussing the theoretical and the practical with me. I extend special thanks to Timothy K. Earle for his longtime interest in my work and for his

Landscapes of Power, Landscapes of Conflict - [PDF Document] (13)


ideas, which have in turn inspired my own. Vance Holliday taught me about the archaeology of soils, Neil Whitehead helped me to evaluate the docu-mentary record, and Robert Ostergren familiarized me with the physical and cultural geography of Scandinavia. I would also like to thank Mark Bassin, who worked to provide me with an understanding of cultural landscape stud-ies, both classic and radical. James Burton advised me on laboratory methods and soil chemical characterization. Other colleagues, both at the University of Wisconsin and other institutions, lent their time to discussing this work: I would like to thank friends and colleagues Susan Kepecs, Sophia Perdikaris, Bernice Kurchin, Janet Romanowicz, William Middleton, and David Yoon.

The first years of this project were guided by two mentors: the ever-helpful Thomas H. McGovern, who nurtured my interest in Nordic archaeol-ogy, and Gregory A. Johnson, whose work and teaching inspired the basic concepts underlying much of the study now in your hands. In addition, Iwould like to thank my Swedish colleagues and advisers: Johan Callmer of Humboldt University, Berlin, and Lars Larsson of the University of Lund, Swe-den, without whose help and support this project could never have been completed, and Berta Stjernquist and Marta Strömberg of the same institu-tion, for their generous sharing of data and insight.

In Denmark, I would like to thank Klavs Randsborg for his initial help in getting me started, and the late Stig Jensen of the Antikvarisk Samling Mu-seum in Ribe for teaching me to dig in Danish soil. In Sweden, my success was largely due to the assistance and support of my friends at the National Board of Antiquities: Bengt Jacobsson, Bengt Söderberg, Håkon Thoren, Katalin Sabo, and Claes Pettersson. Clæs Pettersson of the RAÄ and Måna Ols- son of the Malmö Museum worked in the field with me and were vital to thecompletion of the project; their friendship as well as their archaeological ex-pertise can never be undervalued. My thanks also to the dozens of Scanian farmers and landowners who permitted me to work on their property and provided me with coffee when I had to work in the snow!

My fieldwork was largely supported by the National Science Foundation, which through three NSF Fellowship years and a Dissertation Improvement Grant allowed me to complete my field research. The University of Wisconsin-Madison Graduate School also provided helpful travel funds.

Finally, thanks to Ted, Danya, Brion, and most of all to my husband Nestor E. Zarragoitía, for being so supportive of my work, so willing to dis-cuss all its aspects, and for putting up with my long absences from home.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (14)



Chapter 1 • The Secondary State in South Scandinavia . . . . . . . . . . . .

1.2 Theoretical Background for the Danish State . . . . . . . . . . . . . . . . . . 41.3 The Substrate of State Formation in Denmark . . . . . . . . . . . . . . . . . 6

1.5 Historical Documents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91.6 The Framing of the Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

and Hierarchy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101.6.2 Ethnicity and Unification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111.6.3 Hegemony, Domination, and Resistance . . . . . . . . . . . . . . . 12

1.8 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Chapter 2 • A Multidisciplinary Study of the State . . . . . . . . . . . . . . . . 17

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172.2 The Archaeological Dimension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

2.2.1 The Large Scale: Analysis of the Supraregion . . . . . . . . . 172.2.2 The Intermediate Scale: Analysis of the Region

or Province . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182.2.3 The Smallest Scale: The Analysis of Local

Settlement Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19


. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31.1 Introduction

1.4 Alternative Methodologies for Differently Organized States . . . . 8

1.6.1 Corporate and Network, Heterarchy

1.7 Organization of the Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13


Landscapes of Power, Landscapes of Conflict - [PDF Document] (15)


2.3 The Historical Dimension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

2.3.2 History and Archaeology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212.3.3 Historiography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Types of Historical Document . . . . . . . . . . . . . 23Primary Historic Texts and Their Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 The Self-Chronicles of Scandinavia . . . . . . . . . 252.4 The Geographic Dimension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

2.4.1 Archaeological Perspectives on Landscape . . . . . . . . . . . 282.4.2 Cultural Geography and the Cultural Landscape . . . . . . 292.4.3 Interpreting Archaeological Landscapes . . . . . . . . . . . . . 31

Dynamic Landscapes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

362.5.2 Landscape Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362.5.3 Ideological Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372.5.4 Change in the Power Structure . . . . . . . . . . . . . . . . . . . . . . . 37

2.6 Contextualizing the State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392.8 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

Chapter 3 • Prestate Denmark: The Earlier IronAges . . . . . . . . . . .

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413.1.1 Physical Geography and Cultural Ecology . . . . . . . . . . . 413.1.2 Cultural Ecology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

3.2 The Roman Iron Age: Courses Toward Complexity . . . . . . . . . 44

3.2.2 Interpreting Tacitus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453.3 Political Economy and Social Organization . . . . . . . . . . . . . . . . 483.4 Warfare and the Social Order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513.5 Regions, Boundaries, and Integration . . . . . . . . . . . . . . . . . . . . . . . 52

The Cultural Landscape A.D. 1 to 400 . . . . . . . . . . . . . . . . 533.5.2 Organization of Chiefly Centers . . . . . . . . . . . . . . . . . . . . 543.5.3 Evidence for Other Chiefly Centers . . . . . . . . . . . . . . . . 56Changes in the Balance of Prestige and Trade . . . . . . . . . . . . . 573.6.1 Political Economy in the Germanic Iron Age . . . . . . . . . . 583.6.2 The Age of Emporia: A.D. 700-900 . . . . . . . . . . . . . . . . . . . . 59

2.3.1 Historians. Prehistorians. and the Tyranny of the Historical Record . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

2.5 Studying the Corporate State Through Its

2.5.1 Courses Toward Unification: Changes in Cultural Landscapes, Ideologies, and Power Structures . . . . . . .


3.2.1 Primary Texts: Scandinavian Cultures in the Roman Era . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

3.5.1 Polities of the Roman Iron Age:


Landscapes of Power, Landscapes of Conflict - [PDF Document] (16)


3.7 Polities of the Germanic Iron Age: Courses Toward

3.7.1 Primary Texts: Boundaries and Integration Political Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

in the Germanic Iron Age . . . . . . . . . . . . . . . . . . . . . . . . 623.7.2 Primary Texts: Forms of Rulership . . . . . . . . . . . . . . . 66

3.8 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69


Chapter 4 • The Viking Age in Denmark . . . . . . . . . . . . . . . . . . . . . . 73

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

and Peer Polity Interaction in the Early Viking Age . . . . . . 74

Provinces of Denmark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

4.4.1 Change and Crisis in Viking Age Kingship . . . . . . . . . . 834.5 Strategies for Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

4.5.1 Militarism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 864.5.2 The Transformation of the Law . . . . . . . . . . . . . . . . . . . . . . . 884.5.3 Urbanization and Political Economy . . . . . . . . . . . . . . . . . 90 Urbanization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

4.5.4 Village Organization and Agriculture . . . . . . . . . . . . . . . . . 97

4.1 Introduction4.2


4.4 The Nature of Rulership in Viking Age Denmark . . . . . . . . . . 82

Analysis on the Macroscale: External Forces

Analysis on the Intermediate Scale: The Coalescing Production and Distribution at the Transition to the Viking Age . . . . . . . . . . . . . . .

Agricultural Systems and Rural Settlement in the Late Iron Age . . . . . . . . . . . . . . . . . . . . . . . 97


96 Change in Village Organization During the Viking Age . . . . . . . . . . . . . . . . . . . . . .

4.5.5 The Sacred Landscape: Geographic and Ideological Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Pre-Christian Religion in Scandinavia . . . . . . . 103 The Study and Practice of Old Norse Religion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 The Social and Political Implications The Natural Offering Place and the Built Temple: Personal Versus

of Religion and Religious Change . . . . . . . . . .

Institutionalized Religion . . . . . . . . . . . . . . . . . . 107


Landscapes of Power, Landscapes of Conflict - [PDF Document] (17)

xvi CONTENTS The Construction of the Late Christian Landscape: The Church as a Function of the Town . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 Princes of the Church . . . . . . . . . . . . . . . . . . . . . . 1094.6 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

Chapter 5 • Social Classes in the Viking Age: Contentious Relationships . . . . . . . . . . . . . . . . . . . . . . . 113

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1135.2 Stability and Change in Social Relations, Institutions,

and Political Ideology in the First Millennium A.D. . . . . . . . . . . . 113

5.2.2 The Terms of Social Relations . . . . . . . . . . . . . . . . . . . . . . . 1155.3 The Social Code in Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1175.4 Change in Society and the Terms of Social Relations . . . . . . . . 1195.5 Political Specialists in the Archaeological Record . . . . . . . . . . . . 124

5.5.1 Vorbasse and Omgård: Period I . . . . . . . . . . . . . . . . . . . . . . . 1255.5.2 Vorbasse and Omgård: Period II . . . . . . . . . . . . . . . . . . . 125 Vorbasse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1255.5.2.2 Omgård . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126

Class Conflict in the Early Middle Ages: The Early Historic Record . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1285.6.1 Royal Power and Unification in Viking

Age Denmark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1295.7 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

Chapter 6 • Landscapes of Power andLandscapes of Conflict . . . . 131

6.1 Geographic Approaches to Culture Change in Iron Age Denmark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

6.2 Methods of Geographic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 1336.3 Theories of Locational Geography . . . . . . . . . . . . . . . . . . . . . . . . . . 135

6.3.1 A Discussion of Organizational and Decision-Making Hierarchies . . . . . . . . . . . . . . . . . . . . . . . . . 136

6.3.2 The Jarrestad Region as an Organizational and Decision-Making Unit . . . . . . . . . . . . . . . . . . . . . . . . . . 136

6.4 Rank-Size Analysis and Related Methods of Investigation . . . . 1376.4.1 Explaining Regional Patterns . . . . . . . . . . . . . . . . . . . . . . . . . 1426.4.2 Complementary Statistical Analyses . . . . . . . . . . . . . . . . . 143Patterns of Places: Locational Analysis as an Indicator of Large-Scale, Long-Term Change . . . . . . . . . . . . . . . . . . . . . . . . 146

5.2.1 Viking Age Sociolinguistics . . . . . . . . . . . . . . . . . . . . . . . . 114



Landscapes of Power, Landscapes of Conflict - [PDF Document] (18)


6.5.1 Regional Locational Analysis of Greater Denmark: 147

6.5.2 The State of the State I: Greater Denmark Through Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147

Synchronic and Diachronic Variability . . . . . . . . . . . . . Period I: A.D. 700-850 . . . . . . . . . . . . . . . . . . . . . 1486.5.2.2 Period II: A.D .. 85 0-950 . . . . . . . . . . . . . . . . 1556.5.2.3 Period III: A.D .. 950-1050 . . . . . . . . . . . . . . 1566.5.2.4 Conclusions: Part I . . . . . . . . . . . . . . . . . . . . . . . 159

Through Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1606.5.4 Interpreting Regional Patterns: The Case of Scania . . 163 Conclusions: Part II . . . . . . . . . . . . . . . . . . . . . . . . . 1706.6 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173

6.5.3 The State of the State II: Greater Denmark


Chapter 7 • Reconstructing Cultural Landscapes in Southeast Scania . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177

7.1 The Prehistoric Cultural Landscape: An Overview . . . . . . . . . . 1777.1.1 Reconstructing Landscape in Southeast Scania . . . . . . . . 177

Water Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179Topography, Subsoils, and Soils Associated with Iron Age Habitation Sites

and the Tommarp River Valley Hydrology—Watercourses and Registered Prehistoric Monuments and . . . . . . . . . . . . . . 182

Previous Research in Järrestads Härad. . . . . . . . . . . . . . 184 Phosphate Studies in Archaeology: Discovery, Development, and Use of Phosphate as an Archaeological The Mechanics of Soil Indicator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185

Phosphate . . . . . . . . . . . . . . . . . . . . . . . . . . 1887.1.1.4b Existing Phosphate Data . . . . . . . . . . . 188 Documentary Evidence: Cartographic

Visual Inspection and Surface Collection: Ceramics in Viking Age Denmark . . . . . . . . . . . 190

Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

Landscapes of Power, Landscapes of Conflict - [PDF Document] (19)

xviii CONTENTS Documentary Evidence: Pre-Cartographic Prehistoric Settlement Indicators . . . . . . . 192Place Name Categories: Regional, Provincial,Village. and Parish Names; Farm Names,Field Names. and Terrain Names . . . . . . . . . . 1937.1.1.8a “National” and Regional

Place Names . . . . . . . . . . . . . . . . . . . 1947.1.1.8b Internal Administrative

Place Names . . . . . . . . . . . . . . . . . . . 1947.1.1.8c Harad Names . . . . . . . . . . . . . . . . . . 1947.1.1.8d Parish Names . . . . . . . . . . . . . . . . . 1967.1.1 .8e Village Names . . . . . . . . . . . . . . . . . . 1967.1.1.8f Farm and Field Names . . . . . . . . . . 1987.1.1.8g Terrain Names . . . . . . . . . . . . . . . . . 198

7.2 Fieldwork Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1997.2.1 Sampling Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2007.2.2 Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2027.2.3 Survey Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2027.2.4 Soil Core Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2027.2.5 Controlled Surface Collection . . . . . . . . . . . . . . . . . . . . . . 2037.2.6 Test Excavation Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2047.2.7 Feature Investigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205

7.3 Site Patterning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2067.4 Stratigraphy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2077.5 Laboratory Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2097.6 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210

Chapter 8 • Place, Space, and Experience in



Iron Age Communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213

The Development and Transformation of a Prehistoric Cultural Landscape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213Landscapes of Järrestad’s Roman Iron Age: A.D.. 1-400 . . . . . . 2148.2.1 The Economic Landscape of the Roman Era . . . . . . . . . 2148.2.2 The Political Landscape in Järrestad’s Roman

Iron Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2158.2.3 The Sacred Landscape in Järrestad’s Roman

IronAge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2188.3 The Germanic Iron Age: A.D.. 400-700 . . . . . . . . . . . . . . . . . . . . 219

Internal Integration and Hierarchy in Järrestads Härad . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222


Landscapes of Power, Landscapes of Conflict - [PDF Document] (20)


8.3.2 The Sacred Landscape in the Germanic and Early Viking Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223

8.3.3 Rural Demography and Interior Colonization . . . . . . . 2238.4 The Viking Age in Järrestad . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230

8.4.1 Early Christianization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2378.5 Contextualizing the Landscape and the Record . . . . . . . . . . . 238

to Political Reality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2408.5.2 Interpreting Saga Content for Its Cultural Message . . . 242

8.6 Geography and the Alteration of Everyday Experience . . . . . 243

Cultural Landscape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246

Sociopolitical Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2498.6.3 An Urban Center in Järrestad . . . . . . . . . . . . . . . . . . . . . . . . 251

8.7 Local Christianity in The Late Viking Age . . . . . . . . . . . . . . . . 2558.8 Landscapes in Järrestad A.D. After Restructuring . . . . . . . . . . . 2568.9 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260

Chapter 9 • Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263

9.1 Power and Force, Time and Space . . . . . . . . . . . . . . . . . . . . . . . . . . . 263

8.5.1 Class and Conflict: From Saga and Song


8.6.2 Intensification, Population, and

The Seizure and Appropriation of the

9.2 The Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2649.3 Goals of the Research: Local, Regional, and

Cross-Cultural Study of Alliance, Unification, and State Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264

9.4 Refining Models of Location and Organization . . . . . . . . . . . . . . 2659.5 Increasing Understanding of Demographic Change . . . . . . . . . 2669.6 Advocating the Utility of Soil Chemical Survey . . . . . . . . . . . . . 267

and Territory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2689.8 Extending Cross-Cultural Comparisons . . . . . . . . . . . . . . . . . . . . . . . 269

9.10 Linkages Between Power, Place, and History . . . . . . . . . . . . . . . . 2719.11 The Peopling of the Archaeological Record . . . . . . . . . . . . . . . . . . . 272

9.7 Moving Beyond Simple Ecological Models of Dominion

9.9 Political Development in Light of Some Previous Models . . . . . 270

9.12 Conclusions: Landscape and the Power of Place . . . . . . . . . . . . . . 274

Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285

Landscapes of Power, Landscapes of Conflict - [PDF Document] (21)

This page intentionally left blank.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (22)

Part I Theoretical, Archaelogical, and Historic Background

Landscapes of Power, Landscapes of Conflict - [PDF Document] (23)

This page intentionally left blank.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (24)

Chapter 1

The Secondary State in South Scandinavia


This book is about power: hegemony, domination, centralization, and unifi-cation. It is also about conflict: ethnic strife, resistance, regionalization, and the tenacious persistence of traditional social and political institutions. In one respect it is a case study of developing political complexity in the pre- andprotohistoric contexts of the South Scandinavian Iron Age, focusing on the diverse political and ethnic groups that were becoming Denmark, a society in which a number of interrelated processes resulted in a profound social transformation: the development of a secondary state. The primary focus of this study is the explication of when the unification occurred and how it was achieved. However, the analysis moves beyond the narrower realm of the case study, for the investigation of this sequence has clear implications for many regions and times where striking similarities may be found.

Secondary state formation is a process crucial in human history; al-though far more common to human experience than the establishment of primary or pristine states, it remains a grossly under-studied and underdocu-mented phenomenon. The internal processes leading to such formations, as well as interactions in the regions that produce them, are often turbulent— one need only look at historical cases of core-periphery relationships, hege-mony, and the ethnic unrest of empire-building to appreciate the climate of such episodes, and it cannot be denied that precisely the same powerful and critical processes operated in the past.

Secondary states also have the potential for far greater complexity and internal diversity than primary states (D’Altroy 1992: 1), because they are


Landscapes of Power, Landscapes of Conflict - [PDF Document] (25)


often forged from several distinct social aggregates. The negotiation of power relations and cultural identity between both allied and antagonistic interests becomes an important part of the equation. Entities within the burgeoning secondary state are often resistant to the domination of the center, some-times leading to long drawn-out processes of unification during which rulers must come to terms with internal subgroups and carefully strategize to achieve their goals. The same resistance can, under other conditions or at dif-ferent times, result in violent purges and the removal of local leaders. This can lead to an uneven process of unification, as the strategies used to con-trol the core areas are often inadequate for dealing with outlying regions (D’Altroy 1992:1). The formation of the Danish state reflects many of these cross-culturally observable circ*mstances.


In this book, I investigate the processes through which disparate cultural and political entities, some territorially distant from the core, are incorporated into secondary states. The formation of states has long been a central question in ar-chaeological research, and there are numerous theories that concern increas-ing politically complexity. Among the more familiar are elite control of trade(Rathje and Schiffer 1982; Drennan 1984; Upham 1982), crafts production/sur-plus (Earle 1978; D’Altroy and Earle 1985), a growing need for information pro-cessing and organizational hierarchy (Johnson 1976, 1982, 1983), populationpressure (Cohen, M.N.1973, circ*mscription and warfare (Carneiro 1970),and finally the supervision or organization of intensive agriculture (Wittfogel 1957; Hastorf 1993). While none of these conditions alone appear to be“prime movers” for societies across the board (Feinman and Neitzel 1984), many of these elements, combined together in different ways, are key factors in the pro motion of inequality and increased political complexity.

We can attempt to identify elements important in the development of individual societies and apply techniques for studying them that have proveduseful in many case studies. For example, change in agricultural strategies is often linked to sociopolitical change, and the visible manifestations can be identified: the appearance of large, centralized storage facilities and the or-ganization ofmass labor for agricultural intensification projects. In the sameway, elite economic control might be linked to the large-scale warehousingof surplus trade goods, and the rise of a ruling class might be studied throughthe creation of secular and/or sacred elite architecture. These examples, and a host of other characteristic, idiosyncratic changes that often accompany theshift to a state-level society, are used as proxy indicators for the underlyingchanges in political inequality, political and economic centralization, and the

Landscapes of Power, Landscapes of Conflict - [PDF Document] (26)


development of social and economic classes. These indicators are good and useful ways of looking at change, and because they are frequently visible and quantifiable, are rightly the core of many studies.

Yet to study these changes in Iron Age Denmark one must often do so obliquely, because the formation of the Danish state appears to be accompa-nied by few, if any, of these key indicators. Denmark has innumerable villages and households to study, potsherds and tools to analyze, but for most of the Iron Age, the area is lacking in “typical” indicators of rulership, political power, and class— palaces, cities, storehouses, monumental religious or state architecture, and indigenous texts that record the important political affairs or economic transactions of the state.

Until very recently, the absence of such typical indicators prompted both historians and archaeologists to imagine that before the early Middle Ages (be-ginning ca. A.D. 1075), no state ever formed at all in Denmark, that society in the so-called Viking Age consisted of warlike, fragmented groups similar to the bands that roamed Europe during and after the fall of Rome several centuries earlier (Falco 1964; Heer 1981; Polanyi 1978). Some archaeologists (i.e., Wells 1984:200) believed that the earliest towns of Denmark, founded in the ninth century, were in territories “never ruled by state forms of government” and had no relationship to centralized political and administrative functions.

Within Denmark itself, the emergence of states during this period has long been accepted in the disciplines of archaeology and history (Chris-tiansen 1968, 1969; Hodges and Whitehouse 1983; Randsborg 1980), but per-haps for the wrong reasons. At least two major archaeological studies in the early 1980s (Randsborg 1980; Roesdahl 1982) asserted that a Danish state formed in the earliest parts of the Viking Age-While these were, in their time, remarkable attempts at integrating a large amount of diverse data, they had serious drawbacks, especially in their criteria for describing or identlfying states. Randsborg, for example, defined a state as “a large, stable political unit with a high level of production” (1980:7). It is now clear that states may be large or small: consider a Greek city-state versus the Inka empire. States, es-pecially archaic states, may also be unstable, and tributary states may not have much production at all (Cohen 1978; Claessen and Skalnik 1978; Claessen and van de Velde 1991). Roesdahl based the transition from chief-doms to states on lists of attributes similar to Childe’s indicators of “civiliza-tion” (Childe 1950), now usually considered to be associated with both complex chiefdoms and state-level societies.

If one does monitor and track the state-like attributes—fortresses, store-houses, palaces, and temples—a state seems to appear in the decades be-tween A.D. 1000 and 1050. This “sudden” appearance has been explained with several simplistic theories. One suggests that towns and markets were founded and administered entirely by non-Scandinavian continental intruders

Landscapes of Power, Landscapes of Conflict - [PDF Document] (27)


(Cohen, M. N. 1977:325) who simply set up shop among the backward native peoples. Archaeological evidence belies this; although much contact with other groups is indicated, these places are indigenous. More often cited among these “sudden state formation” theories is one in which the Christian church is credited with the almost overnight development of a monarchy and the state (Cohen 1977:313), yet we will see that although in later times kings and churchmen enjoyed some episodes of mutual support, the church was not established in Denmark until well after a state is observable.

Why is the long-term process of state formation in Denmark so seem-ingly invisible? Social and political geographer James S. Duncan tells us that “not everything that is real and which has causal power can be observed or experienced” (1990: 13). The deficit in Denmark’s “archaeology of the state” is not due to the absence of a state, nor is it a matter of preservation or recov-ery. It is a function of historically contingent social and political processes. The key to this uneven development is in the antagonistic resistance of tra-ditional, horizontally organized institutions to centralization processes.

In his seminal work on social and political complexity, Flannery (1972) stressed that increasing hierarchy is concomitant with increasing centraliza-tion, an idea that has become embedded in many contemporary views of state origins (i.e., Roscoe 1993). If this is true, then Denmark is an unusual ex-ample of state formation. Compared to the states of the Maya, Inka, Minoans, Shang, and many others, there is little evidence of political and economic centralization.Why? Must we, as Feinman (2001) asks, “simply chalk up these differences to unique cultural traditions that have no comparative lessons or theoretical implications” for the study of other sequences? The answer is no: hierarchy and centralization do not necessarily go hand in hand, and the ar-chaeological indicators often considered to correspond directly to state for-mation are only inherent in cases where wealth and power quickly become concentrated in the hands of a few elites, and (not surprisingly) are lacking when states form in different ways. As has recently been demonstrated through studies of the Puebloan Southwest (Blanton et al. 1996) and com-parative analyses of Maya states versus the Teotihuacan polity (Feinman 2000, 2001), there is variety at all levels of political complexity, and all states are not the same in terms of political economy, rulership, and organization. Therefore, they look different “on the ground.”


Let us briefly examine the setting of Danish state formation. Throughout the Early and Middle Iron Age, ca. 500 B.C.–A.D. 700 (Figure 1.1), the region later known as Denmark was composed of numerous stratified, autonomous po-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (28)


Figure 1.1. Chronology for Northern and Western Europe.

litical systems, some with distinct ethnic identities (Figure 1.2). During the Late Iron Age or Viking period (A.D. 700-1074), these polities passed not only from prehistory into history, but from existing as several chiefly societies to unification with an emergent centralized state that developed in a core area in southwestern Denmark. The social code of these Early and Middle Iron Age chiefdoms, described in greater detail below, was one in which leaders and followers had mutual bonds and obligations, and rulers required the assent of an assembly in order to rule, an assembly made up of all free, adult males, who participated in making decisions, and could also depose or assassinate these leaders. In times of external warfare, an overlord was elected from the pool of chieftains, retaining authority only until the threat was neutralized. Such a code discouraged both centralization and unification. Despite all these “checks and balances,” a centralized state eventually emerged.

Denmark thus presents the full formation sequence of a differently orga-nized state: a unification couched in a transition from a corporate mode of gov-ernment to a network, or exclusionary type of rule (Blanton et al. 1996; Feinman 2000, 2001). In a corporate mode, labor, food production, social groups, and even rulership may be controlled through “broad, integrative ritual and ideological means” (Feinman 2000) and extreme differentiation between leaders and followers is suppressed. A network, or exclusionary, mode of gov-ernment differs dramatically in that it stresses “personal prestige, wealth, power accumulation, elite aggrandizement, highly individualized leadership, lineal pat-terns of inheritance and descent, personal networks, longdistance exchange, exotic wealth, princely burials, and the specialized manufacture of status-re-lated craft goods” (Feinman 2000). State building during such a transition gives rise to complications and ramifications far different than in societies where

Landscapes of Power, Landscapes of Conflict - [PDF Document] (29)


Figure 1.2. Denmark in the Late Iron Age.

centralized, network-style power structures already exist. The documentation of corporate states and a comprehensive understanding of their inner workings is a relatively new and understudied area of exploration.


A key issue I wish to address is the problem of studying the rise of such dif-ferently organized states, a topic of interest and debate in recent archaeologi-cal literature (Blanton et al. 1996; Feinman 2000, 2001). With so many of our expectations geared toward the recognition of network-style political organi-zation, such corporate societies have in the past been labeled anomalies, and their seemingly problematic and confusing components debated back and forth with few comprehensive models or theoretical frameworks, and hence, without much success. We have established that corporate states leave a dif-ferent sort of archaeological record than more typical, network-style forma-tions. Rather than merely noting that such atypical polities occur, we must

Landscapes of Power, Landscapes of Conflict - [PDF Document] (30)


learn to recognize the social and political conditions that bring them forth and formulate explicit, proactive research programs to identify and study them.

During the course of this research, a methodology was developed that is specifically geared toward studying and identifying key indicators under these atypical political conditions. While individual rulers, pomp, and ceremony are largely absent from the Danish record, changes in whole landscapes, and the relationship between their institutional elements, can be studied.

In order to comprehend such changes, regional analysis was undertaken at three scales: supra-regional, regional, and local. The local analysis was based on survey of settlement systems in Scania, now a part of southern Swe-den, but until 1658, the easternmost province of Denmark. Site locations and their dynamics through time were established and reconstructed, revealing the rise and fall of centers, changes in their placement, function, and size, re-structuring of the agricultural villages that these sites administered, and the shifting locations of economic nodes. Such processes can be monitored with the use of geographic statistics such as rank-size curves, span of control ra-tios, and boundary shifts measured by changes in Theissen polygons.1


Statistical work was augmented by indigenous and foreign historic documents, sociolinguistic change, and shifts in ideological concepts interwoven in myths, sagas, and laws. Not long ago, we were exhorted by some archaeologists (e.g., Binford 1962, 1968; Clark 1973) to exclude history as a culturally constructed institution that can be deceptive and distracts us from interpreting the past. This rejection of history in archaeology effectively precluded any insights that the written record gives into prehistoric problems. Post-processual approaches viewed social change as hyperhistorical or unique to each time and place, neu-tralizing any headway made in understanding the past through the careful study of cross-cultural patterning. I take a different approach, applying histori-cally specific information and systematic cross-cultural analyses of archaeologi-cal data to the problem of Danish state formation.

As have a number of recent investigations (i.e., see Knapp 1992a), I also borrow the interdisciplinary approach of the Annales historians, in which temporal, spatial, and behavioral variations are studied by using concepts from numerous disciplines. Leone (1 978) noted that there is a strong relationship between Annales methods and archaeological approaches, as both emphasize time, space, and change. Braudel defined three temporal “rhythms”—long, medium, and short-term scales of analysis (1980:25–54). Historical documents

1Rank-size curves and other geographic analyses are discussed in chapter 6.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (31)


give insight into the short and medium perspectives, which in this study are contrasted with the long-term perspective of the archaeological record. The simultaneous consideration of these “separate but parallel sources of informa-tion” (Smith 1992:5l) provides explanations for the complex changes accom-panying the transition to centralization in the study region. In addition to western European texts and official histories, I review indigenous documents, runic inscriptions, sagas, court poems, and travelers’ accounts. These written sources are supplemented by a large body of archaeological data that has been analyzed primarily by Scandinavian and German scholars, data that are not readily accessible outside of northern European institutions and libraries.

These research methods provide a rich database, resulting in a holistic and detailed picture of the development and transformation of regional set-tlement and sociopolitical systems through time, the intertwined processes of the unification of disparate groups and corporate-to-network transitions.


In framing this study, I proposed that imposing a shift from one form of au-thority to another-from corporate to network modes—was easily the most challenging problem faced by elites with ambitions toward centralized power. This would not have been a matter of superimposing a new hierarchy over an old one. Instead, it involved the virtual reinvention of the power structure, from what might be termed a strongly heterarchical politicalmode to a hierarchic system.

1.6.1 Corporate and Network, Heterarchy and Hierarchy

Heterarchy is a useful concept, envisioning the simultaneous operation of both vertical (hierarchic) and horizontal (heterarchic) differentiation (Potter and King 1995: 17). Hierarchically ranked decision making is a familiar enough concept; it incorporates tiers or levels of authority where decisions made on higher levels are spread along regional and local channels though a vertical chain of command over information and power (power to enforce central decisions). A horizontal structure of authority instead “occurs when societal elements are perceived to be functionally distinct but are either un-ranked or occur at equivalent rank within a hierarchic structure” (Potter and King 1995: 17).Within each Danish chieftain’s territory this structure existed. There were separate channels of political decision making: a warlord versus an assembly, functionally distinct bodies yet unranked in the sense that nei-ther warlord-warband nor assembly was ultimately superior to the other.Yet at the same time, a vertical hierarchy operated within the warband, and soci-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (32)


ety was certainly divided along class lines. Heterarchical organization is often a feature of corporate systems.

Corporate power sharing is also seen at the larger scale, in the macro-structure of Late Iron Age rulership: a group of equal status elites who each ruled their own chiefdom and shared power even when acting as a larger en-tity under a single elected overlord. The utility of a concept like heterarchy is to describe and model such societies, where a clear vertical structure oper-ates but is intertwined with other balancing mechanisms that prevent any group or individual from gaining too much power.

In modern democracies hierarchy and heterarchy are, ideally, combined, whereas in pre-modern modes, government is more likely to be skewed to-ward one extreme or another, a supreme elite or a group-oriented decision-making structure (Potter and King 1995). In Late Iron Age South Scandinavia, the hierarchic tendencies of rulers were long subsumed by heterarchically based political and social structures. Over time, this focus shifted, from the rule of many equal elites balanced with a powerful assembly of citizens, to a highly centralized system focused on the person of an individual ruler, under whom the power of the nonelite and lower-level elite was rapidly eroding.

1.6.2 Ethnicity and Unification

One of the most difficult parts of this research has been explaining how a centralized state rises from a political system with such strong preventive measures against centralization, and how a group of federated, autonomous polities becomes a unified entity, even though regional rulers had a substan-tial interest in forestalling the dismantling of the system. It is easy, almost in-tuitive, to imagine that in later Iron Age south Scandinavia, several related groups that occupied obvious districts divided by rivers, islands, or forests, practiced similar religions, customs, and culture. Presented with the threat of conquest, first by the Romans, later by the Franks, they eventually came to-gether as a state for mutual benefit.

Unfortunately, this does not fit anything of what we know from the eth-nographic literature on the unification of diverse groups (Friedman 1989; Bentley 1987). In fact, unification and state formation, even when voluntary, appear to bring out conflict and create ethnic identities that were previously only subconscious (Shennan 1989). Several political entities that unified with the emergent centralized state in southwestern Denmark also constituted dis-tinct ethnic groups. These can be identified by correlating regional archaeo-logical data with the study of the historic record, which is sparse and fragmentary, yet yields important data. Around A.D. 800 (Figure 1.2), when southwestern Danish elites first tried to extend their authority into adjoin-ing regions, the coalescing polity included subgroups known as Scanians,

Landscapes of Power, Landscapes of Conflict - [PDF Document] (33)


Hallanders, Pleicani, North Danes, South Danes (or Sillenders), and Isle Danes (Tschan 1959; Callmer 1991 a).

As will be seen, the archaeological data agree with these ethnographic examples in which, as Skalník (1989) describes, ethnic and other socialgroups “outwit the state” to avoid subsumation within it: research in Scania supports a model of resistance to the state, and since the Scanians were a self-identified ethnic group, it is possible to study ethnic conflict throughout the period. Without evidence for early domestic warfare, how was the subjuga-tion of this resistant province achieved?

1.6.3 Hegemony, Domination, and Resistance

I suggest that in order to unify many autonomous regions, state-building rulers did not at first opt for “conquest” per se, but employed a number of subtle strategies for integration. Many of these strategies involved the manipulation of places. Others involved the manipulation of ideas. The result was a condi-tion of voluntary yet uneasy union that might best be called a hegemony.

Why would a ruler begin subtly rather than with decisive force? Al-though domination depends on direct coercion, hegemony refers to the leadership or control of cooperating groups by one of their number, in his-torical context often referring to the control of a region by one polity in a league. Hegemonic control is a complex interweaving of political-economicconditions and sociocultural values that work to bind together disparate peoples by creating a desire to identify with the most influential group. Domination requires the use of force, whereas hegemony can be promoted through power. The difference between force and power, as defined by Luttwak (1976), is that force involves the use and depletion of resources— warriors, supplies, and money—while power can operate without such con-sumption, and is therefore less costly but often as effective as force itself (Mattingly 1992).

In many cases, a majority power seeks to maintain the status quo, while a small subordinate minority seeks to overthrow it. In the case of Den-mark, we see the exact opposite: rulers are the small group attempting the imposition of an alien network-style system upon a long-standing corporate group that was much larger and not at all powerless. Coercion may not have been a wise choice in such a situation, at least not until other methods were exhausted. Hegemonic control is directly linked to what is perceived as proper, correct behavior, the “natural order of things” within a society. For would-be Danish rulers, always fearful of deposition or death at the hands of the assembly, inducing target populations to think that they were part of the state, rather than attempting to force them into it—might have been an attractive option.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (34)


The strategy of opting for power instead of force worked well in the areas closest to the core, where organizational and hierarchic change is seen clearly and quickly in the archaeological-geographical record. However, loca-tional analyses show that after nearly 100 years of hegemony, central elites of the Early Viking Age were no closer to actually controlling the distant province of Scania than they had been at the outset. Agriculturally rich, densely populated, and valuable for taxation, Scania alone was nearly as large and powerful as the whole of Western Denmark, and it presented a problem for rulers wishing to bring the region under central control. Archaeological and historic evidence indicates Scania's tenacious resistance to abandoning corporate structures and accepting centralized authority. Rulers were forced to try a number of strategies for breaking down corporate systems and suc-cessfully integrating Scania's territory and wealth into the state.

These issues—the problems of differently organized states, domination and resistance, ethnic identity, hegemony and unification—go far beyond a case study, and address problems of hierarchy, centralization, and control in regions as far-flung as the ancient Near East (Stein 1994), Mesoamerica (Blanton et al. 1996; Feinman 1997), the Southwestern United States (Mills 1997; Feinman 2000). My goal in examining Denmark as an example of a long-term sociopolitical transition is to add breadth to current theoretical frameworks and to increase the analytical tools available for comparative study of the less common, decentralized states. With its excellent potential for the recovery of archaeological data and its fragmentary but important documentary record, Denmark provides a model for the multistep process of unification and state formation, grounded in the inherent struggle be-tween corporate- and network-based concepts of rulership.


The chapters of this book are organized into three large parts. The first part—chapters 1 through 3—presents theoretical, archaeological, and historical background. Chapter 1 seeks to establish the precepts and boundaries of the study and its theoretical orientation in the broadest possible manner, famil-iarizing the reader with the concepts and the problems to be approached. In chapter 2, the background to the study of these problems is examined in greater detail, outlining the difficulties inherent in the study of a Danish state, familiarizing the reader with the ethnohistoric and historic record. A sug-gested alternative to traditional studies of late prehistoric social complexity is discussed, the theoretical framework of prehistoric cultural geography. Itprovides a summary of results that may be kept in mind until the study ad-dresses the results in detail. Chapter 3 discusses the physical geography and

Landscapes of Power, Landscapes of Conflict - [PDF Document] (35)


reviews the evidence for social, political, and economic conditions in the pre-Viking Iron Age, the social baseline for all subsequent changes.

The second part of the book—chapters 4 and 5, describes evidence for the fundamental changes that occurred during the Late Iron Age. Chapter 4 discusses Viking Age landscapes using many lines of evidence, and subjects them to interpretation. It also examines the state-encouraged conversion to Christianity, which in Denmark was a peaceful, unforced process that took a long time to implement. While the official adoption of the new religion, be-ginning in the Middle Viking Age, did not affect internal organization, a trans-fer of control of Christian practice, from local to state-level contexts, occurred in the Late Viking Age. The hierarchic nature of Christianity with its single god permitted rulers, already growing more and more powerful, to impose new taxes, laws, and regulations, over and above those required by the secular state.

Chapter 5 identifies class and social change through sociolinguistics. My inquiry into the earliest law codes of Scandinavia had led me to the linguistic research that documents the adoption of a whole new range of social terms, many of which are foreign in their origin, during the precise period of great-est change seen in the study, the Late Viking Age. There is a rare opportunity here to study the evidence for a new, unfamiliar social order through changes in the terms of social relations. The adoption of foreign concepts and symbols is often an indicator of the adoption of new ideologies by a ruling elite (Earle 1991a:7). Indigenous texts indicate that this was occurring over the course of the Viking Age. The chapter also discusses archaeological evidence to cor-roborate the rise of a new, nontraditional ruling class.

Part 3 consists of chapters 6 through 9, presenting the results, analysis, and conclusions that reveal the strategies for unification and the methods by which elites solved the problems of integrating many previously autonomous areas. Chapter 6 presents the locational analyses and the study of large-scaleand small-scale landscape change that are used to overcome the lack of “ad-ministrative” remains and documentary evidence. The two major regions in-volved in the unification, east and west, underwent dramatically different sequences of integration (Thurston 1990, 1996), and a clear set of strategies was employed by rulers during the period of unification in order to accom-plish their goals of fully controlling and exploiting all their territories. Chap-ter 7 describes the archival research and field methods used to reconstruct local landscapes in Järrestad. Chapter 8 examines the changes in political, economic, and social organization seen at the local level and discusses its re-lation to changes in the larger state.

Conclusions about Iron Age Denmark, as well as the implications of this study for other regions of the world and archaeological timeframes form the final chapter, Chapter 9. Changes in landscape and social structure were

Landscapes of Power, Landscapes of Conflict - [PDF Document] (36)


probably not the result of random events or some vague (but typical) histor-ical notion of a normative progression from one thing to another. This re-search highlights the enormity of these changes and the planning that went into them, as well as the attempt to resist them through the persevering power of places, the tenacity of internalized cultural ideologies, and the change-resistant concepts of what is “good, just, and true.” With this current understanding in mind, it is highly improbable that South Scandinavia would have taken the trajectory it did without a specific long-term elite goal of dis-mantling and reshaping their society to an amenable form.


This chapter outlined the problems and approaches to the study of unifica-tion during the Late Iron Age in south Scandinavia. Locational analyses and cultural geographic methodologies on three scales of analysis—local, provin-cial, and regional—can resolve the problematic nature of the archaeological record of central authority, while sparse but important documents can con-textualize the transition to a state. Research questions and methods are summarized, and a summary describes how different parts of Denmark underwent widely divergent courses of integration and that a clear set of strategies was employed by rulers during the period of unification in order to accomplish their goals of fully controlling and exploiting all their territories. The following chapter will detail the problems and solutions for studying the transition to state-level society in Denmark.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (37)

This page intentionally left blank.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (38)

Chapter 2

A Multidisciplinary Study of the State


This chapter describes the three scales of analysis used in the study—thelocal, the regional or provincial, and the supraregional—and the research questions appropriate to each, then considers the main components of the multiple-approach method I have advocated: archaeology, history, and geog-raphy. Since archaeological evidence comprises much of this book, its treat-ment here is brief, but it includes artifact and site distribution, burial evidence, and anthropogenic soil chemical characterization. The textual record is important but cannot be used indiscriminately, nor should one ex-pect history and archaeology to perfectly reflect each other. Yet they can complement each other if used prudently. Finally, in order to overcome diffi-culties inherent in studying a corporate state, I describe what might be termed prehistoric cultural geography, particularly the analysis of cultural landscapes, as the overarching paradigm of this study.


2.2.1 The Large Scale: Analysis of the Supraregion

In this study I argue that archaeological data indicate a multistep process of state formation that becomes highly visible when examined through long-termchange. The scale at which state formation occurs is large; to answer broad questions of integration and control, the appropriate scale of archaeological


Landscapes of Power, Landscapes of Conflict - [PDF Document] (39)


analysis is supraregional. At the largest scale, the study examines the relation-ship of contemporary European political entities with Denmark as it coalesced as a political entity. Secondary state formation often appears to be associated with relationships that develop in prehistoric world systems (Kohl 1988; Schneider 1977). As discussed in chapters 3 and 4, Denmark was peripheral first to Rome, and later to the Franks, and these culture contacts may have acted as catalysts for internal change (Hedeager 1988; Hodges and Whitehouse 1983; Kristiansen 1991). How did such relationships hinder or fuel the course of Danish state formation? The prestate era is characterized by growing trends toward network-style rulership, seen in the striking shifts toward elite-con-trolled long-distance trade, a prestige-goods economy, status and class-ladenburials, elite-attached craft specialization. Regional analysis indicates that as the early state coalesced, its leaders were engaged not only in conflict with neigh-boring states, notably the Frankish Empire, but also embroiled in internal up- heaval. This coincides with changes in the organizational hierarchy of Western Denmark, apparent in locational analyses, indicating increasing stratification and integration in the heartland of the incipient state, while other, more distant areas, such as Scania, remained autonomous and disarticulated with the work-ings of central government.

2.2.2 The Intermediate Scale: Analysis of the Region or Province

As Hastorf notes (1993: 17), many archaeological models of state formation focus on what she terms the being rather than the becoming of states. Be-cause the process of state building in Denmark was lengthy, uneven, and emerged from a society where change from less to more political inequality was being consciously and unconsciously negotiated, looking at the process as an episode of becoming is a profitable paradigm. This becoming, with its uneven development, calls for a different scale of analysis: that of regions within the state. The core and peripheries of the state can be examined in-dependently and compared, giving insight into the differential success elites may have had in integrating new territories and the strategies they used for overcoming organizational problems. At this intermediate scale, we can ask what pressures might have influenced the earliest-incorporated autonomous polities to voluntarily surrender their authority to a centralized leadership. Al-though small-scale raiding was common in the prestate societies, there is no indication that central elites waged large-scale warfare on leaders from ad-joining regions and imposed their own rule. If anything, there appear to be episodes of vicious infighting among the core-area leaders in the early Viking Age (Scholz 1972:94, 96). When the eastern region of Scania was finally an-nexed, how did power-wielding factions in Western Denmark extend their ju-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (40)


risdiction eastwards into Scania and suppress existing political structures? Lo-cational analysis shows that the east and west underwent substantially differ-ent sequences of political and economic centralization, as well as settlement pattern changes. Strategies used in the east appear to be directly related to improving integration and control.

2.2.3 The Smallest Scale: The Analysis of Local Settlement Systems

Finally, the local-scale analysis of a single settlement region permits a human-scale perspective and a sharper focus on the changes in the organization of social, political, and economic systems and the concomitant development and transformation of local cultural landscapes. These detailed and specific data are supplied by the survey of Järrestads Härad in southeast Scania.Throughout the Early and Middle Iron Age (500 B.C.–A.D. 700), Scania, as well as the rest of southern Scandinavia, was composed of a number of ranked, then stratified, autonomous political systems (Hedeager 1987: 30; Kristiansen 1991: 16). The general area of Järrestads Härad appears to havebeen one such system. The local analysis approaches Järrestad as a dynamic and changing human construct under three different sociopolitical and eco-nomic regimes, the Late Roman/Germanic Iron Age, the Viking Age and the Early Medieval eras. Järrestad illustrates, in microcosm, the processes ac-companying state formation.

Evidence for settlement patterns in Järrestad was gathered through the use of controlled surface survey and soil chemical characterization, which forms one component of the research. Inductively-coupled plasma spec-troscopy (ICP) analysis of about 2,000 soil samples was used to map the ex-tent of villages in Järrestad through the distribution of phosphate, a process that will be discussed at length in later chapters. Within each village the dif-ferential distribution of phosphate, as well as artifactual material, sheds light on the internal organization and spatial structure of the settlement. This tech-nical study required the use of one group of methods and theories having to do mainly with archaeometry.

A second group of methods and theories comes into play once the results of the first group are known; these consist of locational-analytical investiga-tions of my own data combined with published findings on neighboring re-gions. The results of the local-level study indicate that villages began as loosely organized communities and became highly regulated, both administratively and spatially. Settlements originate as many places with equal size and function and are transformed into a clear hierarchy of places with differentiated pur-poses. Expanding these regional analyses back up through our three scales, we see that regions that were isolated and detached from the reach of central

Landscapes of Power, Landscapes of Conflict - [PDF Document] (41)


power are integrated through new centers and institutions. There are shifts in boundaries, centers rise and fall, villages are reorganized. These changes, on all levels, can be measured through changes in rank-size distributions, span of control, and levels of administrative hierarchy. New socioeconomic classes are observed through developments in household organization.


2.3.1 Historians, Prehistorians, and the Tyranny of the Historical Record

Until recently, information pertaining to the political development of Euro-pean states came solely from written documents. Thus, the formation of kingdoms, the interaction of nation-states, and the emergence of social and economic forms in Europe during late antiquity (A.D. 1 - 400) and the Early Middle Ages (A.D. 400- 1100) has traditionally, and almost exclusively, been the domain of historians. Writings dealing directly with Denmark consist of sporadic records that span over 1,000 years, from Roman authors to Me-dieval chroniclers. This cumulative documentary evidence from ancient times is now incorporated into modern historical consensus regarding Viking states. The use of archaeology to study these issues was originally only an adjunct to history, a way to verify or support the documentary record. Timothy Champion aptly defined the problem that this creates for archaeologists:

We might similarly think of the tyranny of the historical record in the sense that the programme for the archaeology of the historic period in Europe is set by his tory, and the historic vision of the past. Not only are the materially based studies of archaeology regularly subordinated to those of the literary record, but the en-tire conceptual framework of questions and evidence is limited by historical con-cerns ultimately rooted in a vision of European uniqueness and implied superiority. (Champion 1990:91)

However, over the last few decades, a small group of European scholars made attempts at using archaeological data to illuminate processes totally outside of the scope of history (i.e., Becker 1980; Hedeager 1978; Randsborg 1980). Since the 1970s, archaeology has also yielded evidence that runs counter to some modern historical interpretations, leading many to question the incontestable validity of historical documents. These scholars have in-stead highlighted the limitations of relying solely on the written word (Hodges 1982a, 1989; Hodges and Whitehouse 1983). This does not mean that historic documents are not an important source for archaeologists. The profitable and productive coupling of documentary and archaeological data is imperative to a comprehensive study of the region.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (42)


2.3.2 History and Archaeology

For archaeologists working in eras that produced surviving primary texts, and especially for those working in late prehistory, the useful integration of his-toric documents with archaeological evidence becomes a factor in research strategies. However, as mentioned briefly in chapter 1 , there is a long tradition of the rejection of history in archaeology, led by such notable scholars as Lewis Binford (1962, 1968) and David Clark (1973). Central to this viewpoint is the perception of “proper” archaeological method. Early processual archae-ologists emphasized model building, theorizing, and hypothesis testing, in the fashion of the natural sciences, a groundwork that still underlies the con-struction of many research projects, including my own. The anticipated result is a law or set of laws or generalities about human behavior (Knapp 1992b3). The opposite approach, often considered a tenet of so-called post-processualarchaeology, views behavior as directed by human agency, specific events, and social relations that are historical in origin and may not be subject to any “laws” but are instead unique to each time and place.

Both these paradigms were overly extreme, and we now comprehend that archaeological and textual data, generality versus specificity, are not as mutually exclusive as they once were thought to be. Historical interpreta-tion in the last two decades has become less preoccupied with the single event, a few high-profde personalities, or the philosophies of one or two in-dividuals. Recently, natural scientists have begun to examine “historical as-pects” of their data (Gould 1986), and archaeologists have not been far behind. Trigger (1995), for example, appropriately advocates an approach that integrates the more predictable generalities with historically specific aspects of culture.

Knapp (1992b) noted that while some archaeologists became more in-terested in historic processes, some historians adopted a social science method, referred to as the Annales School of history. Annales historians, such as Bloch and Braudel, rejected surficial, elite-generated history. The first prin-ciple of the Annales method was to take an interdisciplinary approach in which temporal, spatial, and behavioral variation is studied by utilizing con-cepts from sociology, anthropology, economics, geography, and other disci-plines. The second principle was to create a vast, diverse database relating to the past of all classes, gender, age, ethnic, and other groups in a society, at-tempting to reconstruct history from many written sources other than the “official” state versions. Tax records, church rolls, legal documents, marriage and childbirth statistics, merchant records, and any other documents that told the social history of a population were studied for information. This in it-self resembles the archaeological approach to examining cultural change through many types of evidence.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (43)


Braudel (1980:25-54) emphasized the longue durée, or long-term, large-scale analysis of culture change as the centerpiece of Annales history: macro-histories, covering centuries and even millennia were examined; the short-term phenomena and the ephemeral trends of “microhistory” were less important. Later Annales scholars amended this view, seeing “events” as “in-tersections which break patterns,” now considered an integral part of Annales historical explanation (Le Roy Ladurie 1979: 111 – 116). Historians acknowl-edge, as do many archaeologists, that their field is “a composite of different times moving at different speeds” and that they must “first define the object of their research and subsequently determine the time span most relevant for data analysis” (Furet 1984:6–9).

Although Annales history is not in the mainstream of current historical thought, it is a fruitful approach to history in an archaeological context. Leone (1978) noted that there is a strong relationship between Annales meth-ods and archaeological approaches that emphasize time, space, and change. Braudel’s use of different types of temporal “rhythms”—long, medium, and short-term scales of analysis—viewed side by side with archaeological indi-cators of the same type may highlight the correlations between sometimes sketchy mythic, ethno- and protohistories and archaeological records— “sep-arate but parallel sources of information” on a study area (Smith 1992:51). Not every type of social, political, or economic change is equally reflected in the historic and the archaeological records, but many are. Points where they do not co-vary are as interesting as those that do.

2.3.3 Historiography

Foucault (1972) defined history as a discourse, or a “tradition in which rep-resentations of the past, especially textual and pictorial, create knowledge of the past and in a sense also create the reality of the past itself; history deter-mines the scope and content of the past at least as much as the past deter-mines history.” This characterizes the first problem encountered in the historical study of Iron Age Denmark: the historiography of Medieval Europe. Historiography can be defined as the study of how history is written—bywhom, with what agenda, from which perspective. The situation is not dis-similar to the problem of European contact with indigenous peoples in later times in that the historic perspective on Denmark is almost exclusively that of outsiders, in this case usually Frankish, German, or English chroniclers.

It is important to remember that during the period with which the study is concerned, Scandinavia was not a part of “Europe” in the eyes of Medieval chroniclers, but an invading, competitive, hostile entity—much the same as the Moors, Saracens, and the other Infidels of European history—who, much to the chagrin of Western European elites, occupied parts of Europe, or non-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (44)


European Christian outposts (the Holy Land, for example) in the Middle Ages. The communities of these early writers were at first intermittently, and then continually disrupted, first by wandering or invading Germanic tribes, later by Vikings.

The early centuries of recorded European history are thus filled with ac-counts of Scandinavians portrayed as a fierce, unreasoning people who roamed the unknown places just outside the civilized boundaries of the Western world. Devil-worshipping (that is, not Christian) priests sacrificed on their alters many who ventured forth to trade with or convert them. These barbarians lived close to the European milieu, and it is not remarkable that several formal attempts were made to conquer and/or colonialize Denmark, including forays from Rome, by Charlemagne and the Holy Roman Empire, various Medieval German rulers, and the Hansa. However, all were ultimately unsuccessful.

There was constant friction where the European desire to conquer the North rubbed against the outrage and apprehension of the Danes. During the so-called “Viking Age,” despite the clarity with which primary historical doc-uments show aggression on both sides (Chamberlin 1986), the continental view of the “northmen” was that they were the sole aggressors. This view, cur-rent in the minds of eighth- and ninth-century authors, persists in the pre-dominant views held by many of their modern counterparts today. New interpretations, rather than serving to demilitarize or “whitewash” the Vikings, should instead revise the perspective on the equally rapacious ex-pansionism of other European powers. We need at this late date to finally abandon the Eurocentrism of the Medieval West.

I suggest that the wrong questions have been asked of the historic record, the political forms of the late Iron Age misidentified within the his-toric paradigm, and that the methods of regional and landscape analysis pro-posed for studying political structure in this study are one way of avoiding and moving beyond these pointless debates. Rather than rehashing the fa-miliar and largely subjective topics of pagan versus Christian, Viking “bar-barians” versus “civilized” nation-states, the study examines this time of international conflict as the result of an internal power struggle that ensued when Scandinavian rulers attempted to force a continental political model on an unwilling populace. Types of Historical Document

Several types of documents provide historical data for Northern and Western Europe: annals, chronicles, and narratives, usually the product of scribes and clerics in England, Frankia, and elsewhere. Each provides a different type of information.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (45)


In the annal, social and natural events seem to carry equal weight and there appears to be little intention on the part of the unknown annalist to record certain events over others, to “rank” them in terms of importance. Chronicles can be both objective and subjective and sometimes offer per-spective, judgments, ethnocentric viewpoints, yet are not narrative stories; no final point or moral is made. Finally, there are narratives such as Adam of Bremen’s Gesta Hammaburgensis Ecclesiae Pontificum (History of the Archbishops of Hamburg-Bremen), Nithard’s Histories, Saxo Grammaticus’sGesta Danorum, and others. Narrative is the most difficult type of document to interpret. Some of these historical discourses have clearly stated biases and perspectives: Nithard, who chronicled the life of Charles the Bald, son of Charlemagne, beginning in about A.D. 814, wrote: “When you had innocently suffered your brother’s persecutions for almost two years, as you, my lord, know best, you ordered me, before we entered the city of Chalons, to write the history of your time” (Scholz 1972: 129).

Other narratives submerge their true perspective and present the view-point of the hidden narrator as objective. Thus the “events seem to tell them-selves” (White 1987, p. 3), their subjectivity masked with an aura of truth. The central question regarding narrative is, What desire or wish is enacted by the fabrication, or the fantasy that there is a “real story” with some coherence and direction that lurks behind the chaotic historical record? White (1987:4) states that narrative conceits give us a “glimpse of the cultural function of . . . the instrument with which the conflicting claims of the imaginary and the real are mediated, arbitrated, or resolved . . .”

In archaeological terms, this means the legitimation of the actions and ideologies of those who control the inscribing of these great narratives. These “stories” tell the culturally and politically specific elite versions of conquests, religious subjugation, expansionism, and other situations in which there are conflicts, force, or even simply one side or opinion against another. Of course, the more “trustworthy” types of documents—annals and chronicles— can also be deceptive inasmuch as their omissions may stem from unknown motivations on the part of those who inscribed them. They require as much source criticism as the narratives, source criticism meaning the careful eval-uation of the genesis, purpose, scope, and historiography of each document.

Clearly, Late Iron Age Scandinavia was set amid an array of classical and postclassical empires, Medieval kingdoms, dukedoms, principalities, the Papal states, and other political entities that were actively chronicling the his-tory of the world as they saw it. This leaves us with some written references with regard to Scandinavia, but each document must be evaluated for cultural bias, political manipulation, and simple misinformation. Foucault notes (1972:6) that as long as history has existed “documents have been used, ques-tioned, have given rise to questions; scholars have asked not only what these

Landscapes of Power, Landscapes of Conflict - [PDF Document] (46)


documents meant but also whether they were telling the truth, and by what right they could claim to be doing so, whether they were sincere or deliber-ately misleading.” Primary Historic Texts and Their Contributions

History for the south Scandinavian Iron Age begins in the Roman period—the first century A.D.—is recorded sporadically throughout the late Iron Age, and becomes fuller at the end of the study era, ca. A.D. 1200. Therefore, it can give us a part of the picture of change in social institutions and political situations over the longue durée. In this view, when the Royal Frankish Annals are con-sidered, the conflict of the royal personalities of Charlemagne and Godfred of Denmark (the usual topic of historical tracts) is inconsequential, but what can be read between the lines about the form of rulership in A.D. 800 is of import. Do close readings of the texts suggest a sovereign king or a Germanic chief-tain? It is unimportant how Knut (Canute) the Great, at thirteen years of age, boldly conquered the Engish, but it is noteworthy that a child had been made king of a powerful nation, indicating a shift to ascribed status, and salient that English social institutions soon were insinuated into his Danish realm. In read-ing Adam of Bremen’s 11 th-century account of the number of churches in Sca-nia, it is not important that the Scanians were transformed from heathen Odin-and Thor-worshippers into pious Christians, but it confirms the location of vil-lages in that province at about A.D. 1000 when the churches were first built.

Events, those “intersections that break patterns,” also have their place in interpretation. They can help pinpoint the times by which certain changes and transformations were occurring, or had occurred. Thus, we can link ar-chaeological indicators for the unification of Denmark to accounts of diplo-matic meetings between the Danes and the Franks, the founding of urban centers to accounts of military clashes. These two time perspectives, the long term and the momentary, can illuminate some of the questions asked in chap-ter 1 , such as whether social organization changed to accommodate political integration or if conflict arose over change in the sociopolitical order. These texts will be discussed in the sections of the study to which they are relevant. The Self-Chronicles of Scandinavia

There are also indigenous records, the most notable being runestones, mon-uments with inscriptions that record social and political titles, allude to own-ership, property, and inheritance rights, and also vary regionally in form and text style. This allows one to determine if a West Dane, for example, erected a stone in the Eastern Province, and what rights he or she might have had there. The eddas and sagas are epic poems. These frequently record historic

Landscapes of Power, Landscapes of Conflict - [PDF Document] (47)


events, some written contemporaneously to such events, others a century or more later. These must be read with careful source criticism, for they are po- litical documents of the most sophisticated type. Because of specific historic circ*mstances these ethnohistories were often written by dissenters to the royal system, Icelanders who hated kings and loved the older, more corpo-rate ways, and so hold their own special biases. Others were written to glo-rify and support the very same kings that other sagas depict as tyrants. There are also the writings of the skalds, or court poets, who were patronized by royalty and wrote flattering odes to royal personages, often reflecting state-sponsored propaganda and “official history.” There are early Medieval laws that can be searched for hints of earlier conditions, travelers’ accounts that provide the perspective of those familiar with the ways and conditions of the Danes, and early Medieval historic narrative writings from Denmark, Iceland, Norway, and Sweden which actually purport to tell the “history” of these areas during the Iron Age from an indigenous perspective. These are equally subject to the agendas of their authors. There are many relevant documents and a large body of commentary and source criticism about them.

These many types of documents provide us with more than information on rulership, politics, and religion. Part of social reproduction occurs through a process that has been called social memory (Fentress and Wickham 1992). Although the act of remembering is individual in some senses, memories in general are often attached to social groups: the family, the neighborhood, the village, town, assembly, even the farmer-Vikings at the rowing benches of a longboat warship. Maurice Halbwachs, who pioneered the concept of col-lective memory, argued that all memory is structured by group identities; that is, that memories are basically group memories and that it is only through a particular intersection of groups that the memory of the individual is pro-duced (Halbwachs 1925, 1980). This should not imply that the individual is somehow subservient to an overly-reified community will, but it does enable the examination of the social side of an individual’s conscious life.

Fentress and Wickham (1992) state that even the words we use to de-scribe the act (recognize, remember, recall, recount, commemorate, and so on) show that memory can include anything from a “highly private and spon-taneous, possibly wordless, mental sensation to a formalized public cere-mony.” The differentiation between memory as action and memory as representation is important; remembering or commemorating is a physical behavior, while memory is cognitive, a network of ideas. Here, remembering will mean commemoration, the action of speaking or writing about memo-ries, the formal reenactment of the past in a social context.

In the course of the study, I will introduce documents that show con-flicting social memories produced by conflicting groups. There are accounts of assemblies where farmers unite to defy the king, calling on their social

Landscapes of Power, Landscapes of Conflict - [PDF Document] (48)


memory of traditional rights and limitations. These descriptions of gather-ings where verbal recounting and recalling of the past occurred is almost like listening to the expression of social memory. These clash with royally-commissioned, formal attempts to replace the social memory of the farmer with the social memory of the king. These written commemorations de-scribe an invented social order, couched in an imaginary past, where net-work-style power and rulership were not a new, intrusive, and unwanted encumbrance but extended back into primordial times.

These documents are gateways into the collective, social memories of different groups. Yet even while this scant written documentation provides a certain amount of valuable information, it only tangentially supplies the barest of facts, and even these are not always trustworthy. In order to mitigate the lack of history in the writing of this “prehistoric social history‚” the nexus between documentary sources and archaeological material must be carefully navigated. Comparing the historic record with archaeological landscapes can be valuable, because the written history of the Danes often does not match what the archaeological record suggests, indicating an official versus an un-official worldview. In other respects, the two match well. Identifying which social, economic, or political problems give birth to different or coinciding versions of history sheds light on internal conflicts and issues of domination and resistance. What seemingly substantial shifts or reversals are generally ac-cepted by lords and peasants without quarrel? What type of situations are likely to be contested or likely to generate propaganda campaigns? The an-swers to these questions are sometimes unexpected.

In conclusion, the existence of these fortuitous scraps of writing pro-vide details, hints, or allusions to cultural practices that are often missing from the archaeological record. Elusive, yet profoundly important elements, such as concepts of regional identity, internal social conflicts, and ethnicity, become available topics of study, and like other puzzle-pieces that are rou-tinely studied, may be tried and fitted against complementary data. History helps contextualize archaeological observations.


Ultimately, the archaeological changes through time and space, compared with the social and political details gleaned from the textual records, are interpreted through the paradigm of prehistoric cultural geography, a third and final body of theories and explanatory models. A number of useful archaeological frame-works have been adopted and adapted from other disciplines, such as cultural ecology, world systems theory, and, as in the case of this study, cultural geog-raphy. When methodologies and theoretical perspectives developed in other

Landscapes of Power, Landscapes of Conflict - [PDF Document] (49)


disciplines are employed, substantial problems can crop up. The transfor-mation of cultural geography into prehistoric cultural geography is less dan-gerous, for the study of the spatial dimension of human culture is the object of both.

2.4.1 Archaeological Perspectives on Landscape

Within archaeology, the term landscape is associated with several paradigms, including prehistoric cultural ecology and historical ecology. While both these important theoretical frameworks have contributed valuable informa-tion toward understanding culture and culture change, it should be clear that the use of the landscape concept in these frameworks is substantially differ-ent than my own.

Rooted in Steward's early concepts of human interaction with the envi-ronment framed as cultural ecology, prehistoric cultural ecologists viewed landscape as a source of food, a plane divided into territories, and a place to procure raw materials: a thing to be exploited or used to produce things (Steward 1955:40–41). This idea of landscape, enmeshed with Steward's many influential ideas, continues to frnd utility in the work of many contem-porary anthropologists.

Cultural ecologists later added other aspects of culture to Steward's focus on technology and economy (Hardesty 1977, but the view of the land-scape as a largely inert and passive substratum for human activity continued. Humans were seen to interact with the landscape as they made decisions about where to hunt, gather, or cultivate; what items were worth extracting; and whether they were worth the trouble of cooperating with other groups or individuals to procure (Keene 1979; Winterhalder 1981). The landscape is discussed in terms of consisting of patches, grains, and foraging pathways— a supermarket, in which human beings map out aisles, choose brands, and comparison-shop. Although valuable for building behavioral models, such models have been criticized for their mechanistic view of behavior and for their too-literal borrowing from other fields, especially ethology (Yesner 1981; Jochim 1983; Keene 1983).

Many recent studies, while far more sophisticated in their approach and conclusions, retain these earlier ecological concepts and incorporate the term landscape, yet do not differ substantially from earlier ecological studies (i.e., Stafford and Hajic 1992; Camilli and Ebert 1992). They do not fully take up (and are not necessarily meant to take up) how humans shape landscape, or how the landscape, both natural and built, can shape human behavior through its subtle framing of our experience. Historical ecology is also con-cerned with the landscape. This paradigm attempts to look at an historical sequence of ecological changes, with the underlying suggestion of under-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (50)


standing the “sequence and the timing of the causal events that produced them” (Winterhalder 1994: 19).

Most recent studies framed in ecological terms are far more cognizant of the role of landscape in human systems than those of only a decade ago. These are noteworthy works that encompass the study of a full range of human-land relationships. Most are concerned with human interaction with the environment, that is, the natural landscape, and how humans forage in it, farm it, mine it, and control it, how its ecosystems function, what happens if industrialization or climatic conditions change it, and what this means to the human occupants (i.e., Gunn 1994; Schmidt 1994; McGovern 1994).

This focus offers an excellent avenue of investigation into human cul-ture, past and present. However, lest my use of the term landscape cause any confusion, a definition of landscape as it is used in this study is needed. While certain climate shifts and land-use practices will be discussed during the course of the study, it is not the goal of this research to explicate or demon-strate these changes or how they affected Iron Age Scandinavians. Instead, I take my basic definitions of cultural landscapes from the discipline of cultural geography, which focuses on the human construction of the built, conceived, and symbolized landscape, the cultural landscape, and its simul-taneous role in molding our concepts of self and society.

2.4.2 Cultural Geography and the Cultural Landscape

Culture-geographic landscape research attempts to integrate many aspects of culture, using the idea of landscape as the impress of human social relationships, ideas, intentionality, and meaning upon the natural world. Carl Sauer, founder of the Berkeley school of geography, stated that landscape was a cultural expres-sion (1963:26) and that cultural change is recorded in the cultural landscape: that development of one society or its replacement by a second can be read in the succession of landscapes molded by human beings (1963/1925:33). He sug-gested that the combination of “culture” and “time” on the natural landscape pro-duced a number of forms: population density, housing plans and structures, production, land utilization, resource areas, and communication; roads, routes, and boundaries (1963/1925:42–43). In this way cultural geography is a disci-pline that is readily compatible with general archaeological problems.

The theoretical and methodological links between cultural geography and archaeology have not gone unnoticed. Landscape geographer Brian K. Roberts states that

landscape contains within itself the physical remains of many structures, not only castles, great churches, and other buildings, but also boundaries and routeways, settlements and farms, place-names and administrative districts. Only archaeology can reveal with clarity the physical traces of these scenes. (Roberts 1984:224)

Landscapes of Power, Landscapes of Conflict - [PDF Document] (51)


This study employs the concept of a landscape that is an artifact of human actions and intentions, a perspective that is not new to archaeologists who use regional and locational approaches. Yet these approaches, as they are commonly applied in archaeology, are often lacking many of the human-istic elements that cultural geographers attempt to study. In addition to ana-lyzing the physical remains in the landscape, cultural geographers attempt to understand the context of landscape to the people who created it, as op-posed to interpretation through the filter of the researcher’s perception. Dif-ficult as this is, it can be successful, and in this respect, cultural landscape studies address the more elusive questions that the post-processualist para-digm exhorts us to explore.

Like many social scientists, including archaeologists, a number of cul-tural geographers became interested in integrating the objective with the subjective in the 1970s. Then-radical ideas included critical theory and a new awareness of Western biases (Short 1976; Blaut 1980), a revival of revised Marxist interpretation (Anderson 1973; Buch-Hanson and Nielson 1977; Cos-grove 1983), and a desire to integrate the empirical observation of the physi-cal world with the abstractions of those who created it and those who analyze it (Cosgrove 1978; Hufferd 1980).

Some saw human subjectivity as shaping place, each interrelationship unique. Others saw similarities in the human response to landscape stem-ming from the physical needs of humans—our actual use of land as territory, our need to defend against the wild (Appleton 1975), or indeed the human physical shape, the posture, the senses (Tuan 1977). Others (Relph 1976) saw human relationships with land as symbolic expressions of opposition be-tween self—inner identity—and other—outside world. The resolution of the conflict results in the built landscape, a material symbol that represents the process of human-land interaction. Relph (1976: 54–55) stated that the expe-rience of place is a deep biological phenomenon, that places the inner (self, home) in opposition with the outer (everywhere else). Because humans live in social groups, these are not individual symbols but shared perceptions and notions (Cosgrove 1978:68). One recognizes the influence of geographers like Relph in Hodder, especially in his volume The Domestication of Europe (1990). Such older perspectives, as well as more contemporary and better-developed schools of thought, can provide models for geographic or spatial explanations of archaeological phenomena.

Perhaps the greatest difference between cultural geography and archae-ology during this time of paradigmatic shifts is that a substantial attempt at synthesis was made early on. Rather than completely polarizing the disci-pline, the value of both traditional and new approaches in cultural geography was recognized. Cosgrove noted that some basic tensions are inherent in in-terpretations that try to link mind and experience with the objective reality

Landscapes of Power, Landscapes of Conflict - [PDF Document] (52)


of the material world. These tensions are seen equally in archaeology: “At base they are profoundly subjective, while concerned with a material world whose reality cannot finally be ignored” (Cosgrove 1978:67). Cosgrove stated that these theoretical concepts provided useful ideas about the dialectic be-tween ourselves and places, but rejected the phenomenological tendency to become idealist by abstracting collective souls and spirits, ideas and inten-tions, to an unrealistic extent. He writes: “Such abstraction is false: it neither accords with our experience in the world, nor allows us the possibility of understanding the reasons for things” (Cosgrove 1978:70, my italics). This idealism was connected with the failure in cultural geography to link to-gether the idea of the internal dialectic between self and place with the ma-terial world.

For archeological problems, it is perhaps safest to keep highly abstractand phenomenological theories of landscape at a distance, for although such approaches offer interesting perspectives and make us think about important issues, they undeniably pose questions that are currently unanswerable through scientific methods and are unprofitable for prehistoric eras with no links or relationships to any well-understood tradition.

This does not mean that the study of humanistic elements must be abandoned, especially in late prehistoric or protohistoric contexts. Cosgrove suggested that a study linking mind and matter should do two things: study the relationship of landscape development to socioeconomic and political variables and also bring in secondary indicators, such as art and literature, iconography, ritual, social relations and laws, literary works, myths and folk-lore, that illustrate the ideas, motivations, and perceptions of those responsi-ble for change and those who resist it.

2.4.3 Interpreting Archaeological Landscapes

This study employs a cultural landscape approach because it is inclusive, takes into account many aspects of the human past, both empirical and hu-manistic, and is a useful cognitive framework on which to hang relational data. Landscape, in this perspective, is not a background or a stage on which human actions occur, limiting or suggesting certain land-use patterns or ways of life. On the contrary, human-land relationships are a dialectic between hu-mans and their environment, and encoded in the spatial relationships be-tween human-constructed, human conceived, or human-used elements of the landscape, is information that can be studied and interpreted for past re-lationships within and between human systems. This relationship may be un-derstood as the power of place; that is, the role of the natural and built geographic environment in cultural and social processes (Cresswell 1992: 13). In assessing cultural change, several “types” of cultural landscapes may be

Landscapes of Power, Landscapes of Conflict - [PDF Document] (53)


considered: the sociopolitical landscape, the economic landscape, and the sa-cred landscape.

Empirically, in the political landscape are the locations of elite centers and outposts, sites of law making and law giving, the distribution and den-sity of elite government or military presence in a region, and the location and levels of administrative hierarchy. This is not studied by examining evi-dence from single sites but by examining patterns on a regional level. If a pattern of political loci is evident in most of a region but not in a subregion it may indicate either an undetected locus or that the area was administered differently than other areas. Diachronically, the decline of one center and rise of another within the same region may suggest a transfer of administra-tive power from one locale to another. This may be interpreted as func-tional—better access to new types of transportation or technology—or political—a change in rulership reinforced by a shift in the location of im-portant administrative nodes. Of course, it is often the case that more than one factor influences change.

The economic landscape contains the location of raw materials, mar-kets, agricultural and crafts production sites, mints, and tax-collecting activi-ties. Again, it is not necessary to focus on the contents of such sites, but on the way in which these elements are spatially related. This can suggest how strong or weak government intervention was locally, or how well or badly ar-ticulated core areas and hinterlands were. Changes in the nonelite landscape, such as shifts from nucleated villages to dispersed farmsteads and wide-scaleintroduction of intensification strategies, may indicate changes in elite de-mands upon rural subjects. The degree of rapidity or ubiquity of such changes may be interpreted as to how powerful or direct central control ac-tually was or how dependent on the centers—and cooperative with them— outlying areas were.

In the sacred landscape are the monuments of religious activity, whether they are built temples, churches, standing stones, or the sacred groves, lakes, mountains or other natural features perceived as supernatural. Religion under local control is often indicated by dispersed, decentralized locational pat-terns, whereas more structured belief systems are reflected in a more cen-tralized and hierarchic pattern.

Furthermore, where these different landscapes intersect and articulate reveals the relationships between them. In a system with tight central gov-ernment control, there may be a frequent intersection of the economic land-scape and the political landscape. The coincidence or separation of sacred centers with political centers may suggest the amount of articulation be-tween these institutions, or distinguish a locally controlled, small-scale belief system from a state-controlled, politically enmeshed cult. In many systems, the political-administrative hierarchy is directly embedded within the central

Landscapes of Power, Landscapes of Conflict - [PDF Document] (54)


place hierarchy; that is, the individuals who govern on each level, from high-est to lowest, reside in the corresponding geographic places—from capital city to local center. In other systems these may be disembedded from each other. Rulers, lords, and officials may live far from centers, on compounds, es-tates, or manors, where they make decisions that are then operationalized throughout their realms.

Such differing spatial/geographic relationships, synchronic and/or di-achronic, can reveal different strategies employed in different regions, or show change through time in such strategies, probably caused by changes in the internal organization of these relationships. One can conceive of these landscapes as superimposed planes, which like the many layers of a quilt are stitched together or joined at certain places and not at others. Some of these nexes come unstuck over time; others gather more and more connections, piercing through and articulating many layers of cul-tural landscape, concentrating their most important loci at the same pow-erful places.

These changes in the power of places are sometimes a by-product of other elite decisions—a redirection of trade can result in the appearance of new settlements, or conflict with a neighboring ruler may empty out previ-ously populous intervening districts. Such “by-product” changes certainly oc-curred in Denmark. However, the evidence indicates that certain landscape changes were deliberate: planned components of centralizing strategies meant to undermine and alter long-standing cultural patterns that were coun-terproductive to central elite goals.

Scania, a major focus of this study, has been called “a prehistoric country in the process of dissolution” (Strömberg 1977) with a “hidden, but disap-pearing cultural landscape” (Tesch 1980:20). This refers to the fact that al-though Scania is an intensively farmed area today, the fossil Iron Age landscape is still observable in a number of ways. The same is true of other parts of Denmark, where the remarkable preservation of ancient, relict cul-tural landscapes and their internal spatial relationships makes it possible to reconstruct these large-scale relationships through time and across space. Using this unusual and fortunate preservation of ancient patterning is a key to understanding the region’s past.

Finally, to get at more humanistic, subjective aspects of culture, the ar-chaeological record and historical record, if there is one, may contain proxy indicators of contemporary elite and nonelite ideas and perceptions (Duncan 1990). These humanistic cultural productions can then be compared with the more mechanistic study of the material world. Late prehistoric and proto-historic Denmark presents us with such possibilities because there are a number of indicators available, although appropriate care must be taken in considering them.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (55)



After a considered study of the archaeology and historic record of south Scan-dinavia, it might be said that the whole period of state formation in Denmark can be characterized as a royal quest for sovereign power against a very hos-tile system. In itself, this did not prevent the development of a state, but it did affect the physical manifestation of political structure and the material ex-pression of power, hampering centralization and administration, as well as making the ambitions of Iron Age rulers very difficult to attain. As noted in chapter 1, it also presents many problems for archaeologists working in the region because it renders many types of change archaeologically “invisible.” Historically contingent factors, such as intentionally weak central authority and political and economic decentralization, preclude the study of the most-favored types of archaeological indicators, because they prevented rulers from doing/having/constructing many of the things that form the basis of such data.

Although we cannot make a systematic study of many “typical indica-tors” in Denmark, what we can study are whole landscapes and the rela-tionship between their elements. Early in my study of Danish Iron Age site-size and site-type distributions, I recognized that the diachronic change and synchronic variation seen in the archaeological record of places—thatis, geographic loci of all types—was not random, and I hoped to find statis-tical support that these distinct regional patterns and sometimes sweeping locational changes were not accidental. My interest in the relationship be-tween social control and the control and manipulation of places developed over the intervening years. According to Cresswell (1992), the setting, the geographic loci of the human experience, plays a “decisive role in the defi-nition of what is ‘good, just, and true,’ a central function of ideology.” Power relations that connect the individual, the social milieu, and nature occur through place-particular practices, and places are conceptualized partly in terms of the “continuous progression of local events” (Pred 1986). At these settings (places), time-specific and space-specific activities occur. In this way, places and people form a constantly interacting dialectic. People, dur-ing their lifetimes, continually undergo biography formation, the develop-ment of the individual’s self-centered concept of themselves and their place in the world. Biographies form largely in relation to the places where both unusual and commonplace life events occur. Through historically specific events at geographically specific places, both people and places are con-stantly becoming. Becoming, as noted above, is a way of expressing that places are no more static than the people who inhabit, use, or conceptual-ize them. If people are bound together with the places where their biogra-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (56)


phies are formed, then modification, manipulation, and radical or imper-ceptible changes in these places have a profound effect on their perception of the world and their role in it.

The concept of the strong role of places in human practices may simul-taneously seem highly abstract and metaphorical and intuitive or obvious. One the one hand, it is not abstract, but is part of well-developed cultural-ge-ographic argument on the “interwoven relationships between the individual and society, between practice and structure, between agency and structure, between socialization and social reproduction” (Pred 1986:7–8). These inter-acting aspects constantly remake and transform each other in a process called structuration that is historically and geographically specific (Giddens 1979, 1984). Structuration means that the dynamics of society on all levels and scales, from the individual to the state, are contingent not only on the so-ciety’s history but also on the geographic setting (micro and macro, farm-stead to administrative town) where this history unfolds. The manipulation of places can thus be a powerful tool for social control. I will return to this topic—the interrelationship of place, structuration, agency, and biography— in chapter 9.

On the other hand, although the human/place relationship may be im-plicit, it needs to be studied empirically, and one must have the tools and techniques to do this. Since changes in social and political systems are re-flected in locational and organizational change in many areas of the world (Johnson 1980, 1987; Skinner 1977, 1985; Feinman and Nicholas 1990; D’Altroy 1992), I examine changes at regional, subregional, and local scales, not only in the way things were built or constructed or accumulated or pro-duced, but in the location of already extant institutions in the landscape. In this study, by monitoring the physical, tangible responses of the political, economic, and sacred cultural landscapes—responses that are, of course, the results of some type of actions—the otherwise invisible pressures, forces, and policies of elite decision makers can be detected. One can see the efforts of elites to thwart the traditional power structure: throughout the Viking Age and early Middle Ages, places, spaces, and the experience of everyday life were altered in an attempt to undermine old ways and substitute new. One can also see the preexisting systems, the established local rulers, an older pat-tern hanging on tenaciously.

By studying the whole landscape as a constantly changing, culture-bearing and culturally reproduced artifact, locational and landscape analyses permit us to divorce ourselves from the idea that the only way to see the political and economic activities of the state is through the analysis of elite residences, central storage facilities, temples, and other large, obvious, cen-tralized state-sponsored activities. In the latter parts of the Viking Age there are such indicators, and I will highlight and discuss these types of evidence

Landscapes of Power, Landscapes of Conflict - [PDF Document] (57)


when possible. However, to study the transition as uniformly as possible, I instead stress the identification of shifting hierarchic political and economic structures and how they administer geographic regions over time and space.

2.5.1 Courses Toward Unification: Changes in Cultural Landscapes,

These landscape concepts can be used to study Denmark's difficult and hard-to-detect transition from many groups to one, and from corporate to network rule. Although small-scale raiding was common between prestate societies, major warfare related to unification is not seen until long after the state was formed. If so, why did these autonomous polities eventually voluntarily sur-render their authority to a centralized leadership? During what period did this centralizing, state-forming process occur, how long did it take, and how successful was the integration of the outlying polities with the already unified western core? Finally, what were the mechanisms through which the incipi-ent state exercised its authority and consolidated its control over its new ter-ritories? The timing, rationale, and course of this unification have only recently begun to be better understood (Thurston 1996).

2.5.2 Landscape Change

As there is no evidence for widespread internal disturbance at the time of state formation—domestic warfare, violence, depopulation, or mass move-ments of local peoples—I question how the subjugation of Scania could have been achieved. When such radical political and economic shifts are accom-plished in historical and ethnographic examples, combinations of physical and ideological change are usually employed.

In the physical realm an assault was made on traditional systems of gov-ernment and economy, not by eradicating them but by erecting a parallel po-litical and economic infrastructure. Avoiding direct conflict, and the very real threat of assassination or exile associated with it, central elites employed a continuum of least-cost to highest-cost, steplike strategies, predicated on their relative cost to benefit. The least dangerous and expensive methods were tried and exhausted before more expensive strategies, in terms of both financial and political price, were used. After the possibilities of alliance and then hegemony were exhausted with little result, the state induced dramatic changes in central place location, function, and size, creating new places and new landscapes. The state even effected change in the agricultural villages that these sites administered, down to the very structure of the farmstead. In time, the whole organization of the cultural landscape changed, not overtly, but in increments.

Ideologies, and Power Structures

Landscapes of Power, Landscapes of Conflict - [PDF Document] (58)


2.5.3 Ideological Change

The result was a new, centralized hierarchic structure that bypassed corpo-rate authorities by making the traditional places where their activities were carried out redundant. New places preempted the power of older places,whose origins lay in ancient corporate systems. Old systems of power were circumvented, probably without the full comprehension of the public.

Simultaneously, new ideas were introduced into Denmark. Seeking un-precedented control of their subjects, rulers approached the problem from one direction by changing old social forms to new and rebuilding the social structure, a strategy that “allows individuals to gain economically and politi-cally in new ways” (Hastorf 1993: 1).


2.5.4 Change in the Power Structure

To get at these ideological shifts, I examine factors such as elite-generatedchanges in the terms of social relations and shifts in the conception of social classes, both of which are alluded to in the sparse documentary record. I use the premise that ideological changes can parallel changes the “manufactured landscape” (Earle 1991 b:96), and indeed, that these ideological shifts appear to mirror those of the physical environment.

The process of state building in Denmark must finally be understood through the connections between the corporate social code, the multiethnic resistant provinces of the state, and elite manipulation of the landscape. Earle notes that while “seizing control of existing principles of legitimacy” is one political strategy for extending power, “creating or appropriating new prin-ciples of legitimacy” is just as effective (1991a:5). Both new ideas and new places were created in order to bring down the old system and construct a new one.

Although framed by the Danes themselves in terms of social ideology, the reasons for these attempts to reinvent their society were grounded in economic and power relations: rulers wanted more wealth and more control. Since their society’s self-concept did not permit this, central elites attacked the concept and its operating structure rather than its warlords and their armies. Yet this strategy of slow absorption as opposed to conquest was in-herently problematic because it left in place the regional elites and the older, still extant system of governing, paving the way for tenacious resistance. Why give up long-standing, culturally sanctioned autonomy and local control to bow to a central authority? The evidence also suggests that as the state coa-lesced from this soup of neighboring polities, it was attended by a surge in re-gional ethnic identity and subsequent conflicts. Predictably, this slow and cautious method of integration was not always a successful strategy, for a

Landscapes of Power, Landscapes of Conflict - [PDF Document] (59)


number of rulers between A.D. 800 and 1200 were assassinated, not for being despotic but for being too much like kings, asking for taxes and military ser-vice for the purposes of financing and defending the state. In the study area of Scania, this policy led to the ultimate failure of the state to achieve a strong political and economic grip on the region.

In Scania, throughout this integrating process, strategies of domination were met with resistance, seen in the persistence of the old cultural land-scape, where in contrast to the quick reorganization of regions near the core, local chieftains and peasants attempted to control and maintain the tradi-tional geography of marketing, production, taxation, law giving, and other ad-ministrative functions. Long after Scania had been subsumed into the state, Scanian ethnic identity persisted and unrest increased, finally ending in a war of secession. To suppress this rebellion, force was used, but only as a last re-sort (Thurston 1996, 1997, 1999).


A final advantage of a regional-landscape approach is that it points the way to the contextualization of change. The approach is the result of my desire to couch empirical and objective study within social and cultural contexts. In the introduction to their 1988 volume Domination and Resistance, Miller,Rowlands, and Tilley (1 988: 1 -2) critiqued the modeling of growing com-plexity through sterile mathematical models of hierarchy and interaction, and lamented that attempts to integrate such methods with the social dimensions of human culture had been largely unproductive.

Whether or not this is true, abandoning model building completely is unproductive. The nature of prehistoric data necessitates that we fill in the blanks by constructing models, at least to some extent. Yet modeling change and embedding archaeological projects in a cultural context should not be mutually exclusive. My intention in this study is to make strides toward this goal, to examine these questions in a way that would substantially modlfy “objectivist tendencies to work solely towards the production of high-levelcross-cultural generalizations” (Miller, Rowlands, and Tilley 1988: 1).

When compared with political development of states in the Near East, Meso- and South America, and elsewhere, some of the sociopolitical processes in south Scandinavia follow comparable sequences. Prehistoric world systems may commonly have provoked secondary state formation through the impact of highly developed core areas on processes of change in their peripheries (Blanton and Feinman 1984; Chase-Dunn and Hall 1991; Peregrine 1996). Also shared cross-culturally may be the relationship between scalar stress and complexity (Johnson 1982) and elite strategies to integrate

Landscapes of Power, Landscapes of Conflict - [PDF Document] (60)


and exploit their territories, to build complex political structures and main-tain control. These processes do vary widely, and thus merit further docu-mentation, but also contain many parallels that can be compared and contrasted within and outside of a single area and cultural sequence.

Denmark, with its origin in a corporate mode of government, is not unique in its atypical and sometimes confusing indicators. Notable parallels of corporate complex chiefdoms and states are found in the Puebloan soci-eties of the American Southwest (Mills 1997; Feinman 2000), and Teotihua-can (Blanton et al. 1996; Blanton 1998, Feinman 2001), societies that are highly complex yet lack many typical archaeological indicators of centralized governments. Similarly, aspects of Denmark’s evolution are comparable with Mississippian societies, where it is difficult to postulate how a group of elites can come to dominate a society that originally had strong leveling mecha-nisms and horizontal organization (Saitta 1994). Clearly, a better understand-ing of these processes in one context can shed light, through both similarities and differences, on other places and times.

If it is important to make cross-cultural generalizations, it is equally im-portant to study local idiosyncrasies and cultural context. In many ways, it is easier to see the linkages between the socioeconomic, political, and ideolog-ical realms of culture, the connections between the physical and imaginary worlds, the empirical and the subjective, while looking at landscapes, be-cause landscapes are a type of cultural production, unlike regions, which are merely geographic areas. As such, landscape is “integral to both the repro-duction and contestation of political power” (Duncan 1990: 3), a direct re-flection of human-conceived boundaries and administration and also of spatial perceptions, belief systems, power, and conflict. Various institutional landscapes, such as the landscapes of rulership, religion, economy, and sub-sistence, interconnect in a certain way at the beginning of the study period, then change with the passage of time. Implicit in this physical change are shifts in the way people think about the relationships between these institu-tions. In order to include the nonobservable to the greatest extent possible in the context of an archaeological study, landscape changes are compared with other lines of evidence, archaeological, historic, and linguistic, illuminating people’s consciousness of the world they lived in and the conceptualization of authority and jurisdiction.


It is certainly possible to carry out useful archaeological research without ex-amining a single historic text, work of literature or art, or landscape. Why should an archaeologist bother to venture into such territory? It has often

Landscapes of Power, Landscapes of Conflict - [PDF Document] (61)


been stated that it is beyond our capability to recover more than what we can observe (i.e., Binford 1962, 1968). Admittedly, that is sometimes true, but not always. Archaeology is often the study of prehistory; the reading of a record that is missing the underlying social and cultural constructs that ordered and structured the worlds we wish to recover. Interpretations of the past must leap over this significant chasm, and inferences must be made directly from the ma-terial culture we observe and the data based on its quantification. Although some have argued that particularistic, intangible cultural details are unneces-sary to archaeological studies (i.e., Binford 1983), it has become increasingly apparent that the more complex our questions grow, the more problematic a lack of cultural context becomes. Historic and landscape studies can help re-construct cultural context in Iron Age Denmark, and elsewhere.

This does not necessarily mean a journey through an abstracted or in-vented landscape; this study is fundamentally empirical. Duncan (1 990: 13) criticized “self-limiting” empiricism that ignores some of the most interesting and powerful causal factors effecting social phenomena—factors that are largely nonobservable. I, too, eschew self-limiting empiricism, but not em-piricism in general, and while I integrate the visible with the typically unsee-able (that is, according to our expectations of a centralized state), I also do this by using only perceptible evidence analyzed by generally accepted meth-ods.


This chapter first discussed the strategy of the study: using three scales of analysis and multiple lines of evidence. A brief description of the classes of ar-chaeological data used during the course of the research followed. The bi-ases, problems, and advantages of historic texts were then discussed, and the combining of archaeological and historical data sets under the aegis of An-nales School history was recommended. Finally, an alternative paradigm is suggested, that of prehistoric cultural geography, the integration of cultural geographic-landscape theory and method with locational and regional ar-chaeological studies, but with a firm goal of contextualizing change through an exploration of social and ideological phenomena found in the documen-tary record.

The next chapter will focus on southern Scandinavia, from the period beginning in the Roman Iron Age and continuing through the Germanic era. Many inherent conditions in the natural and social environments of the re-gion have important implications for the organization of Iron Age society and the development of the state in the Viking Age.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (62)

Chapter 3

Prestate Denmark The Earlier Iron Ages


This chapter briefly discusses geographic factors important in Iron Age eco-nomic and political organization, then goes on to examine the last centuries of the prestate period in South Scandinavia, when the Roman Empire had a large impact on its northern periphery and profoundly influenced the course of change among Germanic peoples (Figure 3.1). Many aspects of Danish so-ciety that persist into state-building times are rooted in this era; the social code discussed in chapters 1 and 2, based on a military hierarchy and its ide-ology; religious practices; and longdistance trade contacts. In addition, many places founded in Late Roman-Early Germanic times remained active in the study period. Perhaps most importantly for later parts of the locational analy-sis, the settlement regions, social aggregates, and central places of these ear-lier times metamorphose and persist, shaping the factors that are the primary focus of this research: integration, boundaries, and political economy. Cur-rent understanding of these societies is that they were chiefdoms based on prestige-goods economies.

3.1.1 Physical Geography and Cultural Ecology

The area known as Old Denmark is a relatively small land mass, but it has widely varying geography and geology, which form a substrate for patterns of settlement, agriculture, and eventually marketplaces and urbanization. Denmark (Figure 1.2) consists of the peninsula of Jutland, four large islands


Landscapes of Power, Landscapes of Conflict - [PDF Document] (63)


Figure 3.1. Roman Europe in the second century A.D.

(Sjælland, Fyn, Lolland, and Falster) and four-hundred and seventy other small islands, forming an archipelago. Before 1654, Denmark included Scania, Hal-land, and Blekinge (Pleicani). This portion of what is now southern Sweden appears to be seamlessly linked to the rest of the Swedish landmass, but in fact just north of the border of Scania and east of the border of Halland lies a dense, rugged, and impassable forest wilderness, which from the earliest times has formed a boundary nearly as divisive as a mountain range. Further-more, because of its impenetrability, from at least as early as the Viking Age it has been a refuge for outlaws, runaway slaves, and exiles. In the Middle Ages, crossing from Scania to Småland on the overland route was considered almost suicidal. For this reason, Scania, Halland, and Blekinge have always been far more closely associated with Denmark than with regions to the north.

The abundance of islands and the short distances between them led to the evolution of Denmark’s strongly maritime-focused society. Tides in the

Landscapes of Power, Landscapes of Conflict - [PDF Document] (64)


Baltic are minimal or nonexistent , facilitating sea travel. Although winter weather was (and is) treacherous for sailing, during spring, summer, and fall, the interisland routes, never out of sight of land, were usually calm and pre-dictable. In the Iron Ages, these waterways were part of the system of high-ways and roads and were also defensible in places dotted with submerged settings of pilings and stakes.

The archipelagic nature of Denmark in part determined the location of the chiefdoms of the Roman Iron Age. Through time, competing centers were lo-cated on the large islands between Jutland and Scania, that is, Lolland, Falster, Sjaelland, and Fyn (Hansen 1987). They are geographically discrete and therefore defensible, and have fertile soils and good natural harbors (Thrane 1988: 194).

Denmark’s soils are, in the grossest sense, connected to human land-usestrategies. Jutland’s mainly old moraine landscape is a plain of glacial outwash sand that generates highly drained and low-nutrient soils, topped with a vast heathland (Ssmme 1968:116). This soil can support limited agriculture for only short periods of time, a couple of hundred years of shifting cultivation or less (Vad Odgaard 1988). The “walking villages” of Jutland’s Iron Age, sites that weresequentially abandoned and moved to nearby locations, may be in part a reflec-tion of the need to change not only fields but entire field territories every few generations, although social or ritual explanations may be indicated as well. The coasts are bordered with boggy saltmarsh and wet meadows. Only a narrow strip of land on the east coast and small areas on the top of the peninsula are more fertile. In addition, windblown particles eroding out of sandy areas at times covered entire settlements and their field systems, such as the Viking Age settle-ment of Lindholm Høje in northern Jutland (Marseen 1959:66).

3.1.2 Cultural Ecology

Despite little potential for field farming, Jutland is far from economically poor in terms of animal husbandry. The turf is good for grazing, and the bogs and coastal marsh meadows provide abundant hays that can be cut for winter fod-der. These conditions are illustrated archaeologically by the transition from early Iron Age Celtic field farming with fair-sized cultivated fields, to almost exclusive cattle husbandry by the mid- to late Iron Age. The Iron Age sites of Hødde and Grøntoft (Vad Odgaard 1985:121–7) are typical of settlements where cattle production was the major activity, with only small “garden” field systems around the village. Through husbandry and trade, Jutland became a nexus of political power and wealth during the course of the Viking Age. A prehistoric road that bisected the whole length of the peninsula, was, in the Viking Age, called the “Ox Road,” and must have been a major droving route.

The narrow, defensible neck of the Jutland peninsula acted as both a di-rect communication route with the continent and a barrier. The connecting

Landscapes of Power, Landscapes of Conflict - [PDF Document] (65)


aspects of this attachment are seen in the Ox Road, which was also (tellingly) called the Army Road. This road is clearly described in Viking Age documents as having been used to permit traders and travelers to pass between the con-tinent and Denmark. In its defensive role, the narrowness of this isthmus, or stricture, made it well-suited for use in defending against intruders even in its natural state, and it was augmented at various times with artificial earthworks.

Because of the extensive nature of the agricultural system, although Jut-land represents a large land mass, it had fewer areas of nucleated settlement (Callmer 1991 a). The more dispersed settlement pattern associated with hus-bandry and the inability to produce much of a cereal-based surplus may have necessitated the formation of territorially larger, more spread-out political units in order to support the same social structure as that which was found on the smaller, agriculturally richer islands.

In contrast, on the Islands and in Halland, Scania, and Blekinge, new moraine landscapes were rich in clay and other nutrients. Here, cattle farm-ing lost dominance slowly between late pre-Roman Iron Age, and cereal farm-ing rose steadily, with tremendous expansion just at the transition from the Germanic to the Viking periods, ca. A.D. 600-700. There were many large nu-cleated villages with several competing centers during much of the period. Only about one-half of Scanian settlements were periodically dislocated, as opposed to the common Jutland pattern (Callmer 1986:204). Those that moved did not move as often.

Other links with the continent were the Rhine and the Elbe rivers that empty into the North Sea, and the Dvina, Vistula, Oder, Oser, and others, that enter the Baltic from the south. The Dnieper, Dniester, and Volga do not empty into the Baltic, but are only short portages away from these other rivers, and link Scandinavia with the Black Sea, the Caspian, and the Eastern Mediterranean. This combination of access and isolation play a large part in the historical aspects of South Scandinavian development.


During the Roman Iron Age, A.D. 1 to 400, and the Germanic Iron Age, A.D.400 to 700, southern Scandinavian chieftains were transformed into kings, smaller polities joined into larger ones, and political territories expanded, eventually resulting in a single, large political entity. In chapters 1 and 2, I dis-cussed the conflict between chiefs seeking more power and the populace, who wished to limit their authority. There is much evidence relating to how the people kept their rulers in line and how rulers made their inroads into the system. Hereafter, the Roman Iron Age and the Germanic Iron Age may be referred to as RTA and GIA.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (66)


3.2.1 primary Texts: Scandinavian Cultures in the Roman Era

The Germania, authored by Tacitus, a Roman politician and historian, is unar- guably one of the most significant texts relating to prehistoric Scandinavia. Published in A.D. 98 the Germania is now usually called an ethnography, since it describes the customs and organization of the Germanic peoples—includ-ing the Scandinavians—encountered by the Romans. Unlike many other texts, from which concealed information must be recovered by careful reading, Tac-itus directly addresses the topics important to this study, and the prime con-cern in interpretation is his accuracy, not his subject matter.

A great deal has been written about Tacitus, and current scholars concur that he was a careful historian who took his work very seriously. In writing Ger-mania, Tacitus at least partly intended to cast shame on the growing decadence and imperialistic tendencies of Roman society. He admired the Germani, andused their strong points to emphasize Roman weakpoints. Yet he also wished to accurately describe Germanic culture (Benario 1975:34), and he did not distort their culture to an unrecognizable degree. He does idealize some of their per-ceived virtues—absence of corruption, moral values, bravery—but Germania“does not, however, amount to an idealization of the Noble Savage” (Dudley 1968:221). Tacitus readily points out what the Romans considered to be flaws among Germanic warriors, such as swaggering, drunkenness, and indolence.

Tacitus generally had good sources for his information: Pliny the Elder, who had fought in Germania, and the accounts of merchants and travelers who had the latest news from the borderlands (Benario 1975:31). The first 27 chapters of Germania are about customs and social organization, and the next 19 are a cat-alogue of about 70 “tribes” of Germanic peoples (Dudley 1968:220). Modern scholars conclude from Tacitus’s work that the Germans were undergoing rapid social change manifested in the form of private property, the use of money, and the power of the leader of the warband—or comitatus, as the Romans called it (Dudley 1968:22 1). Unfortunately, the term comitatus has acquired neo-Naziovertones in recent times. It has been used by groups of white-supremacist mili-tia who are waiting for the coming of supposed race wars. For them, it denotes their willingness to die for their neo-“Germanic” cause. To avoid any connota-tions with events occurring one to two thousand years after the study period, Ihave used the native term for the warband, drótt, in the majority of this study

3.2.2 Interpreting Tacitus

Either directly through the evidence from artifacts and finds or indirectly through correlation with settlement patterns or social hierarchies implicit in elite grave goods, much of Tacitus appears to be accurate. Most notably, Hedeager (1987, 1992) believes she has identified archaeological evidence in

Landscapes of Power, Landscapes of Conflict - [PDF Document] (67)


Scandinavia for the hierarchic military-political structure that Tacitus describes. Although Germania is a narrative description, and must be interpreted assuch, Roman historians such as Tacitus and Julius Caesar often provide basically reliable and sophisticated accounts of peoples they encountered or interacted with. Tacitus provides a “baseline” for many important aspects of Germanic so-ciety, which should be kept in mind: continuity can be seen in later periods, but the changes, when they do occur, signify substantial sociopolitical trans-formations that are linked with the eventual development of the state.

The term Germani refers to all Germanic peoples, from the Rhine to thetip of Jutland and beyond to the less-known tribes of Norway and Sweden. Tacitus begins by describing a land that includes “broad peninsulas and is-lands of huge size,” where he can only be referring to South Scandinavia (Be-nario 1967:37).

Tacitus describes the institution of the assembly. At such meetings, held on specified days, usually at the new or full moon (Benario 1967:45), “kings,” as Tacitus termed them, were picked or elected, though they had to be of noble birth and once elected did not have limitless or arbitrary power. They made decisions on lesser matters. On greater matters, all free men were con-sulted, and final judgment was in the hands of the masses. These meetings were the antithesis of central authority, and seem uncannily like the ting meet-ings described in later sagas. The participants were not punctual but took two or three days coming together, and then when the crowd seemed amenable, they all sat down fully armed. Religious figures presided, and kings or chief-tains spoke in order of age and perceived degree of nobility and war glory.

Tacitus discusses the punishments for specific offenses that might also be decided there (Benario 1967:46). If fines were involved, part was paid to the king or the “state” (per Tacitus), part to the victim of the crime. The fines were paid in horses and cattle. Provincial or local chieftains were also chosen at the assembly, each one with a retinue of 100 advisers from among his peo-ple, and young men were presented with arms, which Tacitus equated with the receiving of the toga in Rome.

Regarding the form of rulership, it appears that it was rooted in a cor-porate system, for it was not lineal. During the course of the Roman Iron Age corporate rule was changing toward a more network-style system, but in the Early RIA (and certainly the prehistoric Pre-Roman Iron Age), Tacitus writes that although young men of very noble birth were considered to have chiefly potential, they had to attach themselves to an older, wiser leader and follow in his entourage until they themselves achieved enough experience and hon-ors to lead their own men. Hedeager (1992) has described this as ascribed sta-tus, but I would interpret it as something intermediate: a young noble with little talent or chiefly charisma would amount to nothing in this system. His only value would be to father children of his line who might be better

Landscapes of Power, Landscapes of Conflict - [PDF Document] (68)


equipped for leadership. Although membership in a chiefly line placed one in the pool of eligible elites, it is clear that status was partly achieved, at least in Tacitus’ times. There are no 11- or 12-year-old chieftains in the historic or archaeological record of this era, as opposed to the late Viking Age, when we see Erik the Child (exact age unknown), Knut the Great (13 years old at as-cension), Magnus Olofson (11 years old at ascension), and others.

Each chieftain’s warband had specific ranks through which the warriors rose in accordance with the judgment of the leader, and there was much competition to rise to higher ranks and rivalry between chiefs for the largest and strongest following (Benario 1967:46–47). Chieftains were always sur-rounded by as many followers as possible, and powerful chiefs were sought out by Roman embassies, because they had enough prestige to bring wars to an end just by their reputation and presence (Benario 1967:46–47). Tacitus tells us that chieftains had vessels of silver from Rome, given as gifts by Roman emissaries, and that those nearer the border valued gold and silver as money while those further away used barter (Benario 1967:41). This pattern has been noted archaeologically (Hansen 1987, 1990).

Tacitus describes the Germanic battle order, mostly foot soldiers, with a few carrying swords or lances and even fewer elite-only cavalry. This precise configuration has been noted archaeologically because victorious armies sac-rificed the war gear of the losers by tossing it into sacred bodies of water, and the size and makeup of the conquered warband can easily be reconstructed. The infantry was chosen from a pool of all young men, 100 from each can-ton, or “hundred,” from whence comes the still-used term Härad (as in Jär-restads Härad). These young men were called “hundredmen,” which was atitle of honor (Benario 1967:42). For their service the followers were fed, clothed, horsed, and weaponed, with amenities “not luxuriant but abundant.” For this they fought to the death for their chief, but if their province was too peaceful, they sought service with a more active chief.

In order to support this military hierarchy, the people of the district vol-untarily brought offerings of cattle and crops. The more powerful chieftains received tribute in the name of whole neighboring tribes or peoples rather than from individuals, “choice horses, magnificent weapons, decorations, and neck chains” (Benario 1967:48).

Although Hedeager (1992) and Kristiansen (1991) assert that Roman IronAge polities were states, it is difficult to draw this conclusion from the evi-dence. The mobilization of goods to support an elite (Earle 1977) is implicit in Tacitus’s description, in which people of the district voluntarily brought of-ferings of tribute—cattle and crops—to support their local lords. A redistribu-tive economy is clearly active in these polities, where a “sizable fraction of goods and labor is acquired by the political/religious elites and centers for the support of the ruler and elite, and some is allocated back to the maintenance

Landscapes of Power, Landscapes of Conflict - [PDF Document] (69)


to roads, shrines, defense, poor and other activities” (Dalton 1980:3). Also typical of a strongly corporate chiefly polity is the presence of leveling mecha-nisms (Earle 1977): the assembly, the extravagant feasting, and the giving away of treasure to followers. Hierarchic and organizational structures, which will be discussed later, also do not fit a state-level model during this period.


Political economy concerns the methods through which the income of a gov-ernment is organized (Claessen 1991:1). During the study period, south Scan-dinavia underwent a series of economic changes, in both the general economy and the political economy. Early chiefly centers in Denmark most closely re-semble the adaptationist model of political economy, where political leaders intervene in the economy, partly, perhaps, to strengthen and improve the re-sources of the whole society (Claessen and van der Velde 1991a:5), but more often to finance their own activities and that of their government.

The chiefdoms of the Roman Iron Age were based on what is commonly called a prestige-goods economy. Prestige goods are defined as objects that are not necessary for physical survival (these have been called staple goods; i.e., D’Altroy and Earle 1985), but that are fundamental components in social and political organization—absolutely indispensable for the maintenance of social relations (Ekholm 1977:119). An economy based on prestige items only works if one group has a monopoly on the prestige items: the flow of these goods, which are socially indispensable to all members of society, is con-trolled by a limited number of people (Claessen and vanderVelde 1991a: 14). In return for handing over the prestige goods to lower-level elites for their own legitimation, chiefs received tribute in the form of staples: food and draught animals, grain, and other items necessary to support them. The lower-level local rulers in turn extracted this tribute directly from farmers and herders in return for their services as administrators in defense, uphold-ing law, and overseeing rites and ceremonies.

To assure that the availability of prestige goods was limited, elites often ob-tain exotic, imported items as a basis for their system. During the Roman Iron Age, the system was based on imports from Rome and its provinces, notably bronze, glass, and ceramics ( terra sigillata ). Roman goods entered into the cer-emonial and sociopolitical sectors of the Germanic peoples, including the Scan-dinavians, and replaced traditional items as marriage-wealth, in rituals marking important life events, as political gifts and accompaniments to negotiation, as institutionalized payments and tributes, and so on (Friedman and Rowlands 1977; Hedeager 1992), becoming a vital part of the maintenance of social or-ganization. The empire benefited from trade, and probably supplied both

Landscapes of Power, Landscapes of Conflict - [PDF Document] (70)


Roman citizens and the legions with many of the goods it received. Roman records indicate trade for leather, meat, cloth, and probably slaves. While staple goods were leaving Germania, what was coming in were almost wholly elite goods, such as jewels, metal weapons, wine, and imported vessels. It is typical of such economies that staples are traded for elite status markers (Claessen and van der Velde 1991a:15). Monopoly of prestige items increased over time: the distribution of valuable imports as grave goods was limited to about a third of people in the RIA, and later, in the GIA, to fewer and fewer elites. Centers of importation were often sited near shrines, where elites who upheld the sanc-tity of holy places could facilitate trade through their ability to enforce law.

Over the course of several centuries, these goods helped to transform the position of the ruling class (Randsborg 1989a:207) in the following ways: while less complex societies engage in the use of prestige items to indicate personal status within society, political systems based on such goods exist outside of the kinship structure; in fact, they may be directly related to the subverting of kin-ship as the primary form of alliance by replacing kin obligations with socio-political obligations. Reliance on these goods replaces kinship in two ways. First, it necessitates a far wider network of interaction than reciprocal relations between related peoples. A prestige economy requires the control of long distance systems of trade and supraregional interaction spheres that permit the entry of such goods into local economies (Hedeager 1992:88). Second, it cre-ates relationships within local societies of dependence and alliance not based on blood relations. While in some societies only the rulers themselves may dis-play prestige goods, in prestige good economies higher-status individuals se-lectively distribute such vital social objects to their followers and clients as an incentive to obey, a reward for “being part of us.” This payment, and the ability to withhold it, thus excluding offending or disobedient parties from social, rit-ual, and political spheres, allows leaders to control and coerce their followers through nonkin relationships (Hedeager 1992:88). Claessen notes that in ac-cepting the gifts and payments of the ruler, the follower accepts his subordi-nate position and is forever under obligation to the ruler (Claessen 1991 :11), and the social relations between ruled and ruler are cemented.

Not all chiefs had equal access to Roman goods. Originally, goods entered the north when Germani elites were actively sought out by Roman emissariesas representatives of their peoples and gifted with elaborate items (Benario 1967:41). Later, organized importation of such items appears to have been controlled by a limited number of rulers (Hanson 1987). This created a de-pendency on Rome among the highest-level elites who managed to seize con-trol of trade routes and importation. Controlling the entry of goods gave them power over everyone else. Secondary dependence, between rulers of other-wise equal horizontal status, was created between these ruling families and their less-fortunate peers. Within each ruler’s polity, a vertical dependency was

Landscapes of Power, Landscapes of Conflict - [PDF Document] (71)


also established. The minor “provincial” chiefs within each region, as Tacitus’sGermania calls them, were tied to their overlords through the goods; simi-larly, the warband members, who were rewarded richly for their service and loyalty, were the recipients of these goods as well.

The final disposal of these goods was symbolic in nature. Those who controlled the importation of wealth were also able to manipulate its rarity and value through the religious custom of forcing it out of the system. The choicest material was not passed on to future generations but used to gain favor with gods and ancestors by elites who deposited it in bogs and lakes as offerings (Pearson 1984:71). This constant removal of goods from the system kept them rare and coveted, but also necessitated a constant rate of import. The increase over time in Roman goods flowing into Denmark can be docu-mented in the sacrificial bog finds (Randsborg 1989a:215). Gold, silver, bronze, and Roman glass have been excavated as a solid floor of material from many such watery shrines.

The other role of these valuables was as grave gifts. Each type of elite ap-pears to have been laid to rest with a set array of goods that was appropriate to his rank (Hedeager 1992:94). Hedeager’s evaluation of the distribution of weapons and prestige goods shows that a warrior hierarchy was in place by the early RIA, with several lower warrior classes, the hundred-men of the warband, typically displaying bronze imports and spears, the minor elites dis-playing silver ornaments and swords, and the upper-level rulers with more specialized goods, including massive gold grave-wealth, spurs and other horse-gear (only the most elite rode horses), games and gaming pieces, and a complete “set” of weapons. By the late RIA, this organization was no longer reflected so sharply in grave goods. Since the period was more hierarchical and socially structured than previously, it is doubtful that this means the war-rior hierarchy disappeared. Hedeager argues that the hierarchy was so en-trenched it no longer needed to be marked so distinctly (1992:134 –5).

The sacrificial depositing of such valuables indicates that they were not used as currency or coin, but purely as symbolic indicators of status and posi-tion (Hedeager 1988:152). Symbols of rank were not merely the baubles or toys of the upper strata, but items laden with meaning and power, indicative of class and the rights and responsibilities it entailed, and may have been governed by sumptuary rules. This can especially be seen in the functional categories of the imported artifacts. One category of sacrificed import is display wealth: con-spicuous, large or massed items of adornment, gold and silver neck rings, fin- ger rings, brooches, and amulets. Another large component of these sacrificial hoards is ritual drinking equipment, which in both Celtic and Germanic soci-eties was a nexus of magicoreligious ritual and symbolism (Arnold 1999), and in turn, social, political, and economic power. The final category is imported and probably extremely valuable weapons, certainly warrior elite items. These

Landscapes of Power, Landscapes of Conflict - [PDF Document] (72)


weapon deposits, 19 of which are known in western Denmark, began in thePre-Roman Iron Age, before the birth of Christ, but the majority were de-posited beginning at about A.D. 200 and continued until 400, and probably rep resented “the development of rival centers of political power which arose in Denmark in the centuries following A.D. 200” (Jensen 1982:262). Perfectly us-able and highly coveted swords from Roman workshops, along with spears, knives, horse paraphernalia, and miscellaneous material captured from other Scandinavian groups, were ritually broken or “killed” and then submerged, in-stead of being recycled and used by the victors.

This is in direct opposition to several thousand known Danish finds of common Roman coins ( solidii and denarii ) from the second to fifth cen-turies, which are in worn condition, having circulated far longer in the north than in the Roman Empire itself (Hedeager 1988:150). They were apparentlyused as currency for international trade, because local coinage did not begin until the ninth century. They must have been used for trading in nonlocal contexts when products and raw materials were not feasible for exchange, and further emphasize the non-currency role of elite accumulations.

Prestige economies also affect other sectors of society. It is often noted that warfare between chiefdoms is common, almost inescapable, and intensification of production also is often seen, especially in agricultural contexts. Chiefs cover the expenses of government, reward the faithful, and in the sacred sphere, guar-antee peace, prosperity, and fertility through their intercession with the super-natural. But followers are expensive and demand more and more for their loyalty, thus there is a constant tension to acquire more and more income (Claessen and van der Velde 1991:11). This may lead to frequent warfare with neighbors, and competition between polities for control of trade may have been one factor encouraging interelite aggression, creating a demand for more followers and warriors, and therefore a demand for more legitimation through prestige display, in a kind of frenzied spiral of stratification, display, accumula-tion, and sacrifice. Income can also be increased by raising demands on the pro-ducers for surplus. Scandinavian chiefs not only distributed Roman imports to their followers but also supported a full-time warband with meat and drink, and both increasing warfare and agricultural intensification are seen in Denmark.


In Roman Iron Age society, high rank and privilege, along with the right to own and wear prestige goods, was grounded in warfare, and as Hedeager notes (1992:90), warfare is a social phenomenon, governed by rules, rela-tionships, and behaviors. Successful warfare includes leadership, coopera-tion, and organization. But in Germanic societies, overlordship ended when

Landscapes of Power, Landscapes of Conflict - [PDF Document] (73)


warfare ended, and the warlord did not rule society alone in peacetime as in wartime. He again became subject to the assembly as the primary governing body. In Germania, Tacitus gives a clear description of such an episode, where a skilled military leader and culture hero was killed by his own kin as a traitor to the social order, specifically for seizing power after his tenure as overlord was done (Dudley 1968:230).

This occurred in the early part of the Roman Iron Age when confronta-tion was interspersed with peacetime, but during the next three centuries of the RIA, there were more protracted periods of war and constant skirmishing with Roman legions and neighboring polities. Under such conditions it may have been necessary for military leadership to have become continual, thus paving the way for warlord control over leadership activities through the cre-ation of a loyal, and somewhat permanent, warrior class (Hedeager 1992:91).

Warfare among less complex societies requires the warrior to fight for family and clan, and for his own prestige. In Germanic society, incipient cen-tralized hierarchies had formed, and warlords had the ability to control wealth and tribute enough to get prestige goods, ships, weapons, horses, and the like, as well as to control land and labor to feed a military force. Leaders were able to bind a group of warriors to themselves personally, through oaths, allegiances, and the gifting of the group with prestige goods. Instead of fighting for kin, the warrior now fought for a lord; the lord’s interests were the warriors interests, his prestige their prestige. Hedeager (1992:92) states that aside from prestige, motivation came also from economic interests, such as in plunder and tribute and the conquest and incorporation of new terri-tory, as well as from religious and ideological motives. This may have involved dreams, prophecies, or visions that called for warfare, and rituals and sym-bolism surrounding preparations and victories. Furthermore, warfare creates heroes, and cults of hero worship may have arisen, a further reward of pres-tige by the community on its warrior class (Hedeager 1992:93).


Although the polity that controlled the most trade probably had some kind of overarching influence or power over the others (Jensen 1982:261), and its leader may have been most influential, these social aggregates remained sepa-rate at all times during the Roman era, and were not centrally governed by the ascendant group. A direct overlordship uniting groups in alliance, such as those described by Tacitus, was only subscribed to in times of war. Even if war was somewhat continual, the alliance was disbanded in times of peace, and various elite centers, along with their individual sacrificial lakes and bogs, con-tinued to operate independently of each other for the whole Roman period.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (74)


From authors among the Romans and later successor states, we know some of the names of individual peoples and groups, but it is difficult to un-derstand the geographic extent of these chiefdoms. However, it is possible to get a general picture of the regionalized, separate social aggregates that existed between A.D. 1 and perhaps 400, and also to observe that from that point on, smaller polities were continually being subsumed into larger and larger entities.

3.5.1 Polities of the Roman Iron Age: The Cultural LandscapeA .D. 1 to 400

As noted previously, there were few nucleated settlements in the Roman Iron Age, and Tacitus wrote: “The Germani do not live in villages, do not join their houses together, and each has an open area around his stead.” Cultivable land was held in common, the land divided according to rank, and plowlands were changed every year (Benario 1967:53). Thus, during the RIA settlement patterns do not tell us much about the location and extent of chiefdoms. Two lines of evidence—the location of clusters of Roman imports and the elite-warrior burials—are good indicators of competing regional polities or chiefly centers during this period. Grave goods mark the prestige of those accompa-nying them, and gravemounds, which were meant to stand out and be visible to contemporary people, would have been a constant reminder to the living of which families or groups were invested with the right to rule.

In the pre-Roman period, burial rites were uniform and consisted of cre-mation, but during the RIA increasing regionalization is seen in marked dif-ferences from area to area. Burial treatment was dominated by inhumations on Sjælland and central Jutland, while cremation was common on Lolland, Fyn, and southern Jutland (Hedeager 1992:98). The shape, orientation, and construction of the graves varied with territory as well.

The earliest concentration of Roman goods in elite burials was on the island of Lolland (Figure1.2), where a large amount of wealthy burials have been found dating to the first and second century A.D. (Jensen 1982:202). It can therefore be identified as the economic and political epicenter of south Scandinavia during this period.

In the second century, the focus shifted to Fyn, and in the third and fourth centuries, it moved to Sjælland, especially around the district of Stevns. Later, it returned to Fyn once more, beginning in the 500s and con-tinuing through the sixth and seventh centuries (Hansen 1987).

The level of warfare was rising between A.D. 200 and 400, and Hedeager believes that during this time, the polity centered on Sjælland attempted to push westward. In her opinion, though, this attempt failed, as the central focus of wealth and elite burial moved back to Fyn during the 500s and ap-pears to have remained there until the early Viking Age (Hedeager 1978). The

Landscapes of Power, Landscapes of Conflict - [PDF Document] (75)


Fyn polity, from the length of its ascendancy and its leaders’ roles as spokes-men for the Danes in historic records from the earliest Viking Age, seems to indicate that the “overarching power” to which the Sjællanders aspired and failed may have been achieved by the Fyn polity in slightly later times. The amount of wealth on Fyn in terms of total grams of gold per härad was far inexcess of other regions during the same period (Figure 3.2).

Despite these indications that one chiefdom or another was in ascen-dancy at a given moment, the other regions maintained their own autonomy, albeit with less wealth and influence. The islands continued to support re-gional polities, as did Jutland and Scania.

3.5.2 Organization of Chiefly Centers

The internal organization of a Roman Iron Age chiefdom is illustrated by Gudme, a chiefly center on the island of Fyn that has given its name to the archaeological region— the Gudme area. It has been intensively studied, and

Figure 3.2. Roman imports of the terminal Roman Iron Age (after Hansen 1987:320-322) andwealth distribution in the sixth to seventh centuries (after Hedeager 1992:63-64): concentrationon Fyn.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (76)


Major hoard of the Germanic Iron Age, A.D. 400–600

Cemetary A.D. 1–400Contemporary settlement

Harbor-workshops-seasonal market, A.D. 200–400

Major artifact concentration A.D. 200-1000(mostly 400–600)

Late RIA-early GIA farmsteads

Figure 3.3. The chiefly center at Gudme-Lundeborg, Fyn (after Ransborg 1991).

seems to correspond to a discrete social aggregate (Figure 3.3). This district displays complex internal organization, and appears to have been most active between ca. 100 B.C. and A.D. 600 (Randsborg 1990b). The site of Gudme it-self consists of a probable elite residence and many attached crafts work-shops. Probable because, as noted earlier, elites did not live in elaborate structures but the presence of a large wooden hall and the many associated craft production activities is typical of elite centers of the time. Gudme did not consist of any large nucleated settlements like towns, as these did not appear commonly in Scandinavia until a later period, but the site indicates the presence of important local rulers.

The name Gudme, or “place of the gods,” which is of Iron Age pro-venience, indicates an important sacred center, and nearby are former wet-lands that contain one of the richest and largest sacrificial offering places in the region. Wenskus (1985) has suggested that local places of exchange and places of prestige-good importation may have arisen at such sites, such as Helgö (holy island) in Sweden, and Vi (offering place) on Gotland, among oth-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (77)


ers, due to the ban on violence in a holy place. By the end of the Roman era, local elites appear to have taken charge of religious practices (Fabech 1994) and probably the trading activities of such places as well (Thrane 1988:194).

The associated cemetery of Møllegårdsmarken, in use from 100 B.C. toA.D. 400, contains some 2,000 graves, 83 of which contain 133 examples of Roman imports of glass vessels, terra sigillata, and bronzes, as well as manycoins and nearly 1,000 beads (Thrane 1988: 187). These graves, therefore, in-dicate that not only chieftains but a more extensive upper class had access to imports, illustrating the role of prestige goods discussed earlier in binding fol-lowers to leaders. However, Roman goods of the highest quality are restricted to so-called princely graves. These have large components of gold, silver, and horsegear, and also have more goods than others (Hedeager 1988:147). Near Gudme, for example, is the Gudbjerg grave, a rich and elaborate elite eques-trian grave of the early Roman Iron Age. An associated coastal site at a river mouth near Lundeborg, dating primarily to the third and fourth centuries, appears to have been the landing port for goods from across the Baltic (Thrane 1988:188). Here was the port of entry for Roman items.

It is highly likely that the Fyn rulers who emerged in the protohistoric era, of whom the Franks wrote in their Annales, were the descendants of these earlier Fyn chieftains who had fought the hardest, made war the most skillfully upon their rivals, and expanded their influence beyond their local borders. They were the survivors of intense selective forces and were pre-pared for expansion, conquest, and the governance of large polities.

3.5.3 Evidence for Other Chiefly Centers

Because chiefly centers do not stand out on the landscape physically, we do not know the “entire population” of such places, and because even when dis-covered and excavated they do not consist of large imposing buildings, they must be identified by their character rather than by architecture. On Jutland’s western coast the site of Dankirke represents another regional chiefly center. The main site, and several satellite sites grouped around it, date from the first to sixth centuries A.D. (Jensen 1982:210). The major period of ascendancy and wealth was in the latter part of that range. Located on poor agricultural land, wealth was probably based on cattle and trade. The great amount of lux-ury and import goods at this large site has suggested that, as at Gudme, trade occurred here, administered by a local magnate (Jensen 1982:249). For ex-ample, in one burned building, the remains of hundreds of glass vessels were found, indicating a storehouse rather than an area of daily use. Lead scale weights, often an indication of formal trade, were also found at the site.

Another concentration of imports is located in the region of Thy in northwest Jutland, where 46 contemporary villages have been surveyed, dat-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (78)


ing from the first century B.C. to about A.D. 200 (Jensen 1982:212). Another cluster of imports and graves is at the tip of Jutland in what is today Hjsrring Amt, toward Skagen, and on southern and also eastern Jutland, near Randers. There may be others.

In Scania, RIA finds are regionalized in clusters around the coast, espe-cially in the southwest. The site of Uppåkra near Lund (vifot 1936; Stjernquist 1994) is probably a central chiefly site. Excavations in the 1930s revealed cul-tural deposits 3 meters thick, with settlement remains, thousands of sherds of both coarse and fine ceramics dating from the RIA to the early GIA, bronze-working tools, keys of Roman provincial type, and coffers. The finds date back over a long period of time: ornaments from between the third and fourth century, and a large number of fifth century and early sixth century finds of several types. There are also finds from later periods, as late as the late 700s to early 800s. Based on these finds, and on regionalized, stratifiedsettlement patterns extant ca. A.D. 500-550 or directly after the RIA, whichwill be discussed below, it seems apparent that during much of the Roman Iron Age, several competing polities were operating contemporaneously.


For a period of almost 400 years, wealth from Rome was infused into the elite prestige system in Denmark by ruling factions who formed trading relationships with the empire. The south Scandinavian ruling class funneled Roman items northward, controlled the trade routes, and maintained the ports of entry.

Rome was first sacked by the continental Germanic peoples in A.D. 410,but a fragile and shrinking empire continued to exist through the mid-400s,when the Germans and Romans united to fight against Attila. In 476, a peo-ple known as the Heruli, who originated in Scandinavia, led by their “king” Odoacer, overthrew the last vestiges of the western empire and established the Ostrogothic kingdom in Italy. The system for the importation of Roman goods crumbled as Rome weakened, and after the fall of the empire in the last quarter of the fifth century, a period of destabilization occurred in Eu-rope, and these important items ceased to flow northward. The Germanic successor states of Europe—the Italian Ostrogoths, the Visigoths in mediter-ranean France, the Franks in northern France and Germany, the Vandals in Spain, the Suevi in Pannonia, the Lombards along the Danube, and many other groups—scrambled to stabilize, and for a time, trade was disrupted. This situation did not last long; by about A.D. 500, trade and interaction among the Germanic successor states resumed (Randsborg 1989a:210).

Nevertheless, the time between A.D. 410 and 500 must have been stressful for south Scandinavian rulers. Their form of legitimation and their system of

Landscapes of Power, Landscapes of Conflict - [PDF Document] (79)


prestige had been destroyed, and in order to maintain their political control,some interim mechanism must have been needed. It should not be implied thatduring this time elites disappeared or lost their power; rather they had to find new ways to display, maintain, and reinforce that power. Archaeological evi-dence points to a reorganization or refocusing of elite economy in south Scan-dinavia. Precisely at this time, the end of the Roman era, both luxury grave goods and bog offerings ceased, and prestige items disappeared from sacrificial contexts. Instead, they subsequently reappeared in settlement contexts (Rands-borg 1990a: 117). This change in elite use of valuables may reflect new rules about private property and inheritance (Randsborg 1982). If one was born into a chiefly family, one now wished to keep grandfather’s gold neck ring, not throw it into a lake. The goods had left the realm of sacrificial offering and en-tered into the estate of wealth, riches, treasure, that could be passed on to heirs. The orderly transfer of wealth, and the retention and reproduction of so-cial status from one generation to another, was now of primary concern (Christopherson 1982:120). This change is correlated with the founding of many nucleated villages and the sudden appearance of extensive fences and di-visions of land. These appear to be property lines rather than animal enclosures or other farm structures, indicating a major reorganization of land tenure, rural life, and agricultural production, and it is probable that a more formal taxation was introduced at this time. Ownership of land, animals, and staple goods were now equally important to tribute for the support of the chiefly family.

3.6.1 Political Economy in the Germanic Iron Age

Until recent times, a picture of European economies based largely on the his-torical record between the fall of Rome and the end of the Middle Ages was unquestioned. Historian Henri Pirenne’s thesis, first expounded around 1910, and culminating in his influential book Mohammed and Charlemagne in1939, became the primary interpretive framework for the economic transi-tion between A.D. 400 and the Middle Ages. In his model, the Roman organi-zation of Europe, such as the extensive town and market networks and the interregional nature of trade, persisted far into the Medieval period, and as European trade with the Mediterranean was cut off by Moslem expansion, rulers of the Dark Ages slowly developed their regions’ agricultural econ-omies. Up until the 1980s, this model remained the most popular explanation for the economic development of the Dark Ages in Europe.

However, archaeological investigations in the last few decades have re-vised or even reversed the theories on the development of post-Romaneconomies. Archaeological evidence has recently demonstrated that in much of Europe, the sociopolitical structure and market economy of Rome disinte-grated between A.D. 400 and 500, the Roman provinces were transformed

Landscapes of Power, Landscapes of Conflict - [PDF Document] (80)


into regional and subregional systems, and markets largely lost their charac-ter as interregional and long-distance trade centers (Hodges 1982a, 1989; Hodges and Whitehouse 1983).

Although some former Roman towns in southern Europe maintained their urban character, though at a reduced level, western European towns and the markets they fostered were either greatly reduced or largely aban-doned between A.D. 450 and 700. By A.D. 550, the remnants of the many Roman towns nearest to Denmark were mostly deserted, and settlement was diminished to “Episcopal residences, convents, royal castles, and the like” (Randsborg 1989a:219). The Justinian plague also ravaged the continent at this time, reducing the population by as much as one-third. Long-distancecommercial exchange, and with it the interregional market system, col-lapsed. For a while, even after the collapse of the Roman Empire, Danish trad-ing places that had been established flourished (Sawyer 1986a; Randsborg 1991), but by A.D. 600, it appears that the wealth and influence of elite-con-trolled centers like Gudme was waning.

3.6.2 The Age of Emporia A.D. 700-900

With the fall of Rome, the era known as the Dark Ages in medieval Europe began, in Scandinavia called the Germanic Iron Age (approximately A.D.

400-700), and with it came new forms of political and economic organiza-tion. While the longest-distance type of trade in subsistence and commercial goods, typical of the Roman Empire, eroded, local rulers and magnates throughout Europe and Scandinavia maintained local trade in luxuries to so-lidify and maintain political and social relationships with their peers, as well as to obtain the symbols of their rank in local society (Hedeager 1987). Al-though substantially smaller-scale than the economy of the Roman Empire, their efforts were not unsuccessful. Beginning around A.D. 700, trading ten- ters with a new character sprang up throughout western Europe. These sites, commonly called emporia, served as points of entry for elite goods such as precious metals and gems, tableware and glass, wine, textiles, and weapons, from origin points all over western Europe and beyond.

Many emporia have been extensively excavated, such as Ipswich and Hamwic in England; Birka, Ribe, Kaupang, and Hedeby in Scandinavia; Quen-tovic and Dorestad on the Rhine; Staraya Ladoga in Russia; Wollin in Poland; and others (Hodges 1982a, 1982b; Fehring 1991; Clarke and Ambrosiani 1991). Some of the activities at such sites are reminiscent of places like Gudme, but they were qualitatively different from earlier sites that consisted of elite residences with attached specialists.

As already noted, early in the RIA, many autonomous groups coexisted on Sjælland, Fyn, Lolland, Falster, Jutland, and Scania, but during the course

Landscapes of Power, Landscapes of Conflict - [PDF Document] (81)


of the Roman era, Fyn became the wealthiest polity in the region, perhaps having influence over its neighbors. By A.D. 200, elite centers were almost ab-sent in Jutland. Fyn’s good agricultural land, capable of producing a large sur-plus, may have bolstered its economy in Roman times. During the age of emporia, the balance shifted when the people of Jutland managed to over-come their dearth of farmland and rival the wealth of Fyn with a new type of economy that revolved around a new type of settlement: the town, a nucle-ated center of production and trade.

The Frisians, who inhabited the Low Countries, were by far the domi-nant trade power during the Germanic Iron Age. It is probable that the Frisians sought out trade with the Danes for the same goods the Romans had, and that as before Danish elites controlled and administered this trade. The trade routes the Frisians favored were located along the Atlantic coast (en-tirely different from those that had been primary in the RIA (Hodges 1982a:39). The Engish, Norwegians, and Franks also utilized these western routes. Thus, important economic centers in the Germanic Iron Age shifted from the islands and became concentrated in southern Jutland. The volume of trade had also greatly increased over Roman times, when elites could con-tain trading places within their own small compounds, using attached spe-cialists for production.

In the age of emporia, large towns appeared on Jutland with permanent populations of independent craftspeople whose sole purpose was to engage in trade and manufacturing. The remains of their extensive private workshop quarters have been excavated at several sites (Bencard 1978; Feveile, Jensen, and Lunjberg 1989; Muller-Wille 1988a), and to serve and cater to these skilled tradespeople, the towns also had a service sector and supported sub-sistence producers. These places were a loci for international export and im-port, not only for small quantities of riches—precious metals and luxury goods—but for leather, cows, produce, cloth, and utilitarian items like ce-ramics, combs, iron tools, and a host of other necessaries, which are found in a wide variety of settlement contexts in the hinterlands. Rather than directly controlling trade through attached specialists, elites made trade activities tax-able, subject to laws and regulations. Rulers on Jutland, with their trade cen- ters ideally situated on the new trade routes, may have begun to accumulate more wealth and power than those in the older chiefly areas on Fyn. With wealthy markets, it was easier for elites to accumulate followers, and nucle-ation may have made tax collection more feasible.

Artifacts belying elite control are seen in the physical remains of admin-istration and fortification, whose labor and organization only rulers could supply. In Scandinavia, non-defensive boundary earthworks surround some eighth- and ninth-century markets such as Ribe and Löddeköpinge (Feveile, Jensen, and Lunjberg 1989; Ohlsson 1976, 1980), interpreted as defining the

Landscapes of Power, Landscapes of Conflict - [PDF Document] (82)


extent of the elite-administered market, where the laws that regulated trade were in effect. In Ribe and Frisian Dorestad, property and boundary markings between traders indicate careful regulation of space, perhaps for rent or tax-collection purposes. Other markets, such as Hedeby in Denmark, and Ip-swich and Hamwic in England, had protective fortifications.

Fragmentary written records from contemporary Anglo-Saxon England indicate that the revenues earned from controlling such sites formed the basis of the political economy of most northern European rulers during this period, a major change from staple- to wealth-based economies, specifically “cash” wealth. Unfortunately there are no such written records for Denmark, but since the operation and organization of emporia is nearly identical throughout the emporia system, it is likely that obtaining imports and taxing commerce were primary sources of revenue for local leaders in Scandinavia as well (Sawyer 1986a). In return for their tax and toll payments, merchants could expect protection from raiders and thieves, the presence of officials to witness agreements and transactions and enforce the laws of fair trade, and an authority to repair and maintain harbors, wharves, and other facilities.

There are many indications, discussed in chapter 4, that the market town of Ribe in west Jutland, founded around A.D. 720, was under some sort of central elite authority, and that the same authority that controlled Ribe probably initiated several monumental building projects of the same general date. Although these developments span the transition between the GIA and the Viking Age, they represent some of the first indicators of centralized po-litical authority.

Hedeby town, in southern Jutland, was, according to primary docu-ments, founded in the early ninth century by the Danish king Godfred. Around the same time, taxation of towns and trade by the Danish king is ex-plicitly referred to in Frankish documents. These shifts in political economy are key to understanding changes from Roman to later times.


The wealth of the towns is clear archaeologically, and the taxes and other rev-enues they could generate for Danish rulers are explicitly referred to in the historical records of the Franks. Jutland elites of this era had new resources to provide rewards for advisers, client magnates, and their warband or drótt,(Christophersen 1982:115). This could not have pleased the powerful chief-tains on Fyn very much. If not checked or harnessed, Jutland elites would soon command larger forces and amass greater holdings than their peers to the east (Christophersen 1982:123, 129). This may be why the rulers on Fyn

Landscapes of Power, Landscapes of Conflict - [PDF Document] (83)


made a move to the west in their first attempts to incorporate surrounding areas. Jutland was the first area to be subsumed into what was soon to be-come Denmark. When did the small polities of the RIA begin to overtake their weaker neighbors, and how did they do it?

There are some hints in the documentary record, yet they are typically contradictory and difficult to assess. A number of preserved documents from the sixth to the ninth centuries include information on the integration and boundaries of Denmark during later times, but they are not as clear as the so-phisticated Roman, Tacitus, is, and there is much debate on how they should be interpreted. It is not until they are compared with archaeological data that they make sense: the various descriptions of Denmark as a unified land and a fragmented region probably are all accurate, and reflect the long and difficult transition to unification. These primary sources are presented here so that they may later be contrasted against the results of locational analyses, cultural landscape change, and other archaeological evidence.

3.7.1 Primary Texts: Boundaries and Integrationin the Germanic Iron Age

Jordanes was a Gothic monk who was living in Constantinople when he wrote a history of his people entitled The Origin and Deeds of the Goths, orthe Getica, in 551 or 554 A.D. (Goffart 1988:107; Wolfram 1988:15). TheGoths were Germanic peoples who overthrew and established rule in several former Roman colonies. Jordanes’s work is a plagiarism of 16 or more classi-cal and postclassical authors, such as Cassiodorus, an ethnographer who worked in sixth-century Ravenna at the “remarkably high level of ethno-graphic efforts and knowledge at Theodoric’s court” (Wolfram 1988). Jor-danes’s primary value is in preserving earlier works that have since disappeared (Mierow 191 5: 14). His Getica contains some five passages that concern Scandza, a list of 27 peoples and tribes of southern Scandinavia as they probably existed in the sixth century, copied from a 12-volume history of the Goths written by Cassiodorus. Unfortunately, this work has been lost.

A number of modern scholars, most recently the Swedish archaeologist Johan Callmer (1991a), have tried to place the groups mentioned in this gazetteer of peoples and interpret them for political and social organization. The generally accepted conclusions of linguists and philologists, as well as of Callmer (1991a:258–261), are that the geographic locations of some of these peoples are quite clear. For example, the Raumarici and Ragnarici are clearly the districts of Romerike and Ranrike in southern Norway; Bergio is the Scan-ian region Bjäke (spelled Biaerg in A.D. 1250). Fervir is with some certainty the Scanian area of Fiaerae (now Fjäre), and Luothida is likely the region of the River Lödde in western Scania. Hallin clearly refers to all or part of Halland.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (84)


Finnaithae is likely to be Finnveden in central Scania. Jordanes also mentions groups like the Suehans, Gothigoths, Ostrogoths, Dani, and Heruli, which are probably names of large confederations of related peoples (Svear [Swedes] , Goths, Danes, and so on) (Callmer 1991a:262).

There are a number of clear lacunae in the Norse peoples; no mention is made of the Oslo fjord, Sogne, Nord, or Hjörund, though they were impor-tant districts in the Germanic Iron Age (Callmer 1991a:261). Jordanes also seems to stop naming small individual polities when he reaches the bound-ary of what we think of as ancient Scania. It is possible that this is because he lacked information on this area, or that his work has been corrupted or partly lost. Callmer, however, favors Bolin’s theory (1930) that this instead reflects the political divisions of the area accurately. In this model there were several smaller, distinct polities in the northeastern and northwestern parts of mod-ern Scania—such as Bjäre, Fjäre, Finnveden, and Luothida, and others—while all the rest, the southwestern, south, and southeastern portions, were all one polity, known collectively as Scania, in which earlier, smaller units had al-ready been subsumed (Figure 3.4). If this represents the period between 500

Figure 3.4. Settlement areas, according to Callmer, and names of “peoples” in the sixth century A.D., according to Jordanes (after Callmer 1991a).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (85)


and 600 A.D., then it shows that, as postulated, some rulers were ruling larger and larger geographic areas, bringing together smaller social groups into larger political territories. Locational data, discussed later in the study, also point to this.

Although they must have included larger areas like Scania and smaller areas like Bjäre, the typical social aggregate in South Scandinavia appears to have encompassed an area of 20 to 30 kilometers of settled district, mostly along the coasts. Callmer stresses that in the Germanic Iron Age between 400 and 700 A.D., there must have been major and minor social groups, with many possible backgrounds, some stemming from:

long-lasting settlement communities or originally different populations brought together and intermixed as a result of subjugation or migration. Some may be the result of a political development on a regional basis. Both political collapse (disso- lution) and unification may be considered. Social aggregates are to be understood as territorial units . . . . The size of these social aggregates differs as a result of their historical background. (Callmer 1991a 257-8).

On Fyn, several runestones from the ninth century refer to these small subregions. Three known stones refer to godar, “political and religious leaders well known from the Icelandic sagas” (Callmer 1991a:265). Two of the stones name agodi, or leader, of a people called the Næsbor (people of the headland),one found at Flemløse and one at Helnæs (Figure 3.5) Callmer deduces that this group had a well-defined identity and territory in the ninth century and earlier that lay on and around Helnæs. Similarly, the Glavendrup stone refers to a godiof the people called Sølver (silver—probably a terrain name for the nearby bay). These named aggregates were probably two of several such groups on the island, once somewhat independent, later subsumed under one leader yet re-taining an identity as a district. Fabech (1994) suggests that based on Theissen polygons around the weapon sacrifices on Fyn, five groups inhabited the island during the early Roman Iron Age. The Næsbor and Sølver areas of the ninthcentury generally coincide with two of these earlier divisions.

As will be seen in later chapters, some fairly clear subdivisions appear to have existed within the larger region of proto-Scania, if this was indeed one territorial unit as Callmer (1991a) and Bolin (1930) suggest.

Another important source for understanding political integration is the “Orosius” of Alfred the Great, who ruled the Anglo-Saxon kingdom of Wessex from 871 to 899 A.D. . His was the only kingdom in contemporary southernEngland not conquered by Vikings. Alfred promoted the use of vulgar Engish (as opposed to Latin) and ordered a number of English translations of the Latin classics of his time, among them the work of the Spanish churchman and historian Paulus Orosius, Seven Books of History Against the Pagans. Written in the fifth century, it is a history of the world ending in A.D. 417.Alfred, also surrounded by powerful pagans, may have especially wanted to

Landscapes of Power, Landscapes of Conflict - [PDF Document] (86)


Figure 3.5. Runestones referring to low-level social aggregates and settlement regions on Fyn, ca.A.D. 800 (after Callmer 1991a).

make this available to his subjects, and updated the book with new material in his translation. Orosius contained a geography of the known world, to which Alfred added some relevant information on the Danes and northern Europe (Lund 1984:9). In addition, two accounts of the voyages of Ohthere and Wulfstan are transcribed from the narratives of two sea travelers— Ohthere, a Norse chieftain, and Wulfstan, an English or Norwegian traveler who sailed the Baltic.

The anonymous Anglo-Saxon translator tells us that the Danes were di-vided into Sillende (southern Jutland), the South Danes on northern Jutland and the western Islands, and finally the North Danes on the “mainland” of Scania and possibly Sjælland.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (87)


Ohthere’s account says that he came from Halgoland in northern Norway and describes trading expeditions among the Finns and Lapps. He describes Sciringes Heal, the emporia-style marketplace later called Kaupang in south Nor-way, from whence it took him five days to sail south, first with Jutland to star-board, then Sillende, to the town of Hedeby, which is “situated among Wends, Saxons, And Angles, and belongs to the Danes ... on the Port side he had all those Islands which belong to Denmark.” While clearly giving significant divisions within Denmark, he named all these lands as Denmark—that is, although the town and market of Hedeby was in Sillende, it belonged to the Danes.

Of Wulfstan the chronicle says:

He traveled from Hedeby, arriving in Truso after seven days and nights, the boat running under sail the whole way. To starboard he had Wendland, to port Lange-land, Lolland, Falster, and Skåne (Scania). All these lands belong to the Danes.“Then we had Bornholm to port, where the people have their own king. Thenafter Bornholm we had on our port side the lands which are called Blekinge, Möre, Øland, and Gotland, and these lands belong to the Swedes.”

Both of these accounts contain more information on regions not men-tioned here, especially on the Slavs and Estland (Estonia), and the travelers give precise descriptions of their routes that can be easily followed and analyzed on modern maps (Crumlin-Pedersen 1984:30). They refer to conditions in the late ninth century, as these seamen probably visited Alfred between 880 and 890 A.D. Neither of the voyagers mention the division of North and South Danes, which the Anglo-Saxon compiler stresses. Such divisions may still have been used to some extent, chiefly by outsiders like the Anglo-Saxons who knew lit-tle of Danish affairs, but perhaps were not as important as they once had been, especially to those who frequently traveled through Danish territory.

3.7.2 Primary Texts: Forms of Rulership

Historic texts also provide some clues to organization by the ninth century. The Annales Regni Francorum, or Royal Frankish Annals (RFA), were kept during the period from A.D. 741 to 829. This primary source is vital to understanding the transition to the earliest state. As noted in chapter 2, the Franks and the Danes were usually either in a state of war or tension throughout most of the eighth and ninth centuries. Conflicts and diplomatic events between the tworegions are recorded in several instances. This sheds light on individual Danish rulers and their actions, the general powers and reach of Danish rulers, and the extent of the Danish “nation” if, indeed, it can be called this. As Scholz puts it,

The authors of the RFA note only the barest outlines of the world in which they live. Military actions, diplomatic missions, and major political events attract their atten-tion first; yet, in their barren record we catch a glimpse of their universe—physical, social, and spiritual—in which the writers breathed and thought. (Scholz 1972:8)

Landscapes of Power, Landscapes of Conflict - [PDF Document] (88)


Particular reference to the physical extent of Denmark in the eighth to ninth century is made in the A.D. 811 entry. The Annal mentions peace ac-cords between a King Hemming and Charlemagne:

The peace announced between the emperor and Hemming, the king of the Danes,was only sworn on arms because of the severity of the winter, which closed the roads for traveling between the parties. Only with the return of spring and the opening of the roads, which had been closed because of harsh frost, did 12 mag-nates of each party and people, that is, of the Franks and Danes, meet on the River Eider at Heiligen and confirm the peace by an exchange of oaths according to their customs. The nobles on the Frankish side were . . . Count Walach and Count Wigman. On the Danish side were Hankwin and Angantyr, Hemming’s brothers, and in addition, other men distinguished among his people: Osfrid, nicknamed Turdimulo, Warstein, Suomi, Urm, another Osfrid, son of Heiligen, and Osfrid of Schönen (Scania), and Hebbi and Aowin. The envoys of King Hemming, Aowinand Hebbi, came to meet him [Charlemagne, at Boulogne] and brought presents and assurances of peace. (Scholz 1972:94)

The inclusion of Osfrid of Scania appears to mean that Scania was a province of Denmark in A.D. 811. On the other hand, some historians have proposed that Denmark was not united at all, and that it was common for leaders from the various autonomous people among the Danes to each send negotiators to such a meeting, to protect their interests, banding together under an overlord, as in Roman times. This would mean the exact opposite: that stressing “of Schonen” proves their disunity, and that the Franks were simply unaware of how Denmark was internally organized.

The Franks had already had formal diplomatic relations with the Danes, exchanging envoys and messengers, holding councils and assemblies, for nearly 30 years beginning in A.D. 782. Wars and treaties were undertaken, the extradition of fugitives was arranged, and the Franks certainly sound fairly well apprised of conditions in Denmark. It seem unlikely that Charlemagne, emperor of the Holy Roman Empire, did not have spies in the Danish court. In fact, in the entry for the year 823, the annal relates that the next emperor, Louis the Pious, sent two counts, Theothari and Hruodmund, to Denmark to spy out the situation among the Danes. They returned with Archbishop Ebbo of Reims, who was in Denmark making attempts to convert the Danes. The annal states that they “informed the emperor of all they could find out in these lands” (Scholz 1972: 114).

The many tribes of the Slavs, who were also heathens, both allies with and enemies of the Franks, are carefully enumerated in the RFA, which de-scribes them as being separate tribes, peoples, and political entities with kings or dukes, and explains who paid tribute to whom. This lends weight to the ar-gument that the Franks would hardly be ignorant of major fundamental polit-ical divisions in Denmark. The Danes were a neighboring, non-Christianpeople, usually on hostile terms with the Franks, and therefore not excessively

Landscapes of Power, Landscapes of Conflict - [PDF Document] (89)


familiar, but they were no mystery to the Frankish kings. But, I wish to stress, as mentioned in chapter 1, that political unity does not mean administrative or ideological unity. The results of this study show that both interpretations are correct: the Danes were both unified and divided, one polity and many sub-regions, externally identitied as a whole, yet far from whole within.

The Anglo Saxon Chronicle is an important source of both confusion and information in regard to Danish political organization. In 872 two kings of the Danes fought Alfred the Great of England: Bagseg and Halfdan, and a number of Earls (Jarls).When Danish reinforcements arrived on English ground in Sep-tember 875, they were led by three kings: Gudrum (or Gorm), Oskitel, and Hamond. This has led some scholars to agree that Denmark was a fragmented land of chieftains or petty kings and has led others to insist that only one was true king of Denmark, while the others were his vassals (Kroman 1976:166).

The Anglo-Saxon Chronicle illuminates more helpfully in later times, when it discusses the careers of the Danish kings of England: Sven Forkbeard, Knut the Great, Harald and Harthaknut Knutsson. It details how in 10 13 “all the nation re-garded him (Sven Forkbeard) as full king, and the citizens of London submittedand gave hostages” (Whitelock and Douglas 1961:92). In 1017, the entry states:“In this year King Knut succeeded to all the kingdom of England and divided it into four, Wessex for himself, East Anglia for Thorkel, Mercia for Eadric, and Northumbria for Eric” (Whitelock 1961:97). This later “division” may reflect more about the entries of the 870s. The three men who were granted the quar-ters of England were Jarls, not kings, and Knut was their undisputed ruler.

To complicate this matter, the later, indigenous Roskilde Chronicle, re-ferring to this period, states that sometimes “five kings ruled in Denmark, but that sometimes two reigned over all Denmark, sometimes one, and some-times one over England and Denmark” (Kroman 1976: 166).

Thus it appears that during the ninth century when some written records become available, notably the Royal Frankish Annals (entry for A.D.

811) and the Orosius of Alfred (ca. A.D. 890), Denmark consisted of the Jut-land Peninsula above the Eider, the islands of Sjælland, Fyn, Lolland and Fal-ster, and all the small islands, as well as the southern end of the Swedish peninsula, the province of Scania (Lund 1984, 1991). However, other sources are far less clear, and it cannot be assumed from historical documents that Scania, the islands, and Jutland were always an integrated political entity.

The transformation of alliances among the regional social aggregates of the RIA and the GIA into a single entity must have begun at some point, then, between A.D. 500, when competing regionalized polities still coexisted but had recently been somewhat destabilized by the loss of prestige goods from Rome, and the 800s, when historical documents fairly convincingly denote a political unification. For the purposes of this discussion, I define the time of transition as between ca. 500 and 800 A.D.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (90)



Roman Iron Age chiefdoms, based on prestige goods economies, witnessed the increasing stratification of Scandinavian societies before A.D. 400. At the fall of the Roman Empire, chieftains, far from losing power, shifted the em-phasis from prestige display to economic wealth. The right of certain elite lines to rule and control trade was institutionalized by the Germanic Iron Age, when there is evidence of shifting boundaries and the joining of polities into larger territorial units. The encroachment of the Franks created a military and diplomatic alliance between regional leaders, perhaps under the primacy of a Fyn-based elite. Growing centralization is indicated by the appearance of the first centrally controlled towns and the organization of labor for monu-mental construction in and near the Jutland trade routes.

The next chapter provides an overview of the Viking Age and the conflict between the ambitions of rulers and the traditions upheld by their peers and subjects. Many lines of evidence will be examined to provide context for the causes and consequences of large-scale societal change during this period.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (91)

This page intentionally left blank.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (92)

Part II Social, Political, and Economic Change

in the Viking Age

Landscapes of Power, Landscapes of Conflict - [PDF Document] (93)

This page intentionally left blank.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (94)

Chapter 4

The Viking Age in Denmark


Chapter 3 explored the possible nature of leadership in the Roman and Ger-manic periods. The services, as Southall (1991) would term them, that rulers offered their people during these pre-Viking periods were defense, the facil-itation of trade, and their “invisible” protection: they served as religious and ritual leaders. During the Germanic period the political economy shifted from prestige to wealth- or staple-based, that is, from an emphasis on things with symbolic value to things that had cash value. Kings became more per-manent, ruled larger areas, and had more power. However, the role of the leader still rested on successful warfare, personal reputation, the perfor-mance of rituals, and the maintenance of the sacred sector. The Viking Age, A.D. 700 to 1075, presented new problems for rulers.

Early Viking Age kings were caught somewhere between old forms and new. In order to explore the transition from the decentralized, weak rulers of the RIA and GIA to the more powerful kings of the Late Viking Age, this chap- ter examines many threads of evidence. I begin by discussing large-scale, ex-ternal forces that influenced the coalescence of the Viking Age polity in Denmark, then continue with a look at regional, internal, provincial condi-tions. Next, I examine rulership in Denmark and its transformation during the transition to a unified state.

The final sections outline strategies employed by kings to further their goals of centralization: changes in militarism, law, urbanization, religion, and economy.


Landscapes of Power, Landscapes of Conflict - [PDF Document] (95)



In the first half of the Viking Age, several autonomous south Scandinavian peer chiefdoms in loose alliance were transformed into a territorially large polity with a single ruler. Why did such a transformation occur at this partic-ular time? Randsborg, in his seminal work The Viking Age in Denmark (1980) suggests that “disturbances” such as population growth, Frankish pres-sure, and an influx of wealth from trade may have created conflicts between factions on Fyn, one of the large islands in Denmark’s archipelago. It is im-plicit in the historic record that Fyn was the probable power base of early kings such as Godfred and his descendants around the turn of the ninth cen-tury. It was to Fyn that Godfred and his descendants continually are said to have retreated during the upheavals of war with the Franks and among royal contenders. These disturbances on Fyn in turn led to attempts at cooperation and unification in order to diminish conflict (Randsborg 1980:32). As briefly discussed in chapter 1, areas to the immediate east and west—Jutland and Sjælland—were incorporated when the development of new power struc-tures “put pressure” on the central area, and the center then overtook the pe-ripheries, engulfing them and increasing the area of the state. Randsborg traces this through the date and distribution of runic stones.

Conflict with peer polities, especially the Franks, does appear to have been an important factor in unification. By A.D. 760, the Frankish empire was rapidly expanding and swallowing up other Germanic peoples, and was more extensive and too organized to be defeated or held in check by the territorially small polities extant in south Scandinavia in the eighth century (Figure 4.1).

The Danes had ample warning that the Franks represented a threat. In 768, the Emperor Charlemagne began what conventional historians see as one of his most glorious campaigns: the “pacification” and Christianization of the Saxons. The Saxons were a Germanic people, related to the Danes, with similar dialects and related gods, beliefs, and customs. Historical records in-dicate that high-ranking Danes were significantly intermarried with Saxonelites, almost certainly with all the attendant obligations accompanying such relations.

Charlemagne’s first act of religious conversion occurred in 772. After taking and sacking the Saxon castle of Eresburg, the Frankish army was or-dered to destroy a nearby sacred object or monument, a symbolic represen-tation of the Irminsul, or world-pillar, directly cognate with Ygdrasil, the tree that holds up the world in Scandinavian mythology. In both Scandinavian and Saxon religion, trees form a central part of worship. In addition to this mythic tree, the location of shrines was often in holy groves, and sacrifices to Odin

Landscapes of Power, Landscapes of Conflict - [PDF Document] (96)


Figure 4.1. Europe and the Mediterranean in the ninth century (after Randsborg 1991).

were made by hanging sacrificial victims in trees. In the 740s, the “apostle of Germany,” St. Boniface, had attempted to convert the Saxons. As his first mis-sion, he traveled to the great Oak of Geismar in Saxony and tried to chop it down with his own hands—a deed for which he nearly lost his life.

The shrine of the Irminsul apparently contained some of the onlyknown stone buildings of the Saxons, as well as a large treasure-house; it took three days to raze and burn the buildings (Chamberlin 1986:72). The shrine itself is somewhat mysterious; it is unclear whether it was a live tree or a huge carved wooden pillar. Perhaps Charlemagne was inspired by Boniface’s religious zeal, yet the fact remains that aside from destroying the shrine, the Royal Frankish Annals state that he appropriated and “carried away all the gold and silver which he found” (Scholz 1972:48– 49).

For several years after this, both sides raided and plundered back and forth across the Rhine. By 775, an impatient Charlemagne embarked on a pol-icy that was “no less than a policy of genocide . . . for two months he un- leashed the host in a deliberate campaign of terror quite at odds with his normal military strategy” (Chamberlin 1986: 120 –121).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (97)


A kind of resistance movement arose in Saxon regions, led by the Saxon elite Widukind. After raids or episodes of Saxon-Imperial conflict, Widukind often went over the border to his Danish kin for refuge during this period (Falco 1964). The result of the resistance was the so-called Massacre of Ver-den in 782. Leaders—called “nobles” in the Royal Frankish Annals—number- ing 4,500 men, were rounded up and beheaded, all in one day (Scholz 1972:61). Widukind was in Denmark and so escaped this fate. He would have received news of it while a fugitive guest in the court of the Danish king. In comparison, the great war between Rome and Byzantium a few years later produced only 3,000 enemy dead (Chamberlin 1986: 136). Historian Cham-berlin writes:

The twentieth century has put a certain gloss on violence. The century which has seen the incineration of hundreds of thousands of innocent and helpless people at Hiroshima and Dresden, which has seen the slaughter of millions of the defense-less in Kampuchea and Treblinka and Auschwitz, will view the violent death of 4,500 savage young men with a certain detachment. But Charlemagne’s contem-poraries, and commentators for many centuries after, were deeply shocked. (Chamberlin 1986:135–6)

Despite these sentiments, the Saxons were forced to convert to Chris-tianity under threat of death, and the division and transportation of Saxons reached a height in 796 when every third Saxon was transplanted, resettled, and dispersed. The Saxons, as an identifiable population, soon disappeared (Jones 1987:98; Scholz 1972:11). The people who later called themselves ethnic Saxons had little or no direct relationship to this culture. Widukind himself was eventually captured and forced to accept baptism, and was kept by the Franks as an example until his death.

The enmity between the Franks and the Danes grew steadily, mainly over the issue of the harboring of Saxon fugitives (Scholz 1972:15), and beginning in 782 there are many references to negotiations between the Danes and the Franks over the return of “wanted” rebels. Later, when warfare broke out be-tween them, there are many derogatory references to the Danish rulers that clearly are meant to imply the superiority of Imperial authority. Scholz notes:

The Danish kings are represented as haughty and foolhardy potentates who fail to recognize the power and majesty of the Frankish emperor, and references abound to the “the mad king” (Godfred), the “arrogance and pride of the Danish king,” his being “inflated with the vain hope of victory,” his “hypocrisy” and “empty talk.” But the concern with the Danes during these years indicates that the authors of the RFA were not only writing with all the frontiers of the wide empire in mind, but perceived that the Norsem*n constituted a growing threat in their time. (Scholz 1972:15)

The raid on Lindisfarne, during which a number of monks were killed, is usually cited as the first true Viking raid. It occurred in 793. The chronicler(s)

Landscapes of Power, Landscapes of Conflict - [PDF Document] (98)


who wrote the Royal Frankish Annals first noted the Viking raids on Frankish territories in the entries for A.D. 810 and 820. Eurocentric history, based on the writings of early chroniclers in the Frankish Empire, Britain, Ireland, and elsewhere, describes the terror and destruction of these raids while omitting reference to the violent Christian crusade that had preceded these raids by four decades. An effort should be made to demonstrate that both the Norse and the Continental rulers were equally rapacious in the pursuit of their goals. Large-scale, long-term confrontation may be related to a new and long-lasting alliance among the Danes.

It is not difficult to suggest that considerable aggression against the North preceded Viking activities in the south, and that the subsequent, long history of conflict resulted in internal changes in southern Scandinavia. As the southern “bloc” of the Empire sought to expand against the barbarians, a northern “bloc” formed to counter it, led by the rulers on Fyn, Godfred, and his predecessor, Sigifred. Instead of the small-scale, short-term warfare common among the chiefdoms of the Roman Iron Age, incessant friction be-came the norm. This state of hostility, and the levying of large forces to do battle with Charlemagne’s standing armies, consisting of thousands of war-riors rather than the traditional earlier armies of 200 to 300 men, also re-quired centralized leadership. An alliance generated by fear of conquest appears to have operated in late eighth- to early ninth-century Denmark. Holding together a number of smaller polities in an efficient system of war-fare and defense would require changes in the structure of decision making. Providing resources for pan-alliance military needs (shipbuilding, weaponry, beasts of burden, food, and so on) would put stress on this new, large al-liance if it tried to provision itself from only the small hinterland of its origi-nal region on Fyn. Rulers of this northern “bloc” may well have required that all regions in the proto-state intense production and contribute more sta-ples and silver toward a war effort.


Macroregional interactions at the beginning of the Viking Age indicate that several external factors might have been influential in the unification of south Scandinavia. Internal forces also affected the coalescence of the Viking Age polity in Denmark, and a number of substantial alterations in po-litical, economic, and social structure are evident. The correct scale of analysis for investigating questions on the internal factors affecting Danish state formation is the intermediate scale, in which large regions in the state are compared.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (99)


In the introductory chapters, I defined this study as an attempt to understand when and how rulers achieved the unification of several discrete regions and the centralization of their governance. This process has been de-scribed as a question of “how people succeeded in deceiving themselves intoaccepting the rise of the state around and above them, until the point wasreached when they no longer had any choice and had lost the power to re-ject it” (Southall 1991:78). In the context of the Germanic culture discussed in chapter 3, this would not have been easy.

Randsborg (1980:32) is one of the few theorists who has addressed the question of how Denmark became subject to a single, central ruler, and he is far from clear, only offering that when the Fyn rulers sought to expand, “. . . the periphery may have countered through a traditional system of al-liances, but in such cases geography favors the heart of an area” Certainly,any attempts by Fyn rulers to usurp the authority of neighboring chiefs would have been met with strong opposition in the form of warfare or ne-gotiation with individual and allied rulers. Yet upon closer consideration, it is difficult to understand what Randsborg describes as the favorable “heart-land” conditions on Fyn. Even if geography (personified) “favored the heart of an area,” Fyn is only the geographic heart of Denmark in hindsight, when we look at the state at around A.D. 1000. It was not the heart of the vastly larger area of Scania to the east, which dwarfed the small island. Ceramic material and even house construction shows that Scania had more in com-mon with the Baltic region and the Slavs across the eastern Baltic than with much of western Denmark (Hårdh and Wyszomirska-Herbart 1992; Bitner-Wróblewska 1991). Randsborg had already stated (1980:16) that Scania was,in 811, probably “outside the realm,” so he could hardly be implying that Sca-nia formed a subordinate eastern province at that time. Neither was it the heart of Jutland, which looked west for many of its most important ties, forming the eastern fringe of an economic network centered on Frisia and stretching to England in the west and western Norway in the north.

Geography only favors any area, central or not, if it has clear advantages over other areas. Fyn was not in an area superior through sheer natural rich-ness. It had no advantage over Scania, with equal or better agricultural land, a larger expanse, and many more people to produce staples and to fight bat-tles. Fyn had no economic advantage over Jutland, with its trade centers, where “richness” was based on commerce rather than on land. In terms of its placement vis a vis the rest of south Scandinavia, it is clear why conciliation was the probable route to unification. Explaining why they would have rather compromised than fought, Randsborg accurately states: “Fyn [was] here in the middle of such an area, [where] the number of potential opponents would have been largest” (1980: 32).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (100)


Opposition and conflict appear to have produced the Fyn polity. The five distinct social aggregates of the late Roman Iron Age, each with a weapon deposit in its center, indicate at least 200 years of intense warfare be-tween these factions. Out of this fierce competition rose a single ruling Fyn-based group in the early Viking Age. With the Franks poised on the southern border and the areas surrounding Fyn on the north, east, and west having “traditional alliances” in place—which through promised mutual aid sup-pressed the possibility of a Fyn takeover—it would seem that the Fyn polity was in a very poor position to incorporate anyone, except by promising them something they wanted or needed.

I suggest that what they offered was expertise, organization, and leader-ship in warfare. Political geography is the only geography involved here: Fyn had military leaders who were the most experienced and best-organized to lead an offense or defense against the Franks.

Even while the ruling families of Fyn emerged as political leaders of a larger area, in order to access the wealth of the Emporia of Jutland they needed the support and cooperation of elites in other regions. Thus, it is not surprising that the incorporation of outer territories was first at-tempted in a move to the west: Jutland came under the direct control of the Fyn-based rulers. This can be surmised from both the historic record and the archaeological evidence: the founding of Hedeby at the base of Jutland by Godfred, his emendations to the Danevirke wall—a tremendous earth-work many kilometers long that divided the Jutland peninsula from the mainland—and the continual Frankish references to his leadership of the Danes, even though the names of many other prominent Danes were well known to the chroniclers.

The Jelling dynasty of the mid to late 10th century moved the royal base of operations from Fyn to Jutland, perhaps because the earliest trade-routetowns were well established there, and it has also been suggested that the new dynasty moved its center westwards as a rejection of the homeland of an earlier dynasty (Randsborg 1980:33). It was during the tenure of this Jelling polity, with its “capital” on Jutland, that a state, as opposed to a chiefdom, is first clearly seen in the archaeological record. Named after the seat of these kings at the place called Jelling, Jelling-type runestones—which are in essence administrative documents denoting military and political relation-ships and the inheritance rights of those named on them—appear in Jutland and on the islands, showing that their direct power had extended east as far as Sjælland. Finally, in the “After Jelling” period of runestone typology, starting at about A.D. 1000, the influence penetrates further to the east. At this time, a large amount of these new-style runestones appear in Scania and in Jutland contemporaneously (Randsborg 1980:35). Thus, it appears that Scania was

Landscapes of Power, Landscapes of Conflict - [PDF Document] (101)


the last large area to be thoroughly incorporated into the state. The runestone activity strongly supports the trends of change in the influence and span of royal power (Figure 4.2).

Randsborg states that cooperation, unification, and the diminution of conflict appear to have been the favored strategy of early would-be rulers (1980:32), and I have reiterated this several times. But if not by force, then how was unification accomplished? Many archaeological models of the de-velopment of elite power (Upham 1982; Tilley 1984; McGuire 1986) describe elites as having usurped, co-opted, preempted, or controlled the labor of others, and the products of such labor. Although it is possible that the coalescing Danish state usurped some of the most troublesome local rulers by removing their heads from their bodies, further incorporation of Scania and other provinces, where local chieftains would have viewed their auton-omy as an inherent right, seems to have involved some sort of cooperation. What type of political and ideological framework could enable Scanian in-corporation without conquest? What allowed the state to “rise up around them,” perhaps so slowly that it was without their full cognizance?

While outside forces such as the Franks may have been pushing these polities together for common defense, internal forces may have paved the way for rulers to extend their temporary rights into more permanent ones. In the study area, coercive power and surplus labor and goods extraction should not necessarily be conflated. In some economies, surplus labor and surplus goods are held in common: the products and labor of the commu-nity are bestowable where the community feels they are best employed. As noted in chapter 1, this is one of the hallmarks of many corporate polities. Thus, in the earliest times of unification, the late eighth to early ninth cen-turies, it may be that producers, meeting in local assemblies, extracted labor and goods from themselves, allocated its transfer by granting it to elites as a compensation for the elite role in protecting them from enemies, maintain-ing long-distance trade, and regulating, storing, and distributing local prod-ucts as well (Saitta 1994:205). As opposed to a model where rulers of an expanding polity force their rule onto new territories, this is not a directly exploitative or coercive system.

This may be difficult to support when imagining a social setting only through archaeological data and theoretical speculations, but the fairly reli-able ethnography of Tacitus describes a very similar “open contract” with the folc (Swanton 1982:12). Elites were only allowed to rule inasmuch as they served the people, and the people rewarded them with support. This open contract between rulers and subjects was operating in Roman times, and in-volved smaller political units. Sometimes these units came together for a short time and gained support in the same way. If the rule of an overlord was

Landscapes of Power, Landscapes of Conflict - [PDF Document] (102)


Figure 4.2. Runestone change over time, indicating the spread of central (royal) influence (after Randsborg 1980).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (103)


extended, through perceived necessity, to become the rule of a king, the same “open contract” might be assumed by the people or even explicitly ex-tended between the ruler and the assembly.

The notion that different classes and groups within a corporate society each play their strengths and usefulness to each other as leverage toward their objectives, as an alternative paradigm to conquest, takes into account the ability of farmers to give or withhold labor and produce, the ability of middle classes and lower-ranking lordship to extend or withhold their ex-plicit support to rulers, and the ways in which rulers might be obligated to their subjects as opposed to merely exploiting them (Saitta 1994:202).

Hastorf (1990:175) suggests that emerging rulers must convince their subjects that there are advantages in uniting, but that “initial gains and losses,” that is, the trade-offs that people make, do not remain stable. Once change is instituted it may easily escalate. The small loss of autonomy for big gains in protection may, in a generation, become a struggle to hold onto even the most basic former rights or claims. Farmers, fishers, artisans, and other free folc themselves, through their immediate concerns, may have funded and empowered elites who would one day come to rule them absolutely. The col-lective, corporate aspect of RIA and early GIA decision-making would have modified and even disguised the growing power of elites but not prevented its creeping progression. During this process, it would be easy for the ruled still to imagine that there was give-and-take between them and their leaders, when in fact, very little still existed.

Before embarking on a regional/locational/landscape analysis in chapter 5, let us look at Denmark during the Viking Age. In the next sections, several key factors will be examined: the form and nature of rulership, militarism, the advent of urban settlement, changes in political economy and domestic economy, the rise of new social classes, and the adoption of Christianity. Evi-dence indicates that many of these changes may have been elite strategies geared specifically toward integrating the various regions of Denmark.


Because it is so complex and encompasses a transition between one type of rule and another, the exact nature of kingship in Viking Age Denmark is not well-understood. Chronicles and archaeological remains tell of the actions of kings but leave us to speculate on what their exact rights and responsibilities were, how they organized the government of their realms, and what legal rights and privileges they had over their people. Contemporary written sources give only hints, and kingship clearly changed drastically from the ear-lier to the later parts of the Viking period. Royal qualifications and attributes

Landscapes of Power, Landscapes of Conflict - [PDF Document] (104)


might have included an important role in the sacred and mystical: the ability to claim divine ancestry in the pre-Christian era, and later a considerable role in Christian activities. Rulers also commanded large fleets to control the sea routes (both military and economic), and they probably were the greatest landowners in the region. Finally, the distribution of wealth and favor to the retinue was still an important factor (Jones 1987:152).

4.4.1 Change and Crisis in Viking Age Kingship

During the Viking Age, what has been called a crisis or turning point in lead-ership was occurring among the ruling classes (Swanton 1982). Germanic kings were adopting Continental forms of rule because they probably desired the power that their southerly peers had attained and conceivably wished to be members of this profitable political and economic community rather than continue in their role as adversaries. This created problems for Scandinavian rulers that, as described by Swanton, were almost insurmountable: Christian rulers and their subjects who lived on the European continent and in England were used to the idea of a supreme elite. There was a long tradition in which the people obeyed the king, and the local king obeyed the Roman Emperor, and later the Pope, as the final authority.

Also, the proper demeanor for a Christian king was humility: strength in war and in law, but humbleness before God, and piousness, to be an exem-plar to the people of virtuous behavior, chastity, and moderation. Preserved Papal letters address rulers in England from the seventh century on, dis-cussing the king and queen’s clear, expected role in being models of Christ-ian behavior for their people (Blair 1970:60).

Germanic kingship, on the other hand, was manifestly based on very opposite rules: the precise behaviors and morals so abhorred by the Roman Church were those that were the hallmark of a proper Germanic warlord. The king with the ability to swagger and boast personified the epitome of a good ruler, the acquisition of many concubines was not lecherous but vigor-ous, and the consumption of large quantities of food and intoxicating bever-ages at feasts was not considered gluttonous or excessive but part of the offering and accepting of “guest-rights,” through the extension and accep-tance of abundant hospitality. This was not merely polite but a vital part of an accepted political forum. Drink still had magico-religious overtones; wod de-scended upon a man who drank alcoholic beverages in a ritual setting— wodbeing a conflation of inspiration, fury, and madness, an ecstatic state of al-tered consciousness sent by the gods, especially Odin (Wodan).

Before the Danes became Christians, the marriage of Christian ethics with an older Germanic cultural tradition would have been difficult, perhaps barely possible. Even after the Christianization of the royal family, most of the

Landscapes of Power, Landscapes of Conflict - [PDF Document] (105)


king’s subjects remained unconvinced and unconverted for nearly half a cen-tury. Thus, the new Christian Viking rulers, beginning with Harald Blåtand, who converted before his death in A.D. 987, found themselves faced with a paradox of biblical proportions. They had to remain vigorous and powerful in a traditional Germanic sense or be seen as weak and easy to conquer or usurp, while at the same time try to persuade their skeptical followers to join them as humble Christians in order to strengthen their sovereign powers.

The king and his followers were operating on two different systems of ideology, both political and religious. The king attempted to be Christian and pious, leading by example, while the followers still expected a king in the Germanic style. Kings tried to control the physical world by employing an ideology that few others yet believed in, ruling in the name of Jesus Christ, as if this meant the same thing in Denmark that it meant in Frankia or Saxony, where the will of the Christian god, indicated by the Pope’s support, was un-questioned. The Pope extended his approval to Danish Christian kings as an enticement away from apostasy, imagining that papal support was as impor-tant in the North as it was in the Mediterranean. But Jesus Christ and the Pope meant nothing to those whose persons, goods, labor, and taxes that Danish kings wanted to control.

If, on the other hand, the king instead tried to retain the Germanic tra-dition to appease his followers, then he had no right to claim any more than what ancient tradition conferred upon him, and had to leave Christian legiti-mation and aspirations of emulating the power of Christian kings aside. Until both king and subjects agreed on what they believed, there would be no res-olution to this problem.

Rulers of this period offer ample evidence, seen in the historic and ar-chaeological records, of this precarious state. They had moved toward sover-eignty, establishing a base in wealth and land, but rule was still expressed in a chiefly manner, based on personal reputation for luck and skill, and the dis-play of elite goods on their persons, but especially through a continuing per-sonal and reciprocal relationship with the underlords and the warrior elite, or drótt. As in the Roman and Germanic eras, during much of the Viking Age leaders distributed valuable items to their drótt in return for their support and protection. Unlike the prestige rewards of the Roman period, in the Viking Age these items grew to include wealth: land, villages, and cash. In the first half of the Viking Age, followers and their kings were still bound closely together by mutual pacts. To break this relationship—to reduce the war-band’s power and to increase the king’s—would have been an affront to those who protected you and fought for you and who could also very easily have put a knife into you.

This relationship is nowhere more obvious than in the case of the king Harald Blåtand, who became Christian during his reign, and who was over-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (106)


thrown by his son and trusted confidant, Sven Forkbeard, in the late 10th century. Forkbeard, who flaunted his pagan beliefs and persecuted Christian Danes, had no trouble in finding backing from the Danish elite for the war he waged and won against his father. It is probable that Harald’s new, highly centralizing politics more than his religion were behind Sven’s widespread noble support. Harald was deposed because his aspirations ran counter to Germanic forms of rule.

This continued power over the king was not limited to the group of men who closely surrounded him. What of the ruler’s obligations to the people, left over from RIA times? Early Swedish law, recorded in 1020, near the end of the Viking period, stated that the constituency “. . . had the right to elect or reject a king. When he comes to the ting (assembly) he must swear to be faithful to all the Gotar, and he shall not break the true laws of our land’’ (Jones 1987:152). In Viking Age Denmark, when a king was about to be crowned he traveled to the Viborg assembly, where the people had to ap-prove his ascension. In cases where there was conflict between rivals for the throne, it was resolved by the vote of the assembly. In 1074, when the king Sven Estridsson died, he made his second son the heir apparent. But at the Vi-borg assembly, the second son’s claim was rejected in favor of the eldest son. The assembly voted against the royal family and prevailed.

Thus, even as leaders grew in their ability to order and control, the as-sembly—a community-based body that was formed in an ancient, pre-Roman corporate-based society—still wielded a great deal of authority. Most government structures include two types of institutional bodies: the initi-ating body and the executive body. The former makes the decisions on courses of action, the latter carries them out. In pre-modern societies, a ruler—chief, king, emperor, or prophet—is usually the initiator, and some sort of body, such as a military hierarchy, an assembly, or a priesthood, car-ries out or executes these decisions. In traditional Germanic society, the sit-uation was reversed. The assembly made the choices and the ruler carried them out. By the Viking Age, the assembly was still attended by all lords, freeholding farmers, and at least some women, if the sagas are to be be-lieved. Neither farmers nor lords wanted their rights and entitlements re-duced at the expense of the king’s growing power. Bender (1990:255) describes this situation succinctly as “the remarkable capacity of kin-basedsocieties to resist divisions of labor leading to class formation, to resist the lifelong exploitation of one class by another, and to foment rebellions that may threaten or even destroy archaic states . . .”

The result was that political changes that occurred rapidly in continen-tal Europe, within generations, took hundreds of years in the North. Although slow, changes did eventually occur, but these changes could not have hap-pened randomly or without effort. A conscious endeavor to expand royal

Landscapes of Power, Landscapes of Conflict - [PDF Document] (107)


power must have been underway during this period and is not unlikely, since the rulership was carried out by one royal family for over 200 years in the study period, from Gorm the Old in the 930s to Valdemar the Great in 1157.


The following section will describe some of the detectable strategies used by central elites for integrating the provinces of Denmark. Some employed force, others power: the promotion of militarism to facilitate royal control of peo-ple and places, the manipulation of ancient law codes to undermine old rights and obligations, the construction of urban centers, and the adoption of a new religion that could enhance and support wider powers. Previously, many so-called central place services— justice and law, assembly meetings, marketing, ceremonial and religious activities—had been carried out at vari-ous, non-intersecting places in the landscape, leading to a series of disarticu-lated cultural landscapes with non-coincident nodes. At the new towns, services were newly centralized, and helped to empower and finance the state more efficiently. Through the construction of royal centers, the power of those who controlled the older cultural landscapes and their institutions was effectively neutralized.

A series of changes in local subsistence economies also appears to point to state-sponsored centralizing efforts, to regulate rural populations, collect taxes, and increase production. Finally, the ancient religion of the Danes was slowly abandoned—with central encouragement—and was replaced with Christianity. Although early Christianity preserved rather than destroyed old power relationships, late Christianity served the state to improve and rein-force centralization.

4.5.1 Militarism

Archaeological evidence for the organization of massive labor for large-scaleelite military projects is abundant in Viking Age Denmark. A reflection of the geographic extent of the kingdom seen in the runestone activity, canals and earthworks of the Early Viking Age are limited to Jutland and the Islands, while fortresses built simultaneously throughout all Denmark, including Sca-nia, date to the period when After-Jelling runestones too spread into the east.

Monumental projects of the early period include the Kanhave canal, a wide, channelized sea route that divides the entire island of Samsø in order to facilitate the quick movement of a fleet through the archipelago of Den-mark. The Danevirke also belongs to the Early Viking Age. It is a 30-kilometer-long palisade-topped earthwork wall that divides Old Denmark from the

Landscapes of Power, Landscapes of Conflict - [PDF Document] (108)


realms of the Franks and Germans. The earliest phases of the Danevirke and the Kanhave canal, dated dendrochronologically to A.D. 737 and 728, respec-tively, represent substantial amounts of organization of labor and time. A pro-tohistorical Danish king called Angantyr, who ruled in the early eighth century, is mentioned in the contemporary writings of the missionary Willi-brord of the Episcopal See of Hamburg and Bremen, who attempted an early conversion of the Danes. It is likely that Angantyr was associated with these projects. Probably he was a chief or petty king, controlling only Jutland or Sil-lende (southern Jutland), participating in more and more permanent al-liances with other southern Scandinavians.

As part of the later spread of central influence through all of late VikingAge Denmark, a system of fortified military encampments was established throughout the kingdom in the late 10th century (Figure 4.3). They are at-tributed to Harald Blåtand because they are dated dendrochronologically to 960 to 980 A.D., and are all nearly identical in form. Blåtand, as noted, was deposed and killed by his son, purportedly for being Christian, yet clearly for attempting major changes in the political system as well. These forts, or

Figure4.3. Military constructions, ca. A.D. 980.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (109)


Trelleborgs, as they are called after the type-site, were walled, circular, some with a distinct pattern of large longhouses inside, consisting of four “Trelle-borg” houses placed together to form a square with a courtyard, Roads di-vided the circular forts into quadrants. The nearly identical forms and dates seem to indicate highly centralized planning (Randsborg 1980:100). The fortresses are strategically placed to control the whole kingdom, and no sim-ilar idea seems to have preceded them (Roesdahl 1982:154).

It was once believed that each house contained the crew for a longship, and the places were envisioned as a kind of barracks. However, Roesdahl’s analysis (1972) of finds at the Trelleborg site of Fyrkat indicate a different function. Although some halls do indicate barracks, others have no hearths and seem to have been storehouses and stables, smithies, gold and silver workshops, granaries of imported seed grain, weaving huts, and other craft and storage facilities. Coins and trade-weights were also found. The Trelle-borgs were actually royal strongholds, where surplus from taxation and tolls were stored (Randsborg 1980:100), taxation and tolls that had recently been instituted by the king.

4.5.2 The Transformation of the Law

Another royal strategy for instituting change was through manipulation of the law. A large body of orally transmitted laws existed in the Viking Age. The as-sembly was clearly one of the most tenacious traditions in all Scandinavia, and one of its main functions was the giving of law. All Danes, rich and poor, male and female, if supported with the aid of male relatives (according to the sagas), could make complaints or stand up and be heard. As it existed in the Viking Age was witnessed by the missionary bishop Anskar in the ninth cen-tury, the ting had a continuous tradition back into Roman times, as described by Tacitus, and probably originated much earlier. There were small tingplaces where those who lived in a district would meet, and then there were the great Landstings, or regional assemblies. Viborg was the main ting on Jut-land, Odense on Fyn, Lund in Scania, and Ringsted on Sjælland. The fast three places all have pagan sacred connections as well as later church ties. Viborg means “offering place on the hill,” Odense means “Odin’s offering place,” and Lund means “sacred grove.” Viborg, aside from serving as a regional assembly,was the ultimate ting place for the nation: it was here that every king of Den-mark was elected by the people.

These assemblies were not only meeting places for the people to gather and participate in government but also places where the king held official “court” (Randsborg 1980:73). Even after the power to dethrone a ruler had been much reduced, the public still expected to attend the assembly to ap-prove him at an election, and still held the right to debate all types of issues.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (110)


For example, in the Early Medieval Knytlinga Saga, Knut the Holy argued and bargained with farmers at the ting over landrights, taxes, and privileges.

The role of the lawspeaker was one of great importance; this person would recite relevant parts of the law when cases and complaints were being heard. For many centuries the law was passed on orally from genera-tion to generation, but eventually it was taken out of the hands of the lawspeakers and committed to vellum at the order of kings in Denmark, Norway, and Sweden. There are references in the sagas and histories to the initial transcribing of these laws, mostly in the first half of the 1lth century (Norseng 1991:138). This inscribing of law coincides with many new at-tempts at controlling and changing Denmark’s organization and rule. Were laws changed when written down, or largely the same as they had been? How are they related to earlier law?

At one time, anything written in a medieval law code or saga-based ac-count of the law was taken as verbatem. Although the extant manuscripts themselves date to the 1200s and 1300s, the codes were actually written down in earlier times. Snorri Sturleson, the Icelandic historian and author, wrote that Norwegian kings St. Olaf (d. 1030) and Magnus the Good (d. 1047) both had laws committed to writing. Snorri says that King Olaf settled many matters of law himself and had the laws “which Håkon Adalsteinsfostre (d. 960) had given in Trøndelag proclaimed before him.” This code, known as the Norwegian “Olaf” text, is quite old, and parts of it are contained in laws of about 1250 (Norseng 1991:143). These and other medieval laws have often been read closely for clues to law in the Viking Age.

However, this changed when Icelandic law codes became the subject of study. The Islendingabók of Iceland, written in 1122, says that the first Ice-landic law dated to A.D. 930 and was based on a copy of law brought from Norway. Thus, it was assumed that at least a reasonable similarity existed be-tween the laws, and it was hoped that the Norwegian law of 930, a date that is certainly in the midst of the Viking Age, would be generally reflective of Viking Age Scandinavian law in general. The problem is that the earliest Ice-landic code and the Norwegian code it supposedly was based on are com-pletely different (Norseng 1991:140-41).

Modern philologists believe that the Landnámabók and Islendingabókwere intentionally skewed to favor the causes of the first Islanders, who were well-born exiles, and claimed they had been driven from Norway by the King, who was usurping the rights of the old lordship. They rewrote the laws and the historic record of their origin to favor their own claims.

The earliest written codes of Denmark too may have been altered in favor of rulers when they were first inscribed. Since so many other royal strategies in 11th-century Denmark seem aimed at reducing traditionalpowers and rights of the older nobility and the land-owning bonde or

Landscapes of Power, Landscapes of Conflict - [PDF Document] (111)


wealthy farmer, this is not unlikely. For example, some parts of Danish law codes are known to date to Medieval times because they contain references to datable Papal edicts or historical events, yet they also contain fake “ar-chaisms”: they make use of archaic phrases, a hallmark of oral proclama-tion—that is, certain types of repetitive or onomatopoetic phrases and alliteration—or even actual statements in certain paragraphs that the law is “ancient.” These devices for legitimation were, of course, at the disposal of kings and bishops to lend credence to their edicts and codes. With study, some parts of these documents may be assigned to different periods, but some questions on how medieval law reflects older forms of justice and punishments may never be answered.

4.5.3 Urbanization and Political Economy

The following sections discuss the origin of urbanization and how systems of urban production and distribution developed, along with the rights of rulers to intervene in them. In the Roman Iron Age, markets such as are found in contemporary Medieval Europe are not evident. Production of elite goods was carried out by attached specialists. In the late Germanic and Early Viking Age, changes in the economic landscape included new ways of organizing production and the evolution of a hierarchy of urban markets: places where goods left the economies of the local community and entered into the con-trol of a central authority. Urbanization

Unlike other areas of Romanized Europe, in Denmark the town was an un-known phenomenon before the Viking Age. Towns were preceded by elite compounds, while the non-elite utilized seasonal trading places and itinerant crafts workshops, but the character of these earlier places was distinctly dif-ferent. In the Early Viking Age, an urban hierarchy slowly began to develop in Denmark (Figure 4.4).

Early urban sites in Denmark show ample evidence for administration and royal intervention in trade and that they were important political ten- ters. They developed for a variety of reasons, often situated to take advantage of natural features: waterways, good harbors, and defensible land with a connection to their economic advantages, such as: ferry crossings, roads, bridges, fords, trade routes, and fishing grounds (Jansen 1985). The first urban centers were trading towns near the coasts, and the earliest was the town of Ribe, in southern Jutland, or Sillende. In about A.D. 710-720, an area that had previously been a simple farming community by the side of the river

Landscapes of Power, Landscapes of Conflict - [PDF Document] (112)


Figure 4.4. Viking Age-early Medieval Denmark, with some places mentioned in the text.

was systematically demolished, cleared, and divided into regular lots, sepa-rated by ditches running down to the waterside for access to transport (Frandsen and Jensen 1987: 178; Clarke and Ambrosiani 1991 : 53). Each sepa-rate lot displays long-term occupation of a single workshop activity.

This seasonal market with regular, pre-planned, and maintained divisions was administered by a central authority, perhaps the same king—Angantyr or some other—who ordered the construction of the Kanhave canal and the early phases of the Danevirke in the same region. Between A.D. 800 and 850, a deep ditch was dug to delimit the borders of the settlement (Figure 4.5). Not defensive in nature, it was probably a boundary that marked where cer-tain services and laws were in effect, including protection by authority, taxa-tion, and rules of fair trade.

Ribe is first mentioned in the Vita Anskarii, or the Life of Saint Anskar, a hagiography, or biography of a saint, pertaining to that saint’s life and death, activities and miracles. Although clearly religious in focus, these works often contain important historical information. This early missionary to Denmark and Sweden later became archbishop of Hamburg-Bremen, and the docu-ment was written by his immediate successor, Rimbert, in the year 870. One

Landscapes of Power, Landscapes of Conflict - [PDF Document] (113)


Figure 4.5. Excavations at Ribe. 1, 2, 3, 4: excavations of portions of the Viking Age market and town; 5: the locations of the Alt Vall still visible in 17th century maps. At 2, other parts of the wall were excavated, and since 1989 it has been found in other tests (after Feveile et al. 1989: 43).

of its important sections referring to Viking Age Denmark is the description of Ribe. Ribe is referred to as vicus, the Latin word for “harbor town” used in connection with many European towns in the form of “wich” or “wig” or “vig” in Britain and Germany, for example, and “vik” in the north (Randsborg 1980:71). Rimbert told how in this vicus the heathen King Horik gave Anskar a gift of land in the town for the building of a church. In our evaluation of this historic text for archaeological interpretations, what is meaningful here is not that a Christian mission was in progress but that Horik had jurisdiction to give away land in the important trading town. This is indicative of consider-able royal authority there (Jones 1987: 118; Roesdahl 1982:76). In ninth-century Denmark, it is doubtful that any Danes were Christian; therefore,

Landscapes of Power, Landscapes of Conflict - [PDF Document] (114)


Horik must have approved the church more to accommodate and attract traders than to encourage conversion, although such attempts were certainly made by visiting clergy.

Ribe flourished alone for some time. Eventually, other towns took root. In 808 the Frankish annals record the following:

He [King Godfred] ordered the destruction of a trading station, Reric [Mecklen-burg or Lübeck on the Continent] which gave his realm great benefit by the col-lection of taxes. He carried the merchants away with him to Sleisthorp. Here he stayed for several days and decided to protect the borders of his realm with a wall. (Roesdahl 1982:73)

The Danish king Godfred had “owned” or conquered a continental mar-ket town, but felt insecure about its location. He razed it and moved it, along with everyone who worked there, up into Denmark proper, on the banks of the Schlei fjord. The “wall” refers to new additions Godfred made to the Danevirke. The town he founded, Hedeby, is well known archaeologically and has been the focus of extensive excavations since before World War II. Hedeby was remarkable in its strategic placement on the border articulating the lands of the Danes, the Franks, and the Slavs. By A.D. 900 it had become the largest center in western Denmark (Figure 4.6). While the harbor was certainly important, as one would expect in a maritime land, the Weiglesdor, or gate through the Danevirke, had a tollhouse, mentioned by Theitmar of Merseburg in about A.D. 1000, and even more goods appear to have passed through this entry than through the sea route. Traders are said to have landed at Hollingstedt and sent wares overland to the town by wagon (Jansen 1985:189). An extensive road system beginning in Hedeby consists of the main Army/Ox Road (Heervej) and many smaller roads connecting to rural areas; parts of it have been excavated in several places. These major, cobble-paved roads must have been maintained by Godfred’s central au-thority, and that of his heirs, because they traverse many districts and con-tinue on the Islands.

At its greatest extent Hedeby was a walled, 24-hectare site, and by A.D.

1000 it had associated cemeteries with over 10,000 individuals. Based on the length of occupation and the number of graves, a conservative population es-timate of about 1,000 people has been postulated (Randsborg 1980:80). Dif-ferent cemeteries contain different socioeconomic classes, ranging from poor and meagerly furnished to large, rich chamber graves.

The excavated portion of the town has produced only one byre, indi-cating that provisions had to be procured from the surrounding farming communities, several of which are known archaeologically. The town’s economy was centered on production and international trade, clearly

Landscapes of Power, Landscapes of Conflict - [PDF Document] (115)


Figure 4.6. Viking Age Hedeby (after Jankuhn 1986).

represented by the trade goods found there: soapstone and whetstone from Norway; querns, ceramics, and wine barrels from the Rhine; ceramics from the Baltic Slavs; luxury textiles of Continental and even Far Eastern origin; English and Frankish jewelry. Trades practiced there included smithing, tex-tile production, bronze casting, comb making, amber working, leather working, and shoemaking, carpentry, and probably archaeological “invisi-bles,” such as furs and slaves. A slave market in Hedeby is mentioned by mis-sionaries who passed through and rescued Christian slaves by buying them free (Roesdahl 1982:24).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (116)


Coins were minted there in the ninth century, the first in Scandinavia, and in 850 a church was built, probably, like the one at Ribe, for the conve-nience and goodwill of Christian traders who put in there. The size and na-ture of the town, the mint, and the church, are indicative of an important administrative center (Randsborg 1980; Roesdahl 1982). In the mid-ninth cen-tury reign of Horik II, a “Count of Hedeby” is mentioned by the Franks. Countis equivalent to the English Earl (linguistic cognate of the Danish Jarl), a type of elite second only to the king, often mentioned as advisers or administra-tors in historical sources.

By A.D. 900, Hedeby and Ribe were both active centers. The specific site of Hedeby was abandoned in about 1020, probably because its harbor was silting in and the new deep-bottomed trading ships of the times needed a deeper harbor. The town did not disappear, though; the population trans-ferred to a site just across the fjord, Slesvik, and continued, uninterrupted, to function in the same way.

A second tier of urban sites emerged in the state’s core area at around A.D. 950, Århus and Odense, and at around A.D. 1000 more appeared, both in the core and outlying provinces—Viborg, Ålborg, Roskilde, Lund, Tumma-torp, Vä, Helsingborg, and others—most of them with royal connections.These sites were all urban, some of them walled, and although they varied in size from around 5 hectares to 50, they were densely populated with regular preplanned grids of streets. They were scattered throughout all Denmark, with Århus and Viborg on Jutland, Odense on Fyn, and Lund, Tummatorp, Vä, and Helsingborg in Scania. Roskilde, on Sjælland in the epicenter of 11th-cen- tury Denmark, was the royal capital city, for by this time, the peripatetic kings had settled down. A cathedral was built there before A.D. 1040. Roskilde developed by 1050 or 1100 into an urban center that covered some 80 hectares, and by 1200 may have been as large as 120 hectares.

More towns appeared in the Early Middle Ages. Helsingør and Havn (later Copenhagen) were located at a ferry crossings, where distances to des-tinations were shortest. Overland routes were clearly much used, and early towns are spaced from each other by a day’s journey by horse and wagon (Jansen, 1985). Skanör and Falsterbo,Ystad, and Simrishamn, founded at 1200or so, were based on proximity to rich herring fisheries.

Unlike transport and harbor-related sites, towns like Hedeby and Ribe, and especially those in Scania (Lund, Tummatorp, Vä, and Helsingborg) ap-pear to have been royal foundations with the express intention of attracting and regulating trade, administering the province by replacing previous large or wealthy elite settlements like those at Dankirke, Uppåkra, and Vä and inserting others where no such large settlements existed, such as at Tummatorp. Odense grew up around the Viking Age Trelleborg fortresses of Nonnebakken, and therefore also has a royal origin, and the Scanian town

Landscapes of Power, Landscapes of Conflict - [PDF Document] (117)


of Trelleborg also contained one of these fortresses and a contemporary Viking Age settlement of large size. The locational analysis of the urban hierarchy and its changes through time and across space will be presented in chapter 6. Production and Distribution at the Transitionto tbe Viking Age

The development of the rights of kings in Scandinavia to take taxes and tolls and generally intervene in the transactions of merchants and markets can be traced from the fall of Rome up to the 12th century. Sources are archaeologi-cal, numismatic, and linguistic, with some historical sources as well. Kings probably offered protection to merchants to ease their activities and to guar-antee their presence, and trading places developed, such as Gudme on Fyn. Tolls and taxes were an outgrowth of this system, a source of revenue when revenue and what it buys replaced prestige as a way of legitimizing power. Ev-idence from England and Scandinavia indicates that the wealth and stability of later kings was directly correlated to their ability to intervene in trade (Sawyer 1977, 1986a).

Beginning at about A.D. 700, the elite-controlled emporia appear, pointsof entry for goods from all over western Europe and beyond. Emporia in all parts of northern Europe follow similar development in three general stages (Hodges 1982a, 1982b, 1984). In the first, kings established safe, neutral trade sites to attract traders, risking foreign presence on their home soil in order to get their elite prestige items. Early Ribe would be a good example of this, with its administrative and legally protective boundary and its church for foreign traders endowed by the pagan King Horik. In the next stage, kings had more authority and control, and in the archaeological record we see the development of large manufacturing and trading stations like later Ribe, and Hedeby. As noted earlier, extensive excavations and analysis at Hedeby indicate that it was an economic and communications center of superregional importance that served the Baltic and North Seas, acting as a gateway between the continent and the north (Schietzel 1985: 180). “Power politics” played a major role in the founding and in the continuation of production, consumption, and trade.

In the third stage, as kings became sedentary and more powerful, ten- ters became administrative rather than trade-oriented. The first two stages are functions of change in a stratified society, while the last indicates state for-mation, as development into a solar, central place is a conscious effort of the ruler to centralize (Smith 1976b; Hodges 1982a). Ribe and Hedeby are sites similar to what have been termed “solar central places” or locations where “partially commercialized exchange operates through an administered mar-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (118)


ket: usually in incipient states or empires where these centers are principally bureaucratic nodes” (Hodges 1982a:16). Previous centers, where trade was more direct and uncommercialized, either became solar types of central places, or disappeared and were replaced by them.

In about A.D. 800, a series of markets appeared in Scania, their names allwith the linguistic element “köpinge,” which means literally “shopping,” a mar-ket. None of these sites developed into towns or cities, and they existed as functioning markets only into the 900s, when they either reverted to small villages or were abandoned. As opposed to Ribe and Hedeby, which became solar central places, the Scanian köpinge sites disappear. In their place, citieswere founded as administrative centers. This indicates that the köpinge mar-kets may have been organized by a central entity and had a specific purpose that spanned only this period. Much more will be said of these markets in later chapters, it is probable that the redistribution of the coveted products of richly agricultural Scania were behind their foundation (Brattberg 1983).

4.5.4 Village Organization and Agriculture

During the Viking Age, there were a number of changes in village-based rural society. Over time, a hierarchy of villages, in terms of size, function, and po-litical importance, developed a pattern that is observable archaeologically. Agricultural production shifted from swidden farming to a two-course and then three-course rotation. Several “waves” of village foundations occurred during the Iron Age, and around A.D. 1000, a final wave—the creation of small, dispersed agricultural settlements—further increased agricultural in-tensification. It is probable that this intensification was socially motivated, en-couraged by rulers who needed increased revenue. As the state became more centralized, the villages also became more regulated, seen through changes in internal village organization. Agricultural Systems and Rural Settlementin tbe Late Iron Age

Even as late as the final part of the Bronze Age, ca. 800 to 500 B.C., the vil-lage was not yet a common feature of south Scandinavian settlement. In-stead, single farms consisting of one or two houses, occupied by close kin, and the associated structures of the farm were the typical settlement (Fig-ure 4.7a) (Callmer 1991e; Becker 1982, 1983). Tacitus wrote: “The Germanido not live in villages, do not join their houses together, and each has an open area around his stead.” Cultivable land was held in common, the land divided according to rank, and plowlands were changed every year (Be-nario 1967:53). Tacitus states that land was quite abundant and that there

Landscapes of Power, Landscapes of Conflict - [PDF Document] (119)


was no need to “plant orchards, set aside meadows, irrigate” or employ other intensive strategies.

A small number of villages appear in the course of the earlier Iron Age, such as the fully excavated sites Grøntoft and Hødde (Figure 4.7b) (VadOdgaard 1985:121–7), and by the Germanic Iron Age, villages were being founded in large numbers (Figure 4.7c). Place-name research, in large part corroborated by archaeological evidence, suggests that many villages were founded in three discrete phases of settlement dating respectively approxi-mately to ca. A.D. 500 to 700, 850 to 980, and 980 to 1100.

At about A.D. 600 there was a drastic change in Scania’s cultural landscape (Ward and Göransson 1991:172): a tremendous expansion of land under the plow (Regnéll 1991:224). Everywhere, there was a dramatic drop in the pollen levels of plants that indicate pastureland and at about A.D. 700 there was a huge rise in cereal pollens (Gaillard and Berglund 1988). This coincides with the first wave of village foundations, which may be one of the most significant changes in agrarian societies in northern Europe, according to Callmer (1991e:337). It is probable that a combination of technological innovations and social factors was responsible for this shift in land use. Perhaps the most dramatic technological change was the creation of permanent arable.

Permanent arable was created during the later Iron Age through the practice of manuring. The origin of manuring in Scania can be dated to 800 to 500 B.C. (Olsson 1991b:299). However, manuring appears to have been used in a limited fashion, and the time between 800 B.C. and A.D. 600 could be termed a transition period with a long continuum from shifting cultivation to permanent stationary villages. A climatic minimum occurred at this time, beginning in the earlier Iron Age, with its harshest phases after A.D. 500. With a worsening climate, cattle had to be byred and fed fodder for the winter. The cycle of overwintering livestock, foddering through hay gathering and leaf collecting, led to the manuring of the fields with the by-product and an in-crease in the ability to maintain stationary arable lands. In turn, permanent arable land led to the infield-outfield system (Olsson 1991b:299) in which fields directly adjacent to the house-plot are manured and cultivated, while outer uncultivated pastures are used for spring, summer, and fall grazing and contain areas for winter fodder production .

This system implies changes in scheduling, in cooperative work, in intra-village relationships, and of course gives new meaning to the ownership of land. Land now became more valuable with time, as it grew richer from fertil-ization. Because of this, fields were not changed every few generations. If new land was needed, new clearings had to be made in the outlands. This system expanded during the period after A.D. 1000 with an increase in the area of the infields and a decrease in manuring after 1200, but otherwise did not change much between A.D. 1000 and 1600 (Gaillard and Göransson 1991:173).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (120)


Figure 4.7. Village development in Scania (after Callmer 1991a). Change in Village Organization During tbe Viking Age

In Denmark during the Late Iron Age through Early Medieval periods—about A.D. 500 to 1200—rural settlement was characterized by a number of marked waves of village foundations and episodes of colonization (Pamp 1983; Holm-berg 1946), as well as the adoption of new farming technology and crop ro-tation. There is strong evidence that some of these dramatic changes in rural

Landscapes of Power, Landscapes of Conflict - [PDF Document] (121)


settlement and farming strategies were engendered by the increasing powerof central government and the changing social and political relationships be-tween kings and farmers.

The earliest villages, founded between A.D. 500 and 700, before state in-tervention in village organization, each had a core with loosely organizedhouses and farm buildings, surrounded by infields, outfields, and thenwooded areas where forest products were collected.

Between A.D. 850 and 980 a process known as interior colonization oc-curred: the founding of new villages, carved out of the farthest-off part of thevillage periphery. A second wave of villages was founded between the first.

The first villages were settled over a 200-year period, then took another150 years to “fill up” and produce offshoots. However, the second wave of vil-lages did not follow this pattern: founded over a 130-year period, by 980most had plenty of room to grow before sprouting buds, yet the next wavecame almost immediately, between A.D. 980 and 1100.

This third wave of villages almost exclusively end in the suffrx “torp,”meaning “settlement dependent on an older village.” These tiny places—manywith only two or three homesteads—are very numerous. An estimated 3,500torps still exist today, and an equal number have disappeared. Torps werefounded both on good farmland between older places and in the previouslyuninhabited interior hummocky landscape. Then, between A.D. 1100 and1200, settlements ending with the suffix “röd” appear, which indicates aclearing in the forest for a single farm.

We know that at this time, the Danish kings enacted far-reaching newlaws and taxes, and new demands were imposed on manpower for theledung or royal navies. On top of this, in the late 1lth century the Churchadded a tithe that had to be paid in meat and flour to the village priest, thebishop, and the archbishop. These impositions upon the farming classes arereflected in village structure, as it becomes a more and more regulated place.

I would like to suggest that the foundation of tiny agricultural hamlets andsingle farms was directly related to a centrally imposed taxation and demandfor surplus production. Colonization effectively opened up more and morearable land, and pollen analysis shows that woodlands came close to disap-pearing.Why not cultivate new fields from old villages? Robert Drennan (1988)noted that intensive agricultural labor requirements in Late Formative and Clas-sic Maya cities encouraged dispersed settlement. Drennan states: “Each house-hold would want to be adjacent to its infield because of the large amount oflabor that such intensively cropped land requires. The effort of traveling tothese fields, and especially of hauling out refuse for use as fertilizer and haulingin harvested produce, would thereby be reduced.” In Scania, add to the fertiliz-ing and hauling the task of clearing forest and removing boulders and rocks. All this would need to occur quickly to meet demands from the state.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (122)


During the Early Middle Ages, the infield-outfield system was expanded so that every part of the village area was involved in some sort of production. Around the same time, the two-field system was replaced by the three-fieldcrop rotation, far more intensive. This promoted a patchy landscape with many exploitable resources (Olsson 1991b:301). The primitive ard was replaced by the productive wheeled moldboard plow, which eliminated fragments of roots and seeds from plowed land, lowering the occurrence of weeds, a new tech-nology that vastly eased cultivation. Often the area under cultivation was ex-panded at the expense of the pasturelands, and there appears to have been a deficit of manure (Olsson 1991b:300). Low yields just after the Early Middle Ages have been attributed to permanent under manuring combined with cli-mate deterioration (Titlow 1972). In considering the origin and limits of the dense phosphate patches around the prehistoric village sites in Scania, it is im-portant to remember that manuring technology was only intensively used be-ginning in A.D. 500 to 600 and fell off sharply after the 1100s.

What evidence links these changes to state demands for taxes and sur-plus? The first two waves of villages were haphazard and loosely woven. Data from excavated torp hamlets show major divergence. They appear to be laid out on a predetermined plan, a standard land division. Several identically sized lots, or tofts, each with a house in the left corner by the village street, indicate land divisions made according to a standard rule.

Simultaneous with the founding of carefully-laid-out torps and the increase in land under the plow, the older villages were restructured and redefined. Most contracted in size, and their internal organization was altered. Mirroring the layout of the new hamlets, houses moved from scattered and loose, to close and tight, within a new farmyard or toft. The contraction of old Scanian villages between 1000 and 1100 A.D. was up to 60% or more (Thurston 1996).

Early law codes indicate that each farmer had to help outfit and equip longships and man them as warriors. The Danish word for the bench where the farmer-warrior sat in a ship was toft. The linkage of this word in context of taxation through military labor and provisioning and the new domestic residential pattern is unlikely to be coincidental. Mats Ridderspore (1988) be-lieves farmers were required to site their household on a regulated toft, whose size was based on the size of the farm, signifying the farmer’s wealth (Figure 4.8). Thus, royal officials could estimate his contribution in a time of nearly universal illiteracy and lists of what thousands of farmers owned or owed were not feasible. This process of reorganization in rural settlements will be discussed in detail in later chapters.

Simultaneous with these changes came the founding of fortresses and urban centers in Scania. Between 990 and 1000, four Danish administrative towns sprang up. Founded by royal charter, they contained law, tax collection, administration, markets, industries, and crafts. Some even minted royal coins.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (123)


Figure 4.8. Tofts in Lilla Köpinge village 1733 (left) and hypothetical reconstruction of the regulated village tofts. Gray areas are excavated (from Tesch 1993, after Ridderspore 1988).

To summarize, old villages were restructured, new regulated villages founded, agriculture streamlined, and the places of traditional institutional ac-tivities, shrine, assembly, market, and so on that had formerly been in dis-crete locations were conflated in towns. Political changes began to stress the importance of one village over another; places where administration was car-ried out may have become larger in size, creating a hierarchy of village sizes.

4.5.5 The Sacred Landscape: Geographic and Ideological Change

Archaeological finds and primary texts demonstrate that the sacred landscape was also altered coincidentally with other familiar vistas and that changes in religion were engineered by the state. These new practices both reflected and influenced changes in the organization of society, affecting not only the spiritual but the secular affairs of the Danes.

When possible, contextualizing socioeconomic and political change within a belief and value system enables a much more comprehensive un-derstanding of culture change than an analysis of artifact or settlement distri-bution alone. This information is accessible because both elite and common participation in religion has left evidence that can be studied, giving a rounded if fragmentary view of the sacred sector. This will especially be stud-ied as it pertains to changes between the more personal religion, both pagan and early Christian, of the Late Germanic and Early Viking Age and the insti-tutionalized, largely state-sponsored dogma of later Christianity.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (124)


4.551 The Study and practice of Old Norse Religion

There are a number of avenues one may take to interpret and understand Norse religion. There is the linguistic and etymological method—looking at the derivation of words and phrases and their origins for clues to religious and social organization. There is symbolic study that sees the mythic stories of warring gods and heroic figures as symbols of norms, conflicts and processes within society. And there is a historical approach where impor-tant themes and structures on which society hinges can be traced back through time. Some of these approaches give productive insight into Iron Age culture. Pre-Christian Religion in Scandinavia

Much of our knowledge about the details of pre-Christian practices comes from the pens of those who converted, or tried to convert, the Scandina-vians, especially Rimbert’s Vita Anskarii (Life of St. Anskar ), and Adam of Bremen’s Gesta. Since these texts are clearly meant either as hagiographies or at least enshrinements of the deeds of holy men, they contain a “theolog-ical interpretation of history” (Edsman 1990:11). At one time these texts were highly mistrusted; now they are accepted as relatively trustworthy doc-uments of geography, history, and the personalities encountered by the mis-sionaries in the course of their travels and endeavors (Edsman 1990:11). In fact, more is known about the belief system through the preservation of a fairly complete myth cycle than about how rites, ceremonies, or obser-vances were conducted.

Old Norse religion, like other aspects of the culture, is not typical of a state-level society’s belief system. It did not posses a word for “religion” but only the word sidr, meaning “custom.” No original words for “belief or “be-lieve” (in god[s]) exist; the later word for this, trú, is a loan word.“To vener-ate” or “to pray” are nonexistent; the oldest known Norse used only the word bidja, “to ask for” (Edsman 1990:23). No dogma is known, and priests and temples are never mentioned until very late and are probably developments of the late Viking Age, a countermove to the building of churches, rather than the normal way of conducting religious rites.

Then there are the archaeological sites—runes, rock carvings, and offer-ing places—that can be hard to interpret. Offering-places where military gear was ritually broken, in thanks for victory by the warrior class, have already been discussed extensively above. Others yield gold and silver symbols, so-called guldgubbar: small gold foil bits stamped with the pictures of gods and goddesses, each one probably representing a wish for fertility, wisdom, vic-tory, or luck, sacrifices of the noble and wealthy, so that their lineage would

Landscapes of Power, Landscapes of Conflict - [PDF Document] (125)


procreate and their wealth continue into the future. Colleagues in Scania re-port farmer’s shrines where the treasure consists of modest farming tools and wooden bowls and plates that once held offerings of common foods.What do farmers ask of the gods, and what do they give as gifts? They ask for simpler things than lords, and their offerings are subsequently rarely as complex as the intricate guldgubbar found in elite offering places. All these diverse prac-tices are part of Iron Age religion.

One of the main reasons for these vastly different gestures within one belief system is that much of the pre-Christian ritual was of a personal type: private offerings in natural places. There may not have been many hard rules on how to make a plea to the supernatural, although charms and incanta-tions, some of which have been preserved, must have been somewhat stan-dard in form. Although marriages, rites of passage, oaths, punishments, and ordeals were witnessed by the community and probably had established rit-uals, other rites were likely to have been less public.

The well-known passages from Snorri Sturleson’s Heimskringla de-scribing great temples where crowds of farmers and lords participated in rituals—the sacrifice of animals, the sprinkling of blood on idols and tem-ple walls, the feasting on the sacrificial animal, the drinking of memorial toasts or minni to the dead—all this has been convincingly argued to be di-rect interpolations from Old Testament descriptions of Judaic rites put into the saga to emphasize the conflict between Håkon the Good and the hea-then farmers over their pagan practices (Edsman 1990:25–26). Minni ritu-als were actually a Medieval, Christian invention practiced by the Scandinavians and anachronized into pre- and protohistoric times by Snorri. These stories were authored for the benefit of Christians, and al-though they dealt with true events and people from the past, they were made to be understood by those who heard them. This does not mean that rites and rituals did not exist, but merely that the popular conceptions of Viking religion can be deceptive.

Even scholarly conceptions of Norse religion can be baseless and fraught with biases: consider Musset (1967:265), who stated, “Nordic hea-thenism was never a personal faith,” or Baetke (1942:218), who said, “The question [of religion] was not about the individual or his soul or destiny, but about the community.” Ström (1990:374) believes that this is only true if wecompare the typical Viking with a modern Westerner or a Platonic Athenianand that there was much personal faith in old Norse religion (Ström 1990: 375). There are several examples of individuals expressing confidence or trust in promises made by Odin and Thor: “I was confident in believing him [Odin] ” or “Then I should believe in you [Thor] .” There are examples of people becoming angry or disgusted with one god and renouncing their be-lief in favor of another.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (126)


The skaldic poems, composed mainly in the 900s, are an even better source on Viking Age beliefs than saga material, because they were written when Christianity was barely practiced even by kings. In the poem Sonator-rek, or “the irreparable loss of the sons,” made by Egil Skallagrímsson in about A.D. 950, Egil mourns the loss of his sons (Edsman 1990:28). This poem is be-lieved to represent the actual religious testimonial of an old warrior, a Viking, Egil, and clearly shows a personal relationship with the gods. It the poem, Egil rebukes his patron Odin after one young son has died of fever, and the other, 17 years old, has drowned:

I was on good terms with the lordof the spear [Odin], I grew trustful, believing in him, until the friend of chariots, the prince of victory, broke friendship with me.

Although sacrifice was the normal means of personal expression to the gods, there are some examples of actual prayers, both male and female. Women are said to have called on Frigg and Freya when giving birth, while a male begins a prayer with

Frey, who for a long time has been my confidant and received many gifts from me and well rewarded them . . .

One of the best examples comes from an amulet (Ström 1990:377; Lindquist 1987:78–80) in which the bearer, Bofi, inscribed his own name several times, calling particular attention to himself and especially asking for favors. This may represent a type of “appropriate” formula that could be used to speak one to one with gods.

I praise Erka, I am wise May the only son keep evil away from Bofi! May Thor save him with that hammer which beats Am Go into the sea, Am! Fly ugly ill-elf.Yougetnothing fromBofiGods are under him and above him!

Although not many of these prayers are preserved in written traditions, they seem to have been a common practice. In terms of sacrifice, there are numerous examples of personal, as opposed to communal, sacrifices: sacri-fices for personal victory, health, and long life may all be found in Nordic texts (Ström 1975:220 f f ) and in archaeological finds.

Another religious tradition was the útiseta that superficially is described as removing oneself from company and going to sit outdoors. This is decep-tive; it actually represents a personal ritual of going off to gain knowledge of

Landscapes of Power, Landscapes of Conflict - [PDF Document] (127)


“hidden things” through isolation and possibly deprivation, something like a vision quest (Ström 1990:378). This is in contrast with communal rituals such as divination that were performed with many assistants and watchers.

There were many elements of communal focus in pre-Christian religion, but one of the features of the system was the personal relationship between the gods and their worshippers. The Social and Political Implications of Religion and Religious Change

The period spanning the first millennium A.D. may be divided into three phases of religious practice for the purpose of understanding religion in the light of this research. The earliest phase consists of “personal religion,” an era when religion did not seem to be especially controlled by elite (though this did not rule out a role for elites with holy duties). During this time, offerings were small and appear to have been made by individuals (Fabech 1994). There does not seem to have been an organization of religion beyond a local level. An intermediate phase began when local elites appear to have appro-priated the organization and location of religious rites, while the scale of its organization remained fairly small and control still local. Finally, the latest phase involved the interweaving of religion with the state: religious hierarchy became as deep as the secular system, control of belief was lifted up and away from the village, and was centered on royal and urban systems.

In the early Roman Iron Age, sacrifices were made in bogs and lakes and were mostly small objects that were probably offered as personal gifts for personal wishes—the desire for luck, wealth, and children, as they had been for millennia. This practice goes back far earlier than the Iron Age, into the Bronze Age and beyond. Tacitus and Strabo both noted in their writings that sacrifices were made in wetlands.

In the late Roman period these gifts remained focused on water but grew to include large communal offerings of booty and weapons and people (Fabech 1994:169). This reflects the rise in intragroup conflict and in the rise of a military elite hierarchy. By the end of the sixth century, this ended. Bogs were no longer used, and the objects of sacrifice changed. Fabech (1994: 170) concludes that these two changes—the location of rites and type of of-fering—represent a major ideological shift in this transitional time.

Gold bracteates and later foil figures become the offerings of choice, and were now deposited on dry land, and were associated with “important” settle-ments. Procopius, writing in the 6th century, Ibn Fadlan in the 10th century, and Theitmar of Merseburg in the 11th century, describe sacrifices of people and animals that occured on dry land: victims were hanged from trees and then thrown into the scrubland or strangled and burned (Fabech 1994:170).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (128)


Place-names also support this change. Bog finds are rarely associated with sacred names, but the later, dry-land finds are often associated with them: Vä, Gudme, Odense, Helgö, Gudhem, Åsum (Ashem), all are names that indicate pre-Christian holy places. This indicates that older names for shrines in bogs died out with their use, and new place names associated with new traditions remain, even to this day (Fabech 1994:171). The Natural Offering Place and the Built Temple: Personal Versus Institutionalized Religion

Fabech (1994) and Näsman (1988) attribute this religious change to achange in social organization, where chieftains were transformed into kings, and religion was taken out of the vernacular and common practice andbegan to be transformed into an elite-controlled ceremonial system. Places of ceremony associated with these elite—buildings or enclosures—would have replaced the natural shrine. Local magnates of the late Germanic and early Viking Age appropriated the sacred sector, inserting themselves be-tween the folc and the gods.

Fabech argues that in the late 10th century when original, wooden manor churches were built, they were constructed at the locations of the foil figure or bracteate sacrifice spots, at least in all sites where guldgubbar havebeen found to date (1994:174). The church was built by the local magnate in the same spot where the family had previously conducted pre-Christian rites. Slightly later, ca. A.D. 1050 to 1125, stone churches, still standing today, were built on top of these early wooden versions. It is this phenomenon that al-lows us to speculate about the location of the “elite” farm or estate in present-day villages. Thus, the idea of a “manorial” church was not introduced with Christianity at all, but was a long-seated tradition. The big break was between the Roman and Migration eras. Fabech suggests that the early idea of remov-ing the pagan shrine to the elite territory as a means of enforcing control may have been an imitation of Frankish Christian kings who did this within a Christian context, but adapted to the Scandinavian non-Christian milieu.

Although Adam of Bremen (writing in 1075) stated that Odense and Århus were bishoprics in the 950s it is universally accepted that this was a wishful conceit of the Church (Jones 1987). They “appointed” bishops to oversee these areas but no German prelates dwelt there and this fanciful hi-erarchy really represented the conversion attempt rather than any Episcopal presence. As late as 1075, northern Jutland and all of Bornholm were still hea-then lands, so the church was patient, and many compromises were made in order to make conversions and restrain new Christians from apostasy, ac-commodating their belief patterns, syncretizing old and new forms (Johan-nesson 1984:47).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (129)


A major state-sponsored change of religion might seem to be a transition that would lead to unrest and outrage.Yet although the conversion was notforced, and old beliefs persisted, Scandinavians generally embraced Christian-ity, especially with the endorsem*nt of the king, who of course saw many sec-ular advantages in it. Old Norse religion was rather fatalistic: the world wouldone day end with a great battle called ragnarok, the “breaking of the gods” or end of the world, which envisioned not only the death of the deities but of mankind. There was no escape, no way to change this fate, and nothing to come after—no heaven, no hell. All was to cease being. The possibility offered by Christianity of an unending afterlife with a god who was eternal, may have helped convert many pagans from their grim belief system.

In addition, early Christianity was not very different in “feel” from pagan religion. Although the early churches were elite property, the rites were local and all farmers within each lord’s territory came and sat in a little church in their village. Village churches were places of community ceremonies sur-rounding death, birth, and marriage but also personal rites of communion, unction, and confession that were between god and person, with priest as fa-cilitator. One could also pray in this small local place to Jesus Christi, whoembodied the personal aspects of Christian religious practice. Decisions on how church and people interacted and the social reproduction of sacred cus-toms were decided within a few miles from a farmer’s stead.

It might be said, then, that the pagan practice of personal religion was continued in the earliest practice of Christianity. Although there was an Arch-bishop somewhere far away in Germany, for all intents and purposes the vil-lage priest (who was often a local elite) was the highest level of sacred authority and was there to answer to the parishioners. Even the repetitions and recitations of Catholic catechism probably felt familiar from its similarity to performing incantations and singing runes. The idea of Jesus as a watchful figure seeing all may not have been too different from having a personal pa-tron in Frey, Odin, or Thor.

Although the village church and its priest did not disappear in the later Viking Age, these local aspects of belief were soon overwhelmed by other considerations. By the mid-11th century, the control of the sacred sector had been lifted and moved far away, both physically and in the sense of control and power. The Construction of the Late Christian Landscape: The Church as a Function of the Town

When urbanization overtook Denmark, one of the first hallmarks of each town was the construction of large and numerous town churches and monas-teries. Unlike the small village church connected with the old lords, the town

Landscapes of Power, Landscapes of Conflict - [PDF Document] (130)


churches were the recipients of royal gifts of land, mills, estates. The earliest surviving royal donation letter of this sort dates to 1085, in which Knut the Holy gave 52 1/2 bols of land, each bol containing an unknown number of farms, in and around Lund, for the support of the Church, as well as the right to a percentage of royal taxes collected from the towns of Helsingborg, Lomma, and Lund. There were many monasteries, and these and the churches were supporters of the king and the keepers of important records.

As early as the reign of Sven Estridsen (d. 1074), the state was attempt-ing to free itself of its obligations to the Archbishop in Hamburg, since this connection allowed the Germans to insinuate themselves politically into Den-mark (Oakley 1972:44). In 1103, King Erik Ejgod, the son of Sven Estridsen (who might be called the last “Viking” king), traveled to Rome and plunged Denmark into Papal and secular politics. The Pope was deeply embroiled in a long-standing three-way conflict with the Archbishop of Hamburg-Bremenand the German Holy Roman Emperor. The wayward Archbishop in Germany was at that time supporting the cause of a pretender to the Papacy—a so-called “anti-Pope.” Erik found that he was in a position to realize his father’s wish of shaking off the German Archbishop. The Pope, taking the side of the Emperor, chastized Hamburg by creating a new Archbishopric at Lund in Sca-nia and put all of Scandinavia under its control—not just Denmark but Nor-way and Sweden as well. He hoped to diminish the power of Hamburg and give an ally to the Emperor at the same time (Johannesson 1984:45). Princes of the Church

The church in Denmark soon became so enmeshed with state politics that they were virtually indistinguishable. A man named Asser became arch-bishop of Lund in 1104. This period can at best be called protohistoric for the documentary evidence is little better than at A.D. 800.What we do knowis that he came from the greatest magnate family of Jutland, the Thrugotsen clan, which was closely related to the Crown; he was a blood relative of the king (Johannesson 1984:47). Under his rule, the tionde, or tithe of one-tenth, was instituted, supposedly to free the church of dependence on the state yet in fact permitting the church to establish itself as an independently funded partner of the state. Priests and parish churches were each to receive one-third of this tax, to be paid mostly in goods and food directly from their parishioners, and the last third went to the Archbishop. Much of the church’s income was demanded in flour, and since the price of flour rose and rose during this period, the church was enriched by “selling high,” as if successfully playing a stock market. It came to be that the Archbishop could even influence the price of flour. This tenth became a standing tax in Den-mark (Johannesson 1984:46,50).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (131)


Village churches were supported by the tithe but were relegated to in-structing peasants and performing marriages, baptisms, burials, and other rites for the lowest level of society. Local lords remained the overseers of this small-scale Christianity: screened lofts where the lord and his wife could sit undisturbed by the lowly are found in the gable end of many tiny village churches, directly across from and facing the alter. Up a tiny turning stair, in a chamber with alcoves for relics and statuary, they could look, like God, on the village priest and the parishioners. This was all that remained of their one-time complete control over the sacred sector.

In the Late Viking Age/Early Medieval periods, one of the main com-plaints, leading to violent uprisings, was this religious tithe. Farmers in Scania and elsewhere protested it and refused to pay. Thus, it was not the conver-sion to Christianity that produced the majority of unrest but the transition to a state-controlled, bureaucratic church that was perceived to be party to the oppression felt by the angry Danes. In the 1180s, Absalon, a personal advisor to King Valdemar, was Archbishop. During the Scanian uprising he was nearly murdered by angry Scanian farmers and local lords; clearly this usurpation of the sacred by the center did not sit any better with the Scanians than other incursions into older lines of power.

The town of Lund was the Archdiocese of Denmark, and other bish-oprics were located in Roskilde, Viborg, Ribe, and Slesvik, where mints and other central-place functions were also located (Andrén 1985:52). From these bishoprics, and the Archbishopric in Lund, all important decisions in the church were made and proclaimed. They controlled the smaller village churches and also collected the tionde. This change in church organization goes hand-in-hand with urbanization and represents the first discontinuity in the control of religion since the end of the Roman period—the removal of the sacred from the control of local lords and the transfer to the central gov-ernment, in the form of the king and his close association with the new in-stitutionalized Christianity. Thus, the personal, small-scale religion of the Viking Age, both pagan and Christian, became a state-controlled cult.


The chapter began by discussing the correlation between early alliances or proto-unification and military aggression by the Franks. The need to reconcile internal warring factions in order to mount a viable opposition may have added to early integration in western Denmark. Some possible social explana-tions for how unification might be achieved without conflict were offered. Next, the nature of rulership and difficulties in imposing a centralized deci-sion-making hierarchy on a traditional Germanic people were explored. In the

Landscapes of Power, Landscapes of Conflict - [PDF Document] (132)


second half of the chapter, the efforts of rulers toward centralizing and con-solidating power through a number of strategies were presented: changes in the political, the socioeconomic, and the sacred landscapes were widespread.

The location in the landscape of these institutions were un-integrated during the late Roman, Germanic, and early Viking Ages. Political centers, mil-itary outposts, economic hubs, and the places of religious activity existed in separate geographic locations, indicating a decentralized, difficult-to-controlsystem. In the late Viking Age, a sharp discontinuity occured, and fairly rapid changes removed these institutions from local control and conflated them all together in an urban setting. Increasing centralization became the hallmark of the Terminal Viking and Early Middle Ages.

The next chapter will begin to contextualize the archaeological record by couching it within the realm of political ideology. Evidence for the intro-duction of new terms of social relations will be discussed and compared with archaeological evidence for the new social classes to which they referred. It is hypothesized that while new social classes of elites were created, the old ones were far from gone. This unresolved social conflict is one of the factors behind the unification problems of the Viking Age state.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (133)

This page intentionally left blank.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (134)

Chapter 5

Social Classes in the Viking Age

Contentious Relations


It might be expected that there were many difficulties when alliances be-tween peers shifted to a hierarchic system with one supreme elite. This sec-tion will focus on actually identifying the inception and development of a new social class made up of people who were close to the Crown, and who were invested with lordship by the king specifically to carry out the instruc-tions of the Crown and counter the power of the old elite class. There is lin-guistic evidence for these changes as well as archaeological data supporting the rise of a new class. Specific incidents in the historical record indicate how strongly or weakly old Germanic law was in effect at different times between about A.D. 9 and 1200. Coincident with many other indicators discussed al-ready, the traditional relationship between rulers and subjects appears to shift substantially away from its original social and sociolinguistic model.


Language and the way it is used in and among different social classes has long been a topic of anthropological inquiry. Sociolinguistics can provide impor-tant evidence for the Viking Age, when terse runic inscriptions and sketchy


Landscapes of Power, Landscapes of Conflict - [PDF Document] (135)


accounts of elite activity are often the only clues to social relations. Fortu-nately, these types of documents often record terms of social relations. Terms of social relations refers to what people use to name themselves and each other, according to categories that have been constructed by their so-cial groups. These may be religious terms, or political terms, or terms related to work or gender, but they are titles and words that identify people’s place and status within society based on a uniformly understood internal system. These terms are extremely “loaded” in many senses: in historical cases, they were often strictly regulated and imbued with entitlements related to social class and power status. A farmer could not call himself a king or a thegn. A thegn’s household was entitled to extract labor and goods from other house-holds, a bishop could demand a tithe. Even the forms of address between classes were “power-coded” (Trudgill 1983:103–5), for example, a king or lord could call a grown man “boy” while the man must call the king “lord” which is often the same word for god (lord/Lord in English, herr/Herr in Dan-ish), or even call him father, through the use of the term “sire.”

While the terms are only words that signify the power of office or estate, they are, as such, the portion of social class that resides in the ideological realm. Armies and followers are a tangible aspect of power, while the belief system, the social code, and cultural identity rests in an invisible but power-ful sphere of concepts, perception, and ideology almost as fully capable of enforcing institutional rules as men with swords. Such terms change naturally and gradually over time as the systems they relate to evolve and change, but rapid and unusual evo/devolution may indicate that they were manipulated purposefully by one group or another in order to accomplish a goal.

The first part of this section looks at the terms and the relations they de-scribe during the early part of the study period, and follows them through time. The oldest known forms and uses of social terms agree with what is known of the corporate social and legal system of the Roman Iron Age. A few historic examples illustrate how the terms, ideology, and the reproduction of social institutions are interrelated. Between A.D. 800 and 1100, a new set of elite terms of social relations were introduced in Denmark. These terms were related to the creation of a new social group, and served to elevate the new and suppress the old. Later, archaeological evidence for this new social class will be examined.

5.2.1 VikingAge Sociolinguistics

During the Roman Iron Age, the relatively southern Germanic tribes that Tac-itus described and the northern Germanic groups in Scandinavia had very similar social structures, known to us by slightly different names. Specifically, Scandinavian evidence for this system may be researched through terms of

Landscapes of Power, Landscapes of Conflict - [PDF Document] (136)


social relations and their meanings not only through Tacitus but in runic in-scriptions, skaldic poems, the eddas and sagas, and the early laws, covering the period from the 1st century to the 12th century A.D. (Lindow 1976).

5.2.2 TheTerms of Social Relations

Tacitus described the Germanic social institution known to the Romans as the comitatus. The actual word is druhtiz; its leader the druhtinaz (Lindow1976:37) and, similarly, in the Old English of the Germanic Anglo-Saxons,dryhten. In Scandinavia, its close linguistic and cultural cognate is the drótt,which is often referred to in the native literature and chronicles. In addition, Hedeager (1 987, 1992) has combined Tacitus’ description with the Scandina-vian archaeological record, and suggests that this social institution was well es-tablished in southern Scandinavia by the second century A.D.. In its original form, the word drótt and its corollates denoted a relationship between a group of warriors and their leader: not a kin- or lineage-based system, but a personal relationship based on a political hierarchy (Swanton 1982:12).

The drótt was based on sacred mutual bonds and obligations between leaders and followers. In times of conflict it was a warband where neither party was to walk away alive leaving his counterpart still in battle. The war-riors were to die for the warlord, yet he too was expected to die with his men, not use them as a shield or to effect his own escape. In peacetime, fol-lowers received elite status markers, such as neck-rings, Roman glass, and precious metals, that signified and legitimated their elite status. Special favor incurred an increase in rank in the warband. This institution was not only im-portant in political ideology but extended into the sacred sphere. Warriors who fell in battle according to the rules of this system were guaranteed en-trance into the heroic afterlife, and the cultural ideal held that men should rush into battle to die, for if they died in their beds as old men they might be denied a warrior’s reward. The well-known saga Beowulf describes such a warband, and though it is preserved in Old English, the story is set in Scania, in the Germanic Iron Age.

The drótt was ranked by the leader, and there was rivalry and competi-tion among the warriors for favored places, and also between leaders as to who had the best warband (Hedeager 1987:132). In the Roman and Ger-manic periods, all chieftains and petty kings had their own such warband. Al-though the warlord’s power was limited in peacetime, the drótt did not disappear when war ceased.

The drótt was a primary social institution, a behavioral pattern that was of established importance to society and a major element in its cohesion (Lin-dow 1976:2). There is archaeological evidence for the composition and struc-ture of these groups. At Ejsbøl, Denmark, a study of third-century bog offerings

Landscapes of Power, Landscapes of Conflict - [PDF Document] (137)


indicates that the defeated local enemy was an army of about 200 men with spears and shields, 60 of whom had swords. Fifteen elites are indicated, nine mounted. One of these would have been the druhtinaz or chieftain. This is interpreted as the gear of a defeated enemy drótt replete in sacrifice with its

Weapons from a single offering at the Ejsbøl site

Saddle: 9Bridle: 9Spurs: 8-9Sword: 60Dagger: 62Swordbelt: 60Shield: 123+52 fragments Spear: 191 +fragments Pike: 203+f ragments

9 horsem*n 60 footsoldiers with swords, spears and daggers Approximately 130 footsoldiers with spears only

Figure 5.1. Weapon sacrifices throughout south Scandinavia, and their relationship to army size and composition as seen at the Ejsbøl site on Jutland (after Jensen 1982).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (138)


internal ranking and hierarchy. Such a large army does not represent an ad hoc group of peasant recruits, but a full-time warrior group (Hedeager 1987: 132-33). Ejsbøl is not the only site from which the weapon offerings of whole armies have been found: Figure 5.1 illustrates Ejsbøl and other similar sites from the Roman Iron Age.

The title of leader was a title of respect, not of inherited authority. Old kings were called frea, “respected and wise,” while young active kings were called drótten, “vigorous and war-like.” The folc fought to defend the law, not the king, and likewise the concept of treason was not of activities against the ruler but against the law of the community (Swanton 1982:18).

By the early Viking Age, there were clear terms for those in the king’s ret-inue. The dräng indicated a younger elite member of the drótt (Christo-phersen 1982:131). The skeppare appears as the captain of a chiefly or royal vessel, while the himthiki, or housekarls, were the army’s lower-rank, yet still elite, soldiers. Some were loan-words from English lands, where the Danelaw, comprising much of northeastern Britain, was conquered, settled, and ruled by Danish lords, beginning in the ninth century. The word thegnwas one of these; it was the term for an older, more mature elite warrior. These more and less important members of the warband had clear ranks and honors dependent on experience and age. These terms defined fundamental roles in a major social institution, and these concepts were deep-rooted for nearly 1,000 years, from Roman to Viking times.


The earliest example of the removal of over-ambitious elites may also be one of the best, as it is detailed in Tacitus. Arminius (a Romanization of the name Hermann) was the chieftain of the Germanic people called the Cherusci. He was a Roman citizen who reached the rank of Eques (knight) but betrayed the Romans, becoming overlord of an alliance of Germanic tribes. He led his confederates against the Roman general Quinctilius Varus in the Teutoberg Forest in A.D. 9, wiping out three legions and causing the Emperor Augustus to retreat from his attempt to conquer the North. A Roman legion was com-prised of 5,000 foot soldiers and 120 cavalry, as well as additional large num-bers of auxiliary troops and staff. Arminius, therefore, destroyed between 15,000 and 20,000 men. This is no exaggeration; it is well recorded in many Roman chronicles. Suetonius, writing around A.D. 120, tells us that Augustus “took the disaster so deeply to heart that he left his hair and beard un-trimmed for months; he would often beat his head on a door, shouting ‘Quinctilius Varus, give me back my legions!’ and always kept the anniversary as a day of deep mourning” (Suetonius 1979:59).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (139)


Of Arminius of the Cherusci, Tacitus wrote:

Assuredly he was the deliverer of Germany, one too who had defied Rome, not in her early rise, as other kings and generals, but in the height of her empire’s glory, had fought, indeed, indecisive battles, yet in war remained unconquered. He com-pleted 37 years of life, 12 years of power, and he is still a theme of song among barbarous nations, though to Greek historians who admire only their own achieve-ments, he is unknown, and to Romans not as famous as he should be, while we extol the past and are indifferent to our own time. (Benario 1975:86)

Paralleling the finds at Ejsbøl and elsewhere, it is also known historicallythat there were many disputes between the so-called tribes—elite versus elite—which the Romans “encouraged and exploited . . . as the Romans could not defeat the Germans militarily but encouraged them to exterminate each other” (Hedeager 1987: 128). This strategy worked: Arminius’ father-in-law,Segestes, was also a renowned chieftain of the Cherusci, but he allied himself with the Roman Empire and encouraged the Romans to exterminate his son-in-law. Arminius’ brother Flavus was a spokesman for the Roman side. In A.D.

16, Arminius and his brother were representatives for opposite sides. Meeting at the river Weser for negotiations, they “hurled insults at each other across the river” (Dudley 1968:230), Arminius berating his brother for selling the freedom of the Germans “for the wages of slavery, and at a cut rate!”

Ten years after his victory, Arminius was assassinated by his fellow Ger-manic peers in A.D. 19. Tacitus tells us the cause for the murder of this “deliv-erer”: “. . .ironically, his regard for freedom was not absolute. He tried to makehimself king among the Cherusci and was killed by his kinsmen, ‘as the enemy of the freedom of the people’ ” (Dudley 1968:230).

This pattern of assassination of overlords who became too permanent or ambitious is an important recurring theme. The practice of regicide as an approved community action against unsatisfactory leaders continued for a long time.

It is relatively sure, thanks to the clarity and sophistication of Roman writers like Suetonius and Tacitus, that Arminius was assassinated because he was too long and too firmly in power, thereby violating the corporate social and political order. During the next few centuries, no documents illuminate further examples, yet in the ninth century similar events can be interpreted in the same light, though the details are sketchy and documentation is poor. In A.D. 811, as mentioned in earlier chapters, Godfred was in much the same position. He also faced a powerful enemy, the Holy Roman Emperor Charle-magne, and may have been overlord of a federated south Scandinavia rather than king of a truly united polity.

However, Godfred was not in exactly the same position as Arminius and earlier leaders. The Cherusci battled with Rome for a few years before freeing themselves of direct conquest. By the ninth century, the need to organize a

Landscapes of Power, Landscapes of Conflict - [PDF Document] (140)


defense, both military and political, that lasted many decades, transformed Godfred into a new, more sovereign, and less “democratic” ruler. Apparently, like Arminius before him, he did not wish to relinquish the position. Godfred won his war with the powerful Franks in A.D. 810, defeating them all along the Jutland and Frisian coasts in a series of naval battles recorded by the Franks in their Royal annals. One year later, rather than being venerated as military victor, Godfred was assassinated by his own son and other kin. In the Annals, the cause is not made clear, but I suggest that this was the corporate social code in action—the deposing of a leader who has usurped the equality of his peers and no longer answered to the law or the folc. Instead, modeling himself on his enemy, the Holy Roman Emperor, perhaps Godfred attempted to claim sovereign power. Unfortunately for Godfred, the time had not yet ar-rived when a Dane could emulate Frankish political custom, the same Franks who had abandoned or altered their Germanic social code in favor of a Mediterranean model sometime soon after the fall of Rome.

After Godfred’s murder, a power struggle ensued between the elites who were eligible to rule. Eventually, Godfred’s nephew, Hemming, suc-ceeded him, and in 812 he concluded a peace treaty with the Franks. As noted earlier, the diplomatic “team” that negotiated the peace was made up of men from all regions, and included Osfrid de Scanowae (of Scania).

From this point on, a series of close male relations follow each other as king, at least king in south Jutland. Hemming’s son Horik took the throne, and then his son. This may indicate a system at odds with the previous social code, especially when we find that in the mid-ninth century, the king who ascended to the throne was a boy, called Horik, or Erik, the Child. Adolescents and older children became common heirs after this: Knut the Great of Denmark and Mag-nus Olafsson of Norway, two major kings of the 11th century, were 13 and 11 years old, respectively, when they took their thrones. This suggests that the king-ship had altered from an achieved, earned office to an ascribed, dynastic line.

These shifts in the social code, such as the establishment of ascribed rulership and the move away from elected kings, does not appear to have been accomplished through force and large-scale conflict, although the awareness of military might must have been powerful in itself. The manipu-lation of ideas and social relations must therefore be considered as a weapon. There is evidence to support this.


The early Roman era witnessed war rulers who oversaw the ordering of conflict but did not rule over society in peace-time, when the assembly and the lawspeaker governed. By the Germanic Iron Age, kings appear to have

Landscapes of Power, Landscapes of Conflict - [PDF Document] (141)


ruled continually, whether in peace-time or in war-time. The ruler’s retinue was originally made up of lords with whom he had a reciprocal relation-ship. They did not live simply to serve him, he owed something back and was accountable for his actions. These underlords had the right to unseat their elected overlord if he became too powerful and threatened the tradi-tional system.

In the early 11th century, the group of lords and followers around the king appears to have undergone a reorganization of some magnitude. Ar-chaeological, linguistic, and historical sources indicate that the new dynasties began a process of suppressing or eliminating the old-style elites from which they had arisen and created a new group of favored lords. The new lordly class was not connected with the old forms of rule, and, largely owing their existence and sustenance to the largesse of the king, would support him against those peers who had lost or were losing their power and position in the new social and political order.

Old social terms underwent a sharp shift in meaning, and new ones were introduced precisely at the time of this transition. For example, in the early 11th century, the term hird replaced the old term drótt. While hird isindeed a “semantic coalescence” (Lindow 1976:44) of the drótt and other re-lated concepts, it is not a traditional drótt. The term bird, as it is often usedin native writings, indicates a large and structured household of an institu-tionalized nature (Lindow 1976:45). The history of the word itself may shed some light on this. It is an Old English loan word, that came into use in Scan-dinavia while the Danes controlled large parts of England between the 850s and 1040s. In England, in the 6th century, hird means only “household”—anyhousehold, a peasant’s or a king’s. By about 800 it is seen in English use as ex-clusively meaning the king’s household. By the 11th century it can be trans-lated as “court.”

This foreign concept, picked up in a conquered land, may have been readily adopted into Scandinavia by kings eager to alter the existing order. Sven Forkbeard (who ruled A.D. 980 to 1014) and his son Knut the Great (who ruled A.D. 1014 to 1035) were undisputed rulers of England, and both observed and amended the law of that land (Larson 1969; Lawson 1993). The sudden appearance of radically altered terms of social relations, reflect-ing the superior sovereignty that kings were accorded in England, could hardly be an accident.

In early usage, hird was used interchangeably with the word drótt.Later, it clearly described a structure of “officials,” “retainers,” and “servants and guards.” The word drótt fell completely out of usage. There was a hirdlaw, that regulated the precise manner in which hird members must be-have: if a man of higher rank in the hird came into the feasting hall late fordinner, there was a law that stated that his inferiors must make an ap-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (142)


propriate place for him at the table, above and ahead of themselves. If two hirdmen quarreled or fought, clear rules for dispute resolution werelaid out in this law. In Olaf‘s Saga the hird of the king is described in its de-tailed protocol: 60 hirdmen, 30 gestir, and 30 himthiki or housekarls (Lin-dow 1976:57).

New titles appeared: bryte, or “manager of royal land,” tjänsteman, “ad-ministrative official,” and galkare, “debt collector.” The small warband had be-come a large administrative group. By the beginning of the 11th century, England was a part of Denmark’s empire, directly ruled by the same king as one realm for about half a century, which further increased the administra-tive stresses on central leadership and the need to proliferate many types of officials. It also provided ideas for the administration of the kingdom and the partitioning of people into various new classes.

But the addition of terms to the administrative repertoire is not the only change in the terms of social relations. By the early historic period, dräng, which once meant a young elite man of chiefly potential under the old order, meant only a “young man.” Thegn, an Old English loan-word that came into use in the early part of the Viking Age, by the mid-1lth century is translated as a “freeman, a farmer, an older man.” In England the equiva-lent thane retained the original meaning of aristocratic lord until Norman French terms replaced these older titles. One need only remember the Pic-tish King Maelbætha (Macbeth), who ruled in mid-11th-century Scotland, and all his thanes: Thane of Cawdor, Thane of Mercia, Thane of Northum-berland, and so on.

It has been argued by some that the terms dräng and thegn always sig-nified boy and man, and never had any other meaning. Aside from the Engish evidence, their use in numerous runic commemorative stones in specific ref-erence to battles and military hierarchy argues overwhelmingly against this (Christopherson 1982). These very brief and important inscriptions, which established the rights and rule of those lords and magnates and their heirs (Randsborg 1980), were not erected to convey information about the social categories of peasant boys and men. In Denmark, their quite rapid devolution from terms laden with social and political meaning is significant: by the 12th century the words were still used but their meaning was lost or trivialized by the users.

This linguistic change parallels other signs of the changing power of kings, who shifted from being leaders with obligations to their followers, to more powerful, autocratic, network-style figures. The new institution was fully in place by the time of Knut the Great, ca. A.D. 1018. This created a deep organizational and decision-making hierarchy of many levels. With the usually cooperative Church, whose highest functionaries were often cousins to the king during most of the 11th and 12th centuries, this hierarchy became even

Landscapes of Power, Landscapes of Conflict - [PDF Document] (143)


more complex (Figure 5.2). Although it is fairly clear that this represents so-cietal change, I further suggest that these terms were purposely undermined and “de-sanctified” by Viking Age rulers in order to reduce the power and prestige of an unruly old nobility.

The most significant outcome of this situation may be that if this was ac-complished through the manipulation of meaning and political ideology rather than by force, as appears to be the case, at least two sets of elites

Figure 5.2. Early vs. Late Viking Age organizational hierarchy, including some terms of social relations.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (144)


should remain: old and probably disgruntled local lords, and new centralized magnates. Conflict, both personal and administrative, may have occurred be-tween old and new social classes. A new social class, created to facilitate and support the fairly recent, sovereign powers of the kings, may have received the direct benefits of new forms of organization, such as land, money, and en-titlements to local surplus and labor. In fact, the early historic record men-tions a number of such “new” and “foreign” officials, who will be discussed later. These new elites, sent for example into Scania to direct trade, taxation, and other central concerns may have met a group who were already per-forming these administrative tasks.

These displaced Germanic-style local lords were part of an older corpo-rate form of government, perceiving themselves and the elected overlord or king as a group of equal top-ranking elites. They were in no way benefiting from the changing organization, losing both power and prestige. Even if or-dered to relinquish their authority to new officials, old elites would have re-sisted giving up their power for as long as possible. This is not at all unlike what happened after the Norman Conquest, when old Anglo-Saxon lords were suppressed, ousted from any important governing roles, forbidden to speak their language in any official capacity, and forced to speak French. Their ranks soon dwindled as their lands were seized on any excuse and given to Normans. Within only a few generations they were gone, although only some were executed or banished outright. On a similar stage, the state-formationprocess of Denmark occurs. The consequences will be seen in chapter 7.

This situation has been described (Blanton et al. 1996) as a conflict be-tween the corporate form of power, in which power is shared by various groups within a society (in this case the assembly, the warband, the elders, and the lawspeakers) and the exclusionary form of power in which one group monopolizes the sources of power. Although both operate together in many societies and are part of the “fundamental tension that underlies all social formations,” Blanton et al. argue that one or the other form is usually in control. This period in south Scandinavia marks the critical point when one of these forms gave way to the other. This occurred over a long period of time, from the late RIA, when elite first begin to negotiate away the cor-porate power “controlled within the limits set by the prevailing corporate cognitive code” (Blanton et al. 1996:2) through the Viking Age and Early Middle Ages when it was forcibly wrested away. This nearly 1,000-year-longtimespan represents a period of heightened contradiction between these two forms.

Aside from supporting the king through their ability to bring misfortune and death to his enemies, there were other tasks for this hypothetical new elite. I previously described the emergence of town-like central places, Trelle-borg fortresses, minting places, and market sites in Jutland, Scania, and the

Landscapes of Power, Landscapes of Conflict - [PDF Document] (145)


Islands, which would have created the need for specialists in administration of trade, taxation, and transportation. Changes, indicative of elites with new or transitional duties and rights, are seen in the terms of social relations. This phenomenon can also be tested in the archaeological record.


A new class in society, second-tier elites who operationalized the decisions made at the highest level, is evident in new forms of settlement with storage, surplus accumulation, and control of resources markedly different from ear-lier elite sites. These complexes are commonly known as magnate farms or settlements. The term magnate, as I use it here, is commonly used in Scandi-navian archaeological and historical literature to denote this phenomenon that originated in the late 10th century and peaked in the 11th century. This class amassed huge estates and riches not previously seen in the archaeolog-ical record, and their fortunes and families can be followed into historic times. In the following sections, I will illustrate this new form of elite site by examining a couple of typical farming villages and then comparing them with the development of two magnate complexes, Omgård (Nielsen 1980) and Vorbasse (Hvass 1980), where material wealth is far above previous elite accumulations and entirely different in nature.

An example of a typical village of the Viking Age, Trabjerg dates from about A.D. 700 to 1000. During the earlier phase of the settlement, the farms appear to be small with outbuildings exclusively of pithouse type. In the later phase, all the farms grew slightly larger and pithouses were replaced with other small outbuildings. The village remained basically the same during the 300 years it existed at this location (Jorgensen and Skov 1980). Typical house-holds in the village had 63.3% living space and 36.7% associated storage (Thurston 1990, 1996) .

Sædding (Stoumann 1980) is a somewhat larger village, and has been dated to between A.D. 800 and 1100. Covering 3 to 4 hectares, 99 long-houses were excavated, 16 storehouses, and 75 pithouses. Nineteen of the longhouses are large, and 14 of them have evidence of a few stalls. The mixed large and small, stabled and nonstabled complexes seem to indicate that there were at least two economic classes in the village, although the differences were not great. As an example, a well-preserved large complex, called the “three-winged complex,” had 58.7% living area and 41 .3% surplus storage and crafts production area. This household represents the wealthi-est social class indicated in the settlement. The “poorer” households in the settlement are similar to the complex examined at Trabjerg. Sædding, as a village, also remained similar in structure during all phases of its 300-yearViking Age life.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (146)


5.5.1 Vorbasse and Omgård: Period I

In comparison, other sites undergo drastic change. The Omgård site, in its earliest phase (A.D. 700 to 800), originated as a farm community with eleven longhouses (Nielsen 1980), none of which had a greater proportion of out-buildings or wealth, having 78% living space and 22% storage and workshop areas. In its earliest incarnation, it is comparable with Trabjerg and Sædding. It seems to have stood not only by a river but at the crossing of two prehis-toric roadways, which may have played a role in its future development.

At Vorbasse, the first phase of the Viking Age settlement dates from 700 to about 950 A.D. Two farms of similar size, which faced a road that ran through the community, were excavated. Other farms are indicated on the other side of the road, denoting a village of several households. In this early phase, the pattern is slightly different from that of Trabjerg and Sædding. The farmyards of the two fully excavated farms are close in size. The slightly larger southern farm is .5 hectares, while the northern farm is .3 hectares. The two main living quarters are identical in size—247 square meters—and the south-ern has 22 stalls, the northern 20. The southern farm, the larger of the two, has 38.1% living space to 61.8% storage. Therefore, before A.D. 950, surplus accumulation and workshops exceeded living space by about 20%.

Some developments in intermediate socioeconomic classes are now ev-ident. Not only are 20 or more head of cattle unusual for an average Viking Age farm, but the amount of space used for other outbuildings also far ex-ceeds even “rich” farmsteads like the “three-winged complex” in Sædding.

Vorbasse, even at this early time, was fundamentally different from an av-erage village. Elites beneath the king certainly always existed, but like the king himself they were previously not associated with the amassing of stores of wealth. These farms were owned by men richer than the wealthiest farmer. The local lord’s power rested on his ability to levy local forces at in-termittent times, receive tribute, and keep a small warband. Thus the indica-tion of a new type of acquisitive, accumulating socioeconomic group developing at Vorbasse by 800 A.D. represents a break with tradition.

Between A.D. 950 and 980, socioeconomic class development under-went dramatic changes. Although the villages of Trabjerg and Sædding re-mained the same, the sites of Vorbasse and Omgård developed into something different: distinctly wealthier, their village-like character was gone, and in their place, large estates were emerging.

5.5.2 Vorbasse and Omgåld: Period II Vorbasse

By the late Viking Age—the late 10th to early 11th century—the site of Vor-basse “. . . was extended west by an officially laid out settlement area containing

Landscapes of Power, Landscapes of Conflict - [PDF Document] (147)


three farm complexes, of which the southernmost one was very large and must have been the property of a wealthy man” (Stoumann 1980:169). The rest of the village had disappeared, and the southern farm now covered 2.5 hectares, as big as a small village. Two smaller farms were directly to the north. Each of the three farms had a Trelleborg-type house as its main living quarters, identi-cal to those unusual, elite structures found at the fortress of Aggersborg, dating to around A.D. 980 (Stoumann 1980:168–9). At Aggersborg and the other Trelleborg sites, such houses were connected with elite administrators directly associated with the king.

The large farm had stall space for 95 to 100 large animals—that is, cattle, five workshops or storage barns, and two storehouses. There was 14.6% liv-ing and 85.3% storage and workshop areas. In one of the buildings, a bronze-working area is evident. A heating plate, scraps of bronze, bronze filings, and crucibles were found in a trash pit. A series of tiny bronze cutouts of horses with little holes for rivets was found: apparently they were to decorate a small coffer or similar item. In another workshop house, iron bars were found. In all three farms, fine ceramics were recovered.

The two smaller farms seem to have been about equal in size and the use of space similar to the larger southern farm—27% and 14.8% living to 73% and 85.2% storage and workshop. Omgård

At Omgård, the entire area was now one estate that covered at least 7 hectares, almost as large as the largest of villages in Jutland, nearly as large as the town of Ribe. It was completely fortified with a high bank and ditch, topped with a wooden palisade, whose subterranean portion is preserved in the earth (Stoumann 1980:198). It may have guarded the ford over the stream, where a watermill was also located. Mills in this period were associated with elite ownership (Thun 1982). They were used as a form of taxation where those with the royal entitlement to mill kept a portion of the flour as revenue.

A guardhouse and barracks stood near the entrance to the compound, and the construction of the guardhouse/gatehouse is identical in plan with “control post” guardhouses outside the entranceways at the fortresses of Trelleborg (Denmark) and Fyrkat (Stoumann 1980:193). Trelleborg houses are found inside the estate itself. Stoumann (1980:194) states that since the Trelleborg houses are too identical in every detail to those found at the fortresses of Trelleborg (Denmark) and Fyrkat to be coincidental, it is proba-ble that there may have been “master builders,” who worked for the highest social classes, constructing dwellings deemed suitable to their station. If these estates were gifts from the king to favorite new lords, he may have sent his own builders out to establish these elite outposts in rural areas when they were not carrying out royal projects like building ringforts.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (148)


A wagon barn was identified by pieces of axles, wheels, and other ve-hicular spare parts. Three large halls, set together in a U-shaped formation, form a court in the north. One is a dwelling place, while the second is a store-house. The third when excavated proved to be a combined smithy, weaponry, and saddle and bridle workshop. Close to this area is a detached kitchen. In the south are seven outbuildings that contain more smithies, and various crafts workshops. Detached cow sheds contained stalls for 80 ani-mals, and another shed, identified as a horse barn by the narrower width of its stalls, had accommodations for 30 horses.

Living space formed 15% and non-living 85% of the farm, a percentage very similar to those found at the magnate farms of Vorbasse in the same pe-riod. This would indicate either that these settlements were planned or that the entitlements of such magnates were very similar. Figure 5.3 shows the marked differences between the magnate sites and rural villages.

From what is understood of kingship at this time, these do not represent royal estates. There are some indications that lands belonging to the king had

Figure 5.3. Settlement comparison, late period. Magnate sites show unusually high areas of nonliving space.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (149)


a different nature. Although the king was almost certainly a large landowner, the estates he controlled were probably very much like (or even identical to) the Kongelevs or Husabys (meaning, respectively, “king’s inheritance” and “household village”) described in written records from the late 1lth century, or the end of the Viking Age. These were villages and farms scattered through-out the kingdom that owed their produce and products to the king. Aside from being agricultural, these villages usually possessed some craft special-ization on a scale greater than regular villages. The king was entitled to these products, such as surplus grain, cloth, or ironwork, a good example of this being the fully excavated site of Gårdstånga in Scania (Söderberg 1995). Al-though they were geared toward the production of surplus goods, they were villages in character, distinguished sometimes by their names, and the high level of production indicated there. Gårdstånga has several runestones, one of which is an after-Jelling type that uses the words fæller and dräng, twoelite terms of social relations, so it is not improbable that these members of the royal household were in charge of running the farms for the king.

The site of Lejre, near Roskilde, is probably the site where the king resided when he was not moving around. Two longhouses of enormous size have been excavated: in excess of 50 meters long and capable of accommo-dating a large retinue (Christensen 1987a:62, 1987b:251; Madsen 1987:45). The site does not appear to have been able to produce food or goods to feed several hundred hirdmen, servants, advisers, and family members and was probably provisioned from outside holdings: Kongelevs. No accumulationsuch as that seen at Omgård and Vorbasse exists there.


In the very beginning of the study period, evidence points to the leveling mechanisms of the Germanic social code, such as the decentralized and weak power of kings, the power of the assembly and the peer lords. Historic records relating to the assassinations of Arminius and Godfred, as well as the overthrow of Harald Blåtand by his son Sven Forkbeard, discussed in chapter 4, illustrate the perils of attempting a strong central authority.

During the height of theViking period, there were a number of changes: linguistic change signaling transformation of the warband into the royal household, the foundation of ascribed-status dynasties, historic records indi-cating a more unified land through their implications of territorial extent, a more powerful central leadership seen in the building of military installa-tions, the formal taxation of towns and markets, the minting of coins, and a new elite social class that supported these regimes. Yet there are also factors

Landscapes of Power, Landscapes of Conflict - [PDF Document] (150)


that seem to indicate internal disorganization: frequent infighting among elites, records that speak of several kings ruling at the same time, and refer-ences to North-Danes, South-Danes, Isle-Danes, Hallanders, and Scanians. The earliest historic record may shed some light on the power of kings and the unity of Denmark.

5.6.1 Royal Power and Unification in Viking Age Denmark

Knytlinga Saga is a narrative in the truest sense; it was written to glorify the kings of Denmark. It was written around A.D. 1250, probably by the Icelander Ólafur Thordarson, who was a lawspeaker for two terms of office, a teacher, author of books on Latin rhetoric, and a poet (Pálsson and Edwards 1986:18). His understanding of law and his descriptions of assemblies are therefore more or less trustworthy. Knytlinga Saga, while almost certainly meant to flat-ter and legitimize the rule of the Danish kings, contains important references to challenges of royal authority and bears grudging witness to the power of the assembly.

Examples of this include an episode mentioned earlier, when in 1076 Sven Estridsen, after ruling for 29 years, requested that he be allowed to go against the ancient tradition and choose his own heir, hoping to prevent his numerous sons (14 of them) from fighting over the throne. He passed over the oldest, Harald, and named Knut. After his death, the lords elected Harald. Harald lived to rule only four years. In A.D. 1080, Knut was elected king at Vi-borg by general consent.

Under Harald, Denmark supposedly became lawless: “He took the lead so little in government that everyone in the land could do almost as he liked” (Pálsson and Edwards 1986:54). Because of this, the Danes called him Harald the Whetstone—they sharpened their weapons against him. Knut, either mo-tivated to restore law or merely eager to further centralize and enhance the power of his office, made demands for new royal entitlements, increased mil-itary service and taxes, and made harsh laws against pillaging, raiding, and piracy which did not sit well with many. Knut also clashed with farmers over taxes and landuse rights. The outcome was a resurgence of the corporate po-litical tradition. In a violent uprising in 1086, fought by common land-owningmen and led by local lords, Knut and many of his followers were chased across Denmark by a frenzied army of farmers. The king and his retinue were hacked to bits in St. Laurentius Church in Odense town. His brother Olaf was made king in his place.

Whether or not Knut was a good or bad king, it is remarkable that as late as 1086, the free farmers of Denmark felt entitled or even obligated to assas-sinate him. Even later, another of Sven Estridsson’s sons, Niklas the Good, who ruled from 1104 to 1134, was assassinated by a mob of common people

Landscapes of Power, Landscapes of Conflict - [PDF Document] (151)


in Slesvik, because his son had killed a popular lord of the region. This lord, a Jarl (and the king’s nephew), was killed because he was so popular that the Crown prince feared that when his father died this man would be chosen king instead of himself. That this was a valid probability in 1130 is a powerful indicator of the strength of the old laws. His successor, Erik Emune (The Memorable), brother of that murdered lord who inspired such insecurity in the Crown, was elected king. He quickly proved to be “brutal and reckless” and was murdered by his dissatisfied subjects in 1137 when he attended a ting in Ribe (Oakley 1972:51). Finally, in the 1180s, warfare broke out be-tween the West Danes and the Scanians, recounted in primary documents such as Saxo Grammaticus’s history of the Denmark. Issues at the root of re-bellion were taxation, landrights, tithes, and military service. During this episode, known as the Scanian Uprising, the king managed to keep his head, and the farmers finally lost.


A fairly complete impression of Roman Iron Age corporate social order can be extracted from the Roman author, Tacitus. In the next centuries, substan-tial shifts suggest an attempt to overthrow this social order, which inhibited strong central authority. Viking Age rulers appear to have created a new no-bility upon whom they could rely for support rather than challenges. Thus can be inferred from several lines of evidence: the terms of social relations, new forms of elite settlement, and historical accounts of transitions in the re-lationship between rulers and their followers. The very earliest ethnohisto-ries, just after the Viking Age, indicate severe conflicts between local and central elite, especially in Scania and the other eastern province, Halland. The ever-increasing depth of organizational hierarchy brought inherent conflicts between the local and the central, which may be monitored through the study of decision-making hierarchies and the implications of change in their internal organization. The next chapter presents the locational analysis of the Danish polity on the scale of the state as a whole, then between east and west, and finally in the internal organization in Scania. The many indicators of the concerted elite efforts to integrate the eastern province, and resistance to these attempts, is supported by the geographic analysis.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (152)

Chapter 6

Landscapes of Power and Landscapes of Conflict


Thus far, I have argued that much of the Iron Age was characterized by a Ger-manic social code based on corporate decision making and a limited local lordship elected by an assembly. This system was the primary political and ideological framework in south Scandinavia. It was at least as old as the earli-est documentation in the first century, and probably extended back into the pre-Roman Iron Age, which began about 500 B.C.

At some point in the Early Roman era, leaders began to free themselves of kin obligations and form political hierarchies of a military nature. When this was achieved, they moved beyond and attempted to overthrow the code that limited their power and the ability to centralize most major social and political institutions. For several centuries, rulers struggled against this sys-tem and eventually made some headway in suppressing their former peers by creating new lords and disenfranchising the old.

In Denmark, the area immediately surrounding the “homeland” of the rulers became manageable fairly quickly. The greatest obstacle appears to have been the eastern regions, where the folc had a strong sense of their own autonomy, their difference, and the “otherness” of those who ruled them. Dis-tance from the center presented one problem in the administration and con-trol of these eastern areas, but ethnicity seems to have been an equally significant complication.

In terms of western Denmark’s political relationship with its eastern provinces of Halland and Scania, for the first centuries of the Late Germanic-


Landscapes of Power, Landscapes of Conflict - [PDF Document] (153)


Early Viking Age, Denmark functioned as a hegemonic state, a structure where, as Feinman (1998) describes, “peripheral elite have much greater autonomy, and the mechanisms for resource extraction are less formal and standardized.” Feinman continues to describe the type of control found in such states, in that they attempt to integrate the regions that they acquire (and their dissenting elite) with strategies like “exchange and inter-marriage, military threat, frontier garrisons, occasional campaigns [that] involve lower costs for core elite than the construction and supply of a full-fledged governmental infrastructure” (Feinman 2000).

So far we have looked at historic texts and cultural landscapes as indi-cators of change in Danish society. In this chapter, I will address changes ob-served in spatial and locational relationships that are proxy indicators of changes in internal organization. These more mechanistic indictors can be contrasted with the humanistic approach used in other parts of the study. In this light, another way of understanding the development of states is to view increasing complexity as the only difference between pre-state and state so-cieties (Flannery 1972). A political system is an information-gathering and processing bureaucracy, or decision-making hierarchy. The larger and more complex the system, the more parts and levels must be articulated. Only so many parts can be articulated within the shallower power-structure associ-ated with chiefdoms; further complexity must be organized in more levels. This approach, which uses a study of the levels of hierarchy associated with various forms of political organization, was pioneered and used by Wright and Johnson (Johnson 1973; Wright and Johnson 1975). According to this theory, states are:

societies which are primarily regulated through a differentiated and internally spe-cialized decision making organization which is structured in minimally three hier-archical levels [above the average person], with institutionalized provision for theoperation and maintenance of this organization and implementation of its deci-sions. (Johnson 1973:2)

Chiefdoms, in this paradigm, have less than three levels of administrative hierarchy. Administration per se is not necessarily visible archaeologically, but as political systems and economic networks become more hierarchic, a hier-archy of places where these activities are carried out becomes clear (though the actual individual administrators may not be residing physically at these sites). This hierarchy of central places will serve as proxy indicator for internal administrative change. If a whole region is synchronously moving or being moved toward the goal of centralization with little resistance, we would ex-pect that these trends of hierarchization would be more or less uniform across the landscape. If there is dissent, these trends will not be so uniform.

The obstacle to unification was a tradition of equality among peer elites. If each faction within an alliance is of relatively equal power and status, or if

Landscapes of Power, Landscapes of Conflict - [PDF Document] (154)


only slight advantages favor one or another, it would prove difficult for one overlord to make his peers into his subordinates. As discussed in chapter 1, hegemony rather than direct domination was the strategy through which one faction may have seized authority and attempted to hold and expand it. When social relations and political aspirations are so contradictory, transition is chaotic, violent, long, and drawn out. Not surprisingly, the transition re-flected in spatial relationships and central place hierarchies is not uniform, but varied, in Denmark.

This chapter explores the ways in which this difference between the core of the state and the periphery—the areas of stronger versus weaker cen-tral control—affected the development of the cultural landscape. With two very different levels of authority, we might expect to see different culture-geographic developments. It will become clear in the course of the chapter that the east and west did indeed undergo dramatically different courses of locational and organizational change. The “lower cost” of hegemonic rule proved to be ineffective, directly related to the dissent and resistance that central rulers found in the east, and controlling it required a fairly drastic in-tervention into the “power of place.” Instead of letting the rich provinces go their own way, central elite made the investment described by Feinman (1998), constructing and supplying a new infrastructure. This chapter will present arguments as to both when and how this was achieved.


The study uses two geographic analytic methods: the statistical evaluation of changes in cultural geography, that is, regional locational analysis, and the lab-oratory analysis of soil chemistry. These two methods are closely linked in the context of this study: locational analysis relies on the mapping of the cultural landscape, and in many parts of northern Europe, site sizes and locations are best determined through soil chemical characterization, especially for phos-phate content. In some parts of Denmark, large-scale excavations provide data on site size and distribution, especially for towns and markets, and special cases such as the four sites discussed in the preceding chapter. However, in Scania, extensive extant maps from state-sponsored phosphate testing reveal the location and size of a much greater number of sites with accuracy, and re-testing, surface survey, and toponymic data provide these sites with a general temporal assignment. Together, phosphate tests, surface survey, and place-name analyses give an unusually clear picture of settlement dynamics. The use of phosphate is thus the means to an end for the large-scale locational analysis of settlement system structure. Through the combination of these rather dif-ferent types of analyses, insight on prehistoric processes may be gained.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (155)


Locational analysis is a tremendous asset in the context of northwest Eu-ropean prehistory, where its utility is two-fold. First, it bypasses the persistent and culturally biased tyranny of the historical record, discussed in chapter 2, that has often argued against the existence of complex political organization in south Scandinavia during the earlier parts of the study period because of the lack of the supposed “state-like attributes” associated with other state-level societies. Many texts on European history still reflect the perception that northern society was “simple and savage” and that little or no change oc-curred between the “chiefs” of the early Roman era, the “barbarians” of the Viking Age, and the Early Medieval kings, who are usually perceived as weak in power and uncivilized in temperament. Studying sociopolitical and eco-nomic organization from a locational perspective allows the modeling of change free from historical value judgments.

Furthermore, for those who already expect complex social and political structures in this time and region, the locational approach enables analyses of dynamic processes rather than offering static descriptions. Large-scale re-gional analyses have proved to be an invaluable complement to site-specificexcavation (Crumley and Marquardt 1987; Stein and Wattenmaker 1989:2; Falconer and Savage 1995), and add a unique dimension and breadth of per-spective to archaeological problems.

Although regional studies of artifact and monument distribution have been undertaken for southern Scandinavia’s Iron Age (i.e., Hansen 1987; Randsborg 1980) and a number of gazetteers of towns and markets are avail-able (Clarke and Ambrosiani 1991 ; Hodges 1982a), a locational-organizationalanalysis of the prehistoric cultural landscape and its settlement systems has never been performed for any part of southern Scandinavia (with the excep-tion of Thurston 1990, 1996). The introduction of these methods adds a use-ful dimension to the study of the region, adding to the body of knowledge about changes in social and economic organization during a period of signif-icant political change in the macro-region, western and northwest Europe, where far-flung systems appear to have been interdependent in many ways.

In addition to clucidating change in Denmark and other cultures at sim-ilar junctures in organization and transition, I hope to add to the utility of the locational analysis through the use of the unusually fine-grained data at my disposal by studying increments of time of 100 and even 50 years. When in-tervals of 300, 500, or more years are used to compare internal organizational change, shifts appear acute and monolithic. But this is not how settlement, political, or administrative hierarchies usually change. We know what snap-shots of the big shifts look like. How are the incremental ones recognized?

In addition, even when smaller time increments have been used in other studies, they have been used to illustrate change in societies where or-ganizational transitions were fairly rapid and drastic, with major changes in

Landscapes of Power, Landscapes of Conflict - [PDF Document] (156)


periods of 200 years or less. The attempts at change in Denmark met a great deal of resistance, and this study should also present comparative material for those studying other federations and alliances that had difficult transi-tions into unification.

Finally, the changes in archaeological rank-size curves, discussed below, have often been viewed in an historical vacuum. This has produced interpre-tations based only on the curves, which can be rigid and singular: for exam-ple, a convex curve means either non-integration or the pooling of systems. This study will compare the locational data with the many lines of evidence thus far presented, which may give other, more refined and more varied in-terpretations. This may shed some light on other researchers’ more ahistori-cal data and may suggest trajectories that resulted in their research region’s archaeological record.


Chapter 2 discussed a variety of views on the proper role for landscape analy-sis; the following will describe the tools used for interpretation once these landscapes have been mapped. The phenomena surrounding the patterning of economic and political centers, administrative and decision-making hier-archies, and organizational systems appear to be similar cross-culturally(Johnson 1982, 1983) because they are based on scalar stress and the capac-ity of the human brain to process information efficiently. By monitoring such relatively culture-neutral features as span of control, internal integration, boundary changes, and levels of settlement/administrative hierarchy, analyses are less likely to be biased by historical traditions.

In general, archaeological models of state-formation are based on pris-tine or primary states, such as early polities in Mesopotamia, China, South Asia, and Mesoamerica that have been studied through monumental archi-tecture, political iconography, epigraphy, administrative architecture such as elite residences and storehouses, administrative waste such as sealings, state-organized production, and more easily identified political and administrative hierarchies. Locational analysis has been successfully used to illuminate as-pects of state-formation in these areas, and, inadvertently, has become asso-ciated with these particular types of social institutions. The use of locational analysis on other types of political structures, either less complex systems or differently organized states, has therefore sometimes been questioned. The vast differences between political organization in Uruk, for example, and Iron Age Denmark have led some to question the utility of location theory in contributing to explanation in the region (Randsborg 1980:77) but it will shortly be seen that such doubts are miscast.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (157)


6.3.1 A Discussion of Organizational and Decision-Making Hierarchies

There are many valid approaches to understanding the makings and work-ings of a state, and all may be used to examine its development through time. However, since many of the indicators used in typical analyses of state development are not recoverable in north European contexts, we need a working definition of the criteria used to identify a state in the context of this type of study.

A state is a complex decision-making and information-processing hierar-chy. Organizationally, it is a system that has attained three levels of adminis-trative hierarchy over the level of the “common” citizen, seen geographically as the smallest type of settlement in a given region (Falconer and Savage 1995; Feinman 1998; Johnson 1987). This is based on the observation that the scale of a society is linked to its social organization. In cross-cultural stud-ies of societies with two levels of decision-making hierarchy, there is a no-table, sharp cut-off at 2500 +/– 500 population. This appears to correspond to the size in which groups are still able to have direct communication be-tween leaders and followers through verbal communication and memory alone (Kosse 1990). Although this can be extended into higher populations by horizontal decision-making and communication strategies and change in the scale of the decision-making groups, it appears to reach its limit at 3,000 (for the theory of scalar stress; see Johnson 1982). After this, more levels of decision making are needed in order to administer the group effectively. This is perhaps why the minimal estimates for autonomous states given by Ren-frew (1982) also seem to average about 2,500 persons. The type of leadership involved does not seem to be a factor (Carroll 1982). This pattern of more lev-els of decision-making as populations grow and rule becomes more complex can be seen as much in corporate polities, such as the Teotihuacan and Anasazi systems, as it can in network-style, centralized states.

6.3.2 The Järrestad Region as an Organizational and Decision-Making Unit

Järrestad is one of the less populous Härads in Scania, and in the EarlyVikingAge appears to have had a population of over 3,000 persons, based on ac-cepted calculation of population estimates and densities, discussed below. In the Middle Ages, Denmark is estimated to have had a population of about 1 million people. If that represents growth during the Viking Age and Early Medieval period (Sawyer and Sawyer 1993:42), the population of early greater Denmark with or without Scania was at least several hundred thou-sand. It is clear that the population size of the subregions alone already ex-ceeded that which could be served by a two-level hierarchy (two levels over

Landscapes of Power, Landscapes of Conflict - [PDF Document] (158)


the average person), and if the area was unified, it would be far out of the scale of such a system.

One should be able to identify a settlement-size hierarchy—indicative of an administrative decision-making hierarchy—if it existed in the study area through the differential distribution of site sizes, central places versus sec-ondary and tertiary centers, and other places. Given an effective method of administration, decision—making theoretically does not need to be made from an urban context at all. In Denmark, such decisions were probably made from a series of royal estates-not only from small, rural places but from ever-changing localities.

Assuming that a larger site like a central place has a larger population, it is also inferred that it has a larger functional size as well, including apparatus to carry out administrative decisions. In the discussion of urbanization in chapter 4, towns are seen to contain many central-place functions: economic (minting, markets, production, taxing), political (law-giving , assembly places), and religious (monasteries, large churches and cathedrals). In terms of their status as central places, based on these and other activities, they have enormous functional size.

Although in some other regions the functions and activities of centers are still being explored, in Denmark there is less need to guess at what went on at such sites, even in the earlier periods. The wealth of excavation data from towns, market centers, and military hubs combined with the early historic record is unusually substantial in Scandinavia, and lessens the prob-lems of understanding these sites. The fact that Scandinavia’s hierarchical systems—administrative, military, economic, religious—are not articulatedduring much of the GIA and Viking Age, makes locational interpretation of place hierarchies an even more attractive method of analysis because it is in-dependent of these other hierarchic systems.


I have previously mentioned one important statistical method for monitoring change in the organization and integration of regions: rank-size analysis. Rank-size analysis is a tool for monitoring hierarchic structures, boundaries, and integration in settlement systems through the study of variation in site sizes (Falconer and Savage 1995; Johnson 1980, 1981, 1987). The rank-sizecurve charts empirical observations about settlement systems and has been used for nearly 70 years by cultural, economic, and historical geographers (Berry 1967; Carroll 1982), as well as archaeologists, anthropologists, and his-torians (D’Altroy 1992; Falconer and Savage 1995; Holl 1985; Johnson 1980,

Landscapes of Power, Landscapes of Conflict - [PDF Document] (159)


1981, 1987; Randsborg 1991 ; Skinner 1977; Stein and Wattenmaker 1989).Combined with archaeological and historical information, these curves can reveal large-scale, long-term developments.

Rank-size analyses consist of a series of graphs that characterize changes in settlement systems and examine spatial variation at each selected point in time. These first can be compared diachronically; a large area is studied for change through time. Next, this area can be teased apart into several subre-gions and examined synchronically. This permits study of internal differences in regional settlement structures that existed contemporaneously.

Such studies employ the so called rank-size rule: in a well-integratedsystem of central places and their hinterlands, the size of the largest center will be double that of the second-ranked place, triple that of the third, and so forth (Auerbach 1913). This “log normal” curve may be expressed as Rank R = 1/R of the largest settlement. The rank-size curve only charts em-pirical observations about settlement systems, and it does not prove anything.However, it suggests much, which combined with archaeological and histori-cal indicators can shed light on large-scale and long-term developments.

A rank-size curve thus plots the ranking of sites, from largest to smallest, against their sizes, using a double log to aid in interpreting the patterning (Fig-ure 6.1). It shows how settlement systems are put together: it reveals compet-ing sites, lacunae in the expected hierarchy of settlements, unusual patterns of regional networks. A variety of common patterns are observed in modern and historic nations and polities that indicate certain political and economic cir-c*mstances. When similar patterns are observed in prehistoric distributions, some inferences can be drawn about the conditions that may have caused the patterning. One of the most interesting facets of rank-size analysis, known from hundreds of studies of historically known systems in the past few decades, is that the patterns are based only on site size, in hectares or popu-lation if it can be estimated. The larger sites themselves do not have to be as-sociated with elite activities—it makes absolutely no difference if a king lives in a center or on a farm, nor does it matter if there even is a “ruler.” In other words, site-size hierarchies develop on their own and do not necessarily have other types of hierarchies embedded within them, although they may (Carroll 1982). So while they are good for studying centralized polities where several hierarchic structures have their representative apparatus in the same places, they are just as good for studying the differently organized system.

Although the rank-size rule appears to work well in prehistoric distribu-tions, because the economic and political systems of pre-industrialization era states and chiefdoms are so different, archaeologists tend not to make spe-cific inferences between modern Western data and what is revealed in ar-chaeological rank-size distributions (Falconer and Savage 1995:40). Instead of specifics, archaeologists look at deviations from the expected curve and

Landscapes of Power, Landscapes of Conflict - [PDF Document] (160)


Log normal distribution: Settlement of rank "R" = 1/R of Largest Settlement

Figure 6.1. Alternative types of rank-size curve and the formula for the log-normal variation.

make more general inferences about the integration and complexity of the systems they study.

Four key rank-size patterns have been identified by geographers and ar-chaeologists: the so-called log normal curve, the concave curve, the convex curve, and a fourth possibility, the “primo-convex” curve, which is concave at the top and convex at the lower end.

The log normal distribution emerges as a straight diagonal line with aslope of minus one—the second largest site is half the size of the largest, the third is a third as large, and so on—and usually indicates a well-connected, ef-ficiently administered complex chiefdom or state with a highly integrated economy and communication system, a well-articulated political hierarchy, and often (but not always) a long-urbanized society (Berry 1961:582). This pattern has been observed not only in states, but in well-organized chiefdoms as well (Carroll 1982).

Concavity in the curve is called a primate distribution, where one site is much larger than all others and few or no intermediate size sites exist. This curve has been associated with systems where high-level central place func-tions are exclusively carried out at a primate center with unique activities that may revolve around religion, trade, diplomacy, or war (Kowalewski

Landscapes of Power, Landscapes of Conflict - [PDF Document] (161)


1982:65). Second-level central places are missing: the cost of putting goods and services at a regional hub are too high. This often indicates “an extraor-dinary centralization of political and economic functions” (Falconer and Sav-age 1995:40). It also often indicates little investment in regional development by central authorities who administer a region, suggesting an economic strat-egy where administrators seek to extract resources cheaply with little invest-ment in local infrastructure or expansion: maximum exploitation and minimum administrative cost (Johnson 1987:108).

This pattern may also indicate a system was a part of a so-called den-dritic system, where the apparent “capital” of the region is actually a mid-sizesite in a vast system that may have had its true capital far away, such as the re-lationship between provincial centers (like London or Trier) and Rome dur-ing the Roman era (Johnson 1981:171).

In terms of the pattern of places (one huge site with few or no interme-diate centers surrounded by many more small sites than expected) a good ex-ample might be Teotihuacan’s relationship to other sites in the Classic period of the Basin of Mexico or, for a historic example, London’s incredibly primate position among English towns and villages before the Industrial Revolution. Are the patterns the same for the same reasons? Maybe, maybe not. The data must be carefully examined in each case. Because many archaeological datasets examine settlement hierarchies that are little or under-documented, all possibilities must be considered when interpreting prehistoric rank-sizedistributions.

Another deviation from the log normal rank-size curve is the convex dis-tribution, the opposite of the above circ*mstance: there are many more large sites than expected, and no one truly stands out as primary over the others. This is usually interpreted as a settlement system with less than ideal com-munication systems and poor integration of political and economic activities between various sites. It signals that there may be little vertical connection between larger and smaller places (Johnson 1980). However, there are other explanations for a convex distribution. It may simply indicate the pooling of more than one system in the analysis (Johnson 1981). Prehistoric polities were often spatially smaller than modern ones, and it is easy to assume one “region” where there are in fact several distinct systems. This represents a mistake on the part of the archaeologist.

Another possibility is that long-standing patterns of settlement may con-tinue to exist in a region even after the political and economic situation has been altered. Skinner noted this pattern in late imperial China (1977:238–9), a truly unified system but one that had developed as separate regions in prehis-tory, mainly for geographic reasons. Many of the regions when separately ana-lyzed do indeed have a log normal pattern. The former Soviet Union is another good example: a number of previously independent polities patched together

Landscapes of Power, Landscapes of Conflict - [PDF Document] (162)


in name yet basically unintegrated. Even after many decades of supposed uni-fication, their rank-size was extremely convex. Such circ*mstances make a na-tion inordinately vulnerable to breakup; its parts still function independently. As we have recently seen, despite economic hardship most of the former So-viet republics had little trouble reclaiming their independent nationhood.

In the archaeological record, this pattern may indicate recent or poor in-tegration or the traces of independent evolution before state formation. Other possible explanations are that it is an inherently open system (Kowalewski 1982:66) or that the analysis is of a marginal part of a larger sys-tem (Paynter 1983 : 152 –6).

The fourth possible curve is the primo-convex curve, which is concaveat the top and convex at the lower end. This is a special case in which one sees the

superimposition of a centralized or colonially derived system (expressed in the upper curve) on a lower-level system that may be loosely integrated (reflected in the lower curve). This possibility is particularly intriguing since it suggests the si-multaneous operation of two distinct settlement systems in a single region. (Fal-coner and Savage 1995:41)

When assessing the possibility of prehistoric colonialism, conquest, or attempt at imposing centralization on a decentralized region, this is a partic-ularly important pattern.

If settlements must be of known size to construct such a curve, how can the incomplete archaeological record provide such a database? It is unlikely that any database includes all sites once extant. There are two problems to contend with: missing high-order sites and missing low-order sites. Because rank-size graphs are products of logarithmic functions, missing high-ordersites such as large urban centers make major changes in the curve, affecting the interpretation of the most fundamental aspects of organization: inte-grated and unified or not.

On the other hand, smaller sites are the largest group in many distribu-tions; they create variations in the shape and length of the tail and help to illu-minate interesting situations like the primo-convex curve. Because the smallest sites are those most likely to be missed or longdestroyed in the archaeological record, this can be a serious problem (Falconer and Savage 1995:42). Falconer and Savage have devised a way of compensating for the possibility of incom-pleteness by applying a Monte Carlo simulation to their datasets, generating a series of simulated “original populations” of all sites. The simulation may be skewed in favor of injecting more small sites into the hypothetical rank size.

In Scania and the rest of Denmark, a combination of historical and archae-ological data makes it unlikely that the largest sites are absent from the analysis. Rather, what is missing is the complete record of smaller places. In Scania, the phosphate mapping of the entire province largely precludes this problem.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (163)


Phosphate data will be discussed in detail in later chapters, but here suf-fice it to say that any phosphate area over 5 hectares in size has been over-whelmingly demonstrated to date to the late Iron Age–Viking Age, and that the names associated with such areas can give a rough idea of the time of the establishment of the site, as they are directly adjacent to or very near the Early Medieval villages, which are merely the latest version of the prehistoric village. The phosphate patch represents the latest phase of settlement at the abandoned site or portion of the site.

In West Denmark, where there is no large-scale mapping of soil phos-phate to refer to, it is more difficult to assess the smaller places, but a late Viking Age-Early Medieval transition settlement landscape can be approxi-mated by looking at available excavation data and the early medieval villages as they existed up until the Enlightenment. While it would be an enormous task to look at village-level settlement across Denmark, Fyn and Jutland, which provide most of the West Danish sample for this study and represent the two major type of Danish areas (poor heathland versus rich clay islands), should be representative of much of the western realm.

As Johnson noted (1987:109), when comparing many rank-size curves to each other both synchronically and through time it would be helpful to have some type of numerical index with which to assess change. To this end, Johnson devised the Rank Size Index, or RSI. In this formula, “the sum of de-viation of an observed distribution from its associated ‘expected’ log-linearcurve is divided by the sum of deviation of that log linear distribution from its associated maximum convex distribution” (Johnson 1987: 109). The lognormal is calculated with the formula L x (l/R), where L is the largest site (inpopulation or hectares) and R is the rank of each site. The maximum convexcurve simply assumes that all sites are as large as the largest site. The result is that extremely convex observed distributions will have an RSI value closer to 1.0, while those approaching log-linear are closer to 0.0. Increasingly primate distributions have increasingly negative values. Finally, this RSI number does not tell us much about primo-convex curves because they are mixed, but they can be teased apart into separate upper and lower curves.

6.4.1 Explaining Regional Patterns

It is useful to look at an historic example before tackling a prehistoric one. Fig-ure 6.2 (after Johnson 1987) shows rank-size curves for the United States be-tween 1750 and 1850. The bottom part of the diagram reflects the 10 largest cities or towns in the 13 autonomous colonies. These were founded by diverse groups, reflecting British, French, and Dutch concerns, and were organized along different religious, economic, and political principles. In 1776, at the moment of political unification, this curve would have still looked the same. By 1800, the curve was still slightly convex. This reflects regionalization, and

Landscapes of Power, Landscapes of Conflict - [PDF Document] (164)


Figure 6.2. Rank-size curves of the United States A.D. 1750-1850 (after Johnson 1987).

the lingering of residual regional autonomy. By 1850, however, the curve had become primate, and the largest urban site, New York, predominates. The rank-size shift reflects this change. The Civil War occurred during this last pe-riod, ca. 1850, when the various ethnic and regional identities subsumed under the state broke the uneasy bonds of unification in an attempt to secede.

6.4.2 Complementary Statistical Analyses

In conjunction with rank-size analyses, histograms of the site distributions il-lustrate levels of site size hierarchy in the clusters of settlements and give some indication of organization. As mentioned above, middle-range, pre-state

Landscapes of Power, Landscapes of Conflict - [PDF Document] (165)


societies usually exhibit three levels (two levels of hierarchy above the pop-ulace level) while states display four or more (three or more hierarchic levels above the lowest level). The way in which levels of organizational hierarchy were configured in the RIA is somewhat unusual, in that they fluctuated ac-cording to need.

Each warlord and drótt from the Roman through Germanic Iron Age controlled their own portions of south Scandinavia in a series of chiefdoms or petty kingdoms. In times of war, one among them might be elected over-lord of a federation. Although the chiefdoms of this time were complex in terms of social and economic relationships, this amounted to a fairly simple political hierarchy in which levels of administration were fairly shallow (see Figure 5.2).

This results in a variable decision-making hierarchy. When no overlord was in power, there was a two-level administrative hierarchy, like that of a chiefdom, consisting of the druhtinaz on the top level, the lower-level elite (provincial lords) on the next level, and the folc on the bottom. However, thefolc participated in the ting, which was not merely an executive or advisory body, but an initiating arm of governance with the power of decision to make war, to elect rulers, to trade or make treaties. But when an overlord came into play during times of federation, the decision-making hierarchy became deeper: three levels above the populace, usually considered the minimum seen in a state-level society. At this juncture, it may have been that since a cri-sis was at hand (or there would be no overlord), the leader of the federation would have temporary initiating power. It would be improbable that the folcwould demand the right to make decisions during times of conflict; it is im-possible to fight a successful war without a leader making fast and decisive decisions. Thus, there is a shifting hierarchy that fluctuates between a chiefly and a state-level structure, in which power alternates between the leaders and the followers.

With fluctuating hierarchies, resistance to centralization, and jurisdic-tional conflict between old and new elite, it is difficult to tease out what kinds of levels are actually present. To avoid this confusion, histograms are a useful way of examining the clustering of sites in levels based on size, and thus functional size. This will be an important part of assessing regions like the harad divisions of the Viking Age. By examining the histogram of a con-vex structure, one can discern divisions even within what appears to be a non-in t egrat ed area.

A third way of examining change is to identify span of control and mon-itor its development. Span of control indicates the amount of control a central authority has over a region by creating an optimum hierarchic decision-mak-ing structure. It describes, from the top down, how many units are under the control of each decision maker, and conversely from the bottom up, how many units report directly to the node above them. Based on the notion of

Landscapes of Power, Landscapes of Conflict - [PDF Document] (166)


scalar stress (Johnson 1982), optimally any administrator should have no more than five to seven units to control at any given time. Simple decisions can be made easily; add in more and more information and eventually the brain re-ceives too much from too many places and errors occur in greater and greater frequencies (Johnson 1987). In terms of political systems, an administrator can handle the information on just so many units before mistakes occur. In the ab-sence of written record-keeping, as in the Viking Age, this is particularly true. In many examples seen cross-culturally and cross-temporally, the amount of administrative units that can be well-coordinated by one person appears uni-form—about six to seven units at most. Because this represents the biologi-cal-neurological limitations of the human brain rather than culturally particular customs or concepts, cultures with widely differing institutions and organizational forms may be compared (Johnson 1982, 1983).

Very wide span of control indicates little central authority, whereas very tight span indicates strong control. Johnson (1987) notes that although the normal span of control is optimal with one administrative center per each six units of administration, deviations in the pattern may suggest interpretations based on culture-specific historical variations.

In Johnson’s study of the Susiana Plain, and in his discussion of Skinner’s Chinese example, in border regions where foreign elements and proximity to enemies exist, the span may be much narrower, as a stricter control is needed. Conversely, interior agricultural areas often display a much wider span of control. Two factors contribute to this: first, such regions tend to be stable, as farmers are conservative and not usually the element of society as-sociated with upheaval except in unusual circ*mstances. Yet, in applying this to Iron Age Denmark, it should not be forgotten that the Danes had few “pro-fessional soldiers.’’ That is, the farmers were the Vikings, depending on the time of year you caught them. Thus, the traditional idea of the typically pas-sive farmer may not be applicable here. Secondly, the administration of such areas is far less complex than other regions inasmuch as the primary function of administration is to extract taxes, either in products or in coin, or both, de-pending on the system. With fewer things to administer, a single administra-tor can monitor and control many more units.

In China (Skinner 1977), span of control could fall to 2 to 3 sites per controlling unit in border regions, while in agricultural areas is could rise to 18 or more. By looking at archaeological settlement patterns and the di-achronic and synchronic relationships between central places, secondary central places, tertiary central places, villages, hamlets, and single farms in a region, this information on hierarchic decision-making structure can often fa-cilitate useful reconstructions of prehistoric organization.

By the first half of the Viking period, there were probably kings, jarls, and thegns above the populace. By the late Viking era, there appear to be kings, jarls, thegns, the royal officials and managers, local town authorities—many

Landscapes of Power, Landscapes of Conflict - [PDF Document] (167)


strata between the king and the common people. To this is added a second hi-erarchy that usually had a mutually supportive relationship with the Crown: the archbishop, the bishop, the priests at the village level, and also abbots and monks who had certain authority over local people. Most of the high-rankingprelates, such as the Archbishop Absalon in the early 12th century, were sim-ply the noble sons of elite families. They kept armies and treasuries, and fought with swords in battle. When King Niklas the Good fought a losing bat-tle before the end of his reign, no less than five sword- and axe-wielding Dan-ish bishops fell and died in the battle (Oakley 1972:51).

It is clear that the changes in social relations and the growing complex-ity of administration created a much deeper hierarchy of government. How-ever, the old local elite were still a force to be reckoned with. Figure 5.2 illustrates how they may hypothetically have fit in, or rather, not fit in, with this changing hierarchy. At the second and third levels beneath the king, they would have been in direct jurisdictional conflict with new central elite during their development between A.D. 900 and 1000 and thereafter, until the situation was resolved with the bloodshed of the last uprisings. In Sca-nia, and probably Halland, these relationships were at their most strained and contentious.

Finally, it is clear that these models are just that: models, fairly rigid, pre-dictive, systems-oriented. They do not reflect settlement in the real world, which even in its present-day observable reality is difficult to understand. What these models allow us to do is to hang the somewhat tenuous archaeo-logical data on their proverbial framework and present to us the opportunity to contrast the hypothetical with the real. How well the data match or don’t match some variation of these patterns can give another piece of evidence to combine with material culture, social relations, protohistory, and all the other lines of evidence that the study has thus far followed.


In the following sections, the rank-size rule, levels of hierarchy and span of control will be used to interpret changing locational patterning. Denmark as a whole will be examined, and then two of its major sub-regions viewed separately—Scania in the east and Jutland and the Isles in the west—span-ning the period when the regions integrated into the Danish state. It will shortly be seen that the east and west underwent dramatically different se-quences of change and integration during this period. The locational data sets will be presented and discussed within the sociopolitical and economic framework to which I believe they are directly related.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (168)


The study of the longue durée—the long-term regional change and dy-namics of south Scandinavia—illuminates the transformation of loosely fed-erated groups into a more cohesive whole, the effect of centralized rulership on the configuration of the cultural landscape, shifts in system integration and boundaries during unification, and the changes in economic and social institutions as a response to changing organizational needs.

6.5.1 Regional Locational Analysis of Greater Denmark: Synchronic and Diachronic Variability

The log normal rank-size curve that would illustrate an integrated system has been described. What would we expect to see if this were not the case? The view that chiefs or petty kings ruled their own territories throughout Den-mark, perhaps forming a confederacy for warfare or plunder as in earlier times, could be supported by a convex pattern in the rank-size graph, indi-cating the lumping together of several distinct systems, which when sepa-rated would be discernible. We know from Jordanes that in the sixth century many separate “peoples” inhabited south Scandinavia. By A.D. 890 or so, Othere and Wulfstan told a scribe at King Alfred’s court that all of Jutland, the isles, and Scania were “Denmark.” I suggested that by the early ninth century, a political union, in name only and not in any true integrative sense, may have preceded real unification. By real unification, I refer to integration of an ef-fective nature that involves the restructuring of local economic activity for the benefit of the center, the reorganizing of communications to include a much larger network than previously, and local control by central authorities. Through an analysis of new data, and a re-analysis of published material, I will try to illuminate the transition between one type of organization and another.

6.5.2 The State of the State I: Greater Denmark Through Time

The period under study can be divided into four rough phases: Late Germanic Iron Age/early Viking Age, from A.D. 700 to 850; the Middle Viking period from A.D. 850 to 950; the Late Viking Age from A.D. 950 to 1075; and the Early Me-dieval period from 1075 to 1200. These divisions are not arbitrary but are cho-sen because they represent periods of marked change. The earliest time-periodis characterized by the appearance of the first towns in Denmark, as well as the first recorded conflict between Denmark and her neighbors, indicating signifi- cant changes in socioeconomic and political spheres. The final period begins at the death of Sven Estridsson, the last Viking king, whose reign spanned the transition from Viking Age to Medieval and is determined by the fact that Den-mark was developing from tenuous protohistory into an historically docu-mented, Christian, urban kingdom, integrated into the northwestern European

Landscapes of Power, Landscapes of Conflict - [PDF Document] (169)


community. The central periods are where change from one form of organiza-tion to another may be expected to be keenly represented: the founding of new primary and secondary central places, the emergence of new social struc-tures, and the continuing processes of centralization in the economic and po-litical sectors. The Late Iron Age/Early Middle Ages will therefore be examined at four points in time: analyses will be presented for conditions at about A.D.

850, 950, 1050, and 1200. In addition, a special stop-over on the time contin-uum of unification will be made at the period around A.D. 1000. This is the time when the hegemonic state had exhausted its options for integrating Scania with little direct intervention and had established some economic centraliza-tion, yet before the really “costly” modifications of infrastructure were under-taken. This will illustrate just how drastic the change was that marked the difference between this period and the next. Period I: A.D. 700-850

Historical records indicate that by the first decade of the ninth century, all re-gions of south Scandinavia were acting in concert in their dealings with the Franks. It has been noted throughout this discussion how certain features, such as trade centers like Hedeby and Ribe, defensive earthworks, royal power, and apparent incipient centralization, have often been listed in an at-tempt to show a unified “Denmark” beginning in the period between the eighth and ninth century. However, this may be a reflection of the common desire to find the earliest, the oldest, or the biggest, rather than the truth. As tempting as it is to interpret this evidence as an indication of a unified state, as long as there are other explanations an independent test is required.

An alternative explanation is that there were a number of smaller, proto-or early state-like organizations dotting the south Scandinavian landscape and that some of these state-level attributes might in fact be associated with one or more of these polities, and not one overarching state. The Franks were preoccupied with the rulers from Fyn, like Godfred in the ninth century, as-suming perhaps that they ruled all regions of Denmark. It may indeed be that this conflict primed the future royal line and allowed them to co-opt the oth-ers into accepting their overlordship as a permanent institution. This is one possibility. It is also possible that, since Sillende (south Jutland) was the clos-est to the mainland and the most under threat and in conflict, we know the names of the kings wielding power there, and not elsewhere. The land of the Franks and the land of the Danes shared a border only on the neck of Jutland. If Godfred went there to fight the Franks, it might have been assumed that he was king of much more than he was. Other rulers with similar power and ca-pabilities may simply remain nameless, or perhaps are named in the Frankish Annals as the men from the various regions who negotiated peace in A.D. 811.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (170)


In an effort to produce more well-defined archaeological tests than a list of attributes, rank-size distributions were created. If Godfred (and his family) between A.D. 800 and 850 ruled all Denmark we should expect to see an in-tegrated rank-size curve. We know they ruled the southwestern realm, south Jutland (where Godfred established Hedeby) and Fyn (where he appears to have been from). If he ruled only this region, we should expect it to be simi-lar to other regions, also ruled by kings with equal power.

It is difficult to construct rank-size curves for western Denmark in the earliest period that has been suggested for a Danish state. In later periods, there is a deep hierarchy of urban sites to consider. In eighth and ninth cen-turies this is not the case. Urban sites are not necessary for a rank-size analy-sis, only a system of sites that have unequal sizes based on differential function. Hundreds of sites from this period are known in what is now modern Denmark, but because phosphate mapping is underutilized outside of Scania, and excavation or testing must be fairly complete to establish vil-lage size, there are a limited number of sites with which to construct a rank-size curve.

However, as it was noted, the largest sites form the overall pattern of the curve. In terms of the sites themselves, urbanization was just beginning, and the site of Ribe represents its first example, established in around 720 and in full flower by 850. Soon after its founding in A.D. 811, the first phase of settle-ment at Hedeby was also in place, also densely populated, and about 5 to 6 hectares in size. These are the only two “urban” sites in the period. This leaves the rest of the rank-size curve to be constructed from villages, of which few have been excavated to an extent that permits site size estimates. However, there are representative types of villages in several categories of site size.

The 10 to 12 hectare site of Ribe is perhaps the largest place, and was limited in size by its location on a dry rise amid wetlands. Excavated portions show that settlement was very dense. It was undoubtedly a specialized cen-tral place where there was a regulated market, a mint, and an early church, apparently under royal control. Giving it a typical urban density of about 40 persons per hectare, derived from the site of Hedeby by Randsborg (1980)— 24 walled hectares, conservatively estimated as having a population of about 1,000 people—perhaps Ribe had a population of 400 to 500, although this extremely conservative.

The second-ranked place might be Hedeby in the first half of the ninth cen-tury, also densely populated and about 6 hectares in size. Although the site size is close to the size of larger villages, the population was certainly higher than the spread-out farming communities of larger size. Using about 40 persons per hectare in an urban setting, perhaps the population was about 250 persons.

The third would be of villages of 10 farms or more, such as Omgård in the seventh to ninth centuries with 11 farms on about 6 hectares. Using a

Landscapes of Power, Landscapes of Conflict - [PDF Document] (171)


rural standard of 20 people per hectare—that is, 10 people per farm—a nu-clear family and their servants or thralls (Olsson 1991a), these villages might have had populations of about 90 to 100.

The fourth type of villages were of five to eight farms, with 50 to 80 peo-ple, such as the seven to eight farms covering about 3 hectares at Nørre Snede from 700 to 800 (Hansen 1988:195, 198; Hvass 1989:98), Staghøj with eight farms dating between 700 and 800 (no size reported) (Siemen 1988), Vorbasse 6 and 7a, dating to the 7th-8th century with seven farms at about 4 hectares each. The villages of Trabjerg and Sædding have phases dating back to the eighth and ninth centuries. It has not been possible to sort out the number of farms in each phase at these villages but they are respectively 2.5 and 3 hectares. This would seem to put them in the range of seven to eight farms as well. Finally, villages of two to three farms represent the smallest units of residence. There are larger and smaller central places, and larger and smaller villages, possibly a four-level hierarchy, but not certainly.

Using a sample of known, published sites, the rank-size curve is markedly convex, with an RSI of .277, indicating a lesser degree of integra-tion (Figure 6.3a). This may be due to the fact that the densely populated town of Ribe was limited to 10 to 12 hectares due to the surrounding wet-land, whereas some of the larger villages were quite extensive. The crowding and lack of room within the boundary wall can be seen in the fact that the outer area near the wall, used for burials in the 700s, was overrun by habita-tion within only a few decades. I have reconstructed population very conser-vatively, using the 40 persons per hectare; however, it is entirely possible according to Scandinavian archaeologists that this site may have had double this population or more (Söderberg, personal communication, 1996). Be-cause it is not possible to be sure I have used the least speculative calcula-tion. At the end of the Viking Age, the town leapt over the marsh and spread onto adjoining high ground, and can be more easily estimated. In this case, because of these site-specific factors, the population curve is a more reliable indicator. If Ribe is acting as a central place for all these settlements (Hedeby was far less important at this time), the span of control is very wide, indicat-ing loose control or possibly no direct central control.

Historical documents and monumental building projects indicate that by the early 800s there was some form of centralized control present in south-ern Jutland that had the ability to administer trade, taxation, and labor. Though the population curve is convex, it does not indicate competing cen-tral places. The top part of the curve is straight, showing that based on pop-ulation there is hierarchic separation between sites. If this were not the case, we would see hyper-convexity: two or more sites of nearly the same size widening out the top of the curve. But the two central places are not the same size: there is no evidence in the locational analysis that South Jut-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (172)


Figure 6.3. Synchronic variation in rank-size curves and corresponding histograms, ca. A.D. 850.

land/Fyn itself was divided into petty kingdoms or chiefdoms with compet-ing centers. This then represents the probable realm of Godfred and ilk. How-ever, if we look at all of “Denmark,” we do get extreme hyper convexity (Figure 6.3c), with an RSI of over .5.

This is mainly because, as might be predicted from the evidence pre-sented in earlier chapters, the rank-size distribution for sites in Scania alone is extremely convex (Figure 6.3b). Based on accepted interpretations of such curves presented above, this may be interpreted in several ways: Scania itself was internally unintegrated, Scania was not integrated with western Den-mark, or both, or the pattern was caused by historical factors relating to ear-lier settlement patterns before unification.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (173)


As we have seen and considered in chapters 4 and 5, in contemporary records from the ninth century Denmark is referred to as a Kingdom that was comprised of Jutland, the Isles, Halland and Scania, and perhaps even parts of Norway. It seems reasonable, therefore, to accept that by some point in the ninth century Scania was politically a part of Denmark. However, politics do not always reflect real social and economic conditions, and the rank-sizeindex belies true integration.

It should be noted that the difference in the nature of settlement and agri-culture between Jutland and Scania, discussed in chapter 3, is clearly illus-trated by the disparity in size between the royal town centers in Jutland and the large villages of Scania. Ribe, and even Hedeby, were 15 and 24 hectares in size at their largest extent. Villages in Scania range from small, ca. 5 to 6 hectares, to large, with the largest in the 40 to 50 hectare range. The villages of Jutland throughout prehistory are far smaller than those of the richer is-lands and Scania; the trend toward husbandry resulted in a more sparse set-tlement pattern. Furthermore, the poor soil was associated with a strategy of swiddening and most of the villages shifted location every few generations. On the contrary, over 50% of Scanian villages never moved from the time they were founded between A.D. 400 and 500 and the Medieval era, or over 800 years. Those that did move moved only 250 meters or less in most cases. Thus, it is not unusual for the oldest and largest villages to have 30 to 40 farms or more, while Jutland’s villages usually contained about 7 to 10 farms.

These large productive villages in Scania and the Islands were the source of surplus for the rulers who controlled Jutland’s towns, and who moved their base of power to Jutland beginning with the Jelling Dynasty in the 900s. The east-west disparity between site and population size is a reflection of the dis-parity in production, and of course is one of the main reasons why the West pursued the incorporation of the East. One can imagine that the two areas spent much effort negotiating the access of the East Danes to trade goods and imports produced and received in the western towns on Jutland, and the simultaneous West Danish demand for access to the agricultural surplus of Scania.

Following Johnson (1981:167–9), a politically unified area may contain regions with different patterns of settlement organization, depending on how recently they were annexed and what their previous organization was. The distribution for Jutland is convex yet displays no truly competing ten- ters, but the distribution of sites in Scania, in both the histograms and the rank-size graphs, distinctly look like a number of middle-range, mostly three-level hierarchies or prestate systems with competing centers. When mapped back onto the cultural landscape, these large sites each appear to be the largest site in separate, regionalized settlement clusters. Lumped together, they cause the convexity in the rank-size curve (Figure 6.4). This does not mean that Scania was not part of Denmark at this time. Skinner (1977:242)

Landscapes of Power, Landscapes of Conflict - [PDF Document] (174)


Figure 6.4. Settlement regions in Scania with Theissen polygons. a. Harjagers Härad, b. Torna Harad, c.Bara Härad, d.Herrestads Härad, e.Ingelstads Harad, f.Järrestads Härad.

described what he termed residual regionalization in China, where previ-ous, pre-Imperial forms of organization with established and stable central places remained in place, causing convexity in the overall system that is still seen today after thousands of years. Similarly, in Scania one may not neces-sarily see changes in settlement patterns unless there is a political, economic, or military motivation for physical change in the cultural landscape.

Most of the rank-size curves of the six studied Scanian regions have vary-ing degrees of convexity (Figures 6.5). Harjagers, Torna, and Järrestads Häradsare very convex, at RSI .406, .464, and .539. Herrestads Härad at .229 is far less convex. Bara Härad is slightly primo-convex, and far closer to log linear at .169, followed by Ingelstads Härad, which while it has a higher RSI of 373, isalso actually markedly primo-convex. These primo-convex systems indicate that a new type of organization may have been superimposed over the oldest settlement pattern: perhaps an attempt by local elite to centralize. Bara Häradis especially noteworthy. As mentioned in a previous chapter, the large site of Uppåkra appears to be a chiefly center of some magnitude. It appears to be a more highly integrated system than West Denmark during the same time pe-riod, and perhaps Ingelstad is as well. Herrestads Härad appears to most

Landscapes of Power, Landscapes of Conflict - [PDF Document] (175)


Figure 6.5. Synchronic variation in Scanian rank-size curves and corresponding histograms. a. Harjagers Harad, b. Torna Harad, c. Bara Härad, d. Herrestads Härad, e. Ingelstads Härad,f. Järrestads Hårad.

closely mirror the Jutland-Fyn region, comparable with the west-Danish pat-tern seen in Figure 6.3a.

In examining the histograms, however, Harjagers, Torna, Herrestads, and Järrestads Härads, which are all more convex, or similar to the curve of the

Landscapes of Power, Landscapes of Conflict - [PDF Document] (176)


west, there is little clustering in terms of levels of hierarchy. In Harjagers, the top of the histogram is flat; in Torna, three sites of very similar size form the top level of what looks like a two-level hierarchy. In Herrestad, two sites are in the same position; Järrestad has the least separation of all. If theJutland/Fyn histograms in Figure 6.3a are re-examined, it may be noted that there is a large separation in the graph for population, which makes this re-gion more like Bara and Ingelstads Härads. Population is cited because it is more accurate than simple site size, and if it can be estimated it is preferable.

Note that I have not presented any population curves for the Scanian Härads—because they would be identical with site size. There is only one type of population density in Scania—the village or rural density. This brings us to the difference between the east and west that separates the west distinctly from other subregions. Part of its population is agglomerated in urban settlements.

Does this mean that in A.D. 850 the western Danish king was the nomi-nal ruler of all later Denmark? No. But it is likely that whoever ruled the larger region needed an administrative apparatus to do so. These two towns are the only such places with urban/central place functions, and they are in the West Danish territory. This appears to lend some support to the idea that a West Dane was ruling the Denmark that Othere and Wulfstan told of in the 890s.

Span of control from these sites makes it improbable that they effected any administration at all in regions as far off as Scania. The span would be one to at least a dozen just between Hedeby or Ribe and Scania’s regional large-village centers.

65.2.2 Period II: A.D. 850-950

By the end of the century, circa A.D. 900, Hedeby had become the largest ten- ter in Western Denmark, when it was a 24-hectare walled site. Based on the length of occupation and the number of graves, a population of at least 1,000 people has been postulated. For this era, most historians still interpret the ref-erences to “kings” to mean rulers with very limited geographic realms and correspondingly small extent of power, because in the primary sources, the continuation of earlier conditions seems further upheld. As noted previously, the discussion of the Danish conquest of England in the late ninth century in the indigenous Roskilde Chronicle states that there were sometimes “five kings in Denmark, but that sometimes two reigned over all Denmark, some-times one, and sometimes one over England and Denmark” (Kroman 1976: 166). The Anglo-Saxon chronicle similarly states that in 872 two kings of the Danes fought Alfred: Bagseg and Halfdan, and a number of Earls ( jarls), andwhen Danish reinforcements arrived on English ground in September 875, they were led by three Kings: Gudrum, or Gorm, Oskitel, and Hamond.

Kroman (1976:166) argues that the existence of “five kings” does not mean that Denmark was divided into five kingdoms but that Gorm, nephew

Landscapes of Power, Landscapes of Conflict - [PDF Document] (177)


and cousin, respectively, to the previous Kings Horik I and Horik II, suc-ceeded them and was the supreme ruler. Gorm was clearly a leader who ruled from a power base in western Denmark, for he bears a traditional royal name of that region. Thus Gorm would have been the “true” king, and the oth-ers would have been his underlords, high-status vassals of magnitude only second to the king. In this scenario, the division of the kingdom was to facil-itate military organization, as evidenced by the above described arrival of re-inforcements after the A.D. 875 battle of Kestelren in East Mercia. In chapters 4 and 5, the possibility that many “kings” may have sometimes indicated one overlord and a number of chieftains was entertained. Although this explana-tion is plausible, it is not certain. Left with two competing interpretations of the historical record, locational interpretation becomes especially revealing.

True subdivision into five petty kingdoms would be indicated in the lo-cational distribution of central places: five polities in Western Denmark would almost certainly have five centers, five systems to support them, as ap pears to be the case in Scania (Figure 6.5). In this period, Scania, and there-fore the combined rank-size curve, has changed a little but remains very convex with an RSI of .574.

However, dramatic change has occurred in the west: rather than indi-cating division, a high degree of integration is now apparent (Figure 6.6). The population curve is primo-convex with the attached implications of po-litical organizational change, and the RSI is almost perfectly log linear at .009. Even the curve based only on site size has become significantly less convex, with an RSI of .284. The histograms at this period show that a new level of site size has been added, with the development of Hedeby into a major walled 24-hectare site. There is now a large center, a secondary center, large villages, and small villages.

The histograms show remarkable separation of hierarchic levels in both population and size versions. Regarding the theory that many petty kings still ruled western Denmark, no such divisions seem to be indicated by the loca-tional analysis, and moreover, all indications are that this period shows a marked increase in integration.

65.2.3 Period III: A.D. 950- 1050

Before discussing the central places of the third period, there is the symbolic center to consider: the royal monuments and runic inscriptions at Jelling. There, sometime just before 950, King Gorm raised a stone to his wife Thyra, calling her glory, or amender, of Denmark (translation is uncertain). This runic reference to Denmark is the first native, written naming of the kingdom. In the History of King Olaf Tryggvason of Norway, it is mentioned that Gorm (who reigned from about 920 to 950) cleared all competing royal claimants

Landscapes of Power, Landscapes of Conflict - [PDF Document] (178)


Figure 6.6. West Denmark, ca. A.D. 950: rank-size curve and corresponding histogram.

from Denmark, conquered part of Wendland (the Slavs in presentday Poland),and persecuted Christians. A second runestone was raised by Gorm's son Har-ald Blåtand, who was king after him, commemorating his rule and the Chris-tianization of the kingdom. Harald ruled from about A.D. 950 to 980.

Also at Jelling are two monumental earthworks; one mound is a ceno-taph 77 meters across and 11 meters high. The other is a burial mound 65 meters in diameter and 8.5 meters high, the largest burial mound in Den-mark. It held no remains but had rich grave furnishings, male and female. It is believed that Gorm (who died ca. 950) and Thyra (who predeceased him) were buried there by their son, Harald (Roesdahl 1982:172-3). Because he eventually became Christian at about A.D. 970, their bones were apparently later removed to a church. The mound was carefully opened after its con-struction and all the grave goods left alone—but the bones are now missing. In the early wooden stave church at Jelling, excavated in the 1980s, the bones of a middle-aged to elderly man were found, in disarticulated disarray, and clearly in secondary context. This is almost certainly Gorm the Old. This

Landscapes of Power, Landscapes of Conflict - [PDF Document] (179)


site marks the time when the church got its first foothold in Denmark, de-spite 200 years of previous active proselytizing. Even so, the rest of the Danes would not follow their kings until well after the millennium.

There is no habitation at Jelling, so it is considered to be a political mon-ument in what Roesdahl terms “native tradition” even with its Christian con-nections. She speculates that it may have been a nationalist, unifying response to the ever-threatening Germans and Franks (1982:175). Although this leap from earthwork to ideology is rather large, the rank-size distribution for western Denmark at about A.D. 1050 is indicative that from the reign of Harald Blåtand starting in 950 and the start of Sven Estridsson’s era in 1047, the Danish kings ruled a highly integrated kingdom (Figure 6.7). The rank-sizecurves have now dropped into negative numbers, or a primate curve, at – .299 and – .095. The top half of the distribution shows a very concave curve. The histograms have now included Roskilde as the largest place—this

Figure 6.7. West Denmark, ca. A.D. 1050: rank-size curve and corresponding histogram.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (180)


new urban seat of Danish kings encompassed about 50 hectares within its walls—and with Viking Age archaeological material in the area between the city and the fjord harbor included, the settlement as a whole covered more than 80 hectares at A.D. 1050.

Just as significant is the system of fortified military encampments estab-lished all over the kingdom, described in chapter 5. These are found in the core of the state as well as in Scania (Jacobsson 1989, 1990a, 1990b; Ohlsson and Cinthio 1980). These combination military and administrative outposts were very similar and date dendrochronologically to the same 20-year period. The fortresses are not outward-looking but are focused on internal routes and approaches, strategically placed to control the whole kingdom. No similar idea seems to have preceded them (Roesdahl 1982: 154).

Smaller settlements also underwent change at this general time, the first real change since the end of the Roman Iron Age. In Jutland and the Islands, many villages moved a short distance at this time, to the site of the medieval village that is still extant, and lie within a few hundred meters or less of their Viking Age counterparts.

In Scania, at just around A.D. 1050, some villages moved a similar short distance but most of them instead contracted. Ridderspore (1988) noted this phenomenon—the abandonment of the loosely woven Viking Age site and the movement of the community in toward what had been large open space between the dispersed farmsteads. Tesch (1992a, 1992b, 1993) found this phenomenon in every village in Herrestads Härad where he tested all settle-ments that could be located.

In the survey of Järrestad conducted for this study, it was found that on av-erage, the large and medium-size villages contracted by about 60%, while the smaller places contracted, expanded, or stayed the same. The other obvious ex-ception in Järrestad is Tummatorp, one of the royal founded towns, which of course grew enormously from a small torp village to a city. This happened prior to 1050 and so is not reflected in Table 8.1 , which illustrates the amount of change seen in village size in this close survey of a typical Scanian system. In looking at Scanian sites in other Härads where there was contraction, it can beseen that a comparable proportion occurred. Examples are places such as On-nerup, in Torna, which dropped from 15 hectares to 7.5 hectares, and Stora Kopinge in Herrestad, which shrank from 42 hectares to about 20.

65.2.4 Conclusions: Part I

Locational data support the argument that a state did not rise full-blown in the late 11th century under the auspices of the advancing church or any other Late Viking or Early Medieval institution. On the other hand, there is no indi-cation of an integrated, unified state in the ninth century or early Viking Age.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (181)


At about A.D. 850 the rank-size distributions are not evocative of state-like organization, but there is a great deal of activity that points to the be-ginning of centralized control, and historical documents indicate a probable political alliance that really existed only in name. By A.D. 950, the distribu-tion is growing more linear in Jutland, as a class of high-ranking nobles is emerging, controlling great wealth. By A.D. 1050, western Denmark is show-ing a primo-convex curve, and even with Scania and its residual patterns of settlement the kingdom seems fairly united, supported by other data, in the form of centralized building and engineering, inscriptions, and distribution of military sites. The growth of the state is gradual and spans the entire 350 year period.

6.5.3 The State of the State II: Greater Denmark Through Space

Archaeological and historical evidence cited above suggests that the unifica-tion process began in the Jutland/Fyn area and spread east over time (Rands-borg 1980). Even if some aggressive action began this unification, it was not an alliance held together by force, as it is clear that the ruler of the unified state still had to win the approval of his supporters and answer to the people. The tradition of the rule of the ting and the assassination and deposition of many rulers during the whole period, up to and including the 12th century, attests to relatively weak central control and the need for voluntary partici-pation of the provinces in the state forming process. It also demonstrates the reason why central authorities would go to some lengths both to accommo-date the differences and to construct unifying institutions in these areas.

However, to reiterate chapter 5, such voluntary political unification may have created conflict between preexisting local elite and new elite who rep-resented central authority. This may have necessitated a strategy in which the new dynasties suppressed or eliminated the old-style elite that had spawned them, and created a new group of favored lords. The old chiefly class had the right to unseat their elected over-lord if he became too powerful and threat-ened the traditional system. The new lordly class was not connected with the old forms of rule, and owing their existence to the king, would support him against those who had lost or were losing their power and position in the new social and political order.

It has been argued by several scholars that during this transition local no-bles would have resisted centralization, attempting to maintain their local au-thority for as long as possible (Löfving 1984). This would lead to different patterns of culture-geographic change in the more resistant areas. Locational analysis of existing archaeological material, discussed earlier, suggests that the two major regions involved in the unification—east and west—under-went dramatically different sequences of integration. This may indicate that

Landscapes of Power, Landscapes of Conflict - [PDF Document] (182)


the struggle for jurisdiction of the eastern region, and the reluctance of pre-existing elite to relinquish authority, led to the slower integration of Scania with the Danish state.

Change in rank-size relationships, boundaries, and levels of hierarchy show that western Denmark was approaching effective integration by about A.D. 950, and even more so by A.D. 1050. Western Denmark appears to have been growing easier to administer, while at the same time, taken as a whole, Scania was weakly integrated until the very last period thus far examined.

To understand how this integration occurred, Scania must be examined from the early Viking Age, when the west was beginning to come together under dynasties based in Fyn-Jutland, through the time around A.D. 1000 when the west had become integrated but the royal towns were not yet established or were just being established in Scania, and finally at the era of town-foundationsin eastern Denmark. The graph for around A.D. 1000, seen in Figure 6.8, shows

Figure 6.8. Scania, ca. A.D. 1000: rank-size curve and corresponding histogram.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (183)


that the Scanian rank-size curve is very convex, suggesting that a number of competing centers form the top part of the curve. Refer back to Figure 6.4; the Theissen polygons show that when mapped onto the landscape, each large site appears to the center of a smaller system, some appearing to be bounded by rivers on each side, others appearing to lie along and on both sides of a river in its river valley with some kind of no-man’s-land between. The other subregions are similar, but these six areas were chosen because there has been more archaeological work in them and they are representative of the others that have not been included here. The current name and the oldest known name-form of the six Harads studied here are Harjager (Harthakærs), Torna (Thornæ) Bara (Bergio), Herrestad (Haeruæstatha), Ingelstad (In-gilstatha), and Järrestad (Jarllestatha).

In examining these largest sites, starting in the upper left and following the coastline counter-clockwise, we have Löddeköpinge, which has been al-most completely excavated (Ohlsson 1976, 1980; Ohlsson and Cithio 1980). Part of this site dates back to about A.D. 550, though the main occupation began at around A.D. 800 and continued in the same place until the early Mid-dle Ages (Callmer 1986). The next site is at Borgeby, named for the probable late 10th-century ringfort that is located there, but that dates back to be-tween 700 and 800 (Stromberg 1961; Callmer 1986). Uppåkra is the next one; it was discussed in an earlier chapter as the probable seat of the most powerful polity in Scania and likely to be the seat of the probable king of a united Scania. It dates back to the late Roman-Germanic Iron Age and con-tinued up until the time when the royal town of Lund was founded nearby in about A.D. 990. Traveling around the coast to the southeast, Stora Herrestad is the next site. This site and region have been extensively investigated and nearly every village in the system has been excavated to some extent (Tesch 1993; Berglund 1991a). The site dates back to about 600 to 700, and has been occupied continuously since then. In Ingelstads Härad, the site of Örum prob-ably dates to about 600 to 700, and so do most of the other sites surrounding it. Finally, in Järrestads Härad, Simris is the largest site, and was establishedsometime in the late Roman–early Germanic Iron Age, when it appears to have been a chieftain's seat. This will be discussed in later chapters. The names indicate that all these are very old sites, founded as early as the 500s and no later than 700. This is well before any unification has been suggested in this study.

Do these sites represent the central places in prestate polities? The ear-liest Härad division known from the Viking Age match the polities suggestedby the Theissen polygons almost exactly. The sites that form the centers of the polygons are far older than the state, which suggests that these divisions or very similar divisions are much older than the later units and probably formed the basis for the later districts of the Danish state.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (184)


Thus, the condition of Scania in A.D. 850 when the Danish state began to gain impetus and this eastern area was first named historically as part of Den-mark shows a highly regionalized, decentralized area with settlement pat-terns entrenched since the first nucleated settlements were founded, and whose sub-region inhabitants may very well have had a deep sense of their identities as different “peoples.”

A closer look at the rank-size curves for these individual polities in Figure 6.5 show us that in A.D. 1000—before any major changes had much altered the curves as they were in 850 and 950—the distributions are internally far more integrated than Scania as a whole. Some are quite convex, like Järrestad; oth-ers are much closer to log normal and not very different from Jutland in 850, such as Herrestad and Ingelstad. Bara Härad, where Uppåkra’s chiefly center is found, is really statelike, and in A.D. 850 far more complex than anything on Jutland at that time. It appears to have four levels of hierarchy, which supports Kristian Kristiansen’s idea that some of these pre-state polities were stratified societies approaching archaic states (Kristiansen 1991).

Convexity in other of these very old systems is not surprising. When the villages were established, Germanic custom indicates that they ruled them-selves through meeting at the assembly, with the whole community making decisions. A warlord was paramount only in times of war and held a heterar-chic role in balance with the assembly in peacetime. Later the elite were more powerful, yet still were constrained by the code. It is unlikely that they would have initiated the kind of centralization that affected society so much as to recast the settlement pattern. The “Hær” or lord, the “Jarl,” and the man called “Ingel” or Ingulf, combined with “statha” (seat of the lord, seat of the Jarl, seat of Ingel ) named in the last three Härads suggest that regional elitewere the nominal rulers of these systems.

6.5.4 Interpreting Regional Patterns: The Case of Scania

Locational data and social change must be examined in tandem because the evidence supports the idea that they are interrelated. Scania, which was al-most certainly politically unified with Denmark by A.D. 850 and possibly ear-lier (Andersson 1947; Johannesson 1984; Jorgensen 1987), was still operating on a pre-state system of places involved in trade, assembly meetings, religion, and settlement. Performance of related institutional activities remained in their ancient locations much more so than in western Denmark, and this is seen in the residual regionalization, the lag in integration long after actual unification. I interpret this as evidence of administrative conflict. As local and central authorities resolve problems of control, we expect to see sequen-tial changes in the locations of central political places, religious centers, and economic institutions. The mechanisms through which the state consolidated

Landscapes of Power, Landscapes of Conflict - [PDF Document] (185)


its control over Scania involved the manipulation of central place location and function, through the founding of new centers. As noted in the begin-ning of the chapter, the west Danes’ first strategy was to control the east as a hegemonic state, without direct intervention. It must soon have been clear that a more direct form of control was necessary. Little by little, an infra-structure was built.

The first evidence for this comes in the transformation of economic in-stitutions, in the location of markets, their administration, and their system structure, as the exploitation and distribution of vital agricultural products was transferred from a local distribution system to a much wider network. A tier of new markets was established in the 800s, currently interpreted as having been both founded and controlled by central authority (Brattberg 1983). Evidence for this comes from their nearly simultaneous founding, the use of similar names, and their strategic placement to cover the whole province evenly and completely (Figure 6.9a). In each region a site with the suffix köpinge (basically meaning “shopping”) appears, indicating a market settlement of the late EarlyViking Age (Brattberg 1983). Although some set-tlement at these sites may begin as early as A.D. 550, the markets— köpinges—were established in the 800s. All of these markets were locatedapproximately 2 to 3 kilometers upriver from the coast, usually in a mean-der to facilitate defense, as there was too much raiding and feuding to ac-commodate a market in a vulnerable coastal position without heavy losses (Brattberg 1983).

Through excavation of some of these sites, especially Löddeköpinge, itcan be seen that these places had a large functional size that included craft production, trade, regulation, and taxation. Aside from craft activities and production, the presence of many coins, scales, and weights for measuring silver shows that monetary transactions took place there. The collection of taxes appears especially evident since the boundaries around the site were not defensive—they were probably administrative, indicating that within the enclosure certain laws, regulations and taxes were in effect (Ohlsson 1976, 1980; Callmer 1983, 1984). This is similar to what is found at Ribe in the west somewhat earlier, which may have been the model for these new markets.

However, there is one difference: the Scanian markets are interpreted as being primarily export locations. The seasonal nature of the marketplace at Löddeköpinge is seen in the floor layers of pithouses, where occupation lev-els are interspersed with clean wind-borne sand. Earliest Ribe was also a sea-sonal market, and it did not become permanent until the early 800s. This seasonality has led Rausing (1990) to suggest that goods were brought to the sites in fall after harvest and kept until spring, the sailing season, when they were shipped out. With so many products and so much money concentrated

Landscapes of Power, Landscapes of Conflict - [PDF Document] (186)



















Figure 6.9. The development of markets and towns in Scania (after Tesch 1883:72).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (187)


in one place, there must have been elite administrators present to oversee it and a force of soldiers to protect it.

It is probable that the concentration of political and economic power among fewer and fewer western elites over time led to the channeling of grain and other produce, which was not abundant in Jutland, away from the pro-ducers and into the hands of administrators for shipment elsewhere (Christo-phersen 1982:127). The 9th and 10th centuries were also periods when defense against the Franks and Germans, and against internal pretenders to the throne, gave rise to standing armies of large size. They would have needed pro-visioning. The collection of surplus may have been accomplished at köpinge lo-cations. Figure 6.9a shows that in the southwestern corner of Scania, where the hook-shaped peninsula called Fotevik is located, not one but several such sites are located. Documentary evidence from slightly later times tells us that after the small, local fleets of ships had been levied in their traditional, regional har-bors, they converged on Fotevik, which was the primary launching place for the entire Danish navy. The number of collection sites near this location sup-ports the interpretation of the markets as provisioning centers.

In many cases, the köpinges also appear to have been sited where there isevidence of strong, local elite in power since much earlier times. Löddeköpinge is located to serve the region of Harjagers, Torna, and Bara. Nearby is the large chiefly center of Uppåkra, and the Källby burial, a rich princely grave, which is probably linked to the petty kingdom that was centered here. Near Löd-deköpinge itself is located the equestrian boat grave at Lackalanga, also indi-cating an elite presence (Callmer 1987:177–8). Another regional market and production site, Gådsköpinge, near the place where Va was founded, has also been extensively investigated. Nearby, there are also impressive graves, far above average in grave goods, which has suggested to some that this was the seat of a petty kingdom in the eastern part of Scania. Runestones near some of these sites indicate the continued presence of local elites (Callmer 1986).

The large central-state markets are set down among these long-entrenched elite territories with fairly clear purposes; first, to establish eco-nomic control and gain direct access to provincial wealth. Second, in terms of the roles of local elites, the locations were probably chosen both to co-opt and utilize their power in their own districts; that is: (1) they shifted staple economic transactions out of the local harad level and into a national network, and (2) local elites were probably enjoined to help run these mar-kets at the direction of central authorities, not an atypical strategy for cost-conscious rulers, and one often seen in colonial or imperial strategies for provincial rule (D’Altroy 1992).

The timing of the founding of these markets seems to be a period of transition, when the western and eastern Danes appear to have been politi-cally united but their systems were not yet integrated, and it is notable that

Landscapes of Power, Landscapes of Conflict - [PDF Document] (188)


köpinge sites are limited to the poorly integrated province of Scania. Al-though places with the element “köpinge” are found throughout Denmarkand Sweden, these are all the sites of later, high Medieval marketplaces, when the suffix was commonplace. Only in Scania do köpinges date to the early Viking Age (Brattberg 1983; Callmer 1983). Unlike late Medieval köpinges, the Scanian markets also were abandoned before the end of theViking Age, as in the case of Gårdsköpinge, or reverted to regular rural vil-lages, with only the name hinting at their original function, such as Stora Köpinge and Löddeköpinge. This indicates that they came into being to serve a special function, and when that function was no longer necessary they fell into disuse. When further steps toward infrastructure building and central place manipulation were taken that truly integrated Scania with western Den-mark, such sites became redundant and eventually faded.

The rank-sizes at A.D. 1000 show that these marketplaces did not ade-quately integrate the eastern province, and further action was taken. Around the end of the 10th or beginning of the 11th century, four towns sprang up in Scania: Lund, Tummatorp, Vä, and Helsingborg (Figure 6.9b). These all had strong royal presence; in fact, they were royal foundations, and the church here was not that of the local lord but of centralized, institutionalized religion with bishops who were princes or close relations to the royal family. At these centers are also found indications of law, taxation, marketplaces and industry, and at Lund and Tummatorp, the minting of coins was carried out. Economic, political, and religious traditions had previously been practiced outside of central locations. Now, all were conflated at one site. The old elite were com-pletely cut out of the systems that they had once managed.

Although there were small, earlier settlements under some of these places, with the exception of Vä, which had substantial regional prominence in earlier times, they were not centers. The act of establishing new centers in-dicates that for a time at least there were indeed conflicting systems. A con-quest, with utter eradication of opposing elite, commonly results in the less costly method of elaborating old centers rather than the building of new ones. Here, with the exception of Vä, the building of new centers supports the idea that the old centers were still occupied by their local rulers and that the towns were built not only to integrate the area but to counter the power of these old elite.

How does this affect rank-size curves? Figure 6.10 shows that the unified state at around A.D. 1050, after the following had already ocurred: the expan-sion of Roskilde, the use of the köpinges between ca. A.D. 800 and 980, thesubsequent founding of four urban centers in Scania, and finally the further urbanization of the whole state, including Scania. The time also marks the regulation of the village in Scania, and the contraction, or shrinking, of many village sites.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (189)







N Figure 6.10. West Denmark and Scania, ca. A.D. 1050 and 1200: rank-size curves and corresponding histograms

Landscapes of Power, Landscapes of Conflict - [PDF Document] (190)


Finally, the span of control has been narrowed—first when markets were established and then again when towns were founded, lowering the number of settlements administered by a controlling place (Figure 6.11). This whole sequence shows the transition from a very wide-span or loose central control, to a very narrow span of control, where central authority attempted to keep the region on a tight rein by decreasing the span of control and in-creasing the presence of central authority.

The centralizing purpose behind these activities are seen in the following features of cultural geographic change: (a) the simultaneous foundation of marketplaces with similar functions and names; (b) the founding of new cen-ters rather than the elaboration of old ones; and (c) their strategic location for countering extant local elite centers and for full coverage of the province.

Figure 6.11. Changing span of control in eastern Denmark.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (191)


At the same time, a probable combination of village regulation for the purpose of tax and military levies and the state-encouraged conversion to Christianity caused the population to abandon their loosely organized, spread-out villages and resettle densely near churches (Emanuelsson 1985: 38). The new towns also provided new ways of living and must have con-stantly been subtracting population away from the older villages and nucle-ating it around new centers. In this way, the fabric of the cultural landscape was abandoned and rebuilt according to central design. Although the reset-tlement was nearby and the functioning of the cultural landscape little af-fected, this discontinuity may have reinforced the changes in administration. This will be discussed further in later chapters.

In the same period, the late 10th century, came the construction of the Trelleborgs all over the kingdom (Randsborg 1980; Roesdahl 1982), in the core of the state as well as in the study area. Two of these are known archaeo-logically in Scania (Jacobsson 1989,1990a, 1990b; Ohlsson and Cinthio 1980). These are located in the northwest and southwest quadrants; some Swedish archaeologists expect that two more may be found in their respective quarters of Scania. This would have brought a strong central military presence to the province at the same time the towns were established in A.D. 990–1000.

A final look at the rank-size for the whole state in 1200 shows that the problem of integration and control had finally been surmounted, at least in terms of administrative functions (Figure 6.10). The population curve for the entire state is close to log linear. In Scania, more towns rose in this era, and in the western region many cities were founded and chartered. Since kings con-tinued to have problems keeping their subjects from killing them in fits of dis-satisfaction until almost 1200, this success at building infrastructure and effecting unification must have been one of the few aspects that allowed the Late Viking Age/Early Medieval kings to have continuity, to permit the less-hatedson or brother to take up the ruling of the state without utter chaos surround-ing succession, as there had been at some points in the 9th and 10th centuries. Conclusions: Part II

By 1100, the whole of southern Scandinavia was united both in name and in fact as the state of Denmark. Change in the social and political order was first attempted without large-scale intrusion or conflict, and this is reflected in the slow process of integration seen in cultural landscape change.

While Gorm the Old, who ruled from 920 to 950, was apparently from a line of Jutland elite, linguistic evidence suggests his wife was an “East Dane,” the scion of a Sjælland or Scanian family. A number of scholars have suggested that this marriage was a strategy for unification. Gorm’s memorial to her includes a phrase often disputed by epigraphers, two words in the inscription: Dan-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (192)


markar bót. The meaning is unclear. It means that she was either the adorn-ment or the amender of Denmark. If she brought unity between oft-dividedlands, as Lund (1991) believes, “amender” might be reasonable (Jones 1987).

The use of the word “Danmark” in the Jelling monument is relevant to the study of unification. The text was probably inscribed after 983, when portions of southern Jutland in the possession of the Germans were re-conquered by the Danes and the borders made more firm (Anglert 1995:9). It is the fast “na-tive” inscription of the word Danmark, which means “land of the Danes.”

In this inscription, and in other non-Danish sources dating from the late 800s, a fairly clear distinction is made between what Lund calls “Denemearc,” “Tanmarkar,” and “Tanmaurk Ala” (Lund 1991 ; Anglert 1995). Tanmarkar refers to the land that belongs to the Danes, while Denemearc refers to an area where the Danes did not have complete control or authority, especially the large land mass that made up the easternmost regions (Anglert 1991:9). Denemearc is the word used on the Jelling monument. The raising of a stone stating that a region was part of the kingdom, yet at the same time acknowledging that it is yet something a little different than land belonging to the Danes, expresses the in-ternal divisions of the realm as well as a desire on the part of rulers to annex or claim these lands. While chapter 2 noted that contemporary sources indicate that Scania was a part of Denmark as early as the second half of the 800s, by the late 900s the Jelling monument indicates that large parts of the east appear to have still been incompletely melded with the core of the state.

Harald Blåtand, who reigned from 950 to 986, and his son Sven “Fork-beard” Haraldsen, who ruled from 986 to 1014, founded towns and built ring-forts, Knut the Great ruled all of England, part of Norway, and Denmark as one nation between 1018 and 1035 and brought to Denmark many Engish ideas on law, taxation, and administration. Sven Estridsson’s reign from 1047 to 1074 is documented in protohistoric texts as very peaceful in terms of internal con-flict. He was also a town-builder and a church financer. Most of these kings used integrative strategies that were intrusive into local systems, but not vio-lent or militaristic. In addition, Sven Estridsen was almost certainly of Scanian extraction: his mother was sister to Knut the Great, but his father was the shadowy Jarl Ulf, whose origin is unknown. During his battle for the Danish throne against the Norwegian king who sought to seize it, the Knytlinga Saga repeatedly indicates that Estridsen always retreated to Scania to regroup and strategize. Thus, Jarl Ulf almost certainly hailed from Scania. This ethnic con-nection might have made the continuing unification that much easier.

However, in the final stages, conflict became unavoidable and is recorded in the early historical records. Records from the Early Medieval period tell of up heaval and revolt; the animosity between the Scanians/Hallanders and Knut the Holy and his assassination in 1086; and, especially, in 1180–82, the time of the Scanian Uprising, aimed primarily at the centralized state: refusal to pay taxes,

Landscapes of Power, Landscapes of Conflict - [PDF Document] (193)


tithes, and serve military duty (Andersson 1947:412). Saxo Grammaticus is not always a reliable author, but in regard to these events he is generally corrobo-rated by other sources. He wrote of the physical assault and near-murder of royal officials, such as the king’s bailiff and tax collector Åge, the so-called Sjæl- lander elites, Esbern Snare, Sune Ebbesen, and Sakse Thorbernsen, close kins-men who were non-Scanian appointees wielding great power. Their job was to “take up the king’s errands” in Scania (Andersson 1947:413). Violence and protest were also directed toward the foremost royal official in Denmark, the debt-collector Thord, who was only saved from the mob by the archbishop. The accusation that the foreign elite (from Sjælland) had robbed Scania was heard at every assembly, and the demand repeated that they be removed (An-dersson 1947:414). Local noblemen spoke out at the assembly in defense of the farmers who were in open revolt (Andersson 1947:414). The province of Sca-nia had been a part of Denmark for over 300 years, and the elite from the cen-tral royal court on Sjælland were still perceived as literal foreigners.

King Valdemar was forced to sail to the province, and with the archbishop travel from ting to ting, where the public railed at him to remember the “oldlaw” that the king must answer to the people. He eventually gave in and removed a number of the foreign elite (Andersson 1947:415). Most of the other non-Scan-ian magnates soon fled. Several treaties were attempted, but these concessions were not enough for the Scanians. In the end, war resulted, as the farmers and their local leaders clashed with the king (Andersson 1947:416). Saxo describes a number of battles in detail, and the recruitment of a scion of a rival royal fam-ily Harald Olofson, who had been living in political asylum in Sweden. He was the son of a former Scanian noble, referred to in the records as Olof, King of Scania. This “pretender” led the rebellion for some time. In the end, the Scani-ans lost this war in a series of battles where thousands were killed and they were at last reconciled with the central government (Andersson 1947:418–419). After this, the issue of Scanian independence apparently was resolved.

These uprisings in the early historic record indicate a long-time antago-nism between the native Scanians and their new, foreign overlords from mag-nate families on Sjælland, appointed directly from the royal court. Saxo in particular clearly implies this, and his story is confirmed in the Sjælland Chronicle (Andersson 1947:420). The concept of a network-style central power was not recent yet still alien. The Late Viking-Early Medieval transition included more incidents where public disapproval of kings culminated in regicide than did the less documented Early Viking Age. This may reflect that written records grew more frequent, but it may also indicate that people were more dissatisfied as the grip of central authority tightened.

Our knowledge of Danish-Scanian political interactions in early historic times, where a long history of strained and violent relations is indicated, sug-gests that a strong sense of ethnic identity persisted in Scania and that the

Landscapes of Power, Landscapes of Conflict - [PDF Document] (194)


preexisting elite attempted to maintain their authority for as long as possible, through the control of the preexisting central places in the old cultural land-scape, controlling and maintaining the traditional geography of marketing, production, taxation, lawgiving, and other administrative functions.

Thus, while central elites appear to have used intermarriage, prestige gifting, and other social strategies for the purpose of integration and control, because of long-term resistance the actual disruption of old elite systems was necessary, first by inserting new places into the central place system and eventually through the use of direct coercion and force. In this model, calcu-lated steps were taken to slowly undermine and dismantle local elite juris-diction in order to integrate and exploit this rich agricultural region that was densely populated and valuable for taxation and cannon-fodder as well as for staple agricultural produce (Brattberg 1983; Callmer 1987). Between the 10th and the 12th centuries, every ruler of Denmark was a direct descendant in a royal family, from father to son, nephew, or grandson or from brother to brother. It is therefore not remarkable that such a program may have been carried out, perhaps not with decisively timed and organized stages but with a strongly calculated intention.

It has been noted that some states impose hierarchy by expanding exist-ing sites, whereas others add levels of hierarchy by adding levels onto the top (Kowalewski et al. 1989; Johnson, personal communication 1989). Levels were added on top because new locales, new hubs were needed in order to disrupt ancient entrenched patterns, to dismantle the landscapes of everyday life, and to redesign them into a more amenable pattern. In the lower end of the curves that illustrate this chapter there remains a convex bulge of persis-tent local settlement patterns, still exhibiting pre-unification characteristics.

In a period of about 300 years, the power of old places was undercut and the power of the new was developed and transformed as important ele-ments of the cultural landscape—political power, economic activity, and sa-cred concerns—shifted from a disarticulated, loosely controlled set of systems into a coinciding, tightly overseen, and regulated network, located in the state-founded towns that contained the churchly hierarchy, the markets, the law, and the government.


Chapter 6 took shape against the background of the varied lines of evidence discussed in chapters 3, 4, and 5 and measured them against the changes in the cultural landscape of south Scandinavia through time and space, in hopes of combining the disparate bits of archaeology and history into a unified un-derstanding of cultural persistence and cultural change.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (195)


The chapter began with a discussion of geographic and locational analy-sis, stating my expectations for the utility of such studies in the study area be-cause of the unique problems found in northern Europe. Theories, terms, and methods used in the chapter were defined and explained, especially the rank-size analysis and complementary observation of histograms and span of control. In the next sections, these methods were applied to the archaeolog-ical/historical record of settlement in Denmark, first through time from the earliest to the latest phases of study, and then through space, looking more closely at the differences between regions.

Throughout the chapter, locational patterns were integrated with cul-tural data on political and economic change in order to present a more “cus-tomized” analysis than the mere translation of curve-shapes with a limited set of interpretations onto an archaeological dataset.

The chapter strove to illustrate that while the western “homelands” of Danish rulers quickly came under their control, eastern regions with indige-nous elite and a strong ethnic and regional identity resisted unification. In order to bring the eastern provinces into the state, kings first used hege-monic strategies, then economic intrusions, followed by tinkering with cen-tral place location and function and building infrastructure, and finally the quelling of independence with force. These strategies formed a continuum of least-cost to highest-cost, both in terms of silver-and-gold and “cost” in public relations–the slaughtering of one’s own citizenry to put an end to their claims and perceived rights.

The next chapters will move on to the smallest scale of analysis, the local level, in which one of the six studied harad regions, Järrestads Härad, where extensive survey was carried out, will be examined closely for the ef-fect of these broader changes on life in a typical Scanian sub-system.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (196)

Part III Power and Force:

Courses Toward State Integration

Landscapes of Power, Landscapes of Conflict - [PDF Document] (197)

This page intentionally left blank.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (198)

Chapter 7

Reconstructing Cultural Landscapes in

Southeast Scania


Before a landscape such as Järrestad’s can be analyzed, it must be recon-structed. I will identify important components of the cultural landscape in the Järrestad area and comparable neighboring regions, and discuss a method for reconstructing the late Iron Age and early medieval milieu consisting of eight specific criteria. Finally, I will outline the survey strategy and laboratory techniques used to study the region’s settlement and land use history.

7.1.1 Reconstructing Landscape in Southeast Scania

In the Germanic Iron Age and Viking Age, Scania constituted a major portion of Denmark, approximately 20,000 square kilometers. Järrestads Härad, thesubject of analysis on the local level, is an area of approximately 150 square kilometers and is located in the southeastern morainic plain of Scania, which is called Österlen, or “eastern province.”

As noted in chapter 2, the province of Scania has been called “a prehis-toric country in the process of dissolution” (Stromberg 1977) with a “hidden, but disappearing cultural landscape” (Tesch 1980: 20), because submerged in current patterns of settlement and agriculture, prehistoric cultural landscapes can still be detected. The cultural landscape is made up of many diverse as-pects of the natural and built environment, and its evolution must be unrav-elled in a comprehensive manner. A picture of prehistoric settlement—where


Landscapes of Power, Landscapes of Conflict - [PDF Document] (199)


people were and what they were doing—can be reconstructed from a variety of landscape components. They are identified (adapted from Söderberg 1994) through the following procedures:

1. Determination of hydrology during the study period. Habitations ofthe Late Iron Age are usually located close to water sources and wet hay meadows.

2. Examination of the makeup of the subsoil in the region. Althoughmost farming took place on light to heavy clays, settlements are al-most always located on subsoils of gravel, sand, or mo.

3. Location of areas with a suitable topography for prehistoric settle-ment and other land use needs. Habitations usually are associated with areas of plains or plateaus, often at places where they articulate with other landscape types, such as wet meadows, scrub, bogs, hilly terrain, which were all exploited by the village during the course of the year. The exact location of the village is typically on a ridge or is-land of sandy or gravely soil several meters higher than the surround-ing plain.

4. Consideration of known, registered prehistoric monuments suchas medieval churches and the ruins of manors and castles, Iron and Bronze Age mounds that often continued to be used for various purposes (especially as assembly places) by later inhabitants of the region, gravefields, and runestones, to name the most common examples.

5. Familiarity with earlier a rchaeological investigations (unregis-tered) that have been carried out in the area, as well as the location of stray finds and hoards.

6. A study of existing phosphate maps, assuming the likelihood that areas with high phosphate content not attributable to other sources are the location of prehistoric settlements. This serves as a palimpsest for retesting areas of interest with a sampling strategy in accordance with current archaeological survey standards.

7. The visual inspection of all surface areas tested for phosphate con-tent, and the mapping and collection of surface artifacts, which consist mainly of ceramics. Subsurface testing may or may not be con-ducted pending landowner cooperation.

8. Finally, the documentary record of maps, manuscripts, registers, and landbooks should be researched for indications of prehistoric settlements, monuments, land-use, land-ownership and administra-tion, all of which may be preserved in the oldest known maps and written forms of names of districts, villages, fields, and topographic features.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (200)


7.1.1. 1 Hydrology— Watercourses and Water Sources

With few exceptions, Scanian settlements were located close to large or small watercourses (Callmer 1986:187). In addition to the main course of the Tom-marp River, the Järrestad region is rich with wetlands and waterways, andevery village in the study area is either in close proximity to the river or wa-tered by one of many smaller feeders. These smaller watercourses often made up boundaries between village areas, and these boundaries and existence of fords across them are preserved in place names.

However, since the mid- to late 19th century, large drainage and reclama-tion projects have severely altered the region. To understand Late Iron Age hy-drological systems, one must refer to the Skånska Recognosceringkartan orReconnaissance Map of Scania, completed between 1800 and 1820, when wetlands and water features now gone or greatly reduced were still preserved (Tesch 1992a). Figure 7.1 shows the Järrestad region in this map at around181 5, and a simplified wetland distribution diagram is seen in Figure 7.2.

Before drainage, there were extensive wetlands of different types—wet bogs, fens, calcareous wet hay meadows where fodder was cut—and dry bogs and fen woodlands, where stock was grazed and pigs were set to forage. There are still some extensive peat areas in the Tommarp River resource area, notably south and southeast of Tommarp, the Backemosen peat bog. In addi-tion to the use of the peat itself for fuel, bogs have also been used in Scania for hay making far back into the past (Mörnsjö 1969:14). As noted in chapter 3, bogs and wetlands were also used as offering places in certain phases of pre-Christian religion.

The Tommarp River is today in some seasons little more than a stream. However, in earlier times it was considerably larger. This is true of most Scan-ian rivers, some of which are estimated to have lost 96.6% of their water vol-ume over the last 200 years (Ekman 1973:68). Due to some small falls, the Tommarp has probably never been navigable between the settled regions and the coast in the way that the Lödeköpinge river in western Scania once was; there, trading vessels could sail or row all the way upriver to market during the Viking period (Stromberg 1976). It has been suggested that transport on the river may never have gone to the east at all, but toward the west on a long-gone water route referred to in old documents as stora floddiket or the “big river channel” (Hoflund 1921:55). However, there is not much evidence for such a route. By the 20th century this hypothetical artery consisted only of in-terconnecting streams, and it is doubtful that much shipping ever progressed along such a channel. It is far more likely that the well-established land-routeswere the primary method of travel along the Tommarp River Valley.

The falls on the Tommarp are of importance in and of themselves; they were probably the location of mills that had great economic importance,

Landscapes of Power, Landscapes of Conflict - [PDF Document] (201)


Figure 7.1. The Järrestad region, ca. 1815 (Swedish Survey Map of 1815).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (202)






















Figure 7.2. Simplified hydrology of Järrestad region

Landscapes of Power, Landscapes of Conflict - [PDF Document] (203)


both by milling grains into flour and by providing taxes. Royal authorities controlled the mills before 1085, when through a royal gift Knut the Holy gave five mills on the Tommarp to the church. Topography, Subsoils, and Soils Associated with Iron Age

In Scania, about 66% of sites are on soils that locally are sandy or gravely, while the other 34% are on light clays. Villages themselves are often perched on sandy outcrops, with surrounding field systems incorporating richer clay soils. This may be related not to soil preference but to preference for loca-tions by watercourses (Callmer 1986), which were formed by glacial activity and incorporate deposits of glaciofluvial sand (Callmer 1986: 197). It is alsoclear that settlements were often located where several soil types, hence en-vironments, articulated, thus taking advantage of their various resources.

The study of Scanian settlement topography (Figure 7.3) is often based on a tripartite scheme dividing the landscape into three regions: a coastal region with sandy soils and peaty areas that lies below 25 meters in eleva-tion; an outer hilly landscape with clay soils and rolling relief lying between 25 and 75 meters; and an inner hummocky landscape with mixed sand and clay soils and more acid content than the previous zones, lying above 75 meters (Berglund 1991:26). About 73% of settlements are in areas that can be described as the rolling outer plains. Järrestad is located on the south-east coast of Scania, in a region of just such varied relief. The prehistoric vil-lages in the Järrestad area appear to have been established on ridges 5 or 10meters higher than the immediate area, with lighter soils than the sur-rounding clay plains, possibly because of the limitations of the ard plow used in the earlier Iron Age. Heavier soils did not come into cultivation until later in the period.

The FAO designates southeast Scanian soils as an Orthic Luvisol (FAO l981:69), which has and ochric or umbric A horizon, over 50% base satura-tion, and an argillic B horizon, with no other subsurface horizons. These soils are generally well drained and oxidized. In the present, and in many ways except for extent of land under cultivation, the area remains similar to what it was a 1,000 to 1,500 years ago: “a rich agricultural region with level and even clay fields of a pronounced ‘plains’ character” (Ekström 1950:57). (See Figure 7.4.)

The coast itself was important for fishing in the Germanic and Viking Age, evidenced by the bones of cod and herring preserved in Iron Age pit-houses (Tesch 1991b:139). In the Early Middle Ages the herring fisheries of Scania were among the most profitable in all of Europe, propelling the de-velopment of coastal towns.

Habitation Sites

Landscapes of Power, Landscapes of Conflict - [PDF Document] (204)






















Figure 7.3. Topography of the Järrestad region.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (205)


Figure 7.4. Soil distribution in the Järrestad region. Registered Prehistoric Monuments and Previous Research in Järrestads Härad and the Tommarp River Valley

Because of thousands of years of farming, an alarmingly large number of pre-historic monuments in Scania have disappeared; not just in recent times, but Bronze Age has eradicated Stone Age, Iron Age has leveled Bronze Age, and so forth. It must be assumed that the complement of archaeological monu-ments in Scania is much reduced from the original. Still, clear patterns emerge, and regionalized land-use and settlement have followed the same trend for many thousands of years.

In comparison with western Scania, which has been a fertile ground for many dissertation projects, Lund University investigations, and rescue opera-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (206)


tions by the State Archaeologists or Riksantikvarieämbetet (RAÄ), also basedin Lund, the Järrestad area, rather remote from these institutions, has under-gone little research. What research has been carried out is of high quality: be-ginning in the 1970s at the western extreme of the river valley a major investigation was conducted at the settlement of Gårdlösa, an Iron Age villagewith associated burials and a religious shrine. The site was excavated by an interdisciplinary team under the direction of Berta Stjernquist and the study published in three volumes (Stjernquist 198la, 1993a, 1993b). This investi-gation represents the most extensive project in the study area.

Between 1959 and 1978 several small investigations were made in Tom-marp, revealing mainly early and later Medieval materials but with certain Viking period components, and new excavations, as yet unpublished, were undertaken in 1994–1995.

In the village of Järrestad a small excavation was conducted by Marta Strömberg in 1972 (Strömberg 1976), revealing one Late Germanic/EarlyViking Age pithouse smithy that produced large amounts of iron slag. This small project yielded important information, as the smithy appears to have been a specialized, elite operation; finds included mold fragments for ornaments of bronze, unusual in any context (Callmer, personal communi-cation 1995).

In Simris, Stjernquist excavated a large cemetery with associated cham-ber graves from the Roman Iron Age for her dissertation research as well as the grave of equestrian elite nearby (Stjernquist 1955). This important re-search revealed a great deal about the RIA in Järrestad.

In Gislöv, Strömberg carried out some small excavations in 1982, 1983 and 1984, revealing Iron Age pithouses, settlement remains, and ironworking.

In Simrishamn, several rescue excavations of small areas at the former shoreline have revealed some Viking Age material beneath the Medieval lay-ers, though no trace of a permanent settlement at the coast (Jacobsson 1979b; Pettersson 1991). A number of stray finds from the late Iron Age have been collected as well from Gladsax, Gröstorp, Järrestad, Viarp, Simris, Tom-marp, and other places. These previous excavations and finds will be dis-cussed in more detail below. Phosphate Studies in Archaeology:. Discovery, Development, and Use of Phosphate as an Archaeological Indicator

The correlation of the phosphate record with human activity was discovered in Sweden during the 1920s by O. Arrhenius, and his archaeological applica-tions were published in a series of articles from 1931 through 1963. He demonstrated the relationship between phosphorus (P) and human habita-tion and located many prehistoric sites (Provan 1971:37). The region of

Landscapes of Power, Landscapes of Conflict - [PDF Document] (207)


Scania was the object of his early research, as the database was produced for the Swedish Sugar Corporation during inquiries into the chemical content of beet-growing soils (Arrhenius 1934). The purpose of the study was to intro-duce modern agriculture to the farmers of Scania, who had previously not used artificial fertilizers but followed the old fallow system they had used for many centuries. Therefore, the data produced by Arrhenius between 1931 and 1934 was procured from soils uncontaminated with artificial additions of chemical fertilizers. Arrhenius’s study contained careful analyses of the levels of natural phosphorus, nitrogen, potassium, chlorine, and other agriculturally important chemicals in soils, from parent material and natural plant commu-nities. Arrhenius also examined soils cultivated only between the Swedish land reforms of 1810 and the present (1931), where he found no accumula-tion of phosphate in that 100-plus year period.

To collect data, an enormous ambitious field project was carried out be-tween 1931 and 1934, the scope of which would perhaps be impossible today. The scientific data that it produced are compatible with modern re-search methods, so it is invaluable to the study of the area’s prehistory. Some 500,000 soil samples, representing nearly all the agricultural land in Scania, were taken in a systematic grid of one sample per hectare. The product of this tremendous project was a map, a small part of which is reproduced in (Figure 7.5). Using the citric acid method of phosphate extraction that was the only available method in the 1930s, Arrhenius discovered that the Scan-ian background level of phosphate is between 1 and 25 ppm. Higher natural levels were found in a few areas due to phosphorus-bearing bedrock or to high organic matter levels of former swamps and bogs. These areas ranged from ca. 50 to 75 ppm. The map also showed large, discontinuous concen-trations of phosphates in areas outside current human habitation, concentra-tions too high to be natural, containing 200 to 900 parts per million or more. These soils turned out to be human-made, the result of centuries of living on the same spot, accumulating garbage and wastes, which are extremely high in phosphate. The manuring of the infields of farms also elevates phosphate levels. These artifical, human-produced agricultural soils are archaeological features which are called plaggen soils. One cow produces 1 metric ton of manure per month, and this manure contains 100,000 to 150,000 parts per million P2O5 dissolved in citric acid.

On the phosphate map of Scania the concentrations are often, but not always, close or adjacent to the historical village sites. Based on one or two known trials, Arrhenius predicted that these would prove to be archaeologi-cal sites and saw the map as “giving extraordinarily worthwhile indication ofprehistoric settlement” (Arrhenius 1934). While high phosphate levels are present at Neolithic and Bronze Age sites, these sites are usually small—less than 1 hectare. Medieval sites produce phosphates, but unless abandoned

Landscapes of Power, Landscapes of Conflict - [PDF Document] (208)


Figure 7.5. The Arrhenius phosphate map oftheJärrestadregion (afterArrhenius 1934).

they are underneath the current villages, as they are related to refuse in the immediate area of the house. As sedentary villages in Scania are believed to have first come into common occurrence during the Iron Age, it is not sur-prising that most sites between 3 and 40 to 50 hectares have usually proved to be of Iron Age provenience, while smaller areas belong to earlier times.

Because of the existence of the phosphate-map, phosphate is routinely used in Scania as a palimpsest for finding archaeological sites as well as for within-site analyses. Although not every site is visible on the phosphate map due to soil conditions and mapping extent, the map accurately identifies the location of many sites and suggests the pattern of location for finding others. In a synthesis of many previous investigations, Callmer (1986) presents a map and table of 78 sites in Scania where excavated late Iron Age settlements or cemeteries are associated with these high-phosphate areas. Callmer had al-ready demonstrated that the 40 known Scanian Viking Age treasure hoards liewithin or adjacent to them as well (Callmer 1980). This did not include vil-lages with Germanic Iron Age–Viking Age names and/or early churches that have immediately continuous phosphate areas, which should be added to the

Landscapes of Power, Landscapes of Conflict - [PDF Document] (209)


database of phosphate sites with clear Iron Age provenience. In most cases, the distance between the Iron Age villages and the early medieval villages is from 0 to 250 meters away (Callmer 1986). The Mechanics of Soil Phosphate

Although the use of soil chemistry as an aid for archaeological interpretation is not new, it has only recently come into wider use, and although the methodology has been the subject of several favorable critiques in recent lit-erature, it is by no means prevalent outside northern and western Europe. Variation in site formation processes and soil conditions, which have led to conflicting reports about what one may expect to find using phosphate test-ing and how it can be used, has deterred many American archaeologists from using soil chemical methods. This is unfortunate because soil chemical data, when properly emplyed, provides data that can be used in conjunction with conventional studies of artifact distribution, structures, and activity areas. However, it is also a boon to those who work in survey situations (Bethell and Maté 1989:21), especially under landscape conditions that are not con-ducive to the preservation or exposure of conventional archaeological fea-tures. Although in Mesoamerica and other arid regions archaeologists may crunch thousands of potsherds underfoot, in more temperate areas material is often buried. In a sense, soil chemistry transforms the invisible into the vis-ible, and allows archaeological analysis of regions formerly thought difficult or impossible to survey.

Anthropogenic chemical enrichment of soils occurs due to several fac-tors: it can stem from the refuse and wastes of human groups that accumu-late beneath their living sites; it can represent the product of animal husbandry in barns, pens, and on livestock paths; or it can represent the in-tentional enrichment of soils for agricultural purposes (Eidt 1984:29–30). En-richment is variable in all these cases, dependent on the intensity and duration of human occupation or activities (McDowell 1988:247). The utility of phosphate analyses will not be explored here, as it has been amply argued elsewhere (Provan 197 1; Sjöberg 1976; Bakkevig 1980; Eidt 1984; Proudfoot 1985; Bethell and Mate 1989). Existing Phosphate Data

The region and sites that are the focus of this project are shown in Figure 7.6. This map is adapted from Arrhenius and represents the testing done in the 1930s. Only the two highest concentration levels should be considered to be the actual site area; the third and lowest category represents the fields around the site core and are included to give a sense of the village territory. These

Landscapes of Power, Landscapes of Conflict - [PDF Document] (210)





















189Figure 7.6. Phosphate map of the Järrestad region showing all P-areas found during the Arrhenius 1934 survey (after

Arrhenius 1934).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (211)


sites all lie along the course of the Tommarp River and in its immediate vicin-ity. Phosphate patches believed to be of Iron Age and Viking Age date are found adjacent to most of the old villages, with the exception of two villages of Viking Age provenience, Karlaby and Gnalöv; the area around them was not sufficiently tested in early phosphate surveys.

Arrhenius’ 1934 phosphate maps were constructed from samples from within the plow zone, at a frequency of one sample per hectare. As already noted, the older research used the citric acid method of laboratory extrac-tion, which is still an acceptable method, and in modified form is still used today. The results of Arrhenius’s work in the 1930s can therefore be com-pletely integrated into modern datasets. The ppm in Arrhenius can be mul-tiplied by about 6.2 to get the equivalent ppm from the extraction method used in this study.

Local conditions for phosphate testing and surface collection were ex-cellent, as might be predicted from the successful testing and survey of other nearby archaeological project areas like Ystad, Kristianstad, and Hagestad. Very high phosphate levels and clear stratigraphic cultural layers containing several types of artifacts were evident in the test units. Variation in soil chemistry was apparent not only horizontally, permitting the evalua-tion of the spatial extent of the site, but also down-profile, enabling the ver-tical changes to be monitored. Along with other indicators such as demographic and settlement trends, this augments the modeling of change in area of occupation over time. For the most part, the Jarrestad area is a plain with gentle undulations, which decreases the complexity of soil chem-ical stratigraphy. It also minimizes the possibility of errors in establishing contemporary prehistoric living surfaces, which might occur in sloping landscapes. The relative recency of the time period under investigation (ap-proximately 1300 to 900 years BP) minimizes the possibility that large-scalegeologic changes have occurred in the area, and local conditions such as draining, filling, and plowing are the primary types of disturbance that could be found at some sites.

7. I. 1.5 Visual Inspection and Surface Collection: Ceramics in Viking Age Denmark

Ceramics that are generally diagnostic were collected during the Jarrestad survey and used to date the sites tested in the project. The Roman Iron Age is represented by well-made and meticulously decorated wares. The quality of local ceramics deteriorates rapidly in the Germanic Iron Age (400 to 700 A.D.). This change correlates with the significant changes in political organi-zation and political economy during that period. It is probable that locally

Landscapes of Power, Landscapes of Conflict - [PDF Document] (212)


made ceramics went from being important carriers of status messages to being mostly utilitarian, and that status marking was transferred or limited to other mediums. In the late Germanic period and the early Viking Age, localceramics continued to be mostly very crude and poorly made wares, and im-ports from western Europe became status markers. Variation exists in locally made wares synchronically and diachronically, and imported wares are found in several types with various origins. The distribution of local and nonlocal wares may indicate patterns of access to utilitarian and luxury imports, as well as indicate patterns of trade in Northern and Western Europe.

The Slavic connection is of primary importance in Scania and the east-ern parts of Denmark and is directly related to political relations in the Baltic region. Before 970, the eastern Danes traded heavily through Russia, where the Swedes had founded such large cities as Starja Ladoga, Novgorod, and Kiev (Tones 1987:246). After 970, the Russian routes became less productive when Arab silver dwindled. At that time, secondary routes through the Slavic lands were utilized (Jones 1987:265). At just this time, a political alliance is recorded in the RFA; King Harald Blåtand of Denmark married a Slav princess, Tova, daughter of Prince Mistevoj of the Wends, a tribe of Slavs in-habiting the southern Baltic area (Jones 1987: 127). Harald’s son, Sven Fork-beard, also married a Slavic princess. This Danish–Slavic alliance lasted until the 1030s, when the Wends and the Danes became adversaries.

On Sjælland and in Scania, crude, locally-made ceramic wares were pre-sent until about 900, when superior wares appeared. These flat-bottomed,biconical vessels were in some cases imports, but more often local imitations inspired by imported Slavic pots (Roesdahl 1982:123). Often called Baltic or Vendisk (Wendish) ware, these Slavic-style vessels are superior in craftsman-ship and firing. The imported Slavic ware was turned on a wheel, finer-grained, nicely decorated, and well-fired. It also often bears a mark on the base that may be a potter’s mark. The imitation Baltic ware was also more carefully crafted and decorated. Baltic pottery was not only limited to Scania but occurred nearly exclusively after 970 on Sjælland, replacing earlier coarse wares on the island as well. It is also seen on Fyn in large and small sites, and more distantly on Jutland, here occurring mostly in elite contexts. Documentary Evidence: Cartographic Sources

State archives preserve a sequence of maps of Järrestad’s landscape, and it isworth the while of any archaeologist to examine them all, even the more re-cent ones, as the landscape changed very little between the Late Viking Age and the 19th century. Nineteenth-century maps are especially useful in de-termining the location of wet, boggy areas that are left off earlier maps and

Landscapes of Power, Landscapes of Conflict - [PDF Document] (213)


no longer extant on modern ones. These are not only communication routes and resource areas but the probable location of pre-Christian offering places.

State-sponsored mapping projects are usually prompted by the need to record major, region-wide changes. The most recent landscape change oc-curred in the period between 1810 and 1825, when Scania underwent a land reform (Pred 1986). Land had been held for so long in village communities that over the centuries it had been divided and redivided until holdings be-came strips no wider than a plow, and one farmer might own many hectares, which were divided into 50 strips, near and far over the village territory. The state re-allocated the land so that everyone wound up with the same amount they had begun with, but all together in one place, and broke up the villages by placing the individual farms out in the landscape, away from the village center. A regional reconnaissance for the purpose of mapping the whole province was undertaken early in the century and completed in the 1820s. Since Järrestads Härad was administratively remote, its villages had not yetbeen broken up at the time of mapping, which is evident both from the ad-ministrative record and from looking at the map. Each field of a farm in a vil-lage had a numerical designation—i.e., Järrestad 14:1—the village name, the farm number, and the field number. These were first recorded in the 1660s.

The prior, and earliest mapping event, occurred at this time. In the 1660s, Scania, which had been a province of Denmark since the Viking pe-riod, was ceded to Sweden as a result of war (Oakley 1972). At this point, the Swedish state requested an “inventory” of the villages and cities of Scania be-cause it was the best agricultural land in the region, and its productivity and taxability were as important in the 17th century as in the Viking Age.

A very large early modern cartographic project was begun that spanned the late 17th and 18th centuries. At this time, the abovementioned farm num-bers were assigned on paper, although these were presumably adapted from an earlier system. They remain in use today. Through these numbers and the names of farms, the history of villages can be intimately followed where doc-umentation is available. Maps of most of Järrestad’s villages from this time are extant and filed in the Länsarkiv in Kristianstad, where I worked with the originals and obtained copies of them.

7. I. I.7 Documentary Evidence: Pre-Cartographic and Prehistoric Settlement Indicators

No cartographic record earlier than the 17th century exists for any part of Scania. To get at settlement before the 17th century, other material must be used. Before 1660, settlement location appears to have been quite stable from at least the Late Viking/Early Medieval transition until the Rennaissance (from

Landscapes of Power, Landscapes of Conflict - [PDF Document] (214)


about A.D. 1075 to 1660). One of the best methods for understanding settle-ment continuity is through the Romanesque churches of Scania, which were built in the decades around 1100, and for the most part are still standing and used. These stone churches were preceded by wooden stave churches in the same locations, built between A.D. 990 and 1000, when a spate of church building is recorded in written records. A number of investigations in stone churches have yielded the remains of these earlier wooden structures (Pet-tersson, personal communication 1995). Adam of Bremen, one of the most reliable primary sources for the 11th century, wrote in 1070 that there were 300 churches in Scania alone, so we can infer that the location of the late Viking Age village coincides with the church. These churches served as vil-lage “anchors” during the next 1,000 years.

The villages and towns are also documented in church missals and tithe records, tax records, landbooks, and royal decrees from the late 12th century onward, and one can follow their names, and the changes from archaic forms to modern, with some clarity over several hundred years (Strömberg 1976;Kousgård-Sorensen 1979).

Additionally, Norborg (1990) presents a reconstruction of road systemsin Scania at ca. A.D. 1000, which shows Tommarp at the intersection of two important east-west and north-south roads, the former one linking Tommarp and Simrishamn along the general course the modern road takes. This would indicate that the route was through the villages as it is today. The villages that were extant in the late Iron Age were probably linked by smaller roads, and these can be filled in by connecting the dots. Place Name Categories: Regional, Provincial Village, and Parish Names; Farm Names, Field Names, and Terrain Names

Place name investigations form a substantial and important part of research on the prehistoric and protohistoric eras in Scandinavia. Although many studies of Danish place names have been undertaken and are generally applicable to Scanian places as part of the Danish realm, Scania has been scrutinized in sev-eral cases, either through specific references to Scania in more general works on Denmark (Kousgård-Sorensen 1968, 1979; Hald 1950; Skautrup 1944) or in specific studies of Scanian etymology (Pamp 1983,1988; SkO 1958-1993; OÅ 1928–1995). At the University of Lund the Dialekt och Ortnamnsarkivet (Di-alect and Placename Archive ), abbreviated as DAL, is devoted to the study of place names in Scania, Halland, Blekinge, Småland, and Öland, which consti-tute southern Sweden today. The archive has identified every written refer-ence to place names from each known document in Danish and Swedish

Landscapes of Power, Landscapes of Conflict - [PDF Document] (215)


historical archives, as well as foreign references in early times, and every name is filed under its respective province and Härad. Thus one can find each men-tion of a certain village from the 11th century or earlier through the Middle Ages, the Renaissance, and the more recent era. One can find the name of a single agricultural field, or a hill or gully, and trace its records through several centuries. Furthermore, the linguistic research is extensively correlated with archaeological material in Sweden, Denmark, Norway, and the Norse colonies in England, Ireland, France, and elsewhere. Research into place names at the DAL revealed that Järrestad has a rich supply of documents preserving ancient names of all types that aid in landscape reconstruction. “National” and Regional Place Names

Several layers of place names relate to pre- and protohistoric sites in Scania and Järrestads Härad. Through them, an idea of the origin and function of differenttypes of sites can be examined. On the broadest level, Scania itself, or Skåne as it is spelled in Danish, appears to mean “a land wholly or partly surrounded by water.” In Beowulf, a poem from the Germanic Iron Age oral tradition written down during the Dark Ages but dating to perhaps 200 years earlier, it is men-tioned as Scedinigge. The Franks called it “Scanowae” in the 800s; in Ohthere and Wulfstan’s accounts to the Anglo-Saxons in 890 it is Scóneg; and Adam of Bremen called it Sconia. Later, the Germans called it Schönen, which means “fine” or “beautiful” but has no direct relation to the origin of the word itself. Internal Administrative Place Names

Although based on the Hundreds of the RIA, the division of Denmark, in-cluding Scania, into Härads of the larger state is believed to have taken placein the late 10th century, from around 980 and onwards, beginning in the southwest of Scania and moving slowly to the east and north.

Based on the timing of the establishment of royal towns and the men-tions of royal donations of land to the church in the east and northeast, it may have taken until the mid-1lth century for real control to be extended into theJärrestad region, and as long as the late 11th century to take root in thenortheast quarter (Anglert 1995:46–47). Härad Names

Was Järrestad a “region” in prehistory, or is it a modern unit? We can get clues from the names of Härad divisions, first assigned by the Danish Crown in thelate 10th century. These were made in two distinct waves of assignment, one

Landscapes of Power, Landscapes of Conflict - [PDF Document] (216)


of primary names, one a secondary naming. Slightly confusing is that the “secondary” Härads are earlier than the “primary,” as these terms refer tothe name-types, not the time they were incorporated into the state. In the northeast and northwest, Härads appear simply to have taken over older dis-trict names, direct impositions of an outside power on still-relevant units (Anglert 1995:43–45). Those in the southwest and southeast are, in contrast, named after a village or other type of place in the Härad. The former seem to have been named with words for existing, meaningful cultural units, while the latter are named for their administrative central places. Järrestad falls into this second category of Härads.

The Härads in the southeast and southwest of Scania were the first to be annexed to Denmark. Thinking back to chapter 3 it will be recalled that this entire area is believed to be the prehistoric “kingdom” of Scania, which was made up of many earlier polities united into one administrative and po-litical unit at some unknown point in the past, presumably between the end of the Roman and the beginning of the Viking Age, with a medieval folk tradition recording the names of some “kings” of Scania in the period before

Linguistic evidence for this is as follows: all the Härads within this areaare named after villages or localities within them. It appears that any pre-Scanian division names (not boundaries) were long-lost by the time the Danes annexed the area, and so the Danish state government picked the most significant or administratively appropriate locale within each area and named the Härad after it. For example, in five of the Härads discussed in chap-ter 7 there are villages called Torna, Bara, Stora Herrestad, Ingelstad, and Jär-restad, after which the districts were named. This is referred to as secondary naming. These areas would have formed the “districts” within “Scania” that in turn are likely to represent earlier settlement units from the RIA. The names of these RIA units had long been forgotten, but their boundaries are still ob-servable in terms of settlement patterns.

The primary named Härads were not annexed into Denmark until asomewhat later time. These, in contrast, retain meaningful district names that were then used as Härad names. These smaller districts retained their relative autonomy from Scania and then Denmark for a longer period. Finnveden and Fjäre, Bjäre and Luothida are all mentioned as discrete po-litical units in Jordanes in the sixth century; these are also primary Häred names. They also form part of the boundary between the primary and sec-ondary Härad divisions.

In chapter 6 it was noted that the Theissen polygons around the prestate era sites were similar to later Härad divisions, suggesting that the units were much older than the state and might be earlier boundaries. In

A.D. 800.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (217)


secondary Härads like Järrestad, these old divisions were given new names by the central government in the late 900s when the Härads were formed.Järrestad was created as a “Härad” of the state in the late 10th century (ca.980) at which time it was named after its own internal central place. The earliest written reference to the Härad as an administrative division is veryearly as written records go: in 1182 it was first recorded as Ierestedt Herrit (DD1R 3 1:171). Parish Names

Parish names in Scania are associated with the name of the church village and its outlying settlements. Villages in Scania are divided into church villages (kyrkbyar ) and non-church villages. The church villages are usually the older villages, though there are some exceptions, while the newer “torp” settle-ments that represent expansion during the late Viking Age A.D. 980 to 1100 are often lacking a church. Village Names

Village names that contain certain linguistic elements can be well-linkedwith specific periods of settlement foundation in Scania and the rest of Denmark. Although some later settlements have old-sounding names, these can be distinguished quite easily through historical research (Kousgird-Sorensen 1979).

It can roughly be stated that places with the suffixes –löv (–lev ),–stad(sted ),–inge, -lösa, and –ie (which is a corrupted form of hög or høj ormound) date to the period A.D. 500 to 700, while those ending in –by and–berga are from the Middle Viking period, ca. A.D. 850 to 980 (Kousgård- Sorensen 1978; Emanuelsson et al. 1985:37). –Tofta, –arp, or –torp arefrom the Late Viking Age, 980 to 1100; –torp and –arp names have mean-ings that actually refer to the founding of new settlements while tofta in-dicated the regulation and restructuring of the farmstead in late Viking times, as discussed in chapter 5. Names with ryd or röd, which refer to a new röyning or clearing, often date to the earliest Middle Ages ca. 1100 to1200, and the element brott often refers to the breaking of new land in thelater Middle Ages.

The development of villages and towns can be followed through their names (Stromberg 1976; Kousgård-Sorensen 1978). Villages and towns that became extinct, such as Gårdsköpinge on the Kristianstad plain in northeast Scania (Callmer 1983), or greatly diminished, such as Vaä a city that was burned in 1612 and continued on as only a village (Thun and Anglert 1984),

Landscapes of Power, Landscapes of Conflict - [PDF Document] (218)


and Tummatorp, a large city whose functions were transferred elsewhere after the high Middle Ages and also shrank to a village (Thun 1967; Redin 1972), can be documented too.

Between A.D. 500 and 1100, place names connoting new settlements support the pollen and archaeological records by further indicating a period of settlement expansion, with many village foundations. The earlier-namevillages (A.D. 500 to 700) are on the best land, close to waterways and hay meadows, while the later-name villages (A.D. 800 to 1050) are established in the interstices. This suggests a period of slow but steady population growth and budding, rather than village location change (Emmanuelsson 1985: 37). Callmer (1986) has demonstrated that Scanian villages rarely moved their location in any significant sense during the late Iron Age, and those that did can usually be detected and explained, such as the abandonment of sev-eral coastal villages in the 900s and their merging with nearby inland vil-lages. This was likely due to increased piracy and slaving by the Slavs during this time.

Older villages with names in stad, lev, inge, by, and so on are normallymuch larger and often are the church villages. About 1,000 villages in Scania have names ending in torp. About 70% of these are settlements with no more than five farms, contemporary with a church expansion that began in the west and spread east (Anglert 1995:27), and thus date to the 980s through the 1000s. Others are single farms, associated with continuing torp expan-sion into the Early Medieval period. Torp villages that today are very large are the result of modern growth rather than growth in the timeframe of this study. Aside from placenames, archaeological survey and excavations at many village sites have further clarified these phases of origin.

Within the Härad of Järrestad, there are place names from all periodsof settlement foundation. Of the oldest place names of common type— löv/lev meaning “inherited or bequeathed property” and stad meaning “seat” or “stead”—we have Gislöv, Vemmerlöv, Gnalöv, Virrestad, and Jär-restad dating to the A.D. 500 to 700 time period, and of the later settlement wave, with endings of by meaning “new village” and dating to the period of 850 to 980, there are Karlaby and Stiby. From the third settlement wave, dating to the expansion of new settlements between 980 and 1100, with names ending in arp or torp, meaning “new settlement dependent on an-other,” we have villages that include Tummatorp (Tommarp), Tågarp, Viarp, Vranarp, and Gröstorp (Kousgård-Sørensen 1979). Nobbelov seemslike a löv name at first glance but is a corruption of the name Nybølle, which is younger, contemporary with the torp phase. The cases of Simris and Gladsax are “unique”: types of placenames that do not fall into any typical categories. Documentary records from the 12th, 13th, and 14th

Landscapes of Power, Landscapes of Conflict - [PDF Document] (219)


centuries include all the villages in the Härad and show that they are much the same as they are today. Farm and Field Names

Farm names can be helpful in dating and assessing the type of the early vil-lages. While some much more modern farms have old-sounding names, it is not difficult to research them. For example, in the late 19th and early 20th centuries, it was in vogue to name farms with the medieval element “toft” or “tofta.” These places are nowhere to be found in the mid-19th century or ear-lier, and usually make their first appearance in the pastorial rolls of 1904 and 1909. They can thus be assigned the last quarter of the 19th century. On the other hand, a name such as “Gaartofta” (Farm-toft) in Stiby village, JärrestadsHärad, A.D. 1570, or Viarp village’s“Hörlands Tofta” (Toft in the flax fields) A.D.

1701, are likely to indicate a toft dating back to the early medieval or late Viking Age.

Field names too can give clues to appearance, usage, and ownership.Fields containing the element “hall” refer to a bare rockface in the field, such as those where petroglyphs are often inscribed. “Kiille,” or modern Kälde, refers to a spring; Hesthaga to horse meadow; Hörekiilleåker to flax-spring-fields—afield near which flax was soaked and retted in a water source. Others tell of ad-ministrative use. “Hovgaard” refers to the “main” farm, or that owned by the local magnate. “Arv” refers to the inheritance of the area under question, but is often used in association with the gift of land from the king to some retainer or follower. “Arvlund,” a place found in Järrestads Härad, would be a whole small forest, its game and products bequeathed to one person. The element ‘‘präst” or “kyrk” or “munk” combined with åker, äng, gård, or berg (field, meadow, farm, or hill) usually means that the land was owned by the church. A tithe of one-tenth was imposed in Late Viking/Early Medieval times, and the land in Stiby vil-lage called Kircke tiende agirenn (Church-tenth-field) in A.D. 1570 says that thisfield was jointly cultivated by the village to pay the tithe. Terrain Names

Besides naming every field in the village, Scanian farmers named hills, valleys, streams, islands, springs, and any other geographic feature that could be perceived—even bumps in the ground. Some of these names are not very helpful in interpreting the past land use history. For example, the names Torn-dala—thorny dale—or Djupdal—deep dale—are found in nearly every vil-lage. Others have useful meanings: Grodstorps Dige, a dike that formed a boundary between two villages, or Korsse Wad, a ford where one can wade

Landscapes of Power, Landscapes of Conflict - [PDF Document] (220)


across the river. Some can be more revealing; for example, names that con-tain the Old Norse bäken (beacon) found in terrain names like “Bågenhög”(beaconhill) probably refer to high places where signal fires or vårdkasewere lighted to call the ledung, or the naval levy, in the late Viking Age (Hall-berg, personal communication 1995).

More importantly, terrain names give insight into the deep and long-lasting sense of place in the lives of Scanian farmers. Allen Pred in his book Place, Practice, and Structure (1986) first raised the idea of the conceptual landscape of everyday existence for Scanian farmers of the 17th and 18th centuries. He discussed the idea of a built, internalized cultural landscape that defined the course of day-to-day life and formed a part of the discourse between humans and their environment. His research concerned the 19th-century government restructuring of agriculture, entailing enclosure, the breaking up of villages and moving out of farms, and redistribution of land within the village arable. He proposed that this might wholly alter an ancient dialogue between humans and their lands. Although this may sound too phenomenological for many archaeologists to swallow, as research brings one to pore through thousands of words for every bump, dip, and mudhole in a farmer's land and one sees the same bump with the same name in A.D.

1300, 1400, 1500, 1600, and up through 1850 or so when farming was re-organized, the idea does not seem so strange.


The fieldwork done ut på landet, out on the land or in the countryside, to re-construct the cultural landscape of Järrestad, consisted of phosphate survey,surface artifact collection, the excavation of test units, and the study of the documentary record. The data discussed primarily consist of new investiga-tions, recovered in the field in 1992 – 1993, and landscape reconstructions modeled by myself through this fieldwork and archival research. The work of Arrhenius and others will be discussed in relevant context.

Sixteen Iron Age villages are assumed to have existed in the area, based on phosphate patches and associated late Iron Age place names, their place-ment in the landscape, and the proximity of surface finds and graves from previous work in the region. Two of these are in small areas that were only partly or not at all tested by Arrhenius in the 1930s. The site of Karlaby was in an area with large gaps in testing, and only a part of it shows in the 1934 map. However, it was given site number J-9 since it is represented as a phos-phate patch. The size of the Early Medieval village is known. A similar occur-rence is found at the site of Gnalöv near Gislöv. In this case, no phosphate

Landscapes of Power, Landscapes of Conflict - [PDF Document] (221)


testing was done in the site area and there is no record of its prehistoric size. Its early historic size is known. No testing was undertaken around these two site areas where nothing is known of the extent or position of the prehistoric site because this would have involved hundreds of hectares. The remaining 14 sites are seen in the phosphate map where their size (5 to 40 hectares) in-dicates they are later prehistoric sites.

There are also a number of smaller phosphate areas that were assumed to date to earlier eras. They are usually under 3 hectares in size and often lo-cated on the first ridge above a water source or on stony uplands. Arrhenius tested a number of these small sites and found that they often dated to the Neolithic and Bronze Age (Arrhenius 1934). I tested four of them, and came to the same conclusion.

7.2.1 Sampling Strategy

Because the population of “all late Iron Age village sites” is known with some security, the method of stratified sampling was chosen for the project (Read 1975:58–9; Flannery 1976). Since the sites are variable in size and probably function, it was necessary to select different classes of sites for different in-tensities of investigation. Seven of the supposed Iron Age sites were investi-gated, including large, intermediate, and small villages. In addition, four of the smaller areas were investigated in order to verify the assumption that they could be excluded as belonging to earlier time periods.

Through the study of numerous excavations and surveys in nearby re-gions (Stromberg 1981b, 1985; Stjernquist 1951, 1955, 1981a, 1994; Callmer 1983, 1984, 1986, 1987; Ohlsson 1976, 1980) it is clear that larger sites are usually functionally larger (refer to chapter 6), and more likely to display vari-ability. It was thus decided to survey more larger sites and fewer intermediate and small sites.

Furthermore, the larger sites are usually found to be older (Callmer 1987), to have place-names of the oldest type, and to be located near the best soils of the valley bottom. Most of them are about 2 kilometers apart. The younger, smaller sites with later place-names are dotted (1) in between the older sites; (2) in less fertile areas; or (3) in upland areas. Since some stratig-raphy both in archaeological features and in soil chemistry was desired, older sites were more likely to produce such data.

The number of villages, their sizes, and the size-classes they fit into were assessed by myself from the phosphate map of Arrhenius, based on the site-sizedivisions described in Callmer (1986) and discussed in chapter 4. Coverage of the Härad by the 1934 phosphate testing is almost complete. Villages to be in-vestigated and those to be left out were chosen randomly. Seventy-five percent

Landscapes of Power, Landscapes of Conflict - [PDF Document] (222)


or three of the four large villages in the entire survey area were thus included in the survey: the villages of Gislöv, Simris, and Järrestad, leaving Vallby out.Fifty percent of the intermediate-size villages were tested or, two out of four: Gladsax and Gröstorp, leaving out Stiby and Ö.Vemmerlöv. Thirty percent ofthe small villages were selected or, two out of six: Viarp and Ö. Nöbbelöv were investigated in the field; Vranarp, Tågarp, Bolshög, and Virrestad were not.

At Karlaby and Gnalöv, which were not surveyed for reasons mentioned earlier, there are records of these villages names in the Early Middle Ages and their approximate site size at that time is known. Karlaby was an intermediate-size site, similar to Ö. Vemmerlöv, while Gnalöv was very small, similar in size to Bolshög and Ö. Nöbbelöv.

In other nearby regional projects (Figure 7.1) the reliability of the older phosphate maps has been established as being excellent (Tesch 1993; Callmer 1980, 1987). This was borne out in the Järrestad region, so existing site size data could be used to incorporate the remaining, un-resurveyed sites into the statistical locational analysis. Generalizations would mostly be limited to the smallest sites of Iron Age provenience, where it was expected and ver-ified that farming was the primary activity of the inhabitants.

Two important sites were not tested due to their size or the extent of modern settlement. Tommarp, or Tummatorp, as it is called in late Viking Age sources, is an urban center of the latest Viking Age and Early Middle Ages. This is one of the four royal foundation towns discussed in chapter 6. It was largely abandoned by 1500 and is now a small village atop a buried pre-,proto-, and historic site. It was not within the scope of this project to exca-vate or test an urban center, but the extent of the site is clear in Arrhenius’s phosphate map, and parts of Tommarp have been intensively investigated by the Institute of Medieval Archaeology at Lund and its relative extent of occu-pation in the Viking Age and Medieval period is known.

The coastal harbor of the mouth of the Tommarp River today lies under the city of Simrishamn, which has a population of about 30,000, and also could not be tested. But Simrishamn has been the site of a number of RAÄ ex-cavations. As it turns out, there appears to have been little or no habitation at this coastal site in the Iron Age; only a few pits and small middens of Late Viking Age date were found there. This follows the pattern for Scania: harbors are dotted with a few pithouses but presumably there was too much piracy to permit large, prosperous villages on the coasts.

The area of the sites to be re-tested was determined by using the exist-ing phosphate map of Scania and transferring the approximate site bound-aries to the most recent economic maps of the region, maps that denote the location of all buildings, fences, utilities, and other minutiae, and the prop-erty numbers, through which the landowners can be contacted. Figure 7.2

Landscapes of Power, Landscapes of Conflict - [PDF Document] (223)


shows all phosphate areas in the study region as they were mapped by Ar-rhenius. Sites J-1, 2, 3, 4, 11, 12, and 13 are the sites that were re-tested; sites J-5, 6, 7, 8, 9, 10, 14 and 15 are sites of GIA and VA provenience not re-tested;and sites J-16 through J-19 are sites that were re-tested and found to be of Ne-olithic and Bronze Age date. Tummatorp, Simrishamn, and Gnalöv are de-noted by name only.

7.2.2 Data Collection

As surface collection was a large part of the strategy, fieldwork was con-ducted in the fall of 1992 and the spring of 1993 in order to coordinate with the scheduling of various crops, growing, and harvest seasons. In fall and spring, fields were either just harvested, just plowed, or newly sown. This provided excellent visibility at most of the sites. Some of the fields remained fallow in both of the field seasons, covered with dense vegetation, and could not be surface-collected, but were soil-sampled. Soil data from these fallow fields, which were interspersed with plowed, collected fields, could be fitted into the pattern of each village.

7.2.3 Survey Methodology

A methodology was developed for the survey that consisted of three steps: the collection of soil samples from soil cores taken with an Oakfield coring tube, controlled surface collection, and hand-excavated test units. Where pos-sible, interviews with landowners regarding their own, their parents’ and grandparents’, and sometimes their neighbors, finds were conducted.

7.2.4 Soil Core Collection

Existing field divisions from the economic maps of the study area, with num- bers assigned by the Swedish state, were used as surface survey unit divisions in all cases and were transferred to centimeter paper field sheets. These were then taken into the field where notes regarding soil conditions, moisture, and visibility could be noted. The profiles from test units were also recorded on these sheets, as was the distribution of surface artifacts, the collection of which is discussed below. A datum point based on a fixed location in each field was established. The properties were then gridded in 25-meter intervals using tapes.

The grid was marked with flags, and the soil cores were taken at the gridpoints. Fields ranged in size from very small—50 x 50 meters—to large— 300 or more meters in length. An average size was 200 x 150 to 200 meters,


Landscapes of Power, Landscapes of Conflict - [PDF Document] (224)


necessitating eight to nine transects with six to eight cores each, and each core producing three to four samples of 10 centimeters each. Each point in the grid was given a number based on the transect number (T) and the core number (C), and the depth below surface (i.e., 30 to 40 cm). The plow zone was discarded, and soil sample collection began in the upper culture layers.

Furthermore, the first insertion of the core was recorded as “ 1 ” and each division of the core also recorded—as 1.1 , 1.2, 1.3, and so on—based on a 10-centimeter arbitrary division, unless clear breaks in stratigraphy were noted. The second insertion would be recorded as “2 ,” and include 2.1 , 2.2 , 2.3 , and so on. Sterile soil was usually reached at about 60 Centimeters below the sur-face, but there was some variation based on local conditions.

The soil samples were recorded with this system, including the field number based on the Economic Map. A typical core site would produce a provenience series such as Järrestad 14:1 / T2C4 1.1 (25–35), 1.2 (35-45), 2.1 (45–55). The soil profiles from the cores reflect the conditions in the field: the depth of the plowzone (typically 20 to 25 centimeter), areas of buried peat, sometimes the presence of springs or bogs below the surface, and the depth of cultural layers. Features could sometimes be observed by heavy concentrations of flecks of burned clay, rust, charcoal, and tiny bone fragments. This served as a constant check that allowed the exclusion of the plow zone, the feel of the depth and arrangement of the culture layers, the presence of features, and the depth at which sterile clay or sand replaced signs of human activity.

7.2.5 Controlled Surface Collection

The controlled surface collection method was used in those portions of the sites with a surface visibility of 10% or greater. This method utilized piece-plot provenience for the generally light scatters. Ceramics were the artifact best suited for dating the occupations, although metal and bone fragments were often associated with them. The relatively low density of ceramics is due to the nature of the materials and the depth of the deposits. The surface frequency of artifacts is completely analogous with other survey results in Scania for Iron Age materials (Callmer, personal communication 1993). This is because intact deposits begin at about 25 to 30 centimeters below the sur-face, and sometimes deeper. A plowzone, typically about 25 centimeters in depth, results in very little ceramic material making its way to the surface, unless individual farmers plow to a deeper level. The wares known as A4 wares that date to the earlier part of the Viking Age are not only deeply buried (45 to 65 centimeters on average) but are highly friable and can be crumbled easily with light pressure. Those that do wind up on the surface

Landscapes of Power, Landscapes of Conflict - [PDF Document] (225)


quickly disintegrate, leaving a disproportionate amount of A2 wares, harder ceramics dating to post-990. The surface collection of a whole village site might result in only 100 to 200 sherds, and a small site with two or three farms, even fewer. Test units, however, belie the very light scattering of sur-face fmds, and if placed into intact deposits, yield large numbers of sherds, metal, and bone.

The piece-plotted proveniencing utilized the same grid employed for soil core collection. Fields were gridded and flagged for gridpoints first, then survey was conducted by fieldwalking at 5-meter intervals, and arti-facts were flagged. When the soil samples were taken, any previously un-noticed artifacts were flagged at that time. Artifacts were plotted by calculating the distance from the nearest two gridpoints. A description of the surface visibility and contents of each grid unit was logged and mapped on the field sheet, and the artifacts were collected and bagged with their provenience after cores were completed. The results of mapping were ex-amined to identify probable activity areas, and in some cases, to redefine site limits. The maps do not indicate the age of the ceramics on the surface because very few were of the older A4 type, and these isolated sherds were randomly distributed among the Late Viking Age material. If clear concen-trations of artifacts were noted, these areas were chosen for test units. If no clear pattern emerged, the 1934 phosphate map was used and tests were placed in high phosphate areas.

7.2.6 Test Excavation Units

Hand-excavated units were the only means of subsurface investigation (be-yond the conditions that could be noted in the soil cores). The data gener-ated from the excavation units were considered in conjunction with the surface collection, to determine the stratigraphy of artifacts and phosphate distribution. For example, in field X the culture layer extending from 30 to 40 centimeters below the surface was correlated with A2 wares dating to 990 to 1050, while the culture layer from 40 to 55 centimeters below the surface correlated to A4 wares dating to ca. 800 to 950. In some fields, prior published test excavations from earlier projects could be compared with our findings. When the project had proceeded for some time, it could be generalized that certain depths below surface could be correlated with gen-eral archaeological phases, although we continued to test in each field where we could get permission.

A test unit was excavated in each field where such permission was granted. Unfortunately, while many landowners gave permission for survey and soil samples (1-inch-diameter probes easily fitted between rows of crops)

Landscapes of Power, Landscapes of Conflict - [PDF Document] (226)


they were not amenable to test units unless their fields were unplanted. Thus, test units were somewhat sporadic.

Test unit location was chosen, as mentioned above, based on artifact density, or in the absence of artifact density, on Arrhenius’s phosphate map. The intention was to find datable artifacts in situ, hopefully associated with features, enabling some discussion of the earlier and later occupations. For-tunately, in Järrestad and Gislöv, two of the three large sites investigated, ex-cellent sub-surface results were obtained. When it was necessary to choose sites for test pits based on the phosphate map of 1934, especially in Simris where there was virtually no material on the surface, this unfortunately usu-ally resulted in units without actual intact material or features. However, the cultural layers in the soil itself were clear, and conformed to the pattern ob-served in all the villages as test unit and soil core data accumulated.

The dimensions of the hand-excavated units were 1 x 1 meter square. The units were taken down until non-cultural soils were reached. The sub-soils were very clearly delineated, consisting of pure yellow or orange clays or sand, in agreement with the published soil and geologic data on the area. The placement of the units at each site is indicated on the site maps.

The plow zone was first removed in the excavation units. Subsequent levels were excavated in arbitrary 5-centimeter increments, as measured from the present ground surface. The presence of cultural and natural soil horizons was noted, and the contents of each were kept separate. Soil samples were taken from the walls of the units. All soils were processed through the equiv-alent of 1/4-inch screen. Excavation continued until a sterile 20 centimeters was recorded.

The contents, specific description, and general comments for each level were recorded on the field sheets. Upon completion of the unit, a soil profile of at least one unit wall was drawn and if features were noted, photographed. All notes, information, and associated illustrations were kept for reference.

7.2.7 Feature Investigation

The identification of a feature or activity area was based on a distinct con-centration or patterning of cultural debris, the presence of soil anomalies, or both. A soil anomaly was defined as a discrete area with a color, texture, or or-ganic content that was in variance with the surrounding matrix.

Features tended to be large, upwelling artifact distributions, which based on prior fieldwork in Scania could be hypothesized to be pithouses, that contain in their lower levels intact occupation materials, and in their upper layers served as dumps or middens when they were abandoned and other fresh pithouses dug nearby. The dumps and the houses are generally

Landscapes of Power, Landscapes of Conflict - [PDF Document] (227)


contemporary; these pithouses were not used for long periods of time (5 to 10 years). The features were usually not excavated because this would re-quire mechanical removal of the topsoil over large areas of the fields, a task outside the parameters of the survey project. Features were also noted in some test unit profiles, and in one case a pithouse was excavated. The de-scription of each feature and its excavation, if conducted, was recorded on a fieldsheet. The soil from each zone or horizon was bagged separately.


As previously discussed, individual farms were not common during the study period. The village as the unit of analysis can be defined by the much higher than background phosphate area, which in turn contains a number of farms seen as discrete even-higher phosphate areas with surface artifact clusters. Although farms within a community might be 50 to 100 meters apart, the older villages were 2 to 3 kilometers apart as a rule (Callmer 1986), while younger villages may lie 1 to 2 kilometers from other sites. These farms are clearly part of a village structure, although the village is more dispersed than modern examples.

Unlike Mesolithic and Neolithic sites tested by Arrhenius, where high phosphate is concentrated in .5 to 2 hectares, Iron Age villages range from about 5 to 50 hectares of high phosphate. They are also identified by their much more extensive field systems. The fields surrounding early sites are usu-ally only two times as large as the habitation area, while Iron Age plow tech-nology, cooperative labor efforts, and probably increased surplus requirements resulted in villages of 20 hectares having 100 hectares of plowed fields.

Random test units were judged not to be worthwhile: as mentioned ear-lier; one of the features of Iron Age villages is that they appear to be loosely organized, spatially spread out. Farmstead houses were not clustered together along a street until the early Middle Ages. The chances of finding positive test units (those containing artifacts) in the hectares that lie within the village but between the farms is poor. Though I would have preferred to do complete transects of test units across each site, this was not feasible.

Soil chemical characterization was used to determine the extent of the occupation area as well as areas of greater and lesser activity within the site, because phosphate values vary widely across the prehistoric occupation area in association with features such as middens, barns, and housefloors. Phos-phate combined with surface collection of artifacts has suggested a number of different patterns. In this study, they will be designated as artifact scatters, village areas, and off-site areas.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (228)


The first type of area is so-called because of the presence of dense arti-fact scatters, including bone, ceramics, metal fragments such as broken knives, rivets, nails, and unidentifiable scraps, plus broken pieces of quern-stones, whetstones, and spindle whorls. These areas are localized over small areas, usually no more than 250 square meters. These highest phosphate areas are the “tofts,” containing the houseplot and immediately surrounding area. Surface artifacts cluster where sunken floor outbuildings, refuse heaps, middens, and pits were located. These areas invariably have the highest phosphate content on the sites, from 1500 to 4000 pprn, with some areas reaching as high as 6000 ppm. On each site, there are a number of these ar-tifact-scattered, high-phosphate areas.

Another type of area includes what I have designated village areas. These are areas of high phosphate, from about 1000 to 1499 ppm, but low in sur-face artifact density. These are the parts of the village that lie between the farm tofts and their artifact scatters. Usually, a few potsherds can be found in these areas but not in identifiable clusters. Only values above 1000 ppm are considered to be actual habitation areas.

Finally, there are areas designated off-site, because the levels fall off sharply yet are still much higher than the local background phosphate level of 300 to 400 pprn or less. The large areas of 500 to 999 ppm surrounding the village, where almost no artifacts are found, represent the manured fields of the villages, or the vång areas. South Swedish archaeologists view values below about 100 ppm in citric acid tests to be off the direct settlement area. For the purposes of this study, 700 to 800 ppm, based on ICP testing, was the cutoff point for delimiting these outer areas, or about 125 ppm translated to citric acid test figures.

It should be noted that these areas are not areas of “plowed out” phos-phate from later activity because they occur in Scania at sites that have been abandoned and not farmed over, such as at Gårdsköpinge, north of Järrestad.These field areas should not be considered when assessing the size of vil-lages. Although Stiby, for example, may look larger than Simris, this is only be-cause of the extent of the fields as they are seen in the phosphate map. This represents the probable extent of the innermost fertilized fields of the village, the vång or blocks of fields surrounding the settlement.


Test units in Gislöv, Järrestad, and Simris, that produced datable artifacts in situ, indicated a general stratigraphy, seen in Figure 7.7. The cultural layer with no artifacts is identified by its dark color, its “fatty” highly organic

Landscapes of Power, Landscapes of Conflict - [PDF Document] (229)


Figure 7.7. Typical stratigraphy of Iron Age sites in the Järrestad region.

texture, and its high phosphate levels. There is no doubt that features and artifacts exist at this level, but they are rarely encountered. Callmer (per-sonal communication 1995), who has conducted numerous projects in Sca-nia, advised that this level represents an earlier phase of the villages, before A.D. 800, when they were more sparsely occupied and features were less dense. During this phase they can only readily be found by large horizontal excavation. The chance of taking tests or cores into the very earliest fea-tures when the village was relatively new are slim, although there is plenty of phosphate spread around. This is corroborated by the work of Strömberg(1976, 1985) and Stjernquist (1981a, 1993a, 1993b) in Järrestad, Gislöv, andGårdlösa.

The plowzone in the soil cores was discarded, and unless clear strati-graphic breaks were evident in the cores or profiles, samples were divided into 10 centimeter increments. Prehistoric layers usually began at a depth of

Landscapes of Power, Landscapes of Conflict - [PDF Document] (230)


about 25 to 35 centimeteres below the surface, but this was determined in-dividually for each site and area within the site by examining the depth of the plow zone in the test units and cores. Therefore, the upper 10 centimeters of “prehistoric” occupation was usually about 30 to 40 cm, but could vary several centimeters in either direction. This represents the top stratum of undisturbed prehistoric soils probably dated to between A.D. 980 to 1050, ac-cording to artifactual material.

Some 3,500 soil samples were taken in course of the fieldwork. A total of about 1,900 were processed for this study. Several hundred taken on sites that were clearly Neolithic or Bronze Age—J-16, J-17, J-18, and J-19—were not processed in the laboratory, but samples were taken and retained for possible future analysis and comparison between Iron Age and earlier settlements. Other samples were of the sterile subsoil, geological deposits of pure orange-yellow clay or white sand, of a type recorded in published geological reports for southeast Scania. About 1,400 samples made up the “upper culture layer” that determined the extent of sites in the mid to late Viking Age (980 to 1050) for all seven tested villages. Then, 500 samples from the deeper or “lower cul-ture layer” in the larger villages of Gislöv and Järrestad were also examined-dating to approximately 800 to 980—in an attempt to see if any change in site size could be seen.

Although features are sparse in the earlier phase, phosphate is not. At this earlier phase, phosphate was not so concentrated as in the later village phases, so it was hypothesized and borne out that phosphate survey is an ef-fective and inexpensive way to get at site size change when full excavations are impractical.


Soil samples were oven-dried in the laboratory for at least 24 hours. Then, 0.20 grams of well-pulverized soil were weighed into scintillation vials. Sam-ples were numbered and weights recorded. The extraction method is as fol-lows: 20 ml of 2-N hydrochloric acid are added. The samples are then left to digest for two weeks cold, and are shaken to stir the sediments and promote digestion about every other day. After two weeks the samples are strained through filter paper into appropriately numbered test tubes. The filtering pre-vents damage to or clogging of the ICP nebulizer. A fine spray of phosphate-laden HCL passes through this minute glass tube; sediment or organic material can clog the tube and limit the steady flow of the sample, thus pro-ducing skewed results, or microscopic particles can scratch the nebulizer and predispose this very expensive little part toward breakage.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (231)


The ICP (Inductively-Coupled Plasma Spectroscopy) method of analysis is undertaken when the samples are filtered. Standards of high, low, Fe/A1, X (Ti) and Zero are run first for Al, Ba, Ca, Fe, K, Mg, Mn, Na, P, Sr, Ti and Zn to calibrate the spectroscope. If the standards are correct, then a reference is run and recorded. About every hour, or 20 samples, the reference is re-run to as-sure that drift within the ICP readings is at an acceptable level. The remaining soil samples, left after phosphate analysis in the laboratory was conducted, have been retained for future ethnobotanical or further chemical testing.


The chapter began by discussing the conditions of cultural landscape re-search in Scania and identifying specific criteria for studying its change. The current knowledge surrounding these eight criteria was summarized.

The continuity of the cultural landscape makes Järrestad, and Scania asa whole, an ideal subject area for study (Callmer 1987, 1988a). While Iron Age and Viking Age sites in Scandinavia are buried and most surface features have been eradicated by agriculture, on most sites it is possible to surface collect artifacts. Though these are not as ubiquitous as in arid environments, they can accurately be used for dating the occupation. The continuity of peasant farming for almost 1,000 years after the Viking Age right up to the end of World War II, has resulted in the preservation of the buried, aban-doned villages of the Iron Age and the prehistoric agricultural features known as plaggen soils, or man-made agricultural soils, as well as many place names, routeways, and boundaries. The accessibility of these “fossil” landscapes enables the study of the locational and organizational analysis of Järrestads Hårad’s prehistory.

All of the eight lines of evidence outlined in this chapter can be used to establish conditions before and during the unification period, but the most significant for reconstructing settlement are phosphate survey, surface col-lection, and place-name studies, which were the main areas of study for the Järrestad survey in 1992–1993. Through the Järrestad Project survey and Ar-rhenius’s 1930s fieldwork, highly enriched anthropogenic phosphate patches indicated the general size of sites, while place names and artifacts deter-mined the approximate date of foundation. A few extensive projects and sev-eral smaller excavations have been undertaken at some of the sites, elucidating stratigraphy, chronology, and activities. Some sites were seen to be farming communities with few signs of any special activities, whereas oth-ers had evidence of earlier and later central place functions, such as Simris and then Järrestad, and later Tummatorp.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (232)


Chapter 8 will use these data for a final locational analysis on the small-est scale: how the wider trends seen in chapter 6 are evident in this much smaller sub-region. The changes in integrative strategies during unification will be seen to have affected the perceptions and pathways of every peasant, landowner, and lord in the district.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (233)

This page intentionally left blank.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (234)

Chapter 8

Place, Space, and Experience in Iron Age Communities


Human beings experience and transform the natural world to a human world through their direct engagement as reflective beings within its sensuous, material reality. . . . The symbolic appropriation of the world produces distinctive lifestyles and distinctive landscapes which are historically and geographically specific. Seiz-ing and understanding this dimension of human interaction with nature and its role in the ordering of space is the task of cultural geography. (Cosgrove 1983:15)

This chapter, which presents the locational analysis of Järrestad, will try to unravel Järrestad’s cultural landscape in the manner that Cosgrove suggests, beginning with the landscape of the Roman Iron Age and continuing through the early Middle Ages, a period of about 11 centuries between about A.D. 100and 1200, but focusing on the Viking Age between A.D. 700 and 1050. The landscape of Järrestads Härad underwent a marked discontinuity around A.D.

500, the transition between the RIA and the GIA, as villages were first formed and society became less corporate and more concerned with private property, money, and inheritance. However, there was also some perpetuation of old structures, especially in the boundaries of the territory that was the homeland of the local social aggregate. Between A.D. 500 and 980 the continuity of set-tlement is notable, for although there were waves of village foundation that modified the settlement system, the functioning of the cultural landscape was largely unaltered. Then, change was introduced suddenly across Scania, as the central state rearranged extant settlement and added infrastructure with new places and institutions. The landscape of Järrestad was seized—symbolically appropriated—and transformed. The elements of landscape that underwent


Landscapes of Power, Landscapes of Conflict - [PDF Document] (235)


the greatest change suggest that political and economic control was the goal behind the shifts, and that the central government in Roskilde initiated them. The rapidity with which these changes occurred after many sleepy centuries is remarkable, and shows a definitive policy change regarding the governance of the east.


The earliest cultural landscape relevant to this study is that of the Roman Iron Age, during which many patterns important in later times were established. The following sections will examine the economic, political, and sacred land-scapes of Järrestad during this era.

8.2.1 The Economic Landscape of the Roman Era

During the Roman period, the economic landscape had two aspects: a local agricultural economy and an elite long-distance trade economy. The local economy was dependent on herding and dairy production, as seen in reports from Tacitus and others as well as through the preponderance of pastureland over cultivated areas seen in pollen analyses. The amount of land coming under the plow grew steadily throughout the period, but the real increase in fields and decrease in pasture did not occur until the beginning of the Viking period, about A.D. 700.

The paleoenvironmental studies of the Ystad and Gårdlösa Projects de-termined that during the Roman Iron Age, farming consisted primarily of hulled barley cultivation in a one-course rotation, inferred from the weed species present, a system that can lead to “soil impoverishment, the spread of parasites and diseases, and invasion by weeds” (Gaillard and Goransson 1991:171). Archaeologiclly, the typical settlement type was the isolated single or double farm. As mentioned earlier, Tacitus corroborates the archaeological findings and adds some ethnographic information by stating that there were no villages, that cultivable land was held in common with the land divided ac-cording to rank, and that plowlands were changed every year—that is, swid- dening (Benario 1967:53). He continues, telling that land was quite abundant, and there was no need to “plant orchards, set aside meadows, irrigate” or employ other intensive strategies.

With the farming technologies of the time, agriculture was very exten-sive, and the RIA in Scania, like that in western Denmark, probably consisted of small settlements of large single households that moved every few genera-tions within a general geographic territory. Although there appears to have

Landscapes of Power, Landscapes of Conflict - [PDF Document] (236)


been crafts specialists in fine ceramic and metal production, these were mainly attached to elite households. Utilitarian items appear to have been made locally in the settlements of the time. Villages, in any real sense of nu-cleated, agglomerated settlement, were few and far between.

In terms of elite economy, it is clear that Roman imports played a large role in marking status here, as they did in western Denmark. The point of entry of imported prestige goods from Rome was probably not any local place. In chapter 3 it was noted that while there are concentrations of Roman goods in Scania at places like Simris, Valleberga, Borrby, Uppåkra, Öremölla, and other sites where rich burials are found, the elites buried with them are not known to have controlled large, rich crafts production centers and ports like Gudme and its harbor Lundeborg on Fyn. However, in western Scania is the site of Uppåkra, the very large phosphate area in Bara Härad, where ex-cavations clearly show the seat of a chieftain or chiefly line, one who was far more powerful than his Scanian peers (Vifot 1936; Stjernquist 1994). The site is so large that excavations are just now finally being planned, so the exact nature of the site’s components are only partly known from fairly extensive tests. Although there is a definite RIA component, its extent is not well un-derstood, and the majority of the material is from the late RIA/GIA and Early Viking Age. The primacy of Uppåkra in the Late Roman–Early Germanic era may be because it was close to Western Denmark and may have been the con-duit through which imports passed from Fyn and Sjælland and other points west during their periods of primacy in the RIA.

8.2.2 The Political Landscape in Järrestad’s Roman Iron Age

The site of Simris was clearly the central place in Järrestad during much of the Roman Iron Age. Simris was the subject of Stjernquist’s dissertation (1955), which investigated a large cemetery of the Roman Iron Age thought to be related to a “nearby” settlement. Whether this RIA settlement was con-gruent with all or part of later Simris is unknown, but the original settlement at Simris dates to the Roman Iron Age, and the site has been continuously oc-cupied since then.

The name Simris was first recorded in 1133, and comes from the word *Simbr, which may mean “slow-moving stream” (Kousgård Sørensen 1968). It is probable that the river’s name was previously Simris or Simbris, or some-thing of the sort. Tummatorp or Tommarp was not established until A.D. 1000or so, and its use as the river’s name almost certainly dates to the early Me-dieval period when Tommarp was a large, important town.

Simris probably began as an elite farmstead that attracted other inhabi-tants over time. It is doubtful whether it was a village like the early Danish

Landscapes of Power, Landscapes of Conflict - [PDF Document] (237)


sites of Grøntoft and Hødde with numerous households, and in the case ofGrsntoft a possible “headman’s” farm, which was larger and had many more buildings. A nucleated habitation from the RIA before A.D. 500 at Simris has not yet been found. The earliest probable village or settlement on the site of the current village probably dates to around A.D. 500. If earlier nucleated set-tlement did exist there, it may be that the Viking Age/Medieval village covers it completely.

Traces of settlement from the later Roman Iron Age, around A.D. 400, are also found at Gislöv in the study area, and at Gårdlösa up the river, which while outside of the study area is in its direct line of communication. These three places may represent slightly more nucleated settlements in the RIA, but otherwise no villages are known. Graves are all that are found with any regularity, and they are mostly scattered and isolated, probably representing burial places in farmstead territories. During test excavations at Gislöv, in the layers beneath the Viking Age occupation in field 2.1, Strömberg found cremation graves that were dated by radiocarbon, one to just before the time of Christ and two from the Roman Iron Age (Strömberg1985:89–90). Other outlying RIA graves have been found in Gislöv 25.9 (Strömberg 1985: 110).

However, in Simris, a large Roman Iron Age cemetery that was used con-tinuously for several centuries lies 750 meters northwest of Simris church, beyond the boundary of the Viking Age village (Stjernquist 1955). In the ab-sence of settlement remains, the burials give indications of political structure in the Roman era.

The cemetery consists of over 100 burials, mostly dating to the Roman Iron Age between the first and fourth centuries. Male and female burials are represented, some rich, some poorer. Many contain no diagnostic artifacts be-yond dating them to sometime in the Roman era: pots and potsherds, combs, bits of bronze and iron. While they cannot be dated precisely, they occur in groups of six or seven burials that appear contemporary to each other, based on ceramic styles (Stjernquist 1955:31). Then there are several phases of very wealthy burials, accompanied by well-dated artifacts. The groups of poorer graves probably are each correlated with some phase of the richer, well provenienced and dated graves. Two child burials were identified, one with no grave goods, and one with a simple pot and some shells.

In the earliest phase of the cemetery, the warrior class is represented by a single male weapon grave with a sword and some iron fittings. This burial can be dated to the first century. In the next phase, the five datable burials can be identified as second century in date, and are similar to each other in terms of grave goods—several with bronze ornaments, some with combs, three with finer ceramics, two with knives. They show little inequality and

Landscapes of Power, Landscapes of Conflict - [PDF Document] (238)


could not be termed elite burials. However, if any of the undatable, poorer graves belong to this set of burials, these are markedly richer.

In the next period, the third century, six graves are datable, and show much more variation in grave goods. One grave has numerous arrowheads, a comb, and some good-quality ceramics. Two burials have knives and combs; one of these has ornaments of silver and bronze. One grave has a shield, spear, lance, fine ceramics, and a silver ornament, and the richest one has spear and lance points, spurs, shield, a sword with a scabbard ornamented with silver, gold foil, glass and beads, a coffer, along with glass gaming pieces, and loose beads of glass. The discussion in chapter 3 noted that only the high-est level of elite were horsem*n, and horse gear and trappings, such as spurs, are a significant social marker. Gaming pieces also accompany only higher classes of elites. Finally, a rich female grave of the same era contains a silver collar, two silver and two bronze fibulae, a silver pin, and beads of blue glass and gold leaf with pendants of lead and tin, which by the position of the beads and pendants on the head of the skeleton, appear to have formed a di-adem or headdress. This burial also contains hairpins, a knife, a comb, and several very fine pots.

In a slightly later group of graves dating to around the fourth century, there is a hierarchy similar to that found by Hedeager in western Denmark: one “princely” grave, a chamber burial, contains two gold rings, two silver rings, bronze ornaments, gaming pieces of bone and a gaming board, a comb, fine pots, and weapons including a knife, a spear and lance, a sword and scabbard with silver and bronze ornaments, and a shield. The other male grave has more usual warriors equipment: a spear and lance and comb. Two female graves are contemporary, one with bronze pins and brooches and a comb, knife, and scythe. The other is richer, with a silver torque and silver and bronze pins.

The burials in the final two periods appear to represent more than one class of elite and closely parallel the description of the warband: chieftains, members of the retinue, and common soldiers. It is not unlikely that the fe-male burials reflect this social hierarchy, and appear to show at least two classes of elite women kept their households at Simris. The undated graves probably represent servants, slaves, and other members of the household.

The presence of the rich equestrian burial of the third century and the princely burial of the fourth century, as well as the accompanying rich fe-male graves, indicates that the area directly around Simris was the center of a district controlled by an elite class over a period of several hundred years. Gårdlösa, up the river, contains very similar artifact material to Simris and has some rich graves and a possible shrine (Stjernquist 1955:162, 1991). It is fairly distant and it may represent another elite center. Other nearby princely burials of the late Roman era are found at Hammenhög, Skillinge, Valleberga,


Landscapes of Power, Landscapes of Conflict - [PDF Document] (239)


Figure 8.1. Theissen polygons showing probable political-social aggregates of the Roman Iron Age in southeast Scania.

Löderup, Övraby, and Borrby. Figure 8.1 shows a Theissen polygon diagramof these known sites, and they are spaced at fairly regular intervals and may represent other ruling elite groups.

8.2.3 The Sacred Landscape in Järrestad’s Roman Iron Age

Not much can be said of religious practice in Järrestad’s RIA, except that vo-tive offerings in bogs, lakes, and springs were the norm in this era. These can be divided into “warrior-offerings” and “farmer-offerings” (Strömberg 1985: 128). The first implies a ceremony crafted for public consumption, while the second consists of single gifts, usually everyday items.

No great warrior offerings of arms and armor have been turned up in Järrestad—the closest find is to the north in Albo Härad—but this is a factorof the dearth of research in the area, and as there is a clear warrior hierarchy in the burials at Simris, some place of this type must certainly exist. Isolated, personal Roman Iron Age offerings are found sporadically, items of jewelry plowed up in former peat bogs such as the place called Gislövsstjärna; a for-mer bog that has produced objects from the Stone, Bronze, and Iron Ages that have been interpreted as offerings. This may have been a local cult-placewith a long history of continuity (Strömberg 1985:126, 128).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (240)



The discussion of the Roman Iron Age forms an important prelude to the later Iron Age of the region because the link between earlier and later structures isclear. Although there was upheaval during the Migration period after the fall of Rome, local continuity is evident, especially in boundaries and social and po-litical units. The early division in Figure 8. l may be a close approximation of the Roman Iron Age divisions that preceded those larger social units that de-veloped in the Germanic and early Viking Ages. The RIA Theissen polygon area around Simris is almost identical with the area that I have defined as a distinct subregion for the purposes of this study. In fact, Late Viking Age JärrestadsHärad includes the Simris polygon area and most of the one directly to the south (Figure 8.2). Similarly, Ingelstads Härad also appears to be a combinationof the three earlier units that lie in the southeast corner of Scania. The spatial broadening of elite control and the shifting of boundaries into larger-scale poli-ties is evident all over Scania. These polities, larger than the RIA but smaller than those known by the late sixth century from Jordanes, must have been the boundaries of active territories between ca. A.D. 400 and 550. This is when the

Figure 8.2. Theissen polygon map of some Scanian subregions; eight smaller regions have been subsumed into what are later known as Järrestads and Ingelstads Härads.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (241)


boundaries that we later see as Härads must have been formed. Callmer(1991a: 257-8), as quoted in chapter 4, noted that the polities of the GIA con-sisted of many groups of different sizes, probably based on varying historical factors, that might have included admixures of earlier groups combined through conquest or alliance, dissolution of large groups into smaller groups, and some that have existed “as-is” for a long time.

In earlier chapters I discussed the evidence from Jordanes (Callmer 1991a), indicating that by A.D. 550 a social aggregate called “Scania” that in-cluded all of the southwest and southeast may have already existed (Figure 8.3). Järrestad was now a district in a larger polity. Figures 8.1, 8.2, and 8.3 are meant to underscore the idea of boundary change. As local elites developed from the warrior class, gained local power and finally regional control, the boundaries of social-political aggregates shifted accordingly, growing larger and larger with time. In the RIA, eight smaller units comprise an area that by

Figure 8.3. Political and social aggregates of the Early Viking Age: the large polity of “Scania” hypothesized by Callmer (1991a) and Bolin (1930) has subsumed a number of smaller units.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (242)


Figure 8.4. Phosphate map of Järrestad’s earliest villages, ca. A.D. 500-800, after 1934 and1992-1993 surveys.

the early GIA were only two. The kingdom of Scania by Jordanes’s time, ca. 550-600, had perhaps incorporated all the southern regions, and by A.D. 800Scania was, in name, a part of a larger Denmark.

Chapter 6 discussed how this final boundary change occurred in Scania as a whole. Now the smallest analytical unit, the Järrestad region, can be ex-amined. I will shortly introduce rank-size evidence, which of course cannot be utilized for periods earlier than A.D. 500-700. The only absolutely neces-sary prerequisite for rank-size analysis is that villages of differing sizes are ex-tant. One can thus begin to examine boundaries through rank-size curves only when villages appear.

Although there are few known settlement remains of the Germanic era anywhere in Scania—or Denmark for that matter—it is clear from place-names and stray finds that a number of villages were established in this time period. Figure 8.4 shows the locations of these early villages in Järrestads Härad based on 1992-93 survey results and Arrhenius. All P areas that are

Landscapes of Power, Landscapes of Conflict - [PDF Document] (243)


known to be of non-Iron Age provenience, as well as those unlikely to be a part of Late Iron Age settlements, are excluded. The general extent of the vil-lage is indicated by outline, while white areas are developed or destroyed.

8.3.1 Internal Integration and Hierarchy in Järrestad Härad

By the end of the GIA, ca. A.D. 700, villages of several sizes and differentiated political function existed in Järrestads Härad. It is therefore possible to con-struct rank-size curves for early-established villages as they were at the end of the period, when they were at their largest extent before interior coloniza-tion (Figure 8.5), a process briefly mentioned in chapter 4, and described in more detail in chapter 5. This curve reflects the series of large villages, close in size, that have a few smaller neighbors. This creates a very convex curve, RSI .475, indicative of a system without any one site that is the economic/ social/political center of the system.

Simris was the largest place in the GIA. The site appears to be one of the more nucleated villages of the Järrestad area’s GIA settlements, possibly be-cause it appears to be so old. However, it does not appear to have been the central place anymore; the village of Järrestad is of the “stad” type, one of the oldest types of village name associated with settlements originating between A.D. 500 and 800. Evidence from Strömberg’s 1970’s excavations in Järrestadvillage produced artifacts from about 600 to 700 and radiocarbon dates in the 600s (Strömberg 1976). Much about this village, which will be discussedbelow, indicates elite presence from early in its existance.

Although Järrestad may have been the elite enclave, the power of these elites was very loose, indicated by the poor integration of the system. Poor integration does not mean that the system did not work, that there was no law or order, or that is was difficult to trade or raise a defense in wartime. It

Figure 8.5. Rank-size curve for the Järrestad region, ca. A.D. 800.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (244)


means that the power of local lords was restricted and social institutions were not centralized. Based on the social code of the era, in which govern-ment interference was expected to be minimal and warlords were the ser-vants rather than the rulers of their districts, this poor integration indicates that this social system was still in place. Rather than imagining that this was a chaotic community to live in, with an inefficient and cumbersome system for exchanging information and carrying out decisions, one must remember that this was the preferred organizational mode and was probably just the way people wanted it.

8.3.2 The Sacred Landscape in the Germanic and Early Viking Age

During the period in question, it was typical for shrines to be relocated from areas outside of habitation sites to the estates or compounds of elites. While an elite-connected shrine has not yet been found in Järrestad for this time period,it is likely to have been located in the village ofJärrestad, discussed below.

8.3.3 Rural Demography and Interior Colonization

Much of what we know of these earliest villages is actually based on what happened in the next phase of settlement, known as interior colonization (Callmer 1991e:347–8; Porsmose 1988:234ff). This refers to the circ*mstance that the old villages in the early period had large territories, field systems sur-rounding the village core, and pastures and woodlands surrounding those, all utilized by the settlement at the center. As the population rose, villages could have added more and more farmsteads to the territory’s core, but this alter-native was not taken. Instead, parts of that territory that included outfields, pastures, wooded areas, and uplands were forfeited by the village to allow a new village to be founded. Figure 8.6 shows an example of the field and woodland systems that belonged to each village, typical of such village terri-torial structure.

In the example, one can see the village in the center surrounded by plowed fields, and then all of this encompassed by the skogsmark, orwooded land, where fuel, building wood, willow wands for baskets and flet-work, nuts and berries, mushrooms, and other forest products are known to have been gathered in early historic times. Pigs also browsed here, fodder was cut from wet meadows, and small game could be hunted. From this outer land, pieces were carved for new villages.

It may be inferred from the fact that interior colonization occurred at all that villages were not able to grow larger or did not want to split their village and infield lands more than they already had been. Land tenure

Landscapes of Power, Landscapes of Conflict - [PDF Document] (245)


Figure 8.6. Veberöd village, Torna Härad, in 1693, showing the village core, the infields, andoutfields.

practice of the late Iron Age was to divide the land equally between sons, or daughters in the absence of sons. This led to severe problems where not enough was left to each child to make a living with, and some have attributed the Migrations and the colonizations of the Viking Age to this factor, along with rising population and poor agricultural conditions be-

Figure 8.7 shows the vång, or field, area of a village in the 1700s, and the hundreds of small parcels of land. This is in fact what led to the land re-forms of the 19th century that redressed the problem of families farming 20

fore A.D. 700.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (246)


strips of inherited land within the village, each in a different vång and none wider than a horse or ox could need to turn with the plow. This happened because of the repeated division of inherited land. The government stepped in because the time it took to walk from one field to another among 20 far-flung strips of land over many days was considered to be unproductive, per-haps enough to endanger the harvest (taxes).

Figure 8.8 further illustrates this problem. All black strips belong to one household, and clearly they are spread all over the village. It was traditional for each farmer to have a strip in each vång, so that the different qualities of land—poor, fair, and good—were evenly distributed. As far as can be deter-mined, this system was very old, and apparently, as seen in Figure 8.9, may al-ready have been a problem in the large old villages of the Viking Age. Any expansion of the central village area would have encroached on carefully ra-tioned and coveted land already apportioned to landowners. Sons and daugh-ters who stayed in the village would split the already narrow strips between them upon inheriting. It is understandable why new generations were en-couraged to go somewhere else.


Figure 8.7. Bontofta village, Albo Härad, in 1699, showing typical land tenure conditions insouth Scandinavian rural settlements before 1800.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (247)


Figure 8.8. Lockarp village in 1764; all black strips belong to household #1 (after Pred 1986).

Interior colonizations in the Viking Age were made in two ways: distant outfields within the territory of the village land might be used to begin a new settlement or pasture could be encroached upon. Usually the villages in “by” came out of the land further afield, as they came first and could choose areas with their own extensive territories to exploit. Villages in “torp” apparently came out of areas much closer to the old village cores, probably pastureland. “By” indicates a new-founded village, whereas “torp” indicates a new settle-ment that is dependent on an older one. Looking at the later maps, one can imagine that Gröstorp was dependent on Gladsax and Viarp on Simris. It iseven more obvious when near Vemmerlöv is foundVemmerlövs Torp, one ortwo farms dating to perhaps 1100. New villages were unusual after the “torp” period. After this, in the medieval era, many isolated farms were established in clearings, a quite different pattern from that of the Iron Age and Viking Age (Sawyer and Sawyer 1993:43). Some of these became villages later in the High Middle Ages, ca. 1400 to 1556. In the early Middle Ages, new settlements are usually associated with the linguistic elements –ryd or –brott, which mean “clearing in the forest” and “to break,” as in breaking new soil that has never been plowed. In Järrestad, this is exemplified by upland settlements of the

Landscapes of Power, Landscapes of Conflict - [PDF Document] (248)


Middle Ages like Bräkenryd (clearing in the bracken) and Långryd (long clear-ing) that consisted of single farms.

From the time that the first interior colonization began, it is unlikely that the size of the early villages changed very much; this has been suggested by the excavations in the Jutland villages such as Vorbasse, Sædding ,and Trab-jerg, which have shown that village size remained fairly stable between the 8th and 12th centuries (Sawyer and Sawyer 1993:46). Since the later “by” and “torp” villages were established at around A.D. 800 to 980 and 980 to 1000, re-spectively (Sawyer and Sawyer 1993:43), it can be assumed that the oldest vil-lages were at their largest extent (before interior colonization) at about A.D.800 to 900. In the same fashion, villages of the first colonization (in “by”) would have been at their largest extent shortly before the second wave of


Figure 8.9. Iron Age field boundaries, such as these from Himmerland in Jutland, were not too different than those of 17th century. While some farmers hang on to larger plots, many others are only 10 meters wide (after Nancke-Kroghe 1982).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (249)


interior colonization (in “torp”) at A.D. 980 to 1000. In fact, phosphate inves-tigations made for this study, which were performed in order to assess the ex-tent of the lower culture layers in the old villages of Järrestad and Gislöv, show that although some growth occurred and features became more or less dense through time, these large old villages did not alter much between the approximate time period of 800 and 1050.

Figure 8.10. Comparison of lower (a) and upper (b) culture layers at Järrestad.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (250)


Figure 8.11. Comparison of upper (a) and lower (b) culture layers at Gislöv.

Figure 8.10 and 8.11 show a comparison of the extent of Järrestad and Gislov from the upper and lower culture layers. The data for the lower cul-ture layer come from analysis of the soil samples taken from about 45 to 55 centimeters below the surface. Tests done by myself and Strömberg at both these sites indicate that this generally represents an early village “floor” and the different pattern of phosphate distribution seen on this level indicates that a different village phase is being observed. A pithouse that was investi-gated in Järrestad appears to have been dug down from a surface about 50

Landscapes of Power, Landscapes of Conflict - [PDF Document] (251)


centimeters below the modern surface, and produced artifacts in situ, datingto between A.D. 800 and 900.

In Gislöv the same pattern may be noted. Farms are in similar places, but their extent, orientation, and associated features are slightly different. The extent of both villages can be said to have remained generally the same. Although landowner refusal to participate in the study resulted in missing the southeast portion of the village, and it is conceivable that some change occurred here, Arrhenius’s map, which reflects the upper culture layer, shows considerably less activity in the southeast part of the village. It would be expected that, if anything, the village would have grown larger between 800 and 1000. It is unlikely that large areas of early activity occurred in the southeast and disappeared by later times.

8.4 The Viking Age in Järrestad

The Viking Age in Järrestad, A.D. 700 to 1075, is the period when interiorcolonization begins. The map in Figure 8.12 shows that “by” villages were established in large inland areas perhaps not utilized very much by the older villages, perhaps forming the inland edge of their village territories. The three “bys”—Vallby, Stiby and Karlaby (J-8, J-9, J-10)—lie in generally higher elevations and are further from the river and the coast. The inland areas were higher and more wooded in the north of Järrestad where Stiby, and Karlaby are found, while Vallby is in a flat cultivable plain. Vallby is es-tablished where in former times, during the Roman Iron Age, a no-man’s-land boundary existed, and no villages of the Germanic Iron Age were founded in that area. It may be that due to the integration of regions within the Kingdom of Scania in the mid-Germanic Iron Age, this no-man’s-landwas no longer necessary and the former boundary region could be ex-ploited for settlement and agriculture.

Villages in “by” form a middle phase of settlement when there were still substantial unsettled areas a little further afield of earlier settlement clustering. “By” villages were established after the early villages, but before those in “torp,” in the time between 800 and 980. Not surprisingly, villages in “by” are usually larger than “torp” villages, yet smaller than the earliest set-tlements. If we add in these later Viking Age sites (with the exception of Karlaby, size unknown), we come up with a rank-size like the one in Figure 8.13. The curve has grown significantly more convex, with an RSI of .596. The general trend of large equal-size villages and a decentralized settlement system continues.

Simris continued to be the largest site, but as noted earlier, was no longer the region’s central place; the focus of local political power had shifted to the

Landscapes of Power, Landscapes of Conflict - [PDF Document] (252)





E, A














S 231

Figure 8.12. Phosphate map of Järrestad's villages, ca. A.D. 850-980, after 1934 and 1992-1993 surveys

Landscapes of Power, Landscapes of Conflict - [PDF Document] (253)


settlement called Jarllestatha (Järrestad). First written in 1231, Jarllestatha,from the old Norse Jarles Stadhit, which means “seat of the Jarl” (Strömberg 1976), indicates the presence of elites. The Härad was named after this vil-lage, and since the secondary Härads were named after the central place atthe time the state made its administrative divisions ca. A.D. 980, this alone makes it clear that Järrestad was the political center of the district. The village is of the stad-type, foundation dating to the period between A.D. 500 and 700. Jarl, of course, is a well-known social-political term for a high-ranking lord of the Iron Age, culturally and linguistically cognate to the English earl. Jarls were subject only to kings, and royal brothers were often Jarls. It is a term often found on runestones, in sagas, and in the European historical sources, where conquering Viking Age forces were commanded by kings and jarls (Kroman 1976). Sometime between the end of the Roman and the beginning of the Viking Age, a ruling line, probably based in Järrestad, subsumed smaller, earlier polities of the Roman era into a larger unit.

Strömberg’s 1972 excavations shed light on how early this may have oc-curred. Investigations revealed a smithy filled with iron slag and fragments, fragments of casting molds, a semi-manufacture of an armband, and several fragmentary pieces of jewelry forms that are dated to the 700s or 800s, while radiocarbon dates indicate that the site was occupied at least as early as the 600s. In addition to the production of elite ornaments, Strömberg suggestedthat the slag indicated that there was an extensive iron working industry at the site, possibly under the control of the Jarl.

My own fieldwork in 1992-93 indicates that the site of Järrestad has adistinctively different character than the other sites in the region, and Ström- berg’s hypotheses that it was an elite residence with attached village may well be supported. The general P distribution showed two distinct areas. In

Figure 8.13. Rank-size curves forJärrestad, ca. A.D. 980.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (254)


the northern part of the village, widely spaced farmsteads were the norm, with midden heaps and very high P areas coinciding, similar to findings at Gladsax, Gislöv, Ö. Nöbbelöv, Viarp, and Gröstorp. Artifacts were of a settle-ment nature: grinding stones, whetstones, ceramic, bone, and metal. How-ever, in the southern part of the village near the river the pattern was quite different, as were the artifacts.

In the properties directly to the south and southeast of Strömberg’s1972 excavations ceramics of a very coarse nature were high in density, and iron slag of several types was ubiquitous. An associated pithouse, half-exposed by machine cuts from a construction site, was excavated, produc-ing ceramics from around A.D. 800, bone, and iron slag. The presence of slag in the layers of the pithouse indicate that iron production was already occurring at the site in the ninth century. Several sherds may be of early Baltic-style fine imported wares. This agrees with Callmer’s recent reevalu-ation of Strömberg’s excavated materials, and his opinion that Järrestad rep-resents and elite production site and that the name “Jarllestatha” indeed denotes a Jarl’s residence.

Aside from the huge amounts of iron slag and the high-quality ceramics (imported or local, as the case may be), the phosphate distribution in the southern part of the village was different also. As opposed to the northern area where distinct high P (over 2,000 ppm) farm middens were clear, with relatively lower P between them (900 to 1300 ppm), the southern area had no such “peak” areas, but was uniformly high, 1200 to 1800 ppm or so, with no very high (over 2,000 ppm) areas. This suggests a lot of activity, but not of a predicted farmstead pattern. In this way, the artifact collection and the soil chemistry reflect that some other use was being made of this area; as Stromberg suggested, iron-working of some type appears to be probable. The presence of a Jarl is based on lines of linguistic evidence, but large-scalecraftwork or manufacturing is often associated with elite in the south Scan-dinavian Iron Age.

There is actually a folk tradition about a Jarl in the Tommarp area that in some form dates to the later Iron Age. During the Germanic Iron Age, Theodoric, king of the Ostrogoths (d. A.D. 536), stands out as one of the only well-known historically documented figures of the era. Around him and his vassal, Didrik of Bern, a series of sagas is based, originating in the sixth or seventh century, yet becoming popular in Scandinavia only in the 1000 to 1100s. In the 1200s, this saga material was brought together with others in Norse and Icelandic works, and in 1715 was translated into Swedish under the name of “Wilkina-saga.” In this convoluted saga tradition, the Tommarp area and the Jarl are mentioned (Wallin 1950b:28). The story is set among historical figures dating to A.D. 550 to 600 and was composed a bit later. It

Landscapes of Power, Landscapes of Conflict - [PDF Document] (255)


may reflect a tradition of elite occupation in the area. There was no Tom-marp then—it was not established until ca. 980—but from all archaeological and etymological accounts, Järrestad, which lies about 2 kilometers away,was extant at that time.

“A man called Bitterulf was known from Denmark in Skåne. He was a powerful man, and he lived in the place that is now called Tummatorp; his wife was called Oda. She was a Jarl’s daughter from Saxland. They had a sonbetween them called Tiettlef. Bitterulf was the strongest hero and fighter for as far as Danmark’s empire stretched.” The rest of the story can be summa-rized: the Jarl’s son, after many adventures, becomes a soldier of the famous Didrik of Bern. He becomes a knight, marries the sister of Didrik’s wife, and is brother-in-law to a powerful king. He is made a Duke, and master of the city of Drekanfii in the forest of Osning (Wallin 1950b).

Today, two Bronze Age barrows outside the village ofJärrestad are calledJarladösen and Jarlafrudösen, “the Jarl’s tomb” and “the Jarl’s wife’s tomb:documented in place name archives since at least the Middle Ages. Although these mounds date to long before the Iron Age, this denotes a folk tradition that elites lived in the settlement, perhaps dating back to a time when people remembered a Jarl in Järrestad.

The köpinge place, or centrally controlled export marketplace, that wasestablished in the southeast at this time (ca. 800) was the market of Stora Köpinge (“big market”). This was not in Järrestad, but nearby in Herrestad. Inchapter 7 it was suggested that a route from the Tommarp valley going southwest to Stora Köpinge appears to have existed. This was the local node of the first economic intrusion into the Järrestad area by the central state.

In terms of settlements during this time of tentative steps toward inte-grating the Scanian economy with that of the west, there is a slight overlap between the “by” and “torp” phases of settlement. Secondary villages with names in “by” stop appearing at about A.D. 950 to 980, while the torps are seen beginning in A.D. 980 to 1000 and continue to appear until about 1100. I would like to look fast at the time around A.D. 1000, when Tummatorp was perhaps just beginning to expand from a very small settlement into a center. This will correlate with the presentation of Härads at A.D. 1000 in chapter 6,to underscore that while western Denmark was moving into the Middle Ages, becoming more and more integrated and well-administered, the Scanian sub-regions were still lagging in the Germanic Iron Age. This study of Järrestadwill mirror the discussion in chapter 6 on the smallest, local level of analysis.

Figure 8.14 shows the map of settlements in the Tommarp Valley at about A.D. 1000. The exact date of when Tummatorp was appropriated from its founder Tumi or Tomme is not known; but investigations by Thun (1960, 1967, 1982) show that it did begin as a small, regular torp settlement. At A.D.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (256)





E, A














S 235

Figure 8.14. Phosphate map of Järrestad's villages, ca. A.D. 1000, after 1934 and 1992-1993 surveys

Landscapes of Power, Landscapes of Conflict - [PDF Document] (257)


1000 it was either still so, or had just begun to function as a small royal cen-ter. Either way, the configuration of the landscape had not yet been affected, although the control of the region by the west was beginning. The map indi-cates how the torps were established on lands that probably were part of the larger village’s territories. Each of the larger old villages had smaller satellites around it. They had filled in many of the spaces and were much smaller than the villages in “by.” Figure 8.15 shows the rank-size for this time. With the ad-dition of new central places, such as Stora Köpinge, we might assume that ad- ,

Figure 8.15. Rank-size curve forJärrestad, ca.A.D. 1000.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (258)


Figure 8.16. Reconstruction of late Viking Age roads in Scania; Kungalevs are marked with large circles (afterBlomqvist 1951, Söderberg 1995).

ministration of the area was improving, but though the addition of many smaller places changed the shape of the rank-size curve, it is still very convex with an RSI of .539. The royal initiative of founding an administrative center at Tummatorp had just occurred but the site was still emerging. A recon-struction of roads from this period (Figure 8.16) shows that the area proba-bly had good overland communication routes. The place where all the roads converged is Tummatorp.

8.4.1 Early Christianization

Around A.D. 980-1000, Christianity finally began to take root in Scania. The early churches were built of wood and were replaced with stone churches between 1050 and 1150. In 1075, Adam of Bremen wrote that Scania alone had 300 churches. These would have been of the type described in chapter 4, built and maintained by local lords, controlled in a way very similar to the

Landscapes of Power, Landscapes of Conflict - [PDF Document] (259)


Figure 8.17. Villages with churches originating ca. A.D. 1000 in Albo Härad. Also shown, Ingelstads Härad to the south and part of Järrestads Härad to the north.

late pre-Christian shrines. Figure 8.17 shows the distribution of these little vil-lage churches known from their stone versions, but also indicating their ear-liest location ca. A.D. 980 to 1050.


Before moving into an analysis of change during the radical landscape shift that occurred between 1050 and 1100 in Järrestad and elsewhere, I would

Landscapes of Power, Landscapes of Conflict - [PDF Document] (260)


like to couch these transformations in their ideological milieu by looking at those indicators implicit in the discussion in chapter 2. The cultural geogra-pher Cosgrove suggested that a study linking mind and matter should do two things: examine the relationship of landscape development to socioeco-nomic and political variables, and also incorporate secondary indicators, such as art and literature, of the ideas, motivations, and perceptions of those responsible for change and resistance.

Ideology both integrates and divides societies; it controls social relations yet may also be manipulated by both common and elite sectors to gain ad-vantage, foment change, or consolidate existing relationships. Just as the socioeconomic and political arenas are subject to change, so is the ideologi-cal realm. Although this is commonly observed by anthropologists working among contemporary people, it is not often discussed in an archaeological context. This is partly because it may be difficult or impossible to address. The paradigm shift of the 1980s, sometimes called the processualist versus post-processualist debate, raised the question (among others) of the validity of explanation without this obviously important element. The unavoidable re-ality is that the most ancient belief/ideological systems may never be known and are better left alone than construed from speculation. However, the Late Iron Age lies in that zone of protohistory that may be more conducive to such humanistic interpretations.

As discussed in chapters 4 and 5, simultaneous with changes in the po-litical and economic landscapes other shifts were occurring, especially in the area of social institutions, the terms of social relations and social obligations, class, law, ownership, and inheritance. All these sectors are normally cultural carriers of ideology and belief, both helping to reproduce and to preserve cultural values and meaning and reflecting change over time. Political ideol-ogy inferred from the terms of social relations has already been examined; now I want to look at social ideology, the reflection of what is considered “right” and “proper,” what might be called a normative view of social rela-tions. Right and proper to whom, though, is another question. Two views of what was right appear to have been prevailing at once. In the documentary record there is evidence directly paralleling the shift in social terms, an indication of tension between old social forms and new, a struggle between rulers and subjects on all levels of society from slaves to free land-owning classes to local lords. The source material is not abundant and must be read and interpreted carefully, but there appears to be enough to correlate these archaeologically invisible factors with the change seen in the archaeo-logical record. Changes in terms and laws are oblique and tangential ways of getting at ideological shifts. There are less equivocal clues lurking in the doc-umentary record.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (261)


8.5.1 Class and Conflict: From Saga and Song to Political Reality

The poem of the Elder Edda called Rigsthula, or the Song of Rig, was written between 900 and 950, almost definitely in Denmark, and very possibly as a song in honor of Gorm the Old, founder of the Danish dynasty, to be sung be-fore his son Harald Blåtand. It is something of an oddity among the poems col-lected in codices now known as the Elder Edda. It is what Bellows (1968:201) calls a “cultural poem,” that “tries, on a mythological basis, to explain the ori-gin of the different castes of early society: the thralls, the peasants, and the warriors. From the warriors, finally, springs one who is destined to become a king, and thus the whole poem is a song in praise of the royal estate.”

It speaks of how Odin (traveling under the guise of Rig, one of his many names) lies with a slave woman, a free woman, and the wife of a lord, and from these unions spring children. He comes to the house of Ai and Edda (great-grandfather and great-grandmother), who are slaves. After he lies with the slave woman Edda,

A son bore Edda, with water they sprinkled him, With a cloth his hair so black they covered; Thraell (thrall, or slave) they named him, The skin was wrinkled and rough on his hands, Knotted his knuckles, Thick his fingers, and ugly his face, twisted his back, and big his heels

Thraell and his wife, Thir (serving-woman) have many children, such as Fulnir, Fjosnir, and Klur (Stinking, Coarse, and Cattle-Boy) and daughters with names like Dumba, Kumba, and Ökkvinkalfa (Log, Stumpy, and Fat-Legged),“And thence has risen the race of thralls.” One can imagine that this was a source of amusem*nt to the Danish king and his court.

The house they cared for, Ground they dunged, and swine they guarded, Goats they tended, and turf they dug.

Rig (Odin) continues to the house of Mi and Amma (Grandfather and Grandmother) who are free farmers, and gets Amma with child.

A son bore Amma, with water they sprinkled him, Karl they named him; in a cloth she wrapped him, He was ruddy of face, and flashing his eyes.

He began to grow, and to gain in strength, Oxen he ruled, and plows made ready, Houses he built, and barns he fashioned, Carts he made, and the plow he managed.

Karl’s sons were given names like Dräng, Bondi, and Smed (soldier,farmer, and smith). One of Karl’s son’s is called Thegn, but remember that by

Landscapes of Power, Landscapes of Conflict - [PDF Document] (262)


this time Thegn has been trivialized from the title of a lord to that of “older man.” His daughters were called Svanni, Svarri, and Bruth (Slender, Proud, and Bride), “And thence has risen the yeomen’s race.”

Rig continues on his way and comes to the home of Mothir and Fathir (Mother and Father), who are magnates.


Within two gazed in each other’s eyes, Fathir and Mothir, and played with their fingers; There sat the house-lord, wound strings for the bow, Shafts he fashioned, and bows he shaped.

The lady sat, at her arms she looked, She smoothed the cloth, and fitted the sleeves; Gay was her cap, on her breast were clasps, Broad was her train, of blue was her gown, Her brows were bright, her breast was shining, Whiter her neck than new-fallen snow.

Odin lies down with Mothir.

A son had Mothir, in silk they wrapped him, With water they sprinkled him, Jarl he was; Blond was his hair, and bright his cheeks, Grim as a snake’s were his glowing eyes.

To grow in the house did Jarl begin, Shields he brandished, and bow-strings wound Bows he shot, and shafts he fashioned, Arrows he loosened, and lances wielded, Horses he rode, and hounds unleashed, Swords he handled, and sounds he swam.

Jarl has sons called Athal, Mi, and Nithjung (Offspring, Heir, and De-scendant—lots of references to inheritance here), but Jarl’s best son is called Kon the Young, a clear Viking Age pun: Kon Ungr or Konungr means King. Unlike any of the others, Odin/Rig teaches magic runes to Kon, giving him godlike powers.

Young Kon rode forth through forest and grove, Shafts let loose, and birds he lured; There spake a crow on a bough that sat: “Why lurest thou, Kon, the birds to come?

“’Twere better forth on thy steed to fare,and the host to slay.”

“The halls of Dan and Danp are noble, Greater their wealth than thou hast gained; Good are they at guiding the keel, Trying of weapons, and giving of wounds.”

Landscapes of Power, Landscapes of Conflict - [PDF Document] (263)


8.5.2 Interpreting Saga Content for Its Cultural Message

Unfortunately for students of Old Norse literature, this song of social repro-duction and “structuration” (as defined in chapter 1) is on the last page of the codex and is incomplete. It ends just as Young Kon, the mythic embodiment of royalty, is about to become established in his rule. The crow (Odin’s famil-iar Raven?) is inciting Kon to go to make his fortune in the Halls of the Danes (The Skjøldungsaga of the 13th century supplies us with the information that the mythological Rig-Kon married the daughter of Danp of Danpsted, from whence sprang the Danish kings). Too bad for literary critics, but not for ar-chaeologists. We know what happened without hearing the end of the tale.

As Bellows noted, none of this invented mythology had anything to do with Scandinavian traditions. It has no roots in earlier concepts of class or rulership that are clear from the classical and later sources discussed at length in this study. The bringing of Odin into the song is an attempt to es-tablish a pre-Christian version of the “divine right” to rule by claiming a god for a royal ancestor.

This is a literary rendition of the agenda that I have suggested through-out the study—that the ruling class was seeking to eviscerate the power of the free and land-owning classes and impose a new and unpopular type of so-cial order. This poem, if viewed in a vacuum, indicates a belief in a divine ori-gin of social classes whose lines, almost caste-like, may never be crossed. It speaks of a “race of thralls” and a “race of yeoman” and a “race of lords” in clever imitation of true sacred sagas that name the “race of dwarves,” the “race of elves, the “race of giants,” and the “races of the Aesir and Vanir,” the gods. It implies a natural order in which those born into the correct family have a di-vine right to rule their neighbors while others must serve them as warriors, farmers, craftsmen, or slaves.

First, while rulers in Roman era Scandinavia usually came from lordly families, the social classes had a great deal of mobility: a well-born man could fail to lead and be rejected, and an able soldier could rise to be a chieftain. Second, in the upper class, whose members were considered to be equal, there was no “ultimate” royal family but overlords temporarily elected from a pool of their peers, as I have reiterated many times. The Rigsthula states that the destiny of Jarl (the lord) and Kon (the king) were distinctly different. While kingship springs from the warrior class, it supersedes it, and somehow with the approval of Odin gains the right to rule the other lords. This goes hard against earlier tradition.

The poem so clearly seeks to define and legitimize the new social order (which at around 950 was still very much an incursion into thencurrent Ger-manic tradition) that it is no great leap to infer the conflict and internal strife it represents, especially in light of all the other lines of evidence discussed in the

Landscapes of Power, Landscapes of Conflict - [PDF Document] (264)


course of this study. It is an attempt to create a false consciousness, subvert the social memory of several classes of society, and replace it with a social memory constructed by kings, formalized in a way not possible for the farmer or local lord at the assembly: through a written text. The middle class (Karl’s family) and upper classes (Jarl’s family) were being forced into roles hitherto un-known. The poem celebrates the deconstruction of a society with exceptional social mobility, even for slaves who could buy themselves free and were often manumitted, and certainly for freemen who could earn the right to be lords.

Most other songs composed in the 10th century have been lost, and those that do come down to us, carefully preserved in codices like the Edda, contain songs of Creation, the Gods, and cultural heroes like Sigurd Fafnírs-bane and Veland the Smith. The promulgation and spreading of the Rigsthula so widely that it appears in a codex with poems of heaven and hell is unusual. The Rigsthula is a piece of carefully constructed Viking Age political propaganda, “archaized” into a form usually reserved for religious or legendary figures. Some flattering skaldic poet has summarized the story of Gorm’s family line: the development of strong local elite, their emer-gence as regional leaders, and their attempt to break apart the old social order and impose a new and unfamiliar regime, all couched in a pseudo-mythical, aggrandizing theme that elevates these political machinations into the realm of theology.

This poem may be interpreted as yet another sign that the ruling line was attempting to incorporate this revision of class codes into an existing body of belief—in this case, songs and sagas accepted by the public as an-cient tradition. Just as they worked falsely “archaized” sections into the law codes to break down the legal recourse of their enemies, as noted in chapter 4, so it may have been done with state poems. If it was indeed sung before Harald Blåtand as Bellows (1968:201) believes, this was the same king who built the Trelleborg fortresses, introduced the church, and is supposed to have driven from Denmark any remaining petty kings. Yet Harald still had to contend with the old social code and its sanctioning of his assassination at the hands of his son and other pagan Germanic-style elites.


Let us now return to the radical landscape and political changes that began in about A.D. 1000 and continued through the 11th century. The loose, un-regulated village of the earlier Viking Age gave way to the toft system, when land appears to have been organized and allocated based on the farmer’s abil-ity to provide money for the tax-collector and labor for the naval levy. Shortly, we will see what happened in Järrestad as a result of this.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (265)


How was such a change imposed on the people of Denmark? It was not done by the simple proclaiming of new rules and seeing them obediently fol-lowed. It was done through a mixture of disguise—that is, a surficial adher-ence to Germanic traditions—combined with acts designed to inspire terror. Of certain events in 1080, the Knytlinga Saga tells us:

King Knut set out on a progress to all parts of the realm. When he came to Halland he held an assembly at which he spoke himself. A certain farmer who was their leader, and the shrewdest man amongst them, stood up and made a speech saying that they would agree to demands by the king no more severe than those set out in ancient law. After the man had spoken the farmers applauded his speech, saying that they would have things as they were, and would put up with no new duties and impositions from the king. After the noise had died down, the king spoke. “Farmers,” he said, “you’ve done well to let me know that I must accept your laws and expect nothing more. I know, too, that you will allow me the protection of your laws, and then I shall have my privileges independently of you.” They allagreed, and the king continued. “In that case, Hallanders, I forbid you to use the forest I own nearby to graze your pigs and other animals.” The farmers were at a loss for words. They could see that this would not do at all, for the royal forest ex-tends across the whole of Halland and the Hallanders own large numbers of pigs which feed in the oak- and beech-woods, so they decided to agree to let the king have whatever he wanted. The farmers were strictly bound to this arrangement and no other, and at that very hour the farmer who had most opposed the king was put to death. (Pálsson and Edwards 1986:55)

It can be suggested from this cynical episode in the reign of a man that the Danes regarded as a tyrant, from a saga that accounts him a saint in the Catholic Church, that the king (in this case Knut in 1080, but also earlier kings) did actually circulate and attend numerous assemblies. After Knut’s visit to Halland, there were several similar ones in Scania.

When King Knut reached Skaane he held an assembly attended by a large num-ber of people. The farmers thought his demands altogether excessive, and shouted out in unison, refusing what the king asked and declaring that they would never pay the king any tax or duty beyond those prescribed by law. When he could get a hearing, the king spoke. “People of Skaane,” he said, “you’re intelli-gent men. I can tell from the way you’re behaving that you must have heard how my business turned out with the Hallanders. You’ve found a smarter way than theirs to refuse my demands, since I can’t blame any single individual here for your reply. However, I’ll make the same request to you as I made to the Hallan-ders, that you allow me my privileges freely, without interference.” They all agreed, and the king continued. “Everyone must know what belongs to the king here in Denmark,” he said, “and what belongs to the farmers. The king owns all the uncultivated areas, don’t you agree?” All agreed that it was so, and the king said that the uncultivated areas included the sea as well as other uninhabited places. “In that case,” he continued, “I lay claim to the Øresund. Should you refuse me anything I ask of you and fail to give consideration to my requirements, I shall forbid all such fishing as you’ve been able to do in the past.” Once the king had said this, they all realized that it would not do for the farmers to be deprived of

Landscapes of Power, Landscapes of Conflict - [PDF Document] (266)


herring fishing in the Øresund, so they decided, just like the Hallanders before them, to let the king have his way and agreed to whatever he asked, and that was the end of the assembly. (Pálsson and Edwards 1986:56)

While the confrontations over the landscape began at such assemblies with threats and executions, this possibly true, possibly fictionalized anec-dote merely symbolizes a long process unique to the era around A.D. 1000 to 1100, in which farmers and royal officials negotiated changes in the ancient law and land tenure practices that probably included the changes seen in vil-lage organization. It is difficult to imagine how such changes were carried out; clearly the beheading of a prominent farmer and the cowing of the dis-trict’s landholders were not the endpoint in this process but only a start.

In terms of social memory, as described in chapter 2, Halbwachs (1925, 1980) stated that social groups construct their own images of the world by establishing an agreed version of the past and that these versions are estab-lished by communication, not by private remembrance. The act of talking, writing, and sharing memories with others in the context of a social group is what transforms individual memories into social memories. In these docu-ments, accounts of assemblies allow us to see a part of the process of the construction of the social memory of a particular group: farmers and land-holders in Scania. Here, group gatherings served as a forum for verbalizing, thus cementing collectively “approved” versions of the past.

Yet as helpful as they are in allowing us to gain insight into group iden-tity in the Viking Age, the recounting of events at assemblies, such as those described earlier, were not written down by farmers at the time they hap-pened, but by early historians, based on oral tradition. This occurred after the Crown had suppressed its opponents, and the Scanian Uprising was only a past event. Farmers in 1085 could not write down what Knut had said and done to them in Halland and Scania, and could only speak of it. Unrest could only be diseminated verbally, while elite-sponsored construction of a new so-cial memory could be preserved and legitimized by inscribing it on vellum.

The following paragraph describes events that occurred in villages all over Scania on the eve of major land reform in 1808. The land had been held and subdivided in the same fashion from at least the late 1 lth century, or ter-minal Viking Age, until the early 19th century, as described above, until it was fragmented. Then the government stepped in, in order to maximize produc-tion, and made a series of drastic landscape changes:

It is an autumn day in a nucleated agricultural village situated on the peripheral plains of Skåne, Sweden’s southernmost province. Yet again, all of the village’s landholding peasants have assembled together with a state-appointed surveyor and two outside witnesses. An invisible psychological boundary is irreversibly crossed, thereby enabling a revolutionary realignment of the landscape’s visible boundaries. After years of periodic negotiation, open wrangling, internalized

Landscapes of Power, Landscapes of Conflict - [PDF Document] (267)


resentment, and reluctant compromises, an agreement has finally been reached. The once unthinkable is to be enacted. Old fences and walls are to be destroyed. A new local circulation network is to take shape as existing roads and paths are drastically modified or completely erased and new avenues of movement are etched out. . . . In short, the village core is to be physically disintegrated and the spatial organization of the village territory as a whole is to be radically restruc-tured. (Pred 1986:1)

A series of events very similar to these in 1808 must have occurred be-ginning at around A.D. 1000 when significant shifts in Viking Age villages are seen. What sort of effects do such drastic changes, the “physical disintegration and the radical restructuring” of the village actually have on its inhabitants? Is it a mere readjustment, or is it, as I have suggested earlier, a way of under-mining traditional practice and rebuilding according to a centralized design? I will argue that such changes affect the process of social reproduction pro-foundly, and that the manipulation of places, familiar and unfamiliar, everyday and special, is one of the most effective methods rulers have ever devised for controlling and changing society.

8.6.1 The seizure and Appropriation of the Cultural Landscape

In chapter 6, the contraction of Scanian village sizes in the period between 1000 and 1100 was noted, and I stated that based on data to be discussed in later chapters, this reduction was up to 60% or more. This did not occur be-cause farmers became too lazy to walk to the village pond and converse with their neighbors. Ridderspore (1988) suggested that new demands for taxes and manpower for the naval levy were enacted at this time and that each holding was required to site itself on a regulated toft along the street, the size of which would symbolize the wealth of the farmer and make it simple for a royal official to estimate his contribution. Although this sounds odd, it must be remembered that only the Archbishopric in Lund and a few monks were literate at this time; even parish priests were probably illiterate and not ca-pable of written record-keeping. Laws were barely if at all inscribed on vel-lum, and lists of what thousands of farmers owned or owed were not feasible as they were 200 years later.

The taxation and toft regulation theory is widely accepted in Scania as a feasable explanation for the widespread shrinkage of settlements. I would also suggest that this significant change in village size is also directly related to the founding of torps, which may have been a government-sponsoredmove out into the land in order to maximize the amount of acreage being tilled. This implies not only a contraction of village size but a drop in large village populations. Part of the population went to the small torps, and part may have migrated to towns at this time. Table 8.1 shows early size, later

Landscapes of Power, Landscapes of Conflict - [PDF Document] (268)


Table 8.1. Landscapes of Power, Landscapes of Conflict


Percent Percent reduced/ reduced\

Site Early size Late size increased increased

Tummatorp 45.2 45.2 100.0 Simris 34.3 12.5 36.4 63.5 Gislov 32.0 10.9 34.0 65.9 Vallby 30.0 15.0 50.0 50.0 Järrestad 28.7 18.7 65.0 34.0 Gladsax 26.5 10.9 41.1 58.8 Karlaby 25.0 15.9 63.6 36.4 Stilby 23.0 14.8 64.3 35.6 Virrestad 21.8 6.2 28.6 71.3 Ö. Vemmerlöv 14.8 7.8 52.7 47.2 Vranarp 11.9 13.0 109.2 +9.2 Viarp 11.2 8.6 76.7 23.2 Gröstorp 10.1 12.5 123.7 +23.7

Gnalöv 6.0 6.1 101.6 +1.6 Bolshög 4.5 3.9 86.8 13.3 Ö. Nobbelov 4.1 5.5 134.1 +34.1

Tågarp 9.5 6.2 65.6 34.3

size, and percent of reduction in Järrestads villages. The site size of late Kar-laby is known and is similar to the other “by” villages, so I have assigned it a hypothetical early size of 25 hectares, and have let Gnalov remain the same, as other very small sites did. Based on the assumption of 20 persons per hectare in farming villages, the pre-torp era villages would have had a total population of about 4,900 people. Based on their size after shrinking, they should have been reduced to a population of about 2,500 people. Where did these people go? If one estimates the population of the torp villages, these small villages would have taken about 1,000 people away from the older, larger places. The remainder probably nucleated around the new town, dis-cussed below.

So far this study has used hectares for the construction of population es-timates. Each farm toft occupied about .5 hectares, and a population of 7 to 10 people per farm is usually assumed for the Viking Age–Middle Ages; the es-itimates here are at the higher end. This gives us a rural population density 20 per hectare (Olsson 1991a:190), using the maximum number of persons. An idea of how close these population estimates are to reality can be gleaned from looking at maps from the 1600s and 1700s. Although village size and population in 1690 or 1785 was not identical to that in 1100 to 1200, they were similar. Though some village maps are difficult to make out in terms

Landscapes of Power, Landscapes of Conflict - [PDF Document] (269)


of actual houses in the tofts, some are very well preserved. Reproduced in Figure 8.18 is the village of Gladsax in 1792. There are 29 farms in the village. Here we have the actual farms, not just hectares, so the figure of 7 to 10 per-sons per farm is used. This would give a population of 203 to 290. Using

Figure 8.18. Survey map of Gladsax village core in 1792.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (270)


hectares for the precartographic era of Gladsax, a hypothetical population of about 218 is proposed.

Records from Tummatorp in the Middle Ages (Wallin 1955) show that there were usually “gatuhus” in the villages too. These were small “street houses” as opposed to farmhouses, that were occupied by the landless, such as craftspeople, laborers, and the poor. This would add a few more small households in larger villages.

The shifting of many people into newly built torp settlements facilitated administration by breaking up communities with histories of autonomy,putting much of the population in very small places that had no depth of his-tory and no claims of ancient rights and rules. These claims that farmers made to the “old ways,” as illustrated well in the preceding passages from Knytlingasaga, seem to have been an important precedent in the legal sys-tem. If the saga has any merit, the farmers appear to have used precedent to challenge the king on more than one occasion. It is probable that in the eyes of central authority, new villages had no rights to old claims.

8.6.2 Intensification, Population, and Sociopolitical Change

Were the torps founded by central initiative, in order to increase produc-tion? One of the trends observed by settlement archaeologists, palynolo-gists, and linguists is the expansion of settlement in the Germanic and Viking eras. As noted, several waves of village foundation episodes follow each other: one at about A.D. 500 to 700, another around 850 to 980, and another around A.D. 980 to 1100. The following discussion will show how sociopolitical change, demographic change, and agricultural intensification are often linked together.

Intensification differs from simple increase in that it “refers to an in-crease in the productive output per unit of land or labor or to some other fuzed quantity” (Morrison 1994:115). Most authors link population growth to new agricultural intensification technology (both in the use of new tools or new systems of agriculture with old tools). On top of this rise in the number of Danish villages beginning at this time, there are changes in field systems and cropping methods, and new plowing technology is adopted from Conti-nental Europe.

At one time, population pressure was considered the primary causal fac-tor for intensification; inevitable population growth forced an increase in food production (Boserup 1965). This argument was further explored by ar-chaeologists, and in the following decade found many adherents (Sanders and Price 1968; M. N. Cohen 1977). When linked with the migration of Ger-manic peoples away from their homelands, it seemed plausible to apply this model to Scandinavia.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (271)


At the same time, another theoretical perspective was taking shape: the idea that social factors, such as increased elite demands for surplus or tribute may have been involved (Bender 1978, 1981; Blanton 1975). Although in-crease in population is certainly indicated in Denmark’s archaeological record, according to the social hypotheses, cultural factors determine demo-graphic change and are not driven by it. Is this a theoretical thread worth fol-lowing in Scandinavia? Chapter 3 noted that after the fall of Rome at about A.D. 500 a substantial change in the south Scandinavian political economy oc-curred, and elites appear to have shifted from the control of prestige items to the accumulation of wealth.

Simple population pressure models tend to view population as an “un-differentiated mass.” Human demographics are more complex than this, and populations have not only size but structure (Morrison 1994:118). Fertility and mortality must be taken into account, as they are directly related to labor and consumption. At the same time that the political structure was changing around A.D. 500, the Justinian plague probably ravaged Scandinavia as it did Continental Europe, killing up to a third of the people infected: a much smaller population in a landscape of abandoned, empty settlements, that as which is described on the Continent, would have little reason to intensify. However, if increases in elite demands coincided with a drastic rise in mor-tality (that might take several generations to recover from) it is reasonable to assume intensification would have been the only way to produce more crops with fewer people.

This in turn could affect population. Bender states that when labor is re-organized due to intensification (1985:57–59), people often choose to have larger families for economic reasons. In the specific historical case of south Scandinavia, an effort at repopulation and a rise in the birth rate after a plague episode is also possible. Probably, the interaction of population dy-namics and social conditions is at work.

Increase in population density has been linked to intensification, but de-cisions either to intensify or extensify, based on historic and ethnographic observations, are usually economic strategies on the part of the producers, based on local ecological, social, and economic factors (Morrison 1994: 121). Thus demographics, while certainly a part of the picture, are mediated by other proximate factors. Kirch (1984) presented such a model for Polynesia: intensification resulted from a combination of chiefly demands for surplus (the proximate cause) and population growth (the ultimate cause).

In prehistoric Europe, elite political aspirations appear to have been de-pendent on their ability to extract goods and services from the populace (Gilman 1981). In the Roman Iron Age, this was in the form of direct provi-sioning: Tacitus says that the chiefs were supported by gifts from the local populace. In the Germanic and Viking periods, as polities became larger and

Landscapes of Power, Landscapes of Conflict - [PDF Document] (272)


more hierarchic, with substantial bureaucracies to support, this was probably replaced by taxes (in coin and in kind) that went far beyond the feeding of a local lord and his drótt.

8.6.3 An Urban Center in Järrestad

At the same time that the torps were being founded, a new central place suc-ceeded Jarllestatha: the royal city of Tummatorp. Tummatorp began at around A.D. 1000 as a small, late Viking Age torp settlement. Several finds from this era were excavated by Thun (1967), and this occupation appears to have been limited to a small area in the north part of the later town. However, the situation of the small torp in the center of the rich agricultural plain, in a large, deep meander of the river, near the falls in the river that enabled prof-itable taxation through water-mills, made it the ideal location for several kings to exploit. Knut the Great may have founded the town, while Sven Estridsen and Sven’s sons Harald the Whetstone, Knut the Holy, Ulf, Erik the Memorable, and Niklas the Good developed the town. Tumi’s small clearing soon grew into one of Scania’s royal administrative centers.

Extensive excavations of the Late Viking-Early Medieval culture layers, dated by artifactual material to 1075 to 1100 or so, show that the town was extensive even in this period (Redin 1972:68). A bronze industry is indicated by large areas of dense bronze shavings, semi-manufactures of buckles and fit-tings, and associated hearths. In addition, cemeteries from one of the churches and the monastery have produced material that is stylistically and radiocarbon-dated to the 1000s, much earlier than extant documents imply.

Three coins from different periods of the Late Viking Age have been found in different parts of Denmark that bear the legend Tumai, Tuma, or TU. They were minted in the reigns of Sven Estridsen (1047–1074), Harald Hen (1074–1080), and Knut the Holy (1080-1086). Since Tummatorp was a royal property (kungalev) it is not surprising to find that coins were struck at the site, and this activity marks the period when the settlement was first trans-formed into a center. At this time, a village church already existed, and a sec-ond one was soon founded to serve the larger population, a royal church whose land was acquired from the kungalev, or king’s property in the town (Redin 1972:72).

In 1155, a papal letter indicates that a Premonstratensian monastery was already established in Tummatorp, and evidence discussed below indicates that this probably occurred during the late 1000s. In 1161 King Valdemar char-tered this monastery with royal authority, founded two more churches in the town, and donated five of his mills on the river to the monastery. Parts of what may be medieval mill dams have been found near the river's one-time edge. In the charter of 1161, Tummatorp is called villa, which in Latin documents of

Landscapes of Power, Landscapes of Conflict - [PDF Document] (273)


this era signifies a town. That there were already five mills there indicates that the royal center had a great deal of power over the district (Redin 1972:76). If a mill monopoly was functioning, royal authorities must have had administra-tive control over the area in order to enforce the use of the royal-held mills. The fact that there were five mills supports the notion that there was a lot of business for the millers and a lot of revenue for the Crown. The origin of the milling monopoly appears to have been contemporary with the striking of coins, beginning sometime in Sven Estridsen’s time.

In older maps of Tummatorp from the 17th and 18th centuries, about one kilometer west of the town is an area that bears the terrain name of “Tuggeborr,” which indicates a deep harbor (Hoflund 1921:55). Here the river broadens and in earlier times may have spread into a deep stillwater that is now mostly silted in. Before 1920, constructions of worked oak stakes, boards, and planks were found in this area, which may represent the town’s harbor and a milling complex of large proportions.

A seal of St. Knut’s guild, dating to the 1200s, has been found in Tum-matorp. This was a merchant’s guild, indicating that the site was by this time characteristically urban, with enough merchants to warrant a special associ-ation found in many Danish urban places of the time. Around the same time, a lepers’ hospital was built there, a thing that was found only in cities of some importance. By the 1200s, a large iron industry is seen in the discovery of many large pit smelting ovens, slag heaps, and standardized bars of iron from many places around the site. Coins found in the slag pit and ovens date from between 1180 and 1202.

The Late Viking–Early Medieval was a period when royal administrative centers were being founded in many economically important areas, as the long-distance trade of the Viking Age gave way to more local trade of the Mid-dle Ages, and local administration was needed. Tummatorp was in one of these areas (Redin 1972:76). Tummatorp underwent a gradual development over the period, from about 1050 when central place functions began, to 1150, when highly increased royal activity is already in effect, to the 1200s when crafts and trade dominated a large urban center.

The excess population that did not move out to the torp settlements may have been in Tummatorp, and by the mid 1100s, another new coastal town, Simrishamn. The quickly-founded Tummatorp, which lacked a normal town’s period of development, should have had a somewhat lower popula-tion density than west Danish cities, and less than Lund, which contained the Archbishopric and was close to Havn (later Copenhagen) and Roskilde. Turn-matorp may have had the urban density of west Danish towns by 1200, but at 1075 to 1100 a population of 1,000 to 1,200 or so would be more reason-able. This redistributes the population fairly precisely, showing not growth with the new torp settlements but change in nucleation.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (274)


If a royal official ran things at Tummatorp and local intermediate deci-sion makers were acting in the large villages, then the creation of these small places would also have improved the span of control considerably. Fig-ure 8.19 shows a map of settlement after these changes. Two important runestones were found in situ in the village of Simris, dating to the late Viking Age, and give interesting evidence of the village’s occupation during this phase. One is directly datable: it reads “Sigrev raised this stone over Forkun—father to Knut’s dräng Asulv, God help his soul.” This refers to thefather of a warrior who served Canute the Great sometime between A.D.

1014 and 1035. The second stone also dates to around the same time, and reads “Bjarngeir raised this stone over Ravn, his brother, sven (squire or page) with Gun (Ulv) in Svitjod.” These stones are of the after-Jelling type that emphasize relationships between officials, warriors, and lords rather than being simple family memorials. Their style of carving and the names of the men are non-Scanian. These represent “outside” elite who served central authority and lived in the eastern province—the kind of “foreign” lords the Scanians would shortly revolt against.

Figure 8.19. The Järrestad region, ca. A.D. 1050-1 100.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (275)


The radical adjustment of people and places strongly affects the rank-size curves. The curve by 1050 to 1100 has become much more linear (Fig-ure 8.20), with RSIs of .125 and .261 reflecting these changes, and in fact they are somewhat primo-convex, indicating that one system has been su-

Figure 8.20. Rank-size curve for Järrestad, ca. A.D. 1050- 1100, based on population estimate and size in hectares.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (276)


perimposed on another and that the new system may be largely oriented to-ward the cheap extraction of surplus. This would be the era when coins were first being struck in Tummatorp, when the monastery was founded, and when a royal mill monopoly was operating on the river, forcing every-one in the härad to grind their grain in the king’s mills, leaving behind a part of the flour for taxation. The naval levy that assessed military service based on land, land that was counted by village regulation into tofts, was in effect. In Järrestad, six “bäkenhög” terrain features are present in villages of all foun-dation periods, showing a very regular network of signal fires. It is possible that every village that was on a sight-line with those further off containedone of these.


From archaeological evidence, sagas, and early historic material, one can infer that during much of the late Iron Age, elite lived in the villages along with less powerful neighbors. As the period moved toward its close, there are some indications that magnates left the village milieu and established estates outside the community, as part of the trend of developing social classes. The social and economic distancing of the elites from the common people, as the Germanic social code broke down and the Medieval system began, is re-flected in a physical distancing. Class differences were always present in the Viking period, but as royal power grew less reliant on consent of the masses, more centralized and less populist, the lords and magnates who supported and acted in the name of the Crown reflected this separation spatially. Gislövand Östra Nöbbelöv together provide clues to the state of local religion andthe changing roles of lords in Christianity.

Gislöv, first recorded on paper in 1145, combines a man’s name, proba-bly Gisli, and “löv,” inherited or bequeathed goods. This sometimes refers to family inheritance but often to estates granted by a king to an important re-tainer or lord. It is one of the older villages, dating to between A.D. 500 and 700, and its Iron Age extent is very large, but, unusually, it does not have a vil-lage church. Instead, the church is located at Östra (Ö.) Nöbbelöv. Nöbbelöv, a corruption of Nybølle (“new fields”), is contemporary with the torp settle-ment wave.

Gislöv and Ö. Nöbbelöv are in a unique relationship. They are much closer together than villages normally are, and their sizes, relative to the church location, are unusual. According to all existing records that date back to about the 12th and 13th centuries they have always been discrete villages. Surface collection shows that Gislöv was occupied earlier, but Ö. Nöbbelövwas also occupied during the Late Viking Age. Although Gislöv is very large,

Landscapes of Power, Landscapes of Conflict - [PDF Document] (277)


Ö. Nöbbelöv is tiny—today and in the past—but this small site, barely a half akilometer from Gislöv, is home to the local church, while Gislöv has none.This almost certainly indicates the out-movement of a village magnate or local lord, who removed his household from the village and established a sep-arate compound away from the rabble. This occurred at the time when wooden churches were replaced with stone, perhaps around 1075 to 1100. The wooden church in Gislöv would have been abandonned and a new stone church built on the lord’s estate in “newfields.”


By 1200, Tummatorp contained all its high medieval features and was still a kungalev, or king’s town. It contained a merchant’s guild, a hospital, several churches, and active industries in iron-working and other town-crafts. Figure 8.21 shows the location of St. Knut’s merchant’s guilds in early Medieval Denmark; only the largest and most mercantile places contained such guilds.

Figure 8.21. Danish towns with St. Knut’s Guild in the Middle Ages (after Wallin 1949).

Landscapes of Power, Landscapes of Conflict - [PDF Document] (278)


Figure 8.22. Simrishamn, ca. A.D. 1200 (after Andren 1974).

As stated in previous chapters, Tummatorp had become the political, eco-nomic, and religious focus of the region. Simrishamn (Figure 8.22) was also burgeoning at this time, taking advantage of the new and unusual runs of her-ring off the Scanian coasts that was to make Scania the richest province in Denmark for over 200 years.

Archaeological investigations (Jacobsson 1979a, 1979b; Pettersson 1991) have found many of the ubiquitous “clay bottoms” of the herring in-dustry, waterproof basins where tons of fish were dumped and processed in some way (salting or pickling) and prepared for export to every point in on the continent, from Germany to the Mediterranean, to supply Friday fish to Catholic Europe. The largest markets, at Falsterbo and Skanor in southwest Scania, have revealed hundreds of these features, and Simrishamn was among one of the larger herring towns of the 1200s, and so must have had many such features. Thus, by 1200 two towns were present in the Jarrestad area, both with central-place functions and institutions and both containing strong central investment and interests.

Figure 8.23 shows the location of town churches and village churches in the period of 1100 to 1200. The village churches were still attended by local people for their Sunday worship, but church hierarchy, church law, church records, and church finance in terms of tithes, mills, and other entitlements were now centralized in the towns.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (279)


Figure 8.23. Town and village churches ca. A.D. 1100-1200 in Järrestads Härad. Also shown, Ingelstads Härad to the south and part of Albos Härad to the north.

The rank-size by 1200 (Figure 8.24) is a mirror of the state’s primate pat-tern in the same time period as shown in chapter 6. The RSI for population has fallen into the negative numbers, at – .87. The Scanian uprising and its bloody finale had recently quelled any thoughts of independence. The loca-tional analysis of Järrestad’s new spatial arrangement indicates that even the most distant, eastern part of the Eastern Province had come under strict con-trol and was managed and administered in an orderly way, yielding up its taxes and service with no more problems.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (280)


Figure 8.24. Rank-size curve for Järrestad, ca. A.D. 1200 based on population estimate and size in hectares.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (281)



This chapter discussed the transformation of the landscapes of Järrestad from the Roman Iron Age to the early Middle Ages, from about A.D. 100 until 1200. In the political landscape, the diffuse and non-hierarchic structure becomes highly centralized. Political control is transformed from the decentralized Ger-manic structure to a system of sovereign rule: from the corporate to the net-work mode (Blanton et al. 1996).

Similar change is seen in the economic landscape. In the agricultural economy during the course of this long time span, shifting cultivation ends, permanent arable is created, and small kin-based farms are replaced by ag-glomerated villages. Functional differences between places leads to the emer-gence of a settlement hierarchy. At the period around A.D. 1000 to 1050, these loosely organized settlements appear to become highly regulated, spatially and in terms of governance.

In the nonagricultural economy, elite trade in long-distance prestige goods appears in the Roman Iron Age but becomes less prestige-oriented and more grounded in staples when attached specialists are replaced by the craft-workers of the emporia of the early Viking era. In the small study region, the emporium lay at Stora Köpinge, which in turn gives way to a royal town, Tummatorp, in the late Viking Age. In the Early Medieval period, a coastal market town emerges at Simrishamn. In terms of economy and political econ-omy, the Järrestad landscape is a microcosm of Denmark’s development.

Trade also becomes less international in nature during the Viking Age, when the primary concern appears to be draining goods out of the provinces and into the hands of the larger state. The primo-convexity of rank-size curves at about 1050 supports this; two systems are operating at the same time: a primate structure superimposed over an older landscape is seen in this pat-tern, and the intent of diverting economic wealth with minimum cost has been noted for this type of rank-size curve.

During the transition to Christianity, there is considerable continuity as noted by Fabech (1994), for although the belief system changes, religion stays in the control of local lords. At the time of transformation ca. A.D. 1000to 1050, this changes and religion becomes controlled in the towns, the local version of this being Tummatorp. The eight small local village churches were obligated to bow to the will of the “mini-bishopric” at Tummatorp.

In all this there is the unwritten and unrecorded conflict between the old and the new elite, and the Germanic structure of fluid social classes ver-sus the subjugated peasantry envisioned by the Knytlings—as the royal line was called after the conquest of England by Knut the Great. Old lordship built and controlled the village shrines and later churches and oversaw and inter-fered in the local economy, and minor chieftains and village elite were de-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (282)


pendent upon a Jarl or other high-level lord who controlled the whole river valley politically.

When Stora Köpinge intruded into the local economy some of this power and wealth was diverted into central hands. When the cultural land-scape was disintegrated and reformed during village regulation—and, to para-phrase an old Scandinavian saying like lightning from the clear heavens, Tummatorp fell down from the sky out of nowhere (barely missing Jar-llestatha and its local lordship by a scant kilometer or two)—this conflict must have reached a critical point. The Scanian uprising was fomented by lords and farmers from all over the province; no doubt these long-aggrievedformer rulers and independent landowners had compatriots in JärrestadsHärad.

But the power of place now lay in the hands of the king in Roskilde. The transformation and restructuring of the landscape included the transfer of power from the functionally disarticulated, locally-controlled places of the Late Iron Age into the functionally conflated, royally controlled central places of the Late Viking Age and Early Medieval had worked very effectively to en-sure the failure of this rebellion.

The next and final chapter will deal with conclusions and implications of the analysis of Järrestad, Scania, and Denmark for the study of south Scan-dinavia and other cultural sequences in time and space.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (283)

This page intentionally left blank.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (284)

Chapter 9


The following sections summarize the precepts and findings of the study, and also highlight some of the contributions this work has made to local and re-gional archaeology, archaeological methods of analysis, and several aspects of archaeological theory.


The overarching mechanisms of change that can be followed throughout the Danish state formation sequence are the strategies of power versus force. Power is the ability of a ruling group to impose its will while avoiding the costly depletion of resources. This can be achieved through the creation of a hegemony, where political compensations and economic incentives induce allies to become clients, and the creation of false consciousness, where so-ciocultural values promoted by the dominant group bind peer groups to-gether within a “shared” cultural gestalt. Power is cheap: it consists of ideas, values, associations, and ideologies. Force is expensive: it must make use of warriors, supplies, and money. Danish leaders correctly recognized that power might be nearly as effective as force for achieving their ends.

The approach I have taken to the study of power and force is to follow their traces, archaeologically, textually, and geographically, through space and time. The concepts of power and force were operationalized by examining large-scale, long-term change in the organization of society through locational analysis and the interpretation of landscape change, on a variety of scales and at several slices of time, finally focusing on Järrestads Harad as a dynamic and changing human construct under three different sociopolitical/economic regimes. Other indicators included cultural texts such as laws, sagas, and


Landscapes of Power, Landscapes of Conflict - [PDF Document] (285)


ethnohistoric documents. In this way, many centralizing processes that accom-panied state formation are illuminated, and the geographical concepts of region, location, and the “power of place” are integrated with anthropological ideas about change in social organization, political economy, and state formation.


To frame the study in terms of the roles of power and force, four hypotheses were advanced. The first was that despite the Germanic social code, South Scan-dinavia’s relationship with the encroaching empire of the Franks facilitated at-tempts at unification by would-be kings. The second was that voluntary political unification created conflict between preexisting local elite and new elite who represented central authority. The third considered whether or not the struggle for jurisdiction of the region, and the reluctance of preexisting elite to relin-quish authority, led to the slower integration of Scania with the Danish state. Fi-nally, it was hypothesized that while central elites used social and ideological strategies for the purpose of integration and control, because of long-term re-sistance the actual disruption of old elite systems was eventually necessary.

The analysis of greater Denmark showed that political change was rapidly matched by integrative improvements in the administrative structure of the western kingdom, while areas within Scania continued to exhibit the organi-zational and locational features of discrete social aggregates. Their persistence showed that they were not merely vestigial, but still actively functioning as in-ternally cohesive sub-regions. I believe the study supports the model in which unification was achieved through a succession of elite incursions into local sys-tems, beginning with hegemonic controls and social interactions similar to those that cement alliances, such as gifts and marriages, followed by an attempt to dominate market exchange, followed by reorganization of the landscape and an investment in infrastructure. Finally, the most costly strategy, the armed con-flict of the Scanian Uprising, involved the death of thousands of food-growing,craft-producing, warshipmanning citizenry. It was a last resort, long avoided and very costly in terms of manpower and internal public relations.


It is first and foremost hoped that the Järrestad project will contribute to local and regional archaeology through the addition of new, detailed data concerning the study region, and the further illumination of settlement dy-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (286)


namics through time. As an addition to the body of work produced by my es-teemed Swedish colleagues, I intended that it might enable productive gen-eralization about other nearby polities, where similar processes may have occurred, providing a comparative basis for other studies of settlement change and continuity in south Scandinavia.

I also believe that the study adds to the body of knowledge about changes in social and economic organization during a period of significant political change in the macro-region, Western and Northwest Europe, where far-flung systems appear to have been interdependent in many ways. The out-come of these processes in part determined the shape of the cultural land-scapes of Europe today.

I also hope the Danish state formation model will enrich cross-culturalresearch through its value as a comparative case study. This study advocates the use of various types of analytical methods: the statistical evaluation of lo-cational geography, the laboratory analysis of soil chemistry, and the contex-tualization of archaeological remains through the documentary record. It is hoped that through the combination of these rather different types of analy-ses, new insights on prehistoric processes may be gained. Furthermore, as noted earlier in the study, methods such as rank-size analysis, span of control, and study of levels of organizational and administrative hierarchy have never been performed for any part of southern Scandinavia. I hope this type of re-gional analysis will one day become part of the repertoire of northern Euro-pean archaeology.


Rank-size analysis remains one of the most useful types of archaeological meth-ods, and renewed interest in its application has recently been seen (Falconer and Savage 1995; Stein and Wattenmaker 1989). I hope to contribute to the continuing use of this type of analysis, and also hope that I have added some-thing to it. This study makes use of rank-size interpretation on the large scale, which has a long tradition in archaeology, but also on the small scale, with great utility for the interpretation of shifts in power and the flow of political change. The contextualization of this data within the history and ideology of South Scandinavian culture has added another dimension to the interpretation.

The study resulted in some welcome refinements to the method of loca-tional and organization analysis. I hope that the micro-analysis of subregional rank-size curves may provide models against which others may test their own data. In addition, working with rank-size curves covering very small incre-ments of time showed that much could be discovered about internal organi-zation by examining very small degrees of difference.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (287)


Analyses of sub-regions and brief time intervals together revealed that there are different levels and types of rank-size convexity that mean different things. Using histograms to view the levels of hierarchy and the relationship between sites in a convex curve, I was able to distinguish some clear differ-ences between curves that appeared very similar. For example, in examining the convex curves of western Denmark and Järrestad both at about A.D. 850,I found that while both are convex, the histogram of the former shows a clear top level of site both in size and function, with others subordinate to it, while the latter shows many sites of equal size and function and no such nascent hi-erarchy. Furthermore, the interpretation of primo-convex curves as showing two simultaneously operating systems, one superimposed on the other (Fal-coner and Savage 1995), could not have been supported more perfectly than in this Danish example.


An important finding concerned the development of village size and popula-tion through time. Attempts to gain insight into changes in site sizes led to a clearer understanding of the Iron Age land tenure system: it did not necessary follow that sites would expand even when population rose. In many cultures, primogeniture is the main form of land inheritance system, under which the first-born son (or daughter) gets the entire holding from the parent and the others must find their own way. In Denmark, this was not the case. Rather than primogeniture, the land tenure system dictated that the land be equally divided among sons, so that for many generations the village could stay rela-tively the same size with smaller internal divisions. When land became divided into holdings too small so support a family, interior colonization, as described in chapter 8, became the primary factor that affected site size. Villages did not necessarily continually grow in size, because excess population was distrib-uted in new sites. These waves of settlement foundation largely supplanted the process of meaningful site expansion, and the most phenomenal sequence of change in site size is not seen in prehistory as a simple factor of time and demographic trends but in the transition to historic times with the large-scalerestructuring that occurred at the end of the Viking Age.

Even though villages produced buds rather than growing larger, the un-derstanding of prehistoric village size was still vital to the study, for as villages were added through time during waves of foundation, their hierarchic orga-nization and interrelationships changed. The fieldwork assembled informa-tion on the sizes of the majority of the late prehistoric settlements in the harad so that these sites could be collectively analyzed for rank-size integra-tion and levels of settlement, administrative, and marketplace hierarchy.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (288)


The village had previously been a self-contained autonomous world with its residential core, its fields, and its “wilds.” Earlier interior colonization by vil-lages in “by” had utilized areas where this universe could be reproduced in miniature of the original oldest villages. The earlier format had been a satisfic-ing strategy, not fully exploiting the capacity of the land to produce grains but reserving it for forest products and foraging areas. Activities and land-use in the early village was something that was discussed and planned among villagers and then carried out by them. The late Viking Age transformation of all these vil-lage’s micro-hinterlands into more and more tilled fields is both a symbolic and a tangible sign of the loss of autonomy to the will of the state. The latter is a move toward maximizing, and was, I believe, imposed by the state.

Another important piece of evidence was the demographic change ac-companying Late Viking Age settlement change. Population estimates in Jär-restad indicate that the number of people living in the valley did not change significantly at the time of regulation and restructuring but that their loca-tions in the landscape were significantly altered. Some part of the population was nucleating around the regional center, while others were sent to colo-nize outer areas and increase agricultural hectarage. It was not one of the goals of this study to closely examine this phenomenon, but I will note that since the regulation and shrinking of the large old villages is fairly clearly a centrally motivated design, then the shifting of “excess” people into small fringe settlements that did nothing but produce crops may also have been a centrally conceived plan (Drennan 1988). The study supports the idea that the appearance of centers is likely to create nucleation around their various attractions, such as increased access to coveted goods, foods, and services, leaving productive land idle (Feinman and Nicholas 1987, 1990). The change in rank-size curves is merely a reflection of these circ*mstances.


Although I have already stated that I hope locational analysis sees wider use in Scandinavia, I also hope to bring the merits of soil chemical characteriza-tion, especially of phosphate, to the attention of American archaeologists. Many questions still exist for most archaeologists who are unfamiliar with its application. Attempts have been made to use this method in inappropriate circ*mstances (White 1978; Skinner 1987)—for example, on very short-term, sites such as hunting and gathering campsites occupied for a few days or weeks, or on sites that are relatively short term. Five or 10 years of occu-pation may be a long time to occupy a site in some cultures, while in others 250 to 500 years is the norm. This method is best used in situations where really long-term, sedentary occupation is seen. Such methods, when used

Landscapes of Power, Landscapes of Conflict - [PDF Document] (289)


correctly, deserve the attention of a wider English-speaking audience. Al-though this is not the place, perhaps, for lengthy discussions of technical and laboratory methods, I hope to explore this topic in an appropriate forum.


A closer study of surplus production and labor and its relationship to the growing power of the state would be a worthwhile topic for future investi-gations in Denmark. The fact that the state developed mainly in areas lacking in surplus-producing agricultural land and appears to have imported most of its staples from far-off provinces is noteworthy. I argue that contrary to many simple ecological models with implicit expectations regarding the location of states and the availability of productive agricultural land (Meggers 1954; Sanders and Price 1968; Carneiro 1970; Sanders and Nichols 1988), the nexus of power in early Denmark was located in Jutland, the region of poor-est agricultural potential, from which elites coerced, coaxed, or negotiated the importation of surplus from richer, more distant areas. Since the 1950s, the idea that a state must rise in an area capable of producing a rich agricul-tural surplus has been inherent in many archaeological models of state for-mation (Meggers 1954; Steward 1955). These ecological models, which view the distribution of arable land as a limiting factor on the evolution of states, persist strongly today, and form the core of several debates over state and urban origins in Mesoamerica and Amazonia, among other places. The debate between William T. Sanders and various coauthors (Sanders and Price 1968; Sanders and Nichols 1988) and the opposing theories of Blanton (1980, 1983, 1988), Kowalewski (1980, 1982), Feinman and Nicholas (1990b), Fein-man et al. (1985), Flannery and Marcus (1990) is an example of this, as are the contrasting views of Steward (1946–50), Meggers (1954, 1971, 1985), and many current Amazonian anthropologists (Roosevelt 1994:2).

In western Europe, premodern societies where states are poor or corn-pletely lacking in immediately surrounding agricultural land are not unheard of. Besides Denmark, a good example of this is the Papal State, which from the late sixth century until A.D. 1871 , was a patch of land only a few square miles in size, ruled by the Pope and owned by the Roman Church. From this tiny realm, especially between about A.D. 580 and A.D. 1300, all of Europe was ruled, not only by power but by force: the Papal State during the Middle Ages often levied large armed forces, whose soldiers came from distant places, and took sides in secular confrontations. This immensely powerful state had little surplus-producing land; in fact, it had no hinterland whatsoever, and consisted of Rome, including the Vatican, and the surrounding area (Gonzalez 1984:246,

Landscapes of Power, Landscapes of Conflict - [PDF Document] (290)


289, 311). The Pope commanded surplus wheat and other foods from Sicily and other nearby areas at first, and later this “tribute” extended throughout Christendom. Based entirely on an ideological framework, a system of belief, a tiny locus in southern Europe (this location also chosen for purely historical reasons) commanded kings of every nation, and took surplus, gold, and the flower of Europe’s intellect to support its priestly regime. How would this look archaeologically? While the goals of this study do not include entering the contentious debate over whether such a circ*mstance is “possible” (i.e., Sanders and Nichols 1988) it certainly presents a case study that illustrates one part of the discussion. We should be far less surprised to find such alternative types of states among the many possible variations. We should develop and use methods for identifying, studying, and explaining political structures that do not match expectations rooted in our own experience.


As discussed in the introduction to this book, the sociopolitical processes in South Scandinavia that are reflected in locational and organizational change follow sequences similar to those in other regions. Throughout this study I have tried to highlight places where social, political, and economic structure and change are cross-culturally comparable. I have tried to stress the tension between heterarchic and hierarchic modes of sociopolitical structure, and the conflict between corporate and network, or exclusionary forms of rule in Denmark (Blanton et al. 1996) as well as the problems of alliances in holding together diverse and fragmented peoples. The relationship between the state, demographic change, and agricultural production (Feinman and Nicholas 1987, 1990; Feinman 1991) may have important implications for changes seen in Denmark, and the Danish example may shed light on this process for others. The concept of conflict and negotiation between classes (Saitta and Keene 1990; Saitta 1994) is supported by strong evidence in RIA and GIA South Scandinavia: collective, communal ideologies spawn incipient political hierarchies and must painfully negotiate their development and growth. As political systems rise from communal modes, a small subregion may be squeezed between two states. Autonomy must be surrendered in order to gain protection and leadership in such circ*mstances (Hastorf 1990).

Finally, it is interesting to note that what might be imagined as a rad-ical social shift was in actuality far less disruptive. I refer to the centrally-encouraged abandonment of a religious tradition of nearly a millennium’s duration or more. In fact, this transition was smooth, with the same elite main-taining their grip on the control of the sacred landscape and its ceremonial sector. This suggests that when working in an entirely prehistoric setting, one

Landscapes of Power, Landscapes of Conflict - [PDF Document] (291)


cannot assume that widespread replacement of religious iconography and ap-parent changes in belief systems always produced social or political upheaval. Instead, the strongest backlash in Denmark came with the perceived threat of a decline in regional autonomy and the erosion of the power among the landed peasant and lordly classes. This appears to have polarized south Scan-dinavia along ethnic lines and socioeconomic class divisions. Although issues such as belief systems, political autonomy, and class conflict are often implicit in studies of social change, I hope this case study will provide alternative ideas about what types of change are most likely, as noted in chapter 1, to spawn state-produced “official” histories and violent conflict, while other seemingly momentous changes are relatively uneventful.


What is the alternative to the model of state formation presented here? 0th-ers have proposed the Church as the unifying and state-forming body in Den-mark (i.e., Cohen 1977). Yet this research has demonstrated that integration began long before the church was established in Denmark. It has also been suggested that urbanization was the product of foreigners: Frisians, Franks, and later the Hansa, traveled to Denmark and set up trade and towns while the Danes acted as rustic spectators (S. Cohen 1977:314). All but the most conservative of historians gave up on this theory by the 1970s, as a result of the publications of excavations of sites like Hedeby and Ribe that indicated that the trading settlements were clearly in native style. It also has been the-orized that towns arose for diverse reasons in Denmark—some at markets, some at crossing places, some at fortress sites. As regards western Denmark, I agree with this. But the careful and rapid placement of royal foundations— first markets, then towns—at previously unimportant places close enough to earlier power nodes to successfully challenge their authority, were tactics largely limited to Scania. It seems fairly clear, too, that the Scanians did not consider themselves to be Danes, that they abhorred feeling the hand of cen-tral government upon them, and that they probably resisted every attempt to coax or coerce them into cooperation with Jelling and later, Roskilde.

The only remaining alternative explanation is that the incorporation of Scania and the unification of Denmark was all a product of chance, a nebu-lous progression from one thing to another. I hope I have argued convinc-ingly enough that this was not the case, and that left to itself, the corporate structure of rulership and decision making in the Germanic tradition, as re-flected in the landscape, would never have prevailed for so long and then changed so rapidly if it had been free from outside influence.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (292)



When it first occurred to me that the problem of state formation in Denmark would be a topic I might spend several years investigating, I conceived of a plan to analyze the political developments in the region through regional and locational analyses. However, for the multitude of reasons discussed in this study, I needed to modify these methods in order overcome their limitations for Denmark, and also to take advantage of some of Denmark’s more unique opportunities.

In formulating a research design that incorporated the many contexual-izing elements available in the Danish record—documents, literary texts, myths, laws, and so on—I had to maintain empirical, objective research methods while incorporating the dreaded components of intentionality, context, ideology, class, and other bugbears that many archaeologists fer-vently (and understandably) avoid, and some over-enthusiastically embrace and over-imaginatively apply. I had to present the outcome of the research in a way that would convey the possibility of a study attempting to integrate these approaches.

In the beginning of this study, I proposed that the analysis of the cultural landscape was a viable, and even a necessary method of examining pre-historic societies, especially those that do not conform to more familiar or obvious political configurations, for embedded within places and their rela-tionships to each other across both space and time are large and often un-seen chapters in the unfolding of social history. I have been asked how this is different from traditional regional surveys done in archaeological context, and although there are some exceptional regional research projects in the lit-erature, many rely solely the examination of change in settlement systems and artifacts distributions through time. Proponents of a cultural geography that links together power relations, place, and history have called “place” an historically contingent process (Pred 1986). Since history is made of unend-ing actions and interactions at places in the landscape, places cannot be par-titioned from the development and transformation of society.

I have tried to include a rich variety of documentary sources with a re-gional archaeological approach. Through this method, the apparent contra-dictions between history and archaeology are shown to reflect a long and difficult transition that was much more complex than earlier studies have im-plied. Despite the imposition of direct Danish rule during this period, ethno-historical texts such as the Knytlinga Saga (Palsson and Edwards 1986) describe several assemblies where kings and middle-class landowners were in open conflict over royal rights to land and taxes. Archaeology and docu-mentary evidence both indicate resistance by old Scanian elites and their sub-jects. Significantly, the old social code that sanctioned regicide in cases of

Landscapes of Power, Landscapes of Conflict - [PDF Document] (293)


perceived despotism was replaced with the power of the king to enforce the capital punishment of dissenting citizens.

Records from the Early Medieval period tell of upheaval and revolt, aimed primarily at the centralized state. People refused to pay taxes, tithes, and serve military duty (Andersson 1947:412). The physical assault of royal officials who were non-Scanian appointees was common, and demands were made to the king that “foreign” elite (from west Denmark) be removed. By A.D.

1180, the Danes had been attempting to negotiate or coerce the Scanians into submission for 300 years, yet central elites were still perceived as in-truders. Several treaties were attempted, but their concessions were not enough for the Scanians. In 1180–1182, a civil war known as the Scanian Up-rising occurred in Denmark, as farmers and their local leaders clashed with the king (Andersson 1947:416). However, the quest for Scanian indepen-dence was not lost forever. The appendix at the back of this book illustrates recent attempts at representation or secession by proponents of an indepen-dent Scania. The inclusion of Scania in the “official” Fourth World—unrepre-sented peoples—along with other powerless and struggling indigenous groups, shows that the struggle for Scania is far from over, though it has re-emerged in a different and perhaps unexpected form as it has entered the in-formation age. This electronic political insurrection should be a good endnote for anyone who has read this study.



Let us now look at the linkages between power and place and between indi-viduals and institutions as conceived by behavioral geographers such as Torsten Hägerstrand. By borrowing paradigms from a field concerned with ideas so vital to understanding the prehistoric past, I try to move toward my goal, stated above, which was partly to examine aspects of society that ar-chaeologists often consider to be unrecoverable. Yet we can say for sure that a state is made up of individuals whose present is linked to past and future in time and space, thus affecting the course of events. We can articulate this by examining the concepts path and project, concepts that are a way of de-scribing the actions and events that make up an individual’s life and form a continuous stream of activities (Hagerstrand 1982). Each person forms an in-dividual, internalized biography as he or she moves in this continuousstream, which could be mapped over time and space. This is the path.

The project is the idea that any goal or intention is accomplished through a series of simple or complex tasks. These tasks push together the paths of two or more people at specific places in space and at specific times. The paths and projects are carried out through interaction with tangible ele-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (294)


ments of culture: a plow, a sword, buildings, furniture, roads, or raw material (Pred 1986:10).

As individuals pass through life performing specific activities at certain places, all the while forming their own internal biographies, they undergo socialization—the absorption of social rules by individuals, who both per-petuate and change these rules through time. Structuration is the term both sociologists and behavioral geographers have given to the process where the individual, undergoing constant, lifelong socialization, actively shapes the so-cial structures and institutions of society and in turn is shaped by them. In other words, the two are always reforming and “becoming” each other (Gid-dens 1984). The specific social situations and material things that form struc-turation processes are perpetually spelled out by the intersection of individual paths and projects at specific times and places.

Social structure is, in effect, the rules and power relations that are al-ready built into a specific society and its geography and history. Rules and power relations constrain and enable humans but are also created by them. Rules can be spoken or unspoken, formal or informal, written or unwritten. They exist among individuals, groups, classes, and institutions as well as be-tween these categories. They are expressed as behaviors in certain contexts and places. All practices and social activities are concrete interactions in time and space. Thus, structuration is specific to time and to places (Pred 1986:9).

I have continually emphasized that there are methods other than force with which elites can effect change in society—for example, by working to in-fluence social reproduction. Let us look at this a little closer. “Social reproduc-tion” is not a monolithic thing; it is something perpetuated by individuals, the individuals whom we cannot see archaeologically. However, we know that the performance of activities, related to both mundane and special institutions, transform and perpetuate the institutions. Elites can insert themselves into this series of processes and substantially change them. Power relations affect place and space through time because someone or some group decides “what places and structures will be improved and perpetuated, which will stagnate, which will fall into disrepair, and which will be demolished” (Pred 1986:25). Who is it that makes these decisions? This is the key to the control of places, the paths and projects that converge there, and the structuration processes that move people into their future. Ethnohistoric documents such as Knytlinga Saga and its description of the battle between state and subject in the context of the assembly, reveal how the power of social memory and the cultural con-struction of group identity can be manipulated. By examining the available his-torical and documentary evidence, at least some legitimate investigation into the roles of individuals, acting as members of groups, can be made.

The power of one group over another lies in its ability to force others to participate in duties, obligations, or tasks, or parts of tasks, that the powerful

Landscapes of Power, Landscapes of Conflict - [PDF Document] (295)


have defined, or to prevent the participation of others. These tasks are linked to places, and thus compulsion or restriction is expressed in the spatial pat-terns of activity and residence. The disenfranchisem*nt of others can be in-duced through the employment of laws, rules, sanctions, and economic obstacles or, as Pred states (1986:26), “through the mobilization of knowl-edge that has been kept in bureaucratic records or gathered through surveil-lance . . . or through the making of prior claims on limited time resources or through the actual or threatened utilization of force.” In this way, the chain of biography formation, socialization, structuration, and social reproduction permits us to study, obliquely yet meaningfully, the ways in which the indi-vidual, the community, and the state are bound together through relation-ships of power and resistance.


In the Early Roman Iron Age, the cultural landscapes of Denmark were disar-ticulated. The political landscape, economic landscape, and sacred landscape each had their powerful places, but they did not intersect with each other; they lay in different places in the physical environment. This reflects the fact that the social institutions governing these places were also functioning mostly independently of each other. The warlord’s hall was not the site of the assembly or the sacrifice, and farmers did not exchange produce at elite sites or religious shrines. During the course of the Roman era, this began to change. The economic landscape began to articulate with the elite political landscape and the sacred landscape when elites took over the importation of Roman goods and set up centers at holy places, although this only involved the elite prestige goods economy.

In the Germanic Iron Age, the sacred and elite political landscapes began to share nodes to an even greater extent; the old lakes and bogs were aban-doned and offerings began to appear within elite compounds. During the age of emporia, the Early Viking Age, elite political landscapes began to intersect with the long-distance trade economy. Yet the ting, the assembly, remained dis-articulated from elite political landscapes. The sacred landscape, while now under elite management, was still controlled from nodes in an ancient elite landscape, unchanged for many centuries. Domestic exchange—farmer to fisherman to artisan—was still operating under local systems. During the Late Viking Age, the sacred place, the marketplace, and the assembly were re-moved from the last outposts of the disarticulated landscapes of the earlier Iron Age and conflated in new, royal towns. The planes, or layers of landscape, so to speak, had shifted until all the nodes on all levels intersected both in space and time, and in control. The state had wrested control of these institu-

Landscapes of Power, Landscapes of Conflict - [PDF Document] (296)


tions from the old elite. At the same time, the power of the farmer was dimin- ished by the dismantling of centuries-old village landscapes and the founding of many small, new, “power-less” places. Although these practices were protested, much of this occurred without violence or insurrection.

This is what makes place such a powerful tool for societal change. While sending in armies to control an intractable and self-willed people highlights the attempt at manipulation by force, the shifting of the power of place is subtle and likely to go unnoticed. Place-bound structuration is affected by power relations, but the power relations themselves are constantly undergo-ing change through the performance of routine and non-routine practices. Places are human beings’ tangible and physical link to social reproduction, and continuity in the places and spaces they traverse is what carries social structuration into the future. Take away or change these places and the path is broken. Change the practices, paths, projects, and places, as central elites did during the Late Viking Age, and social structures and social reproduction will change themselves. The manipulation of social systems, and economic and political institutions can be insinuated at this point, and although it can be very slow, it can be very effective.

As noted earlier, radical change in the organization of society can be ac-complished through combinations of force, the threat of force, and the cre-ation of false consciousness. During state formation in Denmark, a corporate society was transformed into a far more hierarchic system, a process that met a conscious resistance. Rather than mounting a full-scale war against their rich and well-soldiered eastern province, these safer methods were used by the cen-tral elites, and social transformations necessary to enact these changes were mediated through shifting control of the landscape. If the landscape is not stud-ied as an integrated whole, many parts this process would be invisible.

To conclude, social reproduction, from generation to generation, elder to youth, parent to child, occurs in the performance of activities relating to vari-ous, long-standing institutions. These may be mundane, involving farm, mar-ketplace, and village, or maybe special weekly, monthly, or yearly ceremonial or political activities and institutions, such as the church or the assembly. Per-formance is intimately tied to places where activities occur. Change the places, and you can change social reproduction. The manipulation of places is a pow-erful tool, a subtler, less costly way for a king to subvert the old ways.

Yet political scientist James Scott advocates that there are two kinds of false consciousness—a thick and thin:

The thick version claims that a dominant ideology works its magic by persuading subordinate groups to believe actively in the values that explain and justify their own subordination . . . the thin theory of false consciousness, on the other hand, maintains only that the dominant ideology achieves compIiance by convincing sub-ordinate groups that the social order in which they live is natural and inevitable.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (297)


The thick theory claims consent; the thin theory settles for resignation . . . subordi-nate groups allow elites to control the political agenda and create effective obsta-cles to participation. (Scott 1990:72).

The country folk of Scania and other disenfranchised parts of the Danish state, it seems, were only resigned, and not consenting. Although ultimately the Scanians did rebel and make war on the central state, enough had been altered to assure the failure of such an attempt; it was too little and too late. This was only possible because the subtlety of manipulating place had kept the Scanians off-guard long enough for rulers to build the state around and above them. On the largest scale, this is seen in the founding of new central places and the abandonment of old places of power throughout Denmark but it is just as powerful on the smallest scale, the reinventing of the village, the changing of the farmers’ fields, and the alteration of the paths of daily life. As predicted at the outset of this study, this battle for control of places and their material and symbolic value can be detected in the archaeological record. This should spur us not only to further archaeological study of domi-nation in the political process but of resistance as well.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (298)


Scania—SkanelandCulture Region in Scandinavia and in Europe

Scania, or Skaneland as it is called locally, is approximately 20,000 square kilometers in area. It is situated just south of the Smaland plateau and is sur-rounded by the Baltic Sea, the Oresund, and Kattegatt. The region consists mainly of open plains in the south, with wooded hills along the northern bor-der toward Smaland.

About 1.5 million people live in Skaneland. The main regional city is Malmo with a population of 250,000. Other larger cities are Halmstad, Kris-tianstad, Helsingborg, Karlskrona, and the university city of Lund.

Skaneland consists of three provinces: Skane, Halland, and Blekinge, which belong to Sweden, and the island of Bornholm, which is a part of Denmark. The provinces are governed by Stockholm (or Copenhagen in the case of Bornholm). There is no local administrative entity that governs the re-gion and no official institutions with a regional view of culture and history. The Organisation Skansk Framtid was founded in order to achieve a change in this regional situation. The Organisation Skansk Framtid has branches in Skane, Halland, and Blekinge and has support organizations in Copenhagen and Goteborg. Bornholm has its own independent organization.


Landscapes of Power, Landscapes of Conflict - [PDF Document] (299)


The Organisation Skansk Framtid

Skansk Framtid was founded in 1989. Its purpose is to work towards the restoration of Scania’s very special but rapidly diminishing culture. The final goal is to achieve a fully autonomous cultural status.

The Swedish central state has since the beginning attempted to erase Sca-nia’s cultural identity and regional awareness. In spite of decades of effort by individuals and local organisations this negative situation continues to prevail.

One reason is that very few channels exist today for regional organiza-tions to enhance the Scanian public’s cultural awareness or even make their opinions known. The Swedish central television and radio monopolies keep the doors locked to regional issues. The major newspapers have been very restrictive in dealing with what is sometimes been called the “Scanian prob-lem.” Official institutions disown the issue entirely. No Scanian history is taught in the schools of Scania or Sweden.

The Organisation Skansk Framtid (SSF) came into being in the wake of increasing evidence of frustration and general mistrust of the establishment among Scanians and people in other culturally divergent regions. SSF bases all of its activities on upholding democratic values and on international conven-tions and recommendations set forth by the United Nations, UNESCO, the Eu-ropean Council, and the Helsinki Accord, as well as the goals of minority organizations in Europe and worldwide.

The History

The recorded history of the Scanian people dates back to the fourth century, including a period of 800 years as an important part of Denmark. Scania was an-nexed to Sweden at the peace treaty of Roskilde in 1658. During the years im-mediately following the peace treaty the Scanian people suffered greatly through resistance actions and through actions on the part of the Swedish cen-tral state. Since then, using economic, political, and propagandistic pressures, the Swedish state has followed a policy of “re-nationalization” toward the peo- ple of Scania. Attempts to remodel Scanians into Swedes are still in progress.


The region of Scania fits very well the criteria for the identification of a nation: (a) an indigenous group of people with a common historical back-


Landscapes of Power, Landscapes of Conflict - [PDF Document] (300)


(b) inhabiting a well-defined territory (c) with a history and cultural background different from the surround-

(d) having or have had its own language. ing areas and


Scanian Period

Year 380: Alaric is the first person to claim the title Rex Scaniae.

380-770: There are 15 known Kings of Scania, of which five use the title Rex Scaniae.

About 770: Ivar Vidfamne, King of Scaniae, becomes the first empire-builderof the North.

Before 800: Harald Hildetand, the great King of Scaniae, reigns.

Danish Period

811- 1047: Danish kings rule the land of Scaniae.

Scanian/Danish Period

1047-1330: Through Sven Estridsen and sons there are Scanians on the Dan-ish throne.

1134–1330: Danish Kings rule the land of Scaniae.

The Period of the Holsteiners

1330-1360: Count Johan of Holstein rules the land of Scaniae.

The Commonwealth Period

1332-1360: Scania is in union with Sweden, Finland, and Norway under the King of Magnus Eriksson Smek. He uses the title Rex Scaniae.

The Danish Period

1360- 1376: Valdemar Atterdag rebuilds the state of Denmark and recaptures Scaniae.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (301)


The Guardian Period

1376-1397: Margarethe Valdemarsdotter rules over Denmark and Norway as guardian of her son Olof II and thereafter for Erich of Pommern.

The Period of the Nordic Union

1397-1412: Margarethe Valdemarsdotter (for Erich of Pommern) is the Nordic ruler.

1412-1439: Erich of Pommern is Nordic King.

1439-1448: Christoffer of Bayern is Nordic King.

The Danish Period

1448-1658: Danish kings rule Scaniae.

The Governoship Period

1658-1720: This is the period of governorship in Skaneland, described by the Swedish government as “a domestic but foreign territory.” Scaniae has its own parliament and a guarantee that it will be allowed to follow its own laws.

The Swedish Period

1720-1801: Swedish kings rule Scaniae.

1801-1809: The governorship is reinstated in the province of Skane.

1809-today: The region of Skaneland (Skane) is divided into three separate provinces under Swedish rule: Skane, Halland, and Blekinge. (The Island of Bornholm is under Danish rule.)

The Demands

1. The indigenous minorities within the present borders of Sweden must be officially recognized by the State.

2. The Swedish State must accept that citizenship and nationality are not the same thing.

3. The Peoples of the regions must be permitted to regain a truthful and complete view of their history and cultural background.

4. Sweden must come to terms with its past by satisfying a regional need for a true historical information process in schools and the like.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (302)


5. The State must actively promote cultural relations over the state bor-

6. The State must cease to deny the existence of the Scanian language. 7. Regional symbols, heraldry, and ceremonial objects associated with

the various regions must be given official status and acceptance as important cultural objects and given the same respect shown Swedish cultural objects.

8. The historical institution at the University of Lund must be given an assignment and financial resources for research and publication of the complete and chronological history and cultural background of the people of Scania.

9. The region of Scania must be released from the burden of being re-ferred to as “Southern Gotaland.” Gotaland is a very specific cultural and historical region in itself, which was allowed to “expand” for the sake of the Swedenization program. Culturally and historically it has nothing to do with Scania.

10. The State must, with reference to the concept of nationality and citizenship under point 2 above, reverse its recent decision to in-troduce an official Swedish National Day celebration on the 6th of June, to replace the Swedish Flag Day formerly on the same date.

11. The State shall encourage the people in Scania to erect statues and other official symbols to commemorate Scanian and Danish persons and events important to Scanian history.

12. A recovery program must immediately be put into effect that will replace and return important Scanian and Danish cultural and his-torical works of art, heirlooms, and relics from Sweden to a Danish/ Scanian cultural sphere of influence.

13. The Government and government institutions should act in accor-dance with the Swedish Constitution, first chapter, second clause, fourth paragraph, which states: “Opportunities for ethnic, linguistic, and religious minorities to maintain and develop their own cultural and community life should be encouraged.”

ders in Scandinavia.

The Activities

The organisation has carried out a number of activities during its existence. Some of the major achievements are listed below.

1. Public opinion analysis of the present cultural relations between Scania and Sweden.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (303)


2. A major campaign to persuade the Swedish Government to return heirlooms, relics, and works of art to Denmark. All were stolen from the Scanian/Danish region by the Swedes between 1657 and 1660 and stored in large quantities in various castles and museums in and around Stockholm.

3. Demands have been made to politicians and the teaching profession to reinstate instruction in Scanian history into the Scanian schools.

4. Demands to the Discrimination Ombudsman that his offices must begin to properly deal with the Scanian problem in accordance with the international conventions on cultural rights.

5. The publishing of the “333-year book” to commemorate the 333rd year of the invasion of Scania by the Swedish armed forces.

6. Legal registration of the Scanian flag and panther symbol. 7. A major campaign to reestablish historical regions in Sweden pur-

sued on Government and local levels and in accordance with Euro-pean Communities’ directives.

8. Participation in seminars, public events, and information meetings on all levels.

9. Participation in campaigns to spread information about Skaneland on an international level.

10. Establishment of support organisations in other parts of the country and abroad.




The Hague, 20-26 January 1995 General Assembly Resolution 4


The General Assembly,

RECOGNISING that Sweden now has become a member of the Euro-pean Union and now is attempting to divide Sweden into regions within the structures of the regional policy of the European Union;

Landscapes of Power, Landscapes of Conflict - [PDF Document] (304)


CONCERNED about Sweden’s attempts to divide the historical region of

CONCERNED that Sweden’s activities in dividing the Scanian territory

NOTING that UNPO has stood behind peoples who have worked for the protection of the right to self-determination within their ancestral territories;

Scania into different parts and add some of these parts to adjoining regions;

will cause further serious identity problems among the Scanian people;


EXPRESSES its concern about the Swedish Government’s attempts to di-vide and split the territory of Scania in the present European Union imple-mentation process.

APPEALS to the European Union Commission not to accept irrational re-gional division proposals from the Swedish Government.

VOWS to continue its support for the Scanian people in its struggle for recognition.

OFFICE OF THE GENERAL SECRETARY Javastraat 40A 2585 AP The Hague The Netherlands Tel: + 31 70 360 33 18 Fax: + 31 70 360 33 46 email: [emailprotected]

AMERICAS COORDINATION OFFICE 444 North Capitol Street Suite 846 Washington D.C. 20001-1570United States Tel: + 1 202 637 0475 Fax: + 1 202 637 0585 email: unposf@igc .apc. org

TARTU COORDINATION OFFICE Tiigstr. 78-1 17 EE 2400 Tartu EstoniaTel. + 372 7430 805 Fax: + 372 7430 805

Reprinted with permission from Center for World Indigenous Studies, George Manuel Library, 1001 Cooper, Olympia, Wash. 98502-1 107.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (305)

This page intentionally left blank.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (306)


1974, FAO (Food and Agriculture Organization)-Unesco Soil Map of The World, Unesco, Paris. 1981, FAO-Unesco Soil Map of the World 1:5 000 000 vol. 5 - Europe, Unesco, Paris. Adams, R. M., 1966, The Evolution of Urban Society, Aldine, Chicago. Adams, R. M., 1978, The Origin of Cities, Scientific American, September, pp. 10-17.Adamsen, C., 1982, Bulagergård En Bebyggelse Fra Tidlig Middelalder I Verst Sogn, Mark og

Afzelius, A., 1943, Diplomatarium Danicum, Vol. 2, Part 5, 1299-1305, Munksgaard, Copenhagen. Afzelius, A., 1948, Diplomatarium Danicum, Vol.2, Part 6, 1306-1312, Munksgaard, Copenhagen. Afzelius, A., 1953, Diplomatarium Danicum, Vol. 2, Part 8, 1318-1322, Munksgaard, Copenhagen. Agnew, P., and J. Duncan, 1990a, The Power of Place, Unwin Hyman, Boston. Aguilar, R., and R. D. Heil, 1988, Soil Organic Carbon, Nitrogen and Phosphorus Quantities in

Northern Great Plains Rangeland, Journal of the Soil Science Society of America 52:1076- 108 1.

Ahlbåck, T., 1990, Old Norse and Finnish Religions and Cultic Place-Names, Donner Institute for Research in Religious and Cultural History, Åbo.

Albrectsen, E., 1970, Vikingetidens Odense, Fynske Minder, 1970(123-132).Ambrius, J., 1995, Skånes Historia I Årtal Fra 550 Till 1720: En Handbook [Scania’s History by

Year from 550 to 1720: A Handbook], Ambrius Antikvariat, Malmö. Ambrosiani, B., 1981, Settlement Structure in Viking Age Sweden, in: Economic Aspects of the

Viking Age (D.Wilson and M. Caygill, eds.), British Museum, London, pp. 47-50.Ambrosiani, B., 1985, Aristocratic Graves and Manors in Early Medieval Sweden, Archaeology &

Environment 4: 109-118.Ambrosiani, B., 1989, Domestic Architecture in Europe 500-1 100 AD, Fornvännen 1989(84):

Andersen, H., 1964, Den Ældste Århus, Kuml 1963.Andersen, S.W., 1977, Vikingerne I Lejre, Historisk Arbog fra Roskilde Amt 1977:11-23.Andersen, S., 1991, When Was Regular Annual Taxation Introduced in the Norse Islands of

Britain? A Comparative Study of Assessment Systems in North-Western Europe. Scand.Jour-nal History 16:73-83.

Montre 1982:39-49.

16 1 -162.

Anderson, J., 1973, Ideology in Geography: An Introduction, Antipode 5(3):1-6.Anderson, I., 1947, Skånes Historia, Norstedt & Sons, Stockholm. Anderson, T., 1964, The Problem of Icelandic Saga Origins, Yale University Press, New Haven,



Landscapes of Power, Landscapes of Conflict - [PDF Document] (307)


Andersson, E, 1970, An Ecosystem Approach to Vegetation, Environment and Organic Matter of a Mixed Woodland and Meadow Area, Thesis, Lund University, Lund.

Andersson, H., and M. Anglert, 1989, By, Huvudgård Och Kyrka: Studier I Ystadområdets Medeltid [Village, Estate, and Church: Studies in the Ystad Region’s Medieval Times], Almqvist & Wiksell International, Stockholm.

Andersson, L., 1989, Pirenne, Bolin, and the New Archaeology, Scandia 55(2):325.Andren, E., 1974, Simrishamn: En Gammel Stad På Österlen [Simrishamn: An Old City in

Österlen] , Nordiska Museet, Stockholm. Andren, A., 1983, Stader Och Kungmakt—En Studie I Danmarks Politiska Geografi Före 1230

[Towns and Royal Power—A Study in Denmark’s Political Geography Before 12301, Scandia49(1):3 1-71.

Andrén, A., 1985, Den Urbana Scenen [The Urban Scene], CWK Gleerup, Malmö, Sweden. Andrén, A., M. Anglert, U.-B. Ekstrand, L. Ersgård, H. Klackenberg, B. Sundnér, and J. Wienberg,

1986, Medeltiden Och Arkeologin: Festskrift Till Erik Cinthio [The Middle Ages and Ar-chaeology: Papers in Honor of Erik Cinthio], Institute of Medieval Archaeology, Lund.

ANF, 1883, Arkiv För Nordisk Filolgi. Anglen, M., 1995, Kyrkor Och Hewavalde: Från Kristnade Till Sockenbildning I Skåne

[Churches and Lordship: From Christianization to Parish Formation in Scania], Almkvist & Wiksell International, Stockholm.

AnnDan., 1920, Annales Danici mediiœvi. Appleton, J., 1975, The Experience of Landscape, Wiley, London. Arnold, C. J., 1982, Stress as a Stimulus for Socioeconomic Change: Angle-Saxon England in the

7th Century, in: Ranking, Resource, and Exchange (C. Renfrew and S. Shennan, eds.), Cam-bridge University Press, Cambridge, pp. 124- 131.

Arnold, B., 1999, “Drinking the Feast”: Alcohol and the Legitimation of Power in Celtic Eu-rope, Cambridge Archaeological Journal 9(1):71–93.

Arrhenius, O., 1934, Fofathalten Z Skånska Jordar [The Phosphate Content in Scanian Soils],Sveriges Geologiska Undersökning Årsbok 28:3, Lund.

Arrhenius, B., 1990, Connections Between Scandinavia and the East Roman Empire in the Mi-gration Period, in: From the Baltic to the Black Sea: Studies in Medieval Archaeology (D.Austin and L. Alco*ck, eds.), Unwin Hyman, London, pp. 118-137.

Arrhenius, B., Atkinson, and Miller, 1990, Vegetational Development and Land Use in Vendel and Sutton HOO, Norwegian Arch. Rev. 23(1–2):61–63.

Asingh, P., and S. Jensen, 1983, En Landsby Fra Tidlig Jernalder Ved Farup Nordvest For Ribe, Mark og Montre 1983:5–10.

Astill, G., 1985, Archaeology, Economics, and Early Medieval Europe, Oxford Journal of Ar-chaeology 4(2):2 15-231.

Atkinson, H., 1990, The Boat Grave Studies of Sutton Hoo and Vendel—A Paleoenvironmental Study, Norwegian Arch. Rev. 23(1–2):65–78.

Auerbach, E, 1913, Das Gesertz Der Bevolkerungskontration, Petermanns Mitteilungen 19 13(74).

Augustsson, J. E., 1991, Rural and Urban Craft in Sweden and Denmark 1000–1600 A.D., in: Re-gions and Reflections: In Honor of Marta Strömberg (K. Jennbert, L. Larsson, R. Petré, and B. Wyszomirska-Werbart eds.), Almqvist & Wiksell, Lund, pp. 309-322.

Augustsson, J. E., 1992, Medeltida Husbyggande: Symposium i Lund November 1989 [Medieval House Construction: A Symposium in Lund, November 1989], Almqvist & Wiksell Interna-tional, Stockholm.

Austin, D., and L. Alco*ck (eds.), 1990, From the Baltic to the Black Sea: Studies in Medieval Ar-chaeology, Unwin Hyman, London.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (308)


Austin, D., and J. Thomas, 1990, The “Proper Study” of Medieval Archaeology: A Case Study, in:From the Baltic to the Black Sea: Studies in Medieval Archaeology (D. Austin and L. Al-co*ck, eds.), Unwin-Hyman, London, pp. 43-78.

ÆDA, De Ældste Danske Archivregistraturer. Baetke, W., 1942, Das Heilige Im Germanischen, Mohr, Tubingen. Baigent, E., 1990, Swedish Cadastral Mapping 1628-1700: A Neglected Legacy, The Geographi-

Baker, A., 1970, Geographical Interpretations of Historical Sources, David & Charles Ltd., New-

Bakkevig, S., 1980, Phosphate Analysis in Archaeology—Problems and Recent Progress, Norwe-

Barber, S. A., 1984, Soil Nutrient Bioavailability: A Mechanistic Approach, Wiley Interscience,

Barker, G., 1985, Prehistoric Farming in Europe, Cambridge University Press, Cambridge. Barley, M.W., 1977, European Towns: Their Archaeology and Early History, Academic Press,

New York. Bartholin, T. S., B. E. Berglund, and N. Malmer, 1981, Vegetation and Environment in the Gårdlösa

Area During the Iron Age, in: Gårdlösa: An Iron Age Community in Its Natural and Social Setting I (B. Stjerquist, ed.) CWK Gleerup, Lund, pp. 45-53.

Bartholin, T. S., and B. E. Berglund, 1992, The Prehistoric Landscape of the Kopinge Area—A Re-construction Based on Charcoal Analysis, in: The Archaeology of the Cultural Landscape: Fieldwork and Research in a South Swedish Rural Region (L. Larsson, J. Callmer, and B. Stjernquist, eds.), Almqvist & Wiksell International, Stockholm, pp. 345–358.

Basso, K. H., 1996, Wisdom Sits in Places, in: Senses of Place (S. Feld and K. H. Basso, eds.), School of American Research Press, Santa Fe, pp. 51-90.

Bates, D., 1980, Normandy Before 1066, Anderson, New York. Becker, C. J., 1980, Viking Age Settlements in Western and Central Jutland: Recent Excavations,

Acta Archaeologica 50:89–94.Becker, C. J., 1982, Siedlungen Der Bronzezeit Und Her Vörromische Eisenzeit In Danemark [Set-

tlement in the Bronze Age and Pre-Roman Iron Age in Denmark], Offa 39:53–71.Becker, C. J., 1983, Enkelgård Og Landsby I Danmarks Ældre Jernalder [Single Farm and Village

in Denmark’s Older Iron Age], in: Gårdens Udvikling Fra Jernalderen Til Nyere Tid [TheDevelopment of the Farm from the Iron Age to Recent Times], (H. Thrane, ed.), Odense Uni-versity, Odense, pp. 5-16.

Bekker-Nielsen, T., 1989, The Geography of Power: Studies in the Urbanization of Roman North-West Europe, British Archaeological Reports, Oxford.

Bellows, H., 1968, The Poetic Edda, The American Scandinavian Foundation, New York. Benario, H.W., 1967, Agricola, Germany, Dialogue on Orators by Tacitus (H.W. Benario, trans.),

Benario, H.W., 1975, An Introduction to Tacitus, University of Georgia Press, Athens. Bencard, M., 1969, Grubehuse I Okholm, Mark og Montre 1969:26–36.Bencard, M., 1978, Ribe Through 1000 Years, BYGD, Esbjerg. Bencard, M. 1981, Ribe Excavations 1970- 76, Sydjysk Universitetsforlag, Esbjerg Bender, B., 1978, Gatherer-hunter to farmer: a social perspective, World Archaeology 10:204–222.Bender, B., 1981, Gatherer-hunter intensification, Economic Archaeology: Towards an Inte-

gration of Ecological and Social Approaches (A. Sheridan and G. Bailey, eds.), BAR Inter-national Series, Oxford, pp. 149-158.

Bender, B., 1985, Emergent Tribal Formation in the American Midcontinent, American Antiq-uity 50(1):52–62.

cal Journal 156(1):62–69.

ton Abbot.

gian Archaeological Review 13(2):73–100.

New York.

Bobbs-Merrill, Indianapolis.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (309)


Bender, B., 1990, The Dynamics of Nonhierarchical Societies, in: The Evolution of Political Sys-tems: Sociopolitics in Small-Scale Sedentary Societies (S. Upham, ed.), Cambridge Uni-verisity Press, Cambridge, pp. 247-263.

Bentley, G. C., 1987, Ethnicity and Practice, Comparative Studies in Society and History

Berglund, B., 1981, Comments, in: Gårdelösa: An Iron Age Community in Its Natural and So-cial Setting I (B. Stjernquist, ed.), CWK Gleerup, Lund, pp. 114-131.

Berglund, B., 1991, The Cultural Landscape During 6000 Years in Southern Sweden: The Ystad Project, Munksgaard International, Copenhagen.

Berglund, B., M. Hjelmroos, and E. Kolstrup, 1991, The Köpinge Area: Vegetation and Landscape Through Time, in: The Cultural Landscape During 6000 Years in Southern Sweden—The Ystad Project (B. Berglund, ed.), Munksgaard International, Copenhagen, pp. 109-112.

Bergsten, K. E., 1955, Sydöstskånsk Geologi [Southeast Scanian Geology], Tumathorps Sanct Knuts Gille Tänkebok 1953:11–30.

Berry, B.J.L. 1961, City-size distributions and economic development, Economic Development and Cultural Change 9:573-587.

Berry, B.J.L., 1967, Geography of Market Centers and Retail Distribution, Prentice-Hall, Engle-wood Cliffs, New Jersey.

Berry, W., 1984, An Experimental Model for Early Medieval Settlement in Southwestern Bur-gundy, in: Archaological Approaches to Medieval Europe (K. Biddick, ed.), Western Michi-gan University, Kalamazoo, pp. 245-262.

Bethell, T., and C. Mate, 1989, The Use of Soil Phosphate Analysis in Archaeology: A Critique in Scientific Analysis, in: Scientific Analysis in Archaeology (J. Henderson, ed.), Oxford, Los Angeles, pp. 1-29.

Biddick, K., 1984a, Archaological Approaches to Medieval Europe, Western Michigan Univer-sity, Kalamazoo.

Biddick, K., 1984b, Field Edge, Forest Edge: Early Medieval Social Change and Resource Alloca- tion, in: Archaological Approaches to Medieval Europe (K. Biddick, ed.), Western Michigan University, Kalamazoo, pp. 105-118.

Biddick, K., 1989, The Other Economy: Pastoral Husbandry on a Medieval Estate, Universityof California Press, Berkeley.

Biddick, K., 1990, People and Things: Power in Early English Development, Comparative Stud-ies in Society and History 32:3–23.

Billberg, I., 1987, Från Trattbägare Till Fajans [From Funnel Beaker to Faiance], Malmö Museer, Malmö, Sweden.

Billberg, I., 1989, Inland-Kustland. Bebyggelse Och Resursutnyttjande För Två Medeltida Byar Inom Ystadsområdet [Inland-Coastland. Settlement and Resource Utilization in Two Medieval villages Within the Ystad Area], in: By, Huvuagård Och Kyka [Village, Manor, and Church] (H. Andersson and M. Anglert, eds.), Almqvist & Wiksell International, Stockholm, pp. 15-48.

Billinge, M., 1982, Reconstructing Societies in the Past: The Collective Biography of Local Com-munities, in: Period and Place (A. Baker and M. Billinge, eds.), Cambridge University Press, Cambridge.


Binford, L. R., 1962, Archaeology as Anthropology, American Antiquity 28:217–225.Binford, L. R., 1965, Archaeological Systematics and the Study of Culture Process, American An-

Binford, L. R., 1968, Archaeological Perspectives, in: New Perspectives in Archaeology (S. R. Bin-

Binford, L. R., 1983, In Pursuit of the Past, Thames and Hudson, London. Birkedahl, P., and E. Johansen, 1993, Nikolajbjerget, Skalk 1993(4):3–8.

tiquity 31:203–2 10.

ford and L. R. Binford, eds.), Aldine, Chicago, pp. 5-32.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (310)


Birks, H.J.B., J. M. Line, and T. Persson, 1988, Quantitative Estimation of Human Impact on Cul-tural Landscape Development, in: The Cultural Landscape: Past, Present, and Future (H.H. Birks, H.J.B. Birks, P. E. Kalend, and D. Moe, eds.), Cambridge University Press, Cambridge.

Bitner-Wróblewska, A., 1991, Between Scania and Samland: From Studies of Stylistic Links in the Baltic Basin During the Early Migration Period, Fornvännen 86:225–241.

Bjøm, C. (ed.), 1988, Det Danske Landbrugs Historie I: Oldtid Og Middelalder, LandbohistoriskSelskab og Forfatterne, Odense.

Blackburn, M., 1981, An Imitative Workshop Active During Wthelreds Ii’s Long Cross Issue, in: Studies in Northern Coinages of the Eleventh Century (C. J. Becker, ed.), Det Kongelige Danske Videnskabernes Selskab Historisk-filosofiske Skrifter, Copenhagen, pp. 89-111.

Blair, E, 1970, The World of Bede, Secker & Warburg, London. Blair, E, 1976, Anglo Saxon England, Cambridge University Press, Cambridge. Blanton, R. E., 1975, The Cybernetic Analysis of Human Population Growth, in: Population Stud-

ies in Archaeology and Physical Anthropology (A. C. Swedland, ed.), Society for American Archaeology, Washington D.C., pp. 116-126.

Blanton, R. E., 1976, Anthropological Studies of Cities, Annual Review of Anthropology

Blanton, R. E., 1980, Cultural Ecology Reconsidered, American Antiquity 45(1):145–150Blanton, R. E., 1983, The Ecological Approach in Highland Mesoamerican Archaeology, in: Ar-

chaeological Hammers and Theories (J. Moore and A. Keene, eds.), Academic Press, New York, pp. 221-233.

Blanton, R. E., 1984, A Comparison of Early Market Systems, in: Markets and Marketing (S. Platt-ner, ed.), University Press of America, Ianham, Md.

Blanton, R. E., 1988, Comment on Ecological Theory and Cultural Evolution in the Valley of Oax-aca (W. T. Sanders and D. Nichols), Current Anthropology 29(1):52–54

Blanton, R. E., 1998, Beyond Centralization: Steps Toward a Theory of Egalitarian Behavior in Ar-chaic States, in: Archaic States (G. M. Feinman and J. Marcus, eds.) School of American Re-search Press, Santa Fe.

Blanton, R. E., S. A. Kowalewski, G. Feinman, and J. Appel, 1981, Ancient Mesoamerica: A Com-parison of Change in Three Regions, Cambridge University Press, Cambridge.

Blanton, R. E., and G. Feinman, 1984, The Mesoamerican World System, American Anthropolo-gist 86(3):673–682.

Blanton, R. E., G. Feinman, S. Kowalewski, and P. Peregrine, 1996, A Dual-Processual Theory for the Evolution of Mesoamerican Civilization, Current Anthropology 37(1):1-14.

Blaut, J. M., 1980, A Radical Critique of Cultural Geography, Antipode 12(2):25–29. Blomqvist, R., 195 1, Lundshistoria: Medeltiden (Lund’s History: The Middle Ages), Liber

Läromedel/Gleerup, Lund. Boldsen, J. L., 1990, Height Variation in the Light of Social and Regional Differences in Medieval

Denmark, in: From the Baltic to the Black Sea: Studies in Medieval Archaeology (D. Austinand L.Alco*ck, eds.), Unwin Hyman, London, pp. 181-188.


Bolin, S., 1930, Skånelands Historia I, Malmö, Sweden.Boserup, E., 1965, The Conditions of Agricultural Growth, Aldine, Chicago. Boyer, E, 1990, Tradition as Truth and Communication: A Cognitive Description of Tradi-

Brattberg, S., 1983, Studier Kring Skånes Köpinger—Vikingatida Marknadsplatser [Studies on

Braudel, E, 1980, On History (Sarah Matthews, trans.), University of Chicago Press, Chicago. Bredsdorf, P., 1973, Kortlægning Og Historiske Studier [Map Making and Historical Studies],

tional Discourse, Cambridge University Press, Cambridge.

Skånes Köpinges—Viking Age Marketplaces], Limnhamnia 1983:83–102.

Københavns Universitets Historisk Institut, Copenhagen.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (311)


Brendalsmo, A. J., and G. Røthe, 1992, Haugbrot Eller De Levendes Forhold Til De Døde [Grave-Robbing or the Living’s Relation to the Dead], Meta 1-2:89–119.

Bring, J., 1925, En Boplats Och Begrafnigsplats Från Sten-Brons-Och Jernåldern I Cimbris Socken, Föreningens för Fornminnes och hembugdesvhrd i Sydöstra Skåne, Skrifter II:27–44.

Brink, S., 1984, Ortnamn Som Kålla I Historisk Forskning [Place Names as Sources for Historic Research], Fornvännen 79:165–172.

Brink, S., 1990, Sockenbildning Och Socknnamn: Studier I Äldre Territoriell Indelning I Nor-den [Parish Formation and Parish Names: Studies in Early Territorial Divisions in Scandi-navia], Acta Academiae Regiae Gustavi Adolphi LVII, Uppsala.

Broberg, A., and K. Svensson, 1987, Urban and Rural Consumption Patterns in Eastern Central Sweden AD 1000- 1700, in: Theoretical Approaches to Artifacts, Settlement, and Society (G.Burenholt, ed.), BAR International Series, Oxford, pp. 479-487.

Brumfiel, E., and T. Earle, 1987, Specialization and Exchange in Complex Societies: An Introduc-tion, in: Specialization, Exchange, and Complex Societies (E. Brumfiel and T. K. Earle, eds.), Cambridge University Press, Cambridge, pp. 1-9.

Bruneau, M., 198 1, Landscapes, Social Relations of Production, and Eco-Geography, Antipode

Buch-Hanson, M., and B. Nielson, 1977, Marxist Geography and the Concept of Territorial Struc-

Buttimer, A., 197 1, Society and Milieu in the French Geographic Tradition, Rand-McNally, Chicago. Callmer, J., 1980, Topographic Notes on Some Scanian Viking Period and Early Medieval Hoards,

Callmer, J., 1983, Production Site and Market Area, Meddekznden från Lunds Universitets His-

Callmer, J., 1984, Recent Work at Åhus: Problems and Observations, Offa 41:63–75.Callmer, J., 1985a, Agrar Bebyggelse I Backlandskapet Vid Överganget Till Historisk Tid [Agricul-

tural Settlement in the Interior at the Transition to Historic Times], in: Kulturlandskap—Dens Framväxt Och Förädring [The Cultural Landscape—Its Growth and Change], (G. Regnell, ed.), Lund.

Callmer, J., 1985b, Pragmatic Notes on the Early Medieval Bead Material in Scandinavia and the Baltic Ca. AD 600- 1000, Studia Nad Etnogeneza Slovian Warsaw T1:28–226.

Callmer, J., 1986, To Stay or to Move: Some Aspects of Settlement Dynamics in Southern Sweden, Meddelanden från Lunds Universitets Historiska Museum 1985-86:167-208.

Callmer, J., 1987, Iron Age and Early Medieval Settlement Development in Southern Scandinavia: Some Contemporary and Future Research Perspectives, in: Theoretical Approaches to Arti-facts, Settlement, and Society (G. Burenhult, ed.), BAR International Series, Oxford, pp.

Callmer, J., 1988a, Three Fundamental Perspectives for the Study of Trade and Exchange, in: Trade and Exchange in Prehistory (B. Hårdh, ed.), Lund University Press, Lund, pp.

Callmer, J., 1988b, Slawisch-Skandinavische Kontakte Am Beispiel Der Slawischen Keramik In Skandinauien Während Des 8. Und 9. Jahrhunderts, Bericht der Romisch-GermanischenKommission Band 69:654–674.

Caber, J., 1991a, Territory and Dominion in the Late Iron Age in Southern Scandinavia, in: Re-gions and Reflections: In Honour of Märta Strömberg (K. Jennbert, L. Larsson, R. Petré, and B. Wyszomirska-Werbart, eds.), Almqvist & Wiksell International, Lund, pp. 257-273.

Callmer, J., 1991b, The Viking Age Landscape: Settlement and Society, in: The Cultural Land-scape During 6000 Years in Southern Sweden: The Ystad Project (B. Berglund, ed.), Munks- gaard International, Copenhagen, pp. 82–85.


ture, Antipode 9(2):1–12.

Meddelanden från Lunds Universitets Historiska Museum 1979-80: 132- 149.

toriska Museum 1981-82:135-165.


261 -270.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (312)


Callmer, J., 1991c, Settlement and Landscape in the Bjäresjö Area In the Early and Late Iron Age and Early Middle Ages, in: The Cultural Landscape During 6000 Ears in Southern Sweden: The Ystad Project (B. Berglund, ed.), Monksgaard International, Copenhagen, pp. 184-189.

Callmer, J., 1991d, Settlement and Landscape in the Krageholm Area in the Iron Age, Viking Age, and Early Middle Ages, in: The Cultural Landscape During 6000 Years in Southern Sweden: The Ystad Project (B. Berglund, ed.), Munksgaard International, Copenhagen, pp. 230-232.

Callmer, J., 1991e, The Process of Village Formation, in: The Cultural Landscape During 6000 Years in Southern Sweden: The Ystad Project (B. Berglund, ed.), Munksgaard International, Copenhagen, pp. 337-349.

Callmer, J., 1992a, Villages in the Hummocky Landscape: The Excavations at Gussnava and Skårby and the Surveys in Skårby, Balkåkra, and Bjäresjö Parishes, in: The Archaeology of the Cultural Landscape: Field Work and Research in a South Swedish Rural Region (L. Larsson, J. Callmer, and B. Stjernquist, eds.), Almqvist & Wiksell International, Stockholm, pp. 377-394.

Callmer, J., 1992b, From Bronze Age Dispersed Settlement to Medieval Village in the Krageholm Area, in: The Archaeology of the Cultural Landscape: Field Work and Research in a South Swedish Rural Region, (L. Larsson, J. Callmer, and B. Stjernquist, eds.), Almqvist & Wiksell International, Stockholm, pp. 395-410.

Callmer, J., 1992c, A Contribution to the Prehistory and Early History of the South Scandinavian Manor: The Bjaresjo Investigations, in: The Archaeology of the Cultural Landscape: Field Work and Research in a South Swedish Rural Region (L. Larsson, J. Callmer, and B. Stjern-quist, ed.), Almqvist & Wiksell International, Stockholm, pp. 411-458.

Callmer, J., 1992d, An Iron Age Cemetery at Bellinga, in: The Archaeology of the Cultural Land-scape. Field Work and Research in a South Swedish Rural Region (L. Larsson, J. Callmer, and B. Stjernquist, eds.), Almqvist & Wiksell International, Stockholm, pp. 459-468.

Camilli, E. L., and J. I. Ebert, 1992, Artifact Reuse and Recycling in Continuous Surface Distribu-tions and Implications for Interpreting Land Use Patterns, in: Time, Space, and Archaeo-logical Landscapes (J. Rossignol and L. Wandsnider, eds.), PIenum, New York.

Campbell, Å., 1928, Skånska Bygder Under Förra Hälften Ov 1700-Talet: Etnografisk Studie Over Den Skdnska Allmogens Äldre Odlingar; Hagnader Och Byggnader [Scanian Districts During the First Half of the 1700s: An Ethnographic Study of the Scanian Country-Folk’sOlder Cultivation, Householding, and Building], Lundequistska AB, Uppsala.

Carneiro, R.L., 1963, The Application of Scale Analysis to the Study of Cultural Evolution, Trans-actions of the New York Academy of Sciences 26(2): 196-207.

Carneiro, R. L., 1970, A Theory of the Origin of the State, Science 169:733-738.Carroll, G. R., 1982, National City-Size Distributions: What Do We Know After 67 Years of Re-

Cavanagh, W. G., 1988, Soil Phosphate, Site Boundaries, and Change Point Analysis, Journal of

Chamberlin, R., 1986, The Emperor: Charlemagne, Franklin Watts, New York. Champion, T. C., 1989, Centre and Periphery: Comparative Studies in Archaeology, Unwin

Hyman, London. Champion, T. C., 1990, Medieval Archaeology and the Tyranny of the Historical Record, in: From

the Baltic to the Black Sea (D. Austin and L. Alco*ck, eds.), London, Studies in Medieval Ar-chaeology, pp. 79-95.

Chase-Dum, C., and T. D. Hall, 1991, Conceptualizing Core-Periphery Hierarchies for Compar-ative Study Core/Periphery Relations in Pre-Capitalist Worlds (C. Chase-Dunn and T. D. Hall, eds.), Westview Press, Boulder, Colorado, pp. 5-44.

search? Progress in Human Geography 6: 1-42.

Field Archaeology 15:67-83.

Childe, V. G., 1950, The Urban Revolution, Town Planning Review 21:3-17.Chisholm, M., 1970, Rural Settlement and Land Use: An Essay in Location, Aldine, Chicago.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (313)


Christensen, A. E., 1964, Birka-Hedebymyntene Som Kilde Til Skipets Historie På 800-Tallets[Birka-Hedeby Coins as Sources for Ship History in the 800s), No. skrift 41, Norsk Sjö fartsmuseums, Oslo.

Christensen, P., 1983, Søby—A Viking Age Settlement on Samsø, Journal of Danish Archaeology

Christensen, T., 1987a, 52. Lejre 2.06.01. in: Arkæologiske Udgravninger I Danmark 1986, (O.Olsen, ed.), Rigsantikvarens Arkæologiske Sekretariat, Nationalmuseet, Copenhagen, Den-mark, pp. 62.

2: 149-155.

Christensen, T., 1987b, G1. Lejre, Allerslev, Journal of Danish Archaeology 6:251.Christiansen, A. E., 1968, Denmark Between the Viking Age and the Time of the Valdemars, Me-

Christiansen, A. E., 1969, Vikingatidens Danmark: Paa Oldhistoriske Baggrund, Copenhagen.Christiansen, A. E., 1976, Archbishop Asser, the Emperor, and the Pope, Scand. J History

Christiansen, T., 1981, Archaeology and History: The Viking Fortress of Trelleborg, in: DanishMedieval History: New Currents (N. Skyum-Nielsen and N. Lund, eds.), Museum Tuscu-lanum Press, Copenhagen, pp. 221-222.

Christophersen, A., 1982, Drengs, Thegns, Landmen, and Kings: Some Aspects on the Forms of Social Relations in Viking Society During the Transition to Historic Times, Meddelandenfrån Lunds Universitets Historiska Museum 1981-82:115-134.

Cinthio, H., 1980, The Loddekopinge Investigation iii, The Early Medieval Cemetery, Medde-landen från Lunds Universitets Historiska Museum 1979-80: 112-123.

Cinthio, H., and J. Boldsen, 1984, Patterns of Distribution in the Early Medieval Cemetery at Löd- deköpinge, Meddelanden från Lunds Universitets Historiska Museum 1983-84:116-127.

Claessen, H.J.M., and P. Skalnik, 1978, The Early State: Theories and Hypotheses, in: The EarlyState (H.J.M. Claessen and P. Skalnik, eds.), Mouton, The Hague, Netherlands, pp. 3-30.

Claessen, H.J.M., and P. van derVelde, 1991a, Introduction, in: Early State Economics (H.J.M.Claessen and P. van der Velde, eds.), Transaction, New Brunswick, pp. 1-29.

Claessen, H.J.M., and Pvan der Velde, 1991b, Early State Economics, Transaction, New Brunswick. Clarke, D. L., 1973, Archaeology: The Loss of Innocence, Antiquity 47(185):6-18.Clarke, H., and B. Ambrosiani, 1991, Towns in the Viking Age, Leicester University Press, London. Cohen, S., 1977, The Earliest Scandinavian Towns, in: The Medieval City (H. Miskimin, ed.), Yale

Cohen, M. N., 1977, The Food Crisis in Prehistory, Yale University Press, New Haven. Cohen, R., 1978, State Origins: A Reappraisal, in: The Early State (H.J.M. Claessen and P. Skalnik,

Cosgrove, D., 1978, Place, Landscape, and the Dialectics of Cultural Geography, Canadian Ge-

Cosgrove, D., 1983, Towards a Radical Cultural Geography, Antipode 15(1):1-11.Cosgrove, D., 1989, Power and Place in the Venetian Territories, in: The Power of Place (R.

Cosgrove, D., and E. Daniels, 1988, The Iconography of Landscape, Cambridge University Press,

Craddock, P.T., 1985, The Application of Phosphate Analysis to the Location and Interpretation

Cresswell, T.J., 1992, In Place/Out of Place: Geography, Ideology, and Transgression, Ph.D. dis-

Crumley, C., 1976, Towards a Locational Definition of State Systems of Settlement, American An-

dieval Scandinavia 1:28-50.


University Press, New Haven, pp 313-335.

eds.), Mouton, The Hague, Netherlands, pp. 31-76.

ographer 22:66-72.

Agnew and J. Duncan, eds.), Unwin Hyman, Boston, pp. 104-223.


of Archaeological Sites, Archaeological Journal 142: 361-376.

sertation, University of Wisconsin, Madison.

thropologist 78(1):59-73.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (314)


Crumley, C., 1984, A Diachronic Model for Settlement and Land Use in Southern Burgundy, in: Archaological Approaches to Medieval Europe (K. Biddick, ed.), Western Michigan Univer-sity, Kalamazoo, pp. 239-244.

Crumley, C., and W. Marquardt, 1987, Regional Dynamics, Academic Press, New York. Crumlin-Pedersen, O., 1965, Cog-Kogge-Kaag, Træk Av En Frisisk Skibstypes Historie, Årbog

Crumlin-Pedersen, O., 1984, Ships, Navigation and Routes in the Reports of Ohthere and Wulfs-tan, in: Two Voyagers, Ohthere and Wulfstan, at the Court of King Alfred (N. Lund, ed.), William Sessions, Ltd., York, pp. 30-42.

D’Altroy, T. N., 1992, Provincial Power in the Inka Empire, Smithsonian Institution Press, Wash-ington D.C.

D’Altroy, T. N., and T. Earle, 1985, Staple Finance, Wealth Finance, and Storage in the Inka Politi-cal Economy, Current Anthropology 26(2):187-206.

Dahl, S., 1940, A Survey Map of Skine’s Villages at the Middle of the 1600s, Svensk Geografisk Årsbok 20-25.

Dalton, G., 1980, Anthropological Models in Archaeological Perspective, in: Pattern of the Past (I. Hodder, ed.), Cambridge University Press, Cambridge, pp. 17-48.

Daniel, E., 1986, Description to the Quaternary Maps—Tomelilla So/Simrishamn Sv Ystad No Ör-nahusen Nv. Swedish Geographic Survey, Lund.

Davidson, D., 1986, Anthropogenic Soils and Landforms in Orkney, in: Essays for Professor R.E.H. Mellor (Ritchie, Stone, and Mather, eds.), Aberdeen.

Davidson, H. E., 1972, The Battle God of the Vikings, Ebor Press, York. Davidson, H. E., and P. Fisher, 1980, Saxo Grammaticus the History of the Danes Books I-Ix

Day, L. D., 1987, Landscape Position and Particle Size Effects on Soil Phosphorus Distributions,

DD, Diplomatarium Danicum. de Certeau, M., 1988, The Writing of History (Tom Conley, trans.), Columbia University Press,

des Gautries, J. A., 1954, Les Noms de Personnes Scandinaves en Normandie de 911 à 1066,

Dodgshon, R.A., 1987, The European Past Social Evolution and Spatial Order, Macmillan, London. Donat, P., 1984, Die Mecklenburg—Eine Hauptburg Der Obodriten [The Mecklenburg—A City

DRB, Danmarks Riges Breve. Drennan, R. D., 1984, Long-Distance Movement of Goods in the Mesoamerican Formative and

Classic, American Antiquity 49(1):27-43.Drennan, R. D., 1988, Household Location and Compact Versus Dispersed Settlement in Pre-

hispanic Mesoamerica, in: Household and Community in the Mesoamerican Past (R. Wilk and W. Ashmore, eds.), University of New Mexico Press, Albuquerque, pp. 273-293.

1965 81-144.

and Commentary, Vols. I and II, D.S. Brewer, Cambridge.

Soil Sci. Soc. Am. J. 51:1547- 1553.

New York.

Lund University, Lund.

of the Obodrites] , Akademie Verlag, Berlin.

Drew, K. E, 1973, The Lombard Laws, University of Pennsylvania Press, Philadelphia. Driver, E, 1991, Political Geography and State Formation: Disputed Territory, Progress in

DS , Danmarks Stednavne .Dubeck, I., 1990, Skånska Lov Og Den Europaeiske Baggrund [The Scanian Law and the Euro-

Dudley, D. R., 1968, The World of Tacitus, Secker & Warburg, London. Duncan, J., 1989, The Power of Place in Kandy, Sri Lanka, in: The Power of Place (P. Agnew and

Human Geograpy 15(3):268-280.

pean Background], Historie 18(3):396-420.

J. Duncan, eds.), Unwin Hyman, Boston, pp. 185-201.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (315)


Duncan, J., 1990, The City as Text: The Politics of Landscape Interpretation in the Kandyan Kingdom, Cambridge University Press, Cambridge.

Dunning, N. P., 1989, Intersite Reconnaissance and Soil: In Archaeological Investigations at Sayil, Yucatan, Mexico, University of Pittsburgh, anthropological papers. No. 2, Studies Dur-ing the 1987 Field Season.

Earle, T. K., 1977, A Reappraisal of Redistribution: Complex Hawaiian Chiefdoms, in: Exchange Sys-tems in Prehistory (T. K. Earle and J. Ericson, eds.), Academic Press, NewYork, pp. 213-229.

Earle, T. K., 1978, Economic and Social Organization of a Complex Chiefdom, Museum of An-thropology, University of Michigan, Ann Arbor.

Earle, T. K., 1987, Chiefdoms in Archaeological and Ethnohistorical Perspective, Ann. Rev. An-

Earle, T. K., 1991a, The Evolution of Chiefdoms, in: Chiefdoms: Power, Economy, and Ideology(T. K. Earle, ed.), Cambridge Unversity Press, Cambridge, pp. 1-15.

Earle, T. K., 1991b, Property Rights and the Evolution of Chiefdoms, in: Chiefdoms: Power Econ-omy, and Ideology (T. K. Earle, ed.), Cambridge Unversity Press, Cambridge, p. 71-99.

Earle, T. K., 1997, How Chiefs Come to Power: The Political Economy in Prehistory, StanfordUniversity Press, Stanford.

Edsman, C. M., 1990, Opening Address: Encounters Between Religions in Old Norse Times and Cul-tic Place Names, in: Old Norse and Finnish Religions and Cultic Place-Names (T. Ahlbäck, ed.), The Donner Institute for Research in Religious and Cultural History, Åbo, pp. 9-34.

Egerbladh, T., 1977, Task Difficulty, Ability Level, and Group Size: A Test of the Theory of Social Decision Schemes, Educational Reports, Umeå University.

Eidt, R. C., 1984, Advances in Abandoneed Settlement Analysis: Application to Prehistoric Antbrosols in Colombia, South America, Center for Latin America, University of Wis-consin, Milwaukee.

thropol. 1987:279-307.

Einarson, S., 1957, A History of Icelandic Literature, Johns Hopkins Press, NewYork. Ekholm, S., 1977, External Exchange and the Transformation of Central African Social Systems,

in: The Evolution of Social Systems (J. Friedman and M. J. Rowlands, eds.), Duckworth, Lon-don, pp. 115-136.

Ekman, J., 1973, Early Medieval Lund—The Fauna and the Landscape, Museum of Cultural History, Lund.

Ekström, G., 1950, Skånes Åkerjordsområden [Regions of Cultivated Soils in Scania], Socker-handlingar 6(3).

Emanuelsson, U., 1985, Det Skdåska Kulturlandskap [The Scanian Cultural Landscape], Bok-förlaget Signum, Lund.

Emanuelsson, U., 1987, Skånes Vegetationshistoria [ Skåne’s Vegetation History], Svensk Ge-ografisk Årbok.

Emanuelsson, U., Bergendorff, C., Carlsson, B., Lewan, N., and 0. Nordell, 1985, Det Skdåska Kulturlanskap (The Scanian Cultural Landscape), Bokforlaget Signum, Lund.

Engelmark, R., 1985,Ögras för Tolkning av Agrara System [Weeds for Interpretation of the Agrar-ian System], in: Kulturlandskap—Dens Framväxt och Förändring [The Cultural Land-scape-Its Growth and Change] (G. Regnell, ed.), Lund, Lund Unversity Press.

Engelmark, R., 1992, A Review of the Farming Economy in South Scania Based on Botanical Evi-dence, in: The Archaeology of the Cultural Landscape: Field Work and Research in a South Swedish Rural Region (L. Larsson, J. Callmer, and B. Stjernquist, eds.), Almqvist & Wiksell International, Stockholm, pp. 369-376.

English, PW., and R. C. Maylield, 1972, The Cultural Landscape: Man, Space, and Environment, Concepts in Contemporary Human Geography, Oxford University Press, New York.

Ersgård, L., 1995, Thirteen Essays on Medieval Artifacts, Almqvist & Wiksell International, Lund.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (316)


Fabech, C., 1994, Reading Society from the Cultural Landscape: South Scandinavia Between Sacral and Political Power, in: The Archaeology of Gudme and Lundeborg (P. O. Nielsen, K. Rands-borg, and H. Thrane, eds.), Universitetsforlag i København, Copenhagen, pp. 169-183.

Fabian, J., 1983, Time and the Other: How Anthropology Makes Its Object, Columbia University Press, New York.

Falco, G., 1964, The Holy Roman Empire: An Historic Profile of the Middle Ages (K. Kent, trans.), A.S. Barnes, NewYork.

Falconer, S. E., and S. H. Savage, 1995, Heartlands and Hinterlands: Alternative Trajectories of Early Urbanization in Mesopotamia and the Southern Levant, American Antiquity 60(1):37-58.

Fehring, G. P., 1981, Ein Neuentdeckter Slawishcher Burgwall bei Klempau, Kreis Herzogtum Lauenburg, und Seine Funde, Offa 1981:277-288.

Fehring, G. P., 1990, Origins and Development of Slavic and German Lübeck, in: From the Baltic to the Black Sea: Studies in Medieval Archaeology (D. Austin and L. Alco*ck, eds.), Unwin Hyman, London, pp. 251-266.

Fehring, G. P., 1991, The Archaeology of Medieval Germany: An Introduction (R. Samson,trans.), Routledge, London and New York.

Feinman, G. M., 1991, Demography, surplus and inequality: early political formations in highland Mesoamerica, in: Chiefdoms: Power Economy, and Ideology (T. K. Earle, ed.), Cambridge University Press, Cambridge, pp. 229-262.

Feinman, G. M., 2000, Corporate-Network: A New Perspective on Leadership in the American Southwest, in: Hierarchies in Action: Who Benefits? Paper prepared for the 14th annual Vis- iting Scholar Conference, Center for Archaeological Investigations, Southern Illinois Uni-versity, Carbondale.

Feinman, G. M., 2001, Mesoamerican Political Complexity: The Corporate-Network Dimension, in: From Leaders to Rulers (J. Haas, ed.), Kluwer Academic/Plenum Publishers, NewYork.

Feinman, G. M., 1998, Scale and Social Organization: Perspectives on the Archaic State, in: TheArchaic State: A Comparative Approach (G. M. Feinman and J. Marcus, eds.), SAR Press, Santa Fe, pp. 95-134.

Feinman, G. M., and J. Neitzel, 1984, Too Many Types: An Overview of Sedentary Prestate Soci-eties in the Americas, Advances in Archaeological Method and Theory 7:39-102.

Feinman, G. M., S. A. Kowalewski, L. Finsten, R. Blanton, and L. Nicholas, 1985, Long-Term De-mographic Change: A Perspective from the Valley of Oaxaca, Journal of Field Archaeology

Feinman, G. M., and L. M. Nicholas, 1987, Labor, Surplus, and Production: A Regional Analysis of Formative Oaxacan Socio-Economic Change, in: Coasts, Plains, and Deserts: Es-says in Honor of Reynold J. Ruppe (S. W. Gaines, ed.), Arizona State University, Tempe, pp. 27-50.

Feinman, G. M., and L. M. Nicholas, 1990, At the Margins of the Monte Albán State: Settlement Patterns in the Ejutla Valley, Oaxaca, Mexico, Latin American Antiquity 1990(1):216-246.

Fentress, J., and C. Wickham, 1992, Social Memory, Blackwell, Oxford. Fernlund, S., 1980, Kyrkor i Skåne—en Kulturhistoria [Churches in Skåne—Cultural History],

Signum, Lund. Feveile, C., S. Jensen, and K. Ljungberg, 1989, Ansgars Ribe Endelig Fundet — Rapport Over en Un-

gravning Ved Rosenalle i Ribe 1989 [Ansgar’s Ribe Found at Iast—Report of Excavations at Rosenallé in Ribe 1989, By, marsk oggeest Årsberetning 1988(1):29-53.

Finnestad, R. B., 1990, The Study of Christianization in the Nordic Countries, in: Old Norse and Finnish Religions and Cultic Place-Names (T. Ahlback, ed.), Koteva, Åbo, pp. 256-272.

Fish, S. K., and S. Kowalewski, 1990, The Archaeology of Regions: A Case for Full-Coverage Sur-vey, Smithsonian Institution Press, Washington D.C.


Landscapes of Power, Landscapes of Conflict - [PDF Document] (317)


Flannery, K.V., 1972, The Cultural Evolution of Civilizations, Annual Review of Ecology and Sys-tematics 3:399-426.

Flannery, K. V., 1976, Sampling on a Regional Level, in: The Early Mesoamerican Village (K.V.Flannery, ed.), Academic Press, New York, pp. 131-136.

Flannery, K.V., and J. Marcus, 1990, Borrón, y Cuenta Nueva: Setting Oaxaca’s Archaeological Record Straight, in: Debating Oaxaca Archaeology (J. Marcus, ed.), Anthropological Papers, Museum of Anthropology, University of Michigan, pp 17-70.

Foote, P. and D. Wilson, 1974, The Viking Achievement, Sidgwick &Jackson, London. Foucault, M., 1972, The Archaeology of Knowledge, Tavistock, London. Frandsen, K. E., 1983, Danish Field Systems in the Seventeenth Century, Scand. J. History

Frandsen, L. B., 1989, Trade, Coins, and Foreign Influences During Pre-Viking and Viking Age Ribe, in: Medieval Archaeology Research Group (H. Clarke and E. Schia, eds.), BAR Inter-national Series, Oxford, pp. 37-42.

Frandsen, L., and S. Jensen, 1987, Pre-Viking and Early Viking Age Ribe, Journal of Danish Archaeology 6.

Frandsen, L., P. K. Madsen, and H. Mikkelson, 1989, Byudgravninger Og Bygningarkælogiske Un-dersøgelser I Ribe 1983-89 [Town Excavations and Building Investigations in Ribe 1983-891, By, Marsk og Geest, Årsberetning 1988(1):2-27.


Fried, M. H., 1967, The Evolution of Political Society, Random House, NewYork. Fried, M. H., 1968, On the Concepts of “Tribe” and ”Tribal Society,” in: Essays on the Problem of

Tribe (J. Helm, ed.), University of Washington Press, Seattle, pp. 3-22.Friedland, K., 1981, Denmark’s, Part in the Economic System of Late Medieval Northern Europe,

in: Danish Medieval History, New Currents (N. Skyum-Nielsen and N. Lund, eds.), Museum Tusculanum Press, Copenhagen, pp. 171-178.

Friedman, J., 1989, Culture, Identity, and World Process, in: Domination and Resistance (D.Miller, M. J. Rowlands, and C. Tilley, eds.), Routledge, London, pp. 246-260.

Friedman, J., and M. J. Rowlands, 1977, The Evolution of Social Systems, Duckworth, London. Friis-Jensen, K., 1981, Saxo Gramaticus: A Medieval Author Between Norse and Latin Culture,

Fuglestad, E, 1979, Earth-Priests, “Priest-Chiefs,” and Sacred Kings in Ancient Norway, Iceland,

Furet, E, 1984, In the Workshop of History, University of Chicago Press, Chicago. Gahrn, L., 1986, The Geatas of Beowulf, Scand. J. History 11:95-113.Gaillard, M. J., and B. E. Berglund, 1988, Land-Use History During the Last 2700 Years in the Area

of Bjaresjo, Sweden, in: The Cultural Landscape—Past, Present, and Future (H. H. Birks, H.J.B. Birks, F! E. Kalend, and D. Moe, eds.), Cambridge University Press, Cambridge, pp.

Gaillard, M. J., and H. Göransson, 1991, The Bjaresjo Area: Vegetation and Landscape Through Time, in: The Cultural Landscape During 6000 Years in Southern Sweden—The Ystad Project (B. Berglund, ed.), Munksgaard International, Copenhagen, pp. 167-174.

Museum Tusculanum Press, Copenhagen.

and West Africa, Scand. J History 4:47-74.


Garmonsway, G. N., 1994 (1953), The Anglo-Saxon Chronicle, Charles C. Tuttle, Rutland, Vt. Geary, P. J., 1983, Ethnic Identity as a Situational Construct in the Early Middle Ages, Mitteilun-

Gelsinger, B., 1981, Icelandic Enterprise: Commerce and Economy in the Middle Ages, Univer-

Gerrard, A., 1981, The Stratigraphical Importance of Soils, in: Soils and Landforms (A. Gerrard,

Giddens, A., 1979, Central Problems in Social Theory: Action, Structure, and Contradiction in

gen der Anthropologischen Gesellschaft in Wien, (113): 15-26.

sity of South Carolina Press, Columbia.

ed.), George Allen & Unwin, London.

Social Analysis, University of California Press, Berkeley.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (318)


Giddens, A., 1984, The Constitution of Society: Outline of the Theory of Structuration, Polity

Gilman, A., 1981, The Development of Social Stratification in Bronze Age Europe, Current An-

Gissel, S., 1979, Tracing Old Structures in the Agrarian Cultural Landscape, Especially in Hor-nsherred, Zealand, in: Permanent European Conference for the Study of the Rural Land-scape (V Hansen, ed.), Danish Social Science Research Council, Roskilde, pp. 3-8.

Glimberg, 1982, Udi Wor Kiøbsted Simbritzhaffn: En Kronika Från Vår Sista Danska Tid: 1611-1660 [Marketplace Simbritzhaffn: A Chronicle from our Last Danish Time: 1611-1660], Osterlen Museum, Simrishamn.

Press, Cambridge.

thropology 22:1-23.

Godelier, M., 1978, Infrastructure, Society, and History, Current Anthropology 19:763-771.Goffart, W., 1988, The Narrators of Barbarian History (AD 550-800) Jordanes, Gregory of

Gonzalez, J., 1984, The Story of Christianity, Harper & Row, NewYork. Gordon, B. C., 1978, Chemical and Pedological Delimiting of Deeply Stratified Archaeological

Sites in Frozen Ground, Journal of Field Archaeology 5:331-338.Gould, S. J., 1986, Evolution and the Triumph of hom*ology, or Why History Matters, American

Scientist 74:60-69.Görman, M., 1990, Nordic and Celtic Religion in Southern Scandinavia in the Late Bronze Age

and Early Iron Age, in: Old Norse and Finnish Religions and Cultic Place-Names (T.Ahlbåck, ed.), Donner Institute, Åbo, pp. 329-343.

Grandell, A., 1986, Kredithandel Med Tillhjälp Of Karvstock [Credit Transactions with the Help of the Tally Stick], Fornvånnen 81.

Grebe, K., 1978, Zur Südlichen Importkeramik Des 10 Und 11 Jahrhunderts Im Havelgebiet, Veroffentlichungen des Museums für Ur und Gruhgeschichte, Potsdam 1978(12):219-248.

Griffith, M. A., 1980, A Pedological Investigation of an Archaeological Site in Ontario, Canada I: An Examination of the Soils in and Adjacent to a Former Village, Geoderma 24:527-336.

Griffith, M.A., 1981, A Pedological Investigation of an Archaeological Site in Ontario, Canada II:Use of Chemical Data to Descriminate Features of the Benson Site, Geoderma 25:27-34.

Groth, P., 1990, Vision, Culture, and Landscape, Berkeley Symposium on Cultural Landscape Interpretation, University of California Press, Berkeley.

Grøngaard-Jeppeson, T. , 1981 , Middelalderlandsbyens Opståen. Kontinuitet Og Brud In Den Fynske Agrarbebygglese Mellem Yngre Jernalder Og Tidlig Middelalder [The Origin of Me-dieval Rural Villages. Continuity and Breaks in Fyn’s Agriarian Settlement Between the Late Iron Age and the Early Middle Ages], Fynske Studier XI.

Gunn, J. D., 1994, Global Climate and Regional Biocultural Diversity, in: Historical Ecology: Cul-tural Knowledge and Changing Landscapes (C. Crumely, ed.), School of American Re-search Press, Santa Fe.

Gunneng, H., and B. Strand, 1979, Kvinnans Ekonomiska Ställning Under Nordisk Medltid [The Conditions of Women in the Medieval North], Kvinnohistorisk Symposium, Lund.

Gustafsson, G., 1925, St. Nicolai Kyrka I Simrishamn [St. Nicolai Church in Simrishamn], Förenin-gens För Fornminnes Och Hembugdesvhrd I Sydöstra Skåne, Skrifter II: 130-164.

Haasum, S., 1974, Vikingatidens Segling Och Navigation, Theses and Papers in North-EuropeanArchaeology, Stockholm.

Halbwachs, M., 1925, translation 1980, The Collective Memory (E J. Ditter, Jr., and V.Y. Ditter, trans.), Harper & Rowe, NewYork.

Hald, K., 1950, Vore Stednavne I, Gyldendal, Copenhagen. Hald, K., 1965, Vore Stednavne 2, Gyldendal, Copenhagen. Hammarberg, I., B. Malmer, and T. Zachrisson, 1989, Byzantine Coins Found in Sweden, Kung-

Tours, Bede, and Paul the Deacon, Princeton University Press, Princeton.

liga Vitterhets Historie och Antikvitets Akademien, London.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (319)


Hansen, V., 198 1 , The Study of the Rural Landscape, The Danish Social Science Research Coun-

Hansen, S., 1982, Andersminde—En Vikingetidig Bebyggelse Nær Ribe, Mark og Montre

Hansen, U. L., 1987, Romischer Import Im Norden [Roman Imports in the North], Det Kon-gelige Nordiske Oldskriftselskab, Copenhagen.

Hansen, B. R., 1988, Settlement Change and the Agricultural Structure in the Late Iron Age and Medieval Åland, Geografisk Annaler 70b: 1988: 1.

Hansen, T. E. , 1988, Die Eisenzeitliche Siedlung Bei Nørre Snede, Mitteljütland, Vorläufiger Bericht, Acta Archaeologica 58:171-200.

Hansen, U. L., 1990, Langdistancehandel I Romersk Jernalder-Fra Gaveudveksling Til Male-handel [Long-Distance Trade in the Roman Iron Age—From Gift Exchange to Administrative Trade], Hikuin 16:63-88.

cil, Copenhagen.


Hanssen, B., 1977, Österlen, Institut fîr Folklivsforskning Stockholms Universitetet, Östervåla.Hardesty, D. L., 1977, Ecological Anthropology, Wiley, New York. Hassan, E, 1978, Sediments in Archaeology: Methods and Implications for Palaeoenvironmental

and Cultural Analysis, Journal of Field Archaeology 5: 197-213.Hassan, E, 1981, Rapid Quantitative Determination of Phosphate in Archaeologcal Sediments,

Journal of Field Archaeology 8:384-387.Hastorf, C., 1990, One Path to the Heights: Negotiating Political Inequality in the Sausa of Peru,

in: The Evolution of Political Systems. Sociopolitics in Small-Scale Sedentary Societies (S. Upham, ed.), Cambridge University Press, Cambridge, pp. 146-176.

Hastorf, C. A., 1993, Agriculture and the Onset of Political Inequality Before the Inka, Cam-bridge University Press, Cambridge.

Hastrup, K., 1985, Culture and History in Medieval Iceland, Clarenden Press, Oxford. Hatt, G., 1942, En Sænket og Haevet Jernaldersboplads ved Ringkøbing Fjord, Svensk Geografisk

Årsbok 194:314-329.Hägerstrand, T., 1978, Survival and Arena: On the Life History of Individuals in Relation to Their

Geographical Environment, in: Timing Space and Spacing Time: Human Activity and Time Geography (T. Carlstein, D. Parkes, and N. Thrift, eds.), Edward Arnold, London, pp. 122-145.

Hägerstrand, T., 1982, Diorama, Path, and Project, Tijschrift voor Economische en Sociale Ge-ografie 73:323-339.

Hårdh, B., 1988, Kan Man Aga En Vikingaskatt? [Is There Evidence for a Viking Tax?], Fornvän-nen 83:57-58.

Hårdh, B., and B. Wyszomirska-Werbart, 1992, Contacts Across the Baltic Sea During the Late Iron Age (5th- 12th Centuries), Baltic Sea Conference, Lund, October 25-27.

Hedeager, L., 1978, Processes Toward State Formation in Early Iron Age Denmark, in: New Di-rections in Scandinavian Archaeology (K. Kristiansen and C. Paludin-Müller, eds.), Studies in Scandinavian Prehistory and Early History, Copenhagen, pp. 217-223.

Hedeager, L., 1982, Settlement Continuity in the Villages of Stevns, South-Eastern Zealand—An Archaeological Investigation, Journal of Danish Archaeology 1.

Hedeager, L., 1987, Empire, Frontier, and the Barbarian Hinterland: Rome and Northern Europe from AD 1-400, in: Centre and Periphery in the Ancient World (M. J. Rowlands, L. Larsen, and K. Kristiansen, eds.), Cambridge University Press, Cambridge, pp. 125-140.

Hedeager, L., 1988, Money Economy and Prestige Economy in the Roman Iron Age, in: Tradeand Exchange in Prehistory (B. Hårdh, ed.), Lund University Press, Lund.

Hedeager, L., 1992, Iron Age Societies: From Tribe to State in Northern Europe, 500 BC to AD 700, (J. Hines, trans.), Blackwell, Oxford.

Hedeager, L., and K. Kristiansen, 1981, Bendstrup—En Fyrstegmv Fra Den Romerske Jernalder: Dens Sociale Og Historiske Miljo, Kuml (1981):81-149.


Landscapes of Power, Landscapes of Conflict - [PDF Document] (320)


Heer, E, 1981, The Holy Roman Empire (J. Sondheimer, trans.), Praeger, New York. Herrmann, J., 1985, Hinterland, Trade, and Craftworking of the Early Trading Stations of the North-

Western Slavs, in: The Comparative History of Urban Origins in Non-Roman Europe: Ire-land, Wales, Denmark, Germany, Poland, and Russia from the Ninth to the Thirteenth Century (H. Clarke and A. Simms, eds.), BAR International Series, Oxford, pp. 249-266.

Hillgarth, J. N., 1969, The Conversion of Western Europe, 350-750, Prentice-Hall, Englewood Cliffs, N.J.

Hjelmqvist, H., 1981, Grain Impressions from Gårdlösa, in: Gårdlösa: An Iron Age Community in Its Natural and Social Setting (B. Stjernquist, ed.), CWK Gleerup, Lund, pp. 54-58.

Hjelmqvist, H., 1992, Some Economic Plants of the Prehistoric and Medieval Periods in Southern Sweden, in: The Archaeology of the Cultural Landscape: Field Work and Research in a South Swedish Rural Region (L. Larsson, J. Callmer, and B. Stjernquist, eds.), Almqvist & Wiksell International, Stockholm, pp. 359-368.

Hodder, I., 1983, Reading the Past, Cambridge University Press, Cambridge. Hodder, T., 1990, The Domestication of Europe, Cambridge University Press, Cambridge. Hodgen, M. T., 1974, Anthropology, History, and Cultural Change, University of Arizona Press,

Hodges, R., 1978, Ports of Trade in Early Medieval Europe, Norwegian Archaeological Re-

Hodges, R., 1982a, Dark Age Economics, Duckworth, London. Hodges, R., 1982b, The Evolution of Gateway Communities, in: Ranking, Resource, and Ex-

Hodges, R., 1984, North Sea Trade Before the Vikings, in: Archaological Approaches to Medieval

Hodges, R., 1988, Primitive and Peasant Markets, Basil Blackwell, Oxford. Hodges, R., 1989, Emporia, Monasteries, and the Economic Foundation of Medieval Europe, in:

Medieval Archaeology (C. Redman, ed.), Medieval and Renaissance Texts and Studies, Binghamton, N.Y., pp. 306-320.

Hodges, R., and D. Whitehouse, 1983, Mohammed, Charlemagne, and the Origins of Europe, Duckworth, London.

Hoffmann, E., 1981, The Unity of the Kingdom and the Provinces in Denmark During the Mid-dle Ages, in: Danish Medieval History: New Currents (N. Skyum-Nielsen and N. Lund, eds.), Museum Tusculanum Press, Copenhagen, pp. 95-111.

Hoflund, S., 1921, Gamla Staden Tumathorp Eller Tummerup, Föreningens För Fornminnes Och Hembugdesård I Sydöstra Skåne, Skrifter 136-58.

Holl, A., 1985, Background to the Ghana Empire: Archaeological Investigations on the Transition to Statehood in the Dhar Tichitt Region (Mauritania), Journal of Anthropological Archae-

Holliday, V., 1983, Radiocarbon Ages from the Lubbock Lake Site, 1950-1980: Framework for Cultural and Ecological Change on the Southern High Plains, Plains Anthropologist

Holmberg, B., 1946, Tomt Och Toft Som Appelativ Och Ortnamnselement [Tomt and Toft as Names and Placename Elements], Uppsala University, Uppsala.

Holmberg, B., 1990, Views on Cultic Place-Names in Denmark, in: Old Norse and Finnish Reli-gions and Cultic Place-Names (T. Ahlback, ed.), Donner Institute, Åbo, pp. 381 -393.

Hørby, K., 1981, The Social History of Medieval Denmark, in: Danish Medieval History: New Currents (N. Skyum-Nielsen and N. Lund, eds.), Museum Tusculanum Press, Copenhagen,


view 11: 2.

change (C. Renfrew and S. Shennan, eds.), Cambridge University Press, Cambridge.

Europe (K. Biddick, ed.), Western Michigan University, Kalamazoo, pp. 193-202.

ology 4:73-115.

28( 101): 165-182.

pp. 36-53.Hudson, F. S., 1970, A Geography of Settlements, McDonald Evans Ltd., London. Hufferd, J., 1980, Toward a Transcendental Human Geography of Places, Antipode 12(3):18-23.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (321)


Hulthén, B., 1978, Keramiktillverkning I Kv Tankbåten IYstad [Ceramic Production in TankbåtenBlock in Ystad], Ystadiania XXIII:103-114.

Hvass, S., 1980, Vorbasse: The Viking-Age Settlement at Vorbasse, Central Jutland, Acta Archaeo- logica 50: 137-172.

Hvass, S., 1983, Vorbasse: The Development of a Settlement Through the first Millenium AD, Journal of Danish Archaeology 2:127-136.

Hvass, S., 1989, Rural Settlements in Denmark in the First Millenium AD, in: The Birth of Europe: Archaeology and Social Development in the First Millenium AD (K. Randsborg, ed.), Analecta Romana Instituti Danici, Rome, pp. 91-99.

Hyenstrand, Å., 1983, Reflexions on Communication Technique, Society, and Raw Materials Dur-ing the Viking Age, in: Society and Trade in the Baltic During the Viking Age (S.-O.Lindqvist, ed.), Gotlands Fornsal, Visby, pp. 63-68.

Hyenstrand, Å., 1987, Fasta Fomlämningar Och Archaeologiska Regioner [Remaining Prehis-toric Monuments and Archaeological Regions], Rapport No. 1984:7, Riksantikarieämbetet (National Board of Antiquities of Sweden).

Ingers, E., 1943, Bonden I Svensk Historia [Farmers in Swedish History], Lantbruksförbundets Tidskriftaktiebolag, Stockholm.

Ionnisyan, O., 1990, Archaeological Evidence for the Development and Urbanization of Kiev from the 8th to the 14th Centuries, in: From the Baltic to the Black Sea Studies in Medieval Ar-chaeology (D. Austin and L. Alco*ck, e&.), Unwin Hyman, London, United Kingdom.

Isaksson, S., 1980, Önsjo Härad, in: Skånes Ortnamn Series A Bebyggelsenamn, Vol. 14, Gleerup, Lund.

Jacobsson, B., 1979a, Bebyggelselämningar Medeltid Tomterna 11- 18 kv. Sumaros, Simr-ishamn, Skåne [Settlement Remains Report, Medieval Site 11-18 Sumaros Block Simrin-hamn], Riksäntikvarie Åmbetet, Stockholm.

Jacobsson, B., 1979b, Bebyggelselämningar, Vikingatid-Medeltid kv. Ålen Simrishamn, Skåne [Settlement Remains Report, Viking Age—Medieval, Ålen Block, Simrishamn] , Riksantik-varie Åmbetet, Stockholm.

Jacobsson, B., 1987, Oden Ett Vikingakvarter I Staden Trelleborg (Oden: A Viking Age Block in the City of Trelleborg), Ale-Historisk Tidskrift För Skåneland 2( 1987): 1-10.

Jacobsson, B., 1989, Trelleborgen På Gång! Riksantikarieämbetet UV-Syd Trelleborgs Museum. Jacobsson, B., 1990a, Visst Har Det Funnits En Borg I Trelleborg, in: Popular Arkeologi 8:7-10.Jacobsson, B., 1990b, En Skånsk Ringborg, Fornvännen 85:53-54.Jacobsson, B., and L. Wallin, 1986, Trelleborg Under Vikingatid Och Medeltid, Riksantik-

varieäbetet, Lund. Jansen, H., 1985, Early Urbanization in Denmark, in: The Comparative History of Urban origins

in Non-Roman Europe: Ireland, Wales, Denmark, Germany, Poland, and Russia from the Ninth to the Thirteenth Century (H. Clarke and A. Simms, eds.), BAR International Series, Oxford, pp. 183-216.

Jansen, H. M., T. Nyberg, and T. Riis, 1977, Danske Byers Fremvekst Og Udvikling I Middelalderen [Danish Town Origins and Development in the Middle Ages], in: Middelaldersteder [Me-dieval Cities] (G. A. Blom, ed.), Universitetsforlaget, Oslo, pp. 9-90.

Jennbert, K., L. Larsson, R. Petré, and B. Wyszomirska-Werbart, 1991, Regions and Reflections: In Honor of Märta Strömberg, Almqvist & Wiksell International, Lund.

Jensen, G. E, 1979, Place-Names and the Scandinavian Settlement in England, in: Permanent Eu-ropean Conference for the Study of the Rural Landscape (V. Hansen, ed.), Danish Social Science Research Council, Roskilde, pp. 119-122.

Jensen, J., 1982, The Prehistory of Denmark, Methuen, New York. Jensen, S., 1990, Handel Med Dagligvarer I Vikingetiden [Viking Age Trade in Basic Commodi-

ties], Hikuin 16:119-138.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (322)


Jensen, J. S., K. Bendixen, N. K. Leibgott, and E Lindahl, 1992, Danmarks Middelalderlige Skat-tefund C. 1050- 1550 [Denmark’s Mediaeval Treasure-Hoards c. 1050-1550], Det Kgl. Nordiske Oldskriftselskab, Copenhagen.

Jochim, M., 1983, Optimization Models in Contex In: Archaeological Hammers and Theories (J. A. Moore and A. S. Keene, eds.), Academic Press, NewYork.

Johannesson, G., 1971, Skånes Historia [Skåne’s History], Bonniers, Stockholm. Johannesson, G., 1984, Skåne, Halland & Blekinge: Om Skånelandskapens Historia [Skåne,

Halland & Blekinge: On Skåne’s Landscape History], Norstedt & Sons, Stockholm. Johnson, A.W., and T. K. Earle, 1987, The Evolution of Human Societies from Foraging Groups

to Agrarian States, Stanford University Press, Stanford, Calif. Johnson, G.A. 1973, Local Exchange and Early State Development in Southewestern Iran, An-

thropological Papers, Museum of Anthropology, University of Michigan, Ann Arbor. Johnson, G. A., 1976, Information Sources and the Development of Decision-Making Organi-

zations, Conference paper presented at Social Differentiation and Interaction, SUNY Bing-hamton, Binghamton, N.Y.

Johnson, G. A., 1980, Spatial Organization of Early Uruk Settlement Systems, in: Colloque L’Ar-qéologie de L‘Iraq du Debut de I’Epoque Neolithique, CNRS, Paris, pp. 204-263.

Johnson, G.A., 1981, Monitoring Complex System Integration and Boundary Phenomenon with Settlement Size Data, in: Archaeological Approaches to the Study of Complexity (S. E. Van Der Leeuw, ed.), Univeristy of Amsterdam Albert Egges van Giffen Insitiuut voor Prae—un Protohistorie, CINGVLA VI, Amsterdam, pp. 143-188.

Johnson, G.A., 1982, Organizational Structure and Scalar Stress, in: Theory and Explanation in Ar-chaeology (C. Renfrew, M. J. Rowlands, and B. A. Segraves, eds.), Academic Press, New York.

Johnson, G. A., 1983, Decision-Making Organization and Pastoral Nomad Camp Size, Human

Johnson, G. A., 1987, The Changing Organization of Uruk Administration on the Susiana Plain, in: Archaeological Perspectives on Western Iran: From Prehistory to the Islamic Conquest (F. Hole, ed.), Smithsonian Press, Washington D.C., pp. 107-140.

Johnston, R. J., 1979, Geography and Geographers: Anglo-American Human Geography Since 1945, John Wiley and Sons, NewYork.

Jones, G.R.J., 1984, The Multiple Estate: A Model for Tracing the Interrelationships of Society, Economy, and Habitat, in: Archaological Approaches to Medieval Europe (K. Biddick, ed.), Western Michigan University, Kalamazoo, pp. 9-42.

Ecology 11(2):175-199.

Jones, G., 1987, A History of the Vikings, Oxford University Press, Oxford. Jones, M., 1991, The Elusive Reality of Landscape: Concepts and Approaches in Landscape Re-

search, Norsk Geografisk Tidskrift 45:229-244.Jorgensen, S., 1987, Danmarks Kongemagt Og Dens Fodsel [Denmark’s Royal Power and its

Birth], Aarhus Universitetsforlag, Århus. Jorgensen, L., and T. Skov, 1980, Trabjerg: A Viking Age Settlement in Northwest Jutland, Acta Ar-

chaeologica 50:119-136Joukowsky, M., 1980, A Complete Manual of Field Archaeology, Prentice Hall, Englewood

Cliffs, N.J. Jørgensen, B., 1979, Field-Names on Demesne Land, in: Permanent European Conference for

the Study of the Rural Landscape (V. Hansen, ed.), Danish Social Science Research Council, Roskilde, pp. 123-128.

Jørgensen, O., 1985, Alfred Den Store Danmark’s Geografi: En Undersøgelse Af Fire Afsnit I Den Gamle Orosius [Alfred the Great of Denmark’s Geography: An Investigation of Four Sections in the Old Orosius], Odense University Press, Odense.

Kaul, E, 1985, A Settlement Site of the Later Iron Age at Vallensbæk Near Copenhagen, Journalof Danish Archaeology 4:157-163.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (323)


Kaul, E, 1993, Tørkens Fortjeneste, Skalk 1993(3):8-10.Keay, S., J. Creighton, and D. Jordan, 1991, Sampling Ancient Towns, Oxford Journal of Arch.

Keene, A. S., 1979, “Economic optimization models and the study of hunter-gatherer subsistence settlement systems,” in: Transformations: Mathematical Approaches to Culture Change (A. P. Vayda, ed.), Natural History Press, Garden City, N.Y.

Kelm, R., 1993, Mölleholmen. Die Untersuchung Des Fundmaterials Und Der TopograpbieEiner Spätwikingerzeitlich-Frümittlealterlichen Inselsiedlung Is Südschweden, Uppsats i Arkeologi, Lunds University.

Kempke, T., 1981, Die Keramik Van Warder, Kreis Segeberg Und Ihre Stulllung In Ostholstein, Offa 1981:289-32 1.

Kempke, T., 1988, Zur Chronologie Der KeramikVon Starigard/Oldenburg, Bericht der Römisch-Germanischen Kommission Band 69.

Killion, T.W., 1989, Intensive Surface Collection of Residential Clusters at Terminal Classic Sayil, Yucatan, Mexico, Journal of Field Archaeology 16:273-294.

Kimes, T., C. Hasselgrove, and I. Hodder, 1982, A Method for the Identification of the Location of Regional Cultural Boundaries, J of Anthropological Arch. 1(2): 113-131.

Kirby, A., 1993, Power/Resistance: Local Politics and the Chaotic State, Indiana University Press, Indianapolis.

Kirsch, E., 1979, Neue Grabfunde Aus Spätsluwischer Zeit Im Bezirk Cottbus, Veroffentlichun-gen des Museums für Ur und Gruhgeschichte, Potsdam 12:249-252.

Kirch, P.V., 1984, The Evolution of Polynesian Chiefdoms, Cambridge University Press,Cambridge.

Klindt-Jensen, O., and H., Andersen, 1963, Det Ældste Århus. Kuml 1963 Knapp, A. B., 1992a, Archaeology, Annales, and Ethnohistory, Cambridge Universtiy Press,

Cambridge.Knapp, A. B., 1992b, Archaeology and Annales: Time, Space, and Change, in: Archaeology, An-

nales, and Ethnohistoly (A. B. Knapp, ed.), Cambridge University Press, Cambridge, pp. 1-22.Kohl, P. L., 1981, Materialist Approaches in Prehistory, Ann. Rev. Anthropol. 89-118.Kohl, P. L., 1988, The Use and Abuse of World Systems Theory: The Case of the “Pristine”West

Asian State, in: Archaeology in the Americas and Beyond (C. C. Lamberg-Karlovsky, ed.), Cambridge University Press, Cambridge, pp. 219-240.

Kolchin, B. A., 1989, Wooden Artifacts from Medieval Novgorod, Part I. Text (A.V. Chernetsov, trans.), BAR International Series, Oxford, United Kingdom.

Konrad, V. A., R. Binnichsen, and V Clay, 1983, Soil Chemical Identification of Ten Thousand Years of Prehistoric Human Activity at the Munsungun Lake Thoroughfare, Maine, Journal of Ar-chaeological Science 10:13-28

Kosse, K., 1990, Group Size and Societal Complexity: Thresholds in the Long-Term Memory, Journal of Anthropological Archaeology 9:275-303.

Kousgård-Sørensen, J., 1968, Danske Sø og ànavne (Danish Lake and River Names) I:A-D. Kousgård-Sørensen, J., 1979, Place Names and Settlement History, in: Names, Words, and Graves:

Early Medieval Settlement (P Sawyer, ed.). Kousgård-Sørensen, J., 1990, The Change of Religion and the Names, in: Old Norse and Finnish

Religions and Cultic Place-Names (T. Ahlbäck, ed.), Donner Institute, Åbo, pp. 394-403.Kowalewski, S.A., 1980, Population-resource balances in Period I of Oaxaca, Mexico, American

Antiquity 45(1)151-165.Kowalewski, S.A., 1982, Population and agricultural potential: Early I-V, in: Monte Alban’s Hin-

terland, Part I: Prebispanic Settlement Patterns of the Central and Southern Parts of the Valley of Oaxaca, Mexico, Prehistory and Human Ecology of the Valley of Oaxaca, Vol. 7


Landscapes of Power, Landscapes of Conflict - [PDF Document] (324)


(K. Flannery and R. Blanton, eds.) Memoirs, 15. Museum of Anthropology, University of Michigan, Ann Arbor.

Kowalewski, S. A., 1983, Boundaries, Scale, and Internal Organization, Journal of Anthropolog- ical Archaeology 2:32-56.

Kowalewski, S. A., G. M. Feinman, L. Finsten, R. E. Blanton, and L. M. Nicholas, 1989, Continuity and Change in Public Architecture: Periods I Through V of the Valley of Oaxaca, Mexico, in: Monte Alban’s Hinterland, Part II: The Prehispanic Settlement Patterns in Tlacolula, Etla, and Ocotlán, the Valley of Oaxaca, Mexico, University of Michigan, Museum of Anthropol-ogy, Memoir 23, Ann Arbor, pp. 409-447.

Kristiansen, K., 1991, Chiefdoms, States, and Systems of Social Evolution, in: Chiefdoms: Power, Economy, and Ideology (T. K. Earle, ed.), Cambridge University Press, Cambridge, pp. 16-43.

Kroman, E., 1976, Det Danske Rige I Den Aeldre Vikingtid [The Danish State in the Older Viking Age], Roskilde og Bagger, Copenhagen.

Krongaard Kristensen, H., 1988, Udgravningerne Ved Viborg Søndersø [Excavations At Viborg Søndersø], Hikuin 1988(14):9-22.

Kroon, S., 1989, Canute the Holy’s Lost Deed of Gift of May 21,1085: An Antedated Twelfth Cen-tury Document, Scandia 55(2):329-330.

Kurchin, B., 1984, The Effect of Roman Colonization on the Votadini, Master’s thesis, Hunter College, City University of New York.

Lagerqvist, L., 1982, Sverige Och Dess Regenter Under 1000 År [Sweden and Its Rulers During 1000 Years], Bonniers, Stockholm.

Langston, J., 1988, The Two Traditions of Geography, Historical Geography, and the Study of Landscapes, Geografisk Annaler 70B: 17-25.

Larson, L. M., 1969, The King’s Household in England Before the Norman Conquest, Green-wood Press, NewYork.

Larsson, G., and B. Starrin, 1984, Decision Making in Stressful Conditions: A Model Based on the Coping Perspective, FOA Report No. C 55064, Havadavdelning, Forsvarets forskn-ingsanstatt (National Defense Research Institute).

Larsson, L., J. Callmer, and B. Stjernquist, 1992, The Archaeology of the Cultural Landscape. Field Work and Research in a South Swedish Rural Region, Almqvist & Wiksell Interna-tional, Stockholm.

Larsson, M., and D. Olausson, 1992, Archaeological Field Survey—Methods and Problems, in: TheArchaeology of the Cultural Landscape: Field Work and Research in a South Swedish Rural Region (L. Larsson, J. Callmer, and B. Stjernquist, eds.), Almqvist & Wiksell Interna-tional, Stockholm, pp. 473-480.

Lauritsen, A., 1974, Volden Omkring Vikingetidens Odense [The Walls Around Viking Age Odense], Fynske Minder 1974: 140- 153.

Lawson, M. K., 1993, cnu*t: The Danes in England in the Early Eleventh Century, Longman,London.

Lb, Landbok, (archive). Le Roy Ladurie, E., 1979, The Territory of the Historian, University of Chicago Press, Chicago. Leone, M., 1978, Time in American Archaeology, in: Social Archaeology (C. Redman, ed.), Aca-

demic Press, NewYork, pp. 25-36.Leone, M., 1981, Archaeology’s Relationship to the Present and the Past, in: Modern Material

Culture: The Archaeology of Us (R. A. Gould and M. B. Schiffer, eds.), Academic Press, New York, pp. 5-14.

Leone, M., 1982, Some Opinions about Recovering Mind, American Antiquity 47(4):742-760.Lewis, A., 1978, The Northern Seas: Shipping and Commerce in Northern Europe AD

300-1100, Octagon, New York.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (325)


Liebgott, N. K., 1978a, Danske Fund Af Møntdateret Keramik Ca. 950-1450 [Danish Finds of

Liebgott, N. K., 1978b, Keramik Fra Vikingetid Og Middelalder [Ceramics from the Viking Age

Liebgott, N. K., 1989, Dansk Middelalder Arkæologi, G.E.C. GAD, Copenhagen. Lilja, M., 1990, Österlen, Med Simrishamn I Centrum [Österlen with Simrishamn in Its Center],

Simrishamn Turist och Fritidsnämnd Grafotryck, Simrishamn. Linderholm, J., and E. Lundberg, 1994, Chemical Characterization of Various Archaeological Soil

Samples Using Main and Trace Elements Determined by Inductively Coupled Plasma Atomic Emission Spectrometry, Journal of Archaeological Science 21:303-331.

Lindow, J., 1976, Comitatus, Individual and Honor: Studies in Northern Germanic Institu-tional Vocabulary, University of California Publications in Linguistics, Berkeley.

Lindquist, S. O., 1974, The Development of the Agrarian Landscape on Gotland During the Early Iron Age, Norwegian Archaeological Review 7(1): 6-32.

Lindquist, I., 1987, Religiosa Runtexter 3, Lund University, Lund. Linnington, R. E., 1972a, A Summary of Simple Theory Applicable to Magnetic Prospecting in Ar-

chaeology, Prospezioni Archeologiche 7:9-59.Linnington, R. E., 1972b, Topographical and Terrain Effects in Magnetic Prospecting,.Pospezioni

Archeologiche 7:61-84.Lippi, R. D., 1988, Paleotopology and Phosphate Analysis of a Buried Jungle Site in Ecuador, Jour-

nal of Field Archaeology 1585-97.Lopez, R. S., and I.W. Raymond, 1990, Medieval Trade in the Mediterranean World, Columbia

University Press, New York. Lorch, W., 1952, Das Erkennen Des Relickt-Charakters Von Waldakpellen Aus Den Phosphatgehalt

Des Bodens, Zeitschrift fur Wurttembergische Landesgeschichte 11:246-253.Lofving, C., 1984, Förändring I Maktstrukturen Kring Ostra Skagerrak Vid Kristendomens In-

forande—En Tvävetenskaplig Untgångspunkt For En Arkeologisk Studie [Changes in Power Structure Around the East Skagerrak at Christianity’s Introduction—An Interdisciplinary Starting Point for Archaeological Study], Fornvännen 79:93-103.

Coin-Dated Ceramics Ca. 950- 14501, Nationalmuseet, Copenhagen.

and Middle Ages], Nationalmuseet, Copenhagen.

LSLb, Lunds Stifts Landebog. Lund, N., 1981, Viking Age Society in Denmark—Evidence and Theories, in: Danish Medieval

History: New Currents (N. Skyum-Nielsen and N. Lund, eds.), Copenhagen Museum Tuscu-

Lund, N., 1984, Two Voyagers to the Court of King Alfred: The Ventures of Ohthere and Wulfs- tan, Together with the Descriptions of Northern Europe from the Old English Orosius, William Sessions Ltd., York, United Kingdom.

Lund, N., 1991, “Denemearc,”“tanmarkar But” and “tanmaurk Ala,” in: People and Places in North- ern Europe AD 500-1600 (I. Wood and N. Lund, eds.), Boydell Press, Woodbridge, pp.

Lundqvist, L., E. Rosengren, and J. Callmer, 1993, En Fyndplats Med Guldgubbar Vid Slöinge, Hal-land, Fornvännen 88(2):65-70.

Luttwak, E., 1976, The Grand Strategy of the Roman Empire: From the First Century BC to the Third, Johns Hopkins University Press, Baltimore.

Lyons, T., R. Hitchco*ck, and W. Wills, 1980, Remote Sensing: Aerial Anthropological Perspectives: A Bibliograpty of Remote Sensing in Cultural Resource Studies, National Park Service, United States Department of the Interior, Washington D.C.

lanum Press, pp. 22-35.


Madsen, H. J., 1969, Om Grubehuse [On Pithouses], Mark og Montre 1969:19-25.Madsen, H. J., 1987, Germanic Iron Age and Viking Age, in: Arkœologiske Udgravninger I Dan-

mark 1986 (O. Olsen, ed.), Rigsantikvarens Arkæologiske Sekretariat, Nationalmuseet, Copenhagen, pp. 44-46.

Landscapes of Power, Landscapes of Conflict - [PDF Document] (326)


Madsen, P. K., 1988, Små Krukker Fra Ribe [Small Vessels from Ribe], Hikuin 14:79-94.Madsen, I? K., 1990, By, Marsk Og Geest 1, Den Antikvariske Samling i Ribe, Ribe. Mahler, D., 1985, Ragnesminde: A Germanic Early Viking Age House Site in Eastern Sjælland,

Journal of Danish Archaeology 4: 164-167.Mahndahl, A., 1974, En Puzzelbit I Oxie’s Historia [A Piece of the Puzzle in Oxie’s History], in:

Kring Malmohus, Malmö Museum, Malmö. Malmer, B., 1966, Nordiska Mynt Fore År 1000 [Nordic Coins Before the Year 10001, Acta Ar-

chaeologica Lundensia Volume 8(4). Malmer, B., 1990, What Does Coinage Tell Us About Scandinavia in the Late Viking Age? in: From

the Baltic to the Black Sea: Studies in Medieval Archaeology (D. Austin and L. Alco*ck, eds.), Unwin Hyman, London, pp. 157-167.

Mann, M., 1986, The Sources of Political Power; Vol. 1: A Histo y of Power from the Beginning to AD 1760, Cambridge University Press, Cambridge.

Marseen, O., 1959, Lindholm Hoje, Kuml 1959:53-68Martens, J., 1989, The Vandals: Myths and Facts About a Germanic Tribe of the First Half of the

1st Millennium AD, in: Archaeological Approaches to Cultural Identity (S. J. Shennan, ed.), Unwin Hyman, London, pp. 57-66.

Mattingly, D. J., 1992, War and Peace in Roman North Africa: Observations and Models of State-Tribe Interaction, in: War in the Tribal Zone (R. B. Ferguson and N. L. Whitehead, eds.), School of American Research Press, Santa Fe, pp. 31-60.