Global Perspectives on Landscapes of Warfare 1646420993, 9781646420995

Citation preview

[128.104.46.206] Project MUSE (2024-03-01 20:28 GMT) UW-Madison Libraries

Global Perspectives on

Landscapes of Warfare

Global Perspectives on

Landscapes of Warfare

EDITED BY

HUGO C. IKEHARA-TSUKAYAMA AND JUAN CARLOS VARGAS RUIZ

U N I V ER SI T Y P R ESS O F C O L O R A DO Louisville

ED I T O R I AL D E L A U N I V ER SI DAD D EL  MAG DAL ENA Magdalena

© 2022 by University Press of Colorado Published by University Press of Colorado and Editorial de la Universidad del Magdalena University Press of Colorado 245 Century Circle, Suite 202 Louisville, Colorado 80027 Editorial de la Universidad del Magdalena Carrera 32 No. 22-08 Santa Marta D.T.C.H. Colombia. Código Postal No. 470004 All rights reserved Manufactured in the United States of America The University Press of Colorado is a proud member of the Association of University Presses. The University Press of Colorado is a cooperative publishing enterprise supported, in part, by Adams State University, Colorado State University, Fort Lewis College, Metropolitan State University of Denver, Regis University, University of Alaska Fairbanks, University of Colorado, University of Denver, University of Northern Colorado, University of Wyoming, Utah State University, and Western Colorado University. ∞ This paper meets the requirements of the ANSI/NISO Z39.48–1992 (Permanence of Paper) ISBN: 978-1-64642-099-5 (hardcover) ISBN: 978-1-64642-211-1 (ebook) https://doi.org/10.5876/9781646422111 Library of Congress Cataloging-in-Publication Data Names: Ikehara-Tsukayama, Hugo C. (Hugo César), editor. | Vargas Ruiz, Juan Carlos, editor. Title: Global perspectives on landscapes of warfare / edited by Hugo C. Ikehara-Tsukayama and Juan Carlos Vargas Ruiz. Description: Louisville : University Press of Colorado, [2021] | Includes bibliographical references and index. Identifiers: LCCN 2021036672 | ISBN 9781646420995 (hardcover) | ISBN 9781646422111 (ebook) Subjects: LCSH: Warfare, Prehistoric. | Landscape archaeology. | Fortification—History. | Siege warfare—History. | Military art and science—History. Classification: LCC GN799.W26 G56 2021 | DDC 303.6/6—dc23/eng/20211028 LC record available at https://lccn.loc.gov/2021036672 Cover photograph by Hugo C. Ikehara-Tsukayama

Contents

1. Landscapes of People at War Hugo C. Ikehara-Tsukayama and Juan Carlos Vargas Ruiz 3 2. The Neolithic and Chalcolithic Fortified Settlements of the Northernmost Eurasian Taiga Viktor A. Borzunov (translated and adapted by Igor V. Chechushkov) 26 3. Winter Is Coming: Is Bronze Age “Fortification” Always Fortification? Igor V. Chechushkov 55 4. Communities, Violence and Fortification: A Study of Longshan Landscapes James T. Williams 79 5. Process of Warfare and Its Landscape in Protohistoric Japan Takehiko Matsugi 101 6. Inscribing Power on a Landscape: The Case of Co Loa in Vietnam Nam C. Kim and Russell S. Quick 118 7. Fortified Roads as Communication and Defense Networks in the Ancient Near East Tiffany Earley-Spadoni 139

8. The Fortified Settlements in the Saint-Dié-des-Vosges Basin during the Iron Age and the Roman Empire, France Lizzie Scholtus 159 9. Late Woodland Cultural Adaptations in the Lower Missouri River Valley: Archery, Warfare, and the Rise of Complexity Kerry Nichols 180 10. Warfare, Landscapes, and Social Complexity in the Pre-Hispanic Llanos of Colombia Juan Carlos Vargas Ruiz 201 11. Changing Threats: Early Fortifications and Regional Politics in Coastal Peru, 2500–1500 bp Hugo C. Ikehara-Tsukayama 224 12. Warfare and Alliance: A Human-Scale GIS Analysis of Defensive Alliances in the Colca Valley, Peru Lauren Kohut 247 13. Defensive Landscapes: Problems, Patterns, and Potential Elizabeth Arkush 273

vi

CO N T EN T S

Index

291

List of Contributors

301

Global Perspectives on

Landscapes of Warfare

1 War does not only incur immediate human and material losses; it forces individuals, families, and communities to change their perspectives and ways of life, and it fundamentally alters landscapes. For people who can migrate, war may mean leaving places their families have inhabited over generations and continuing their lives within other societies, often with unfamiliar cultural norms, a lower status, and different social roles. In search for safety, some people arrive in regions with different ecological and physical settings, requiring new approaches to practices, material culture, meanings, and interaction with the environment. For those who do not migrate, war means adapting to a new life—one shaped by fear and possibly scarcity and famine, hard borders, and banned territories—or being subject to practices that would be unacceptable in other situations, such as abuse or loss of freedom. The scars left by wars go beyond psychological. Conflict, violence, and fear can be fixed and materialized in landscapes. In designing defenses, communities move residences, build fortifications, invest resources, create alliances, and negotiate with human and nonhuman beings for help. The histories of how territories were appropriated and transformed by communities at war offer insight into how built landscapes not only reflect what happened but also influence generations to come. We present in this volume eleven cases of transformed landscapes, of different geographic origin, time depth, social complexity, and historical context.

Landscapes of People at War Hugo C. IkeharaTsukayama and Juan Carlos Vargas Ruiz

https://doi.org/10.5876/9781646422111.c001

3

This chapter briefly reviews how the main topics of warfare and landscapes have intersected in archaeological literature, how the physical manifestations of violence and conflict have become permanent features in landscapes, and how the chapters in this volume contribute to a better understanding of the topic. LANDSCAPES

Through archaeological studies of landscapes, we consider a wide range of questions and approaches—from those related to settlement patterns to symbolic and experiential approaches—that have been used in order to understand and explain past human geographies (Anschuetz et al. 2001; Bradley 1998; Knapp and Ashmore 1999; Moore 2005; Parsons 1972; Tilley 1994). These approaches differ in how they view the relationship between people and their social and natural environments. Some archaeological and anthropological approaches have focused on explaining environmental influences on how people obtained food and other resources, how people distributed themselves in a territory, how they organized themselves and interacted with other groups, and even how their religious beliefs were shaped, in adaptive terms, to keep their sociocultural system in balance. Landscapes were modeled in these terms especially but not only during the apogee of the New Archaeology. The postprocessual critique cast doubt on many of the assumptions that drove archaeological research until the 1970s and promoted a theoretical agenda asking for reflexivity, new epistemologies, individual volition, and practices (Hodder and Hutson 2003; Shanks 2008). Some of these new questions have shaped the way archaeologists understand space and study landscapes today. First, archaeologists were interested in the role of humans as agents of change in opposition to social structure (Dobres and Robb 2000). For instance, people were no longer considered passive beings adapting to predetermined environmental conditions; it was acknowledged that environments were in constant transformation and that people were active agents on it (Blume and Leinweber 2004; Crumley 2017; Hayashida 2005; Roberts et al. 2017). People have contributed to species extinction, transformed species (domestication), and modified environs (niche construction) to fit to their own needs. The view that most landscapes are anthropogenic was considered by many researchers for a long time, but during the last few decades this concept has been explicitly stated and even have become a subject of archaeological investigations. The notion of place—locations meaningful to people due to certain historical, identarian, and experiential circumstances linked to the construction of 4

H U G O C . I K EH A R A -T S U K AYA M A A N D J UA N CA R L O S VA RG A S R U I Z

individual and/or collective memories and practices—has been used in contrast to the notion of space—an abstract, objective, and quantifiable quality of spatial extension, a set of relationships between the subjects and objects and the positions everyone plays. The concept of place not only represents a location of physical activity; it also refers to the behavioral settings happening on it or in reference to it (Bradley 1998; Tilley 1994; Whitridge 2004). The landscapes studied by archaeologists are both manifestations of how people interacted with other people and nature, and how they have assigned meanings to these places. In the Central Andes, for instance, Inka landscapes integrated incredible transformed places with meanings linked by mythical stories and a ceremonial system (e.g., Bauer 2000; Kaulicke et al. 2004; Kosiba 2015; Santillana 2012; Taylor 1987) whose details have reached to us through early colonial records. This “reading” of past landscapes has been practiced in some national archaeological traditions since the early twentieth century (e.g., Tello and Miranda 1923) but also in nonwestern views of existing landscapes (Reid et al. 2014; see also Kim and Quick, chapter 6 in this volume). A. Bernard Knapp and Wendy Ashmore (1999) observe that three processes interplay in the conferral of meaning to places in landscapes. Certain locations (including those without human modification) became places of special cultural significance because they are associated with specific social practices and experiences or are articulated within narratives of how people view their world, forming part of what Knapp and Ashmore call conceptual landscapes and ideational landscapes, respectively. Some places perpetuate or fix meanings through the physical transformation of their topography, the third process resulting in constructed landscapes. While some constructions, such as monuments, are highly visible, other subtler modifications can have powerful meanings, too. Landscapes are not fixed but subject to constant change and reinterpretation, because both natural settings and culture are in constant flux. This flux allows archaeologists and other students of the past to reconstruct ancient landscapes through time; if landscapes were fixed and static, the remains from the past would be indistinguishable from the present. The constant change enables us to consider the historically specific forces, conditions, and contexts through which landscapes have been transformed. In this sense, landscape scale is integrative because it allows us to study human activities within their local historical context (Crumley 2007). Local history matters also matters because people occupy territories that, most of the time, were already modified by their antecessors. Landscapes are not only the result of people interacting with their social and natural L A N D S CA P ES O F P E O P LE AT WA R

5

environment at one moment in time; they are the medium that makes human actions possible (Anschuetz et al. 2001, 161; Giddens 1984), a form of “structure” derived by multiple cumulative past actions that condition the decisions of generations to come (Arkush 2011, 12). In this way, built landscapes reinforce the path dependency in local history and memories of social phenomena (Tilley 1994, 30). Some postprocessual critiques have also enriched current archaeological studies on settlement patterns. Although not all of these studies have engaged in the symbolic and more experiential approaches, this kind of study forms an important part of the archaeological understanding of ancient built landscapes. For instance, siteless survey (Dunnell 1992; Dunnell and Dancey 1983; Peterson and Drennan 2005) is among the most important methodological developments related to the studies of regions. The use of archaeological sites as bounded units of observation and analysis has been considered a limitation to the consideration of landscapes as spaces where people’s movements are fluid and whose activities do not always leave discrete and evident traces such as buildings or high-density clusters of artifacts. Another example of alternative perspectives about landscapes is historical ecology, a framework focused on the interaction between people and environment in historically specific contexts, highlighting human agency, the long-term effects of human actions on the environment, and the need for the collaboration of specialists from multiple disciplines to understand and explain how landscapes are constructed through time (Crumley 2017; Hayashida 2005; Meyer and Crumley 2011). Despite its obviousness, the common ground of all landscapes studies that is important to emphasize is the presence of people and the effects of their actions on the land, whether we are focused on natural resources, natural features, monuments, or dwellings. The recent renewed interest on demography (Bouquet-Appel and Bar-Yosef 2008; Drennan et al. 2015) is very relevant because it allows archaeologists to understand how changes in population might or might not relate to cultural processes that modified landscapes. One way to study the relation between people and landscapes has been the economic dimensions of these interactions. From this perspective, the study of landscapes is related to the use, appropriation, and modification of land and its resources by human communities through time (Metheny 1996). People’s investment (labor, resources, and social relationships) materializes and accumulates in the form of infrastructure (buildings, agricultural facilities, public spaces, fortifications, etc.), which can be transmitted, inherited, disputed, enhanced, or destroyed over time. Landscapes, then, become a critical resource for the negotiation of power relations in human societies as well as a way for 6

H U G O C . I K EH A R A -T S U K AYA M A A N D J UA N CA R L O S VA RG A S R U I Z

the archaeologist to approximate to the goals behind each construction (Earle and Doyel 2008). Landscapes are also physical manifestations of power relations within society. Some places may symbolize the power of specific groups, especially those territorial referents that are involved in the construction, reinforcing, and recreation of social identities. The strategic modification of landscapes can be a medium through which to communicate the importance, influence, strength, and capabilities of some groups to the rest of the society, including enemies, or can modify or reinforce the way in which groups are perceived and conceptualized by others (Branton 2009). These contrasting approaches are not mutually exclusive; collectively, they provide a more complete understanding of the multiple dimensions in which landscapes evolve together with the people on them (Anschuetz et al. 2001; Fisher and Thurston 1999). In this volume we are interested in how landscapes have been appropriated and modified by communities at war. While there is a strong emphasis on the built/constructed aspect, these landscapes were shaped by perceptions of fear and threat, which were influential in the (re)definition of social boundaries and communities’ identities. The cases in this volume permit comparison of regions with contrasting ecologies and topographies, of communities with different historical trajectories and at different socioeconomic situations, and, because the contributors were trained in different archaeological traditions, of different ways in which space and landscapes are studied. LANDSCAPES OF WARFARE

The origins of war, as well as the ultimate and proximate factors that spark violence, have been extensively treated in multiple publications (Allen and Arkush 2006; Arkush 2011; Armit 2011; Chapman 1999; Guilaine and Zammit 2005; Keeley 1996; Kelly 2000, 2005; LeBlanc 2006; Thorpe 2003). In general, war is differentiated from other kinds of violence, such as domestic violence or personal revenge, because it has been defined to signify the exchange of violence between social groups (Kelly 2000; Thorpe 2003). This broad definition of warfare includes a wide range of actions, from small-scale raids of tribal societies to the highly organized, large-scale, and highly destructive encounters of modern armies (Keeley 1996). Abundant historical, ethnographic, and archaeological evidence has demonstrated that war is more complex than it was initially considered to be in anthropological models (Carneiro 1970, 1998; Wilson 1987). Understanding warfare requires acceptance of the fact that the exchange of violence between L A N D S CA P ES O F P E O P LE AT WA R

7

people is intimately related to other aspects of life. Material conditions (such as resources and environment), local and regional politics, social structure and culture—the most common cited causal factors—are not mutually exclusive, but operate together and influence each other, not always in the same way, in each moment of increased conflict (Arkush 2011, 7). In this complex matrix we cannot underestimate the agency of individuals, their personal histories, feelings, perceptions, interests, and goals, which, in certain circumstances, can change history (Flannery 1999). The combination of these factors in a region could determine the ways in which people build, appropriate, and transform their landscapes. Variation in warfare can be explained by the political systems and aims of the groups in conflict (Arkush 2011). Julie Solometo (2006) categorizes the observed variability into six interrelated dimensions: social distance, social scale, tactics, goals, frequency and predictability of engagements, and duration of war. Social distance affects how destructive war can be; it is expected, for instance, that related communities do not combat until the extermination of the other. The size of warring parties may affect the scale of investment of defending populations: to face large armies, people might build massive defenses such as ramparts and ditches, for instance. The tactics and the technology used in each confrontation are related to the reasons and goals for which wars are waged: territorial expansion, slave raiding, resources control, warlords’ competition, and so forth. These elements influence the degree of violence incurred to enemies, how frequent and predictable attacks are, and how people prepare to defend themselves. If attacks are rare and predictable, people may not need strong protections; however, if attacks are frequent and unpredictable, communities might choose to concentrate within fortified settlements. Finally, the duration of violent interaction between groups may be shaped by several other factors, from the impetus of war leaders to live in constant war (benefiting from it, Carneiro 1998) to the capacity of certain polities for supporting longterm investment in the military. Because defensive strategies depend on how a threat is perceived, analyzing how landscapes were fortified allows us to reconstruct how war was waged in specific historical moments. This approach has been used by several authors of the present volume, some more explicitly than others (see Ikehara-Tsukayama, chapter 11 in this volume). Because the scale of the fighting party could be a strong factor in the success in combat, wars encourage the formation of political factions and alliances (Redmond 1994). These groups can unify groups to face the menace of a large enemy (Ikehara 2016), but they can be also instrumental in breaking the power parity between competing polities in a region (Allen 2008; Arkush 8

H U G O C . I K EH A R A -T S U K AYA M A A N D J UA N CA R L O S VA RG A S R U I Z

2011). People could have multiple and changing allegiances, which is reflected in the fact that a single region’s communities could be involved simultaneously in multiple, nested, and overlapped alliances or coalitions. We cannot study these macrocommunities with a focus solely on sites; we require a multiscalar approach to identify the scale and shapes of these macrocommunities (Arkush 2011). Built Landscapes

Despite that participation in war is a group activity, there are different impetuses for social cooperating based on whether engagement in warfare is defensive or offensive. Paul Roscoe (2013) argues that when facing a threat, people are more willing to cooperate in defensive strategies than in the organization of attacks. Organization of attacks requires other more powerful motivations or a centralized command, the latter more likely to occur within hierarchical complex societies. It is not surprising, then, that the most prominent and identifiable material evidence of warfare is defensive infrastructure such as fortifications (Arkush and Stanish 2005; Keeley et al. 2007). Multiple interests may converge in the location and design of fortifications. On the one hand, because we can count fortifications among the most expensive communal projects—in terms of resources, labor, and time— people tend not to invest more than needed to protect themselves (Arkush and Stanish 2005). Therefore, the scale of fortifications is a proportional representation of the degree and kind of threat perceived by the builders. These structures, however, not only provide obvious defensive advantages to their occupants, but can be a form of monumentality that conveys signs of community identity, power, and wealth (Arkush and Ikehara 2019; Armit 2007; Lock 2011; O’Driscoll 2017; Trigger 1990). Then, changes in fortification patterns can be used to track transformation in how warfare was conducted and how warfare and power were related over time. Fortifications, moreover, are likely to be built only once conflict increases in scale, intensity, or frequency. The simplest way to defend a community— clustering people in large settlements—facilitates the rapid organization of defensive parties; also, all other things being equal, a much larger fighting party is the most obvious advantage in a confrontation. Moreover, this unbalance between attacking and defending parties was thought to be enough to discourage the attacks. In part because of this defensive strategy, dense towns are formed and separated by relatively unoccupied zones known as buffer zones or no-man’s-lands (Wilcox and Haas 1994). Building up defenses L A N D S CA P ES O F P E O P LE AT WA R

9

usually means adding fortifications to already clustered populations. The simplest ones are ditches and palisades, while more complex ones involve nested and overlapping defenses (ramparts, towers, fortresses, etc.), including the formation of regional defensive systems. Additionally, because the organization of defenses are usually aligned with existing sociopolitical formations, the analysis of how people defend themselves can offer insight into how people organize locally and regionally (Arkush 2011; Arkush and Ikehara 2019). The scale of fortifications is, in most cases, directly related to the size of the labor pool. Fortifications reflect the capacity of social institutions to manage and coordinate larger groups of people. Another important element in the constitution of landscapes during war is battlefields. An substantial number of encounters may occur around fortresses and settlements, but sometimes these combats could occur in battlefields, in which material evidence is less conspicuous, and it is difficult to include them in reconstructions of past warfare landscapes. Written records are an important means of identifying these locations; archaeological data—including the spatial distribution of combat implements (weapons, armors, etc.), trenches, and injured human remains—also afford ways of identifying these important places. However, the mere existence of battlefields is a subject of debate. The battlefield, as well as the idea of armies of professional soldiers, must be understood as the by-product of certain cultural expectations and rules of combat and purposes of wars (Carman 1999). In contrast with the sporadic nature of combats, defensive infrastructure, once built, becomes a permanent element in landscapes. This observation is especially relevant for those facilities built with very durable materials. That archaeological observation can be made of these facilities, in the present, is a testament to this permanence. These modified spaces restrict the movement of people and their descendants, segregating and stratifying communities, conditioning their options, and affecting their daily lives. The Collas from the Peruvian Titicaca basin (500–950 bp) offer a good example of lifestyles constrained by war infrastructure. It has been noted that the aggregation of Colla communities in pukaras (large, fortified settlements) forced the Colla to adopt risky economic strategies (Langlie and Arkush 2016). Fortifications provided such an advantage that conquering neighbors was very difficult, and regional political consolidation was never achieved by local Colla lords. Fortifications perpetuated conflict between neighbors until the Inca conquest of the region (Arkush 2011). Eventually, the Inca Empire forced the Colla to move to the valley bottom, abandoning their ancestral towns (Arkush 2011; Stanish 1997), maybe as a strategy to reduce their independent ethos and bellicosity. 10

H U G O C . I K EH A R A -T S U K AYA M A A N D J UA N CA R L O S VA RG A S R U I Z

Roads, pathways, and causeways are forms of infrastructure that can have serious impacts in warfare, because they can facilitate communication and the movement of fighting forces. Large empires such as the Inca expanded their road system in each conquest, effectively mobilizing armies and supplies (Hyslop 1984). In smaller-scale societies, causeways were both used to control distant subjugated settlements through the mobilization of military forces (Spencer and Redmond 1998) and for the defense of distant towns against the action of enemies. Checkpoints on roads could be useful for guarding territories and for providing early warnings in case of attacks or invasions. In the Venezuelan lowlands, the Caquetío built large and monumental causeways that radically transformed the natural topography, creating an interconnected landscape of fortified towns, and integrating large polities that controlled hundreds of square kilometers (Spencer 1994). In chapter 7 in this volume, Earley-Spadoni introduces and discusses the role of communication routes in shaping landscapes of the ancient Near East. Places

Landscape transformations during wartimes also involves the articulation of new meanings. Some places can be considered dangerous because of easy exposure to enemies’ attacks. These empty areas may be results of the formation of buffer zones or no-man’s-lands (see chapters 4 and 10 in this volume the chapter by Williams and Vargas Ruiz), zones that remain “empty, underutilized, or unutilized and fallow” (LeBlanc 2006, 445) because hostilities between two or more groups are concentrated in these locations. These unutilized areas can be reintegrated later into production systems if the region is pacified (Le Blanc 2006). However, some places remain dangerous because they were perceived as enchanted, haunted, or possessed by invisible forces, spirits, and beings allied with current or former enemies. Production of cultural meaning and symbols associated with specific places can be viewed as an alternative arena for the creation and manipulation of ideological power in the context of war. Religious and war practices converge in certain places, such as the luakini (war temples) of Hawaii (Kolb and Dixon 2002) or the ritual structure inside the fortress of Chankillo in coastal Peru (Ghezzi 2006). As Elizabeth Arkush and Charles Stanish (2005) remind us, war and ritual are not exclusive: people carried out rituals seeking to be favored in combats, to express gratitude for victories, to acquire enemies’ power, and so on. Moreover, war can be read as a confrontation between nonhuman beings or carried out under nonhuman sponsorship (Nielsen 2009). L A N D S CA P ES O F P E O P LE AT WA R

11

The permanence of landscape elements makes them strategic symbolic resources for building and consolidating authority and power, and people competed for them (Snead 2009). The powerful meanings attributed to some places can be reclaimed later, even after the long abandonment of these places. Memories of past confrontations and successes can be instrumental to the legitimization of political discourses and identities (Rowlands 1993). For instance, in chapter 6 in this volume, Kim and Quick describe the reoccupation of Co Loa’s fortified capital by later royal dynasties as a way to claim connections to an autochthonous powerful past. War memorials are a different kind of symbol, dedicated not only to mourning warriors and soldier but also to commemorating victors and remembering victims of confrontations. War memorials constitute the historical memory of a community inscribed in landscapes; they embody a force that calls to mind social relationships and that makes visible the richness of warfare and history. These places utilize the memory of the dead for political, social, and moral motivations, and they legitimize the act of war through generations (Clarke 2010). Unexpected Consequences of War

Many decisions made during periods of warfare have had long-lasting and unexpected consequences for subsequent generations. Some places (ruined towns, battlefields, etc.) can be permanently abandoned because they became places for remembrance of suffering and death or because surviving populations resettled elsewhere in the aftermath of war. Even after war ends, these areas may continue to be considered dangerous places thanks to the memory of threat preserved in stories and myths. The buffer zones or no-man’s-lands created between competing polities often became areas where human activities were reduced, and these became optimal locations for the recovery of wild species, especially those intensively exploited by people, such as game species. This has been observed in the Korean Peninsula’s Demilitarized Zone (Brady 2012; Kim 1999), have been interpreted from the Lewis and Clark accounts of nineteenth-century western North America (Martin and Szuter 1999), and may have been common in preindustrial societies at war (see Vargas Ruiz, chapter 10 in this volume). In the process of the (re)appropriation of landscapes, walls have a special importance. While the abandonment of lands and the aggregation of population can put boundaries on local defensive communities, walls provide tangible evidence of social boundaries and differentiation, and community power and solidarity, as well as segmentation that persists even during more 12

H U G O C . I K EH A R A -T S U K AYA M A A N D J UA N CA R L O S VA RG A S R U I Z

peaceful times (Arkush 2014; Arkush and Ikehara 2019; Arkush and Stanish 2005; Ikehara and Arkush 2018; Lock 2011). Moreover, walls may form part of the physical structure that shapes social processes. As mentioned by EarleySpadoni (chapter 7 in this volume), defensive concerns were behind the recurrence of occupation in the same locations of the landscape and the construction of perimetric walls around settlements, both leading to the development of tells in the Near East in the past. Linguistics provides useful insights into the normalization of fortifications as part of settlements and landscapes. Many words for “city” or “town” in modern European languages share similar etymology concerned with defense: Proto-Germanic burgz (fortification, stronghold, or fortified city) and Proto-Slavic gordъ (enclosure, fortification, or castle), which derived from Proto-Indo-European bʰerǵʰ (high/lofty, hill/mountain) and gʰerdʰ- (to enclose),1 respectively. The defensive attribute of settlements described in these words may have emerged during times when warfare was so pervasive that fortifications became a basic element in any town or city. The current usages of these words do not necessarily recall the idea of defensibility, as its original usage fixed, because they have been assigned to places that are not or may have never been fortified in their past. Our current definition of cities and town do not include the idea of protection against the attack of enemies. The process of association between defenses and towns should be related to the groups speaking these proto-Indo-European languages. Recent advances in prehistoric population genomics support the hypothesis that they were pastoral groups from the Eurasian steppes (Allentoft et al. 2015; Anthony 2007; de  Barros Damgaard et al. 2018). In chapter 3 in this volume, Chechushkov explores the historical context of these proto-Indo-European-speaking communities of the Sintashta-Petrovka culture, while Earley-Spadoni examines the use of “city walls” to refer to the Near East city of Uruk in the Epic of Gilgamesh. Additionally, Williams, in chapter 4, describes a similar use of the suffix of “wall” as part of the script for “city” during Early Shang times in ancient China. OVERVIEW OF THIS VOLUME

The chapters within this volume have been organized into two sections: Old World and New World. Despite the problematic assumptions associated 1. See American Heritage Dictionary Info-European Roots Appendix (https://www .ahdictionary.com/word/indoeurop.html) and the Proto-Indo-European Root Extension (https://rex.iling.spb.ru/). L A N D S CA P ES O F P E O P LE AT WA R

13

with these terms, we have separated the cases from the Americas because they correspond to social trajectories that developed independently from the rest of the world. In the Old World, there is always the possibility that innovations arrived at specific regions by cultural transmission, exchange, and/or migrations (see discussions in Borzunov, Kim and Quick, chapters 2 and 6 in this volume, respectively). However, similarities in sociopolitical forms, economic institutions, and technology related to warfare between the Old and New Worlds must be considered independent developments, and hence independent subjects of inquiry. Our purpose, however, is both to compare cases or sociopolitical trajectories, and to understand the historical and ecological contexts in which these war landscapes were created, transformed, and made legacy for next generations. For this reason, we invited colleagues specializing in different regions of the world in order to include a variety of cases to help illustrate the many ways in which built landscapes were used to face conflict and violence. Previously published work on this subject has focused, in some cases more explicitly than others, on fortified landscapes (Arkush 2011; EarleySpadoni 2015; Hill and Wileman 2002; Kim 2013; Kolb and Dixon 2002). After all, the study of fortifications has been, together with human remains, the most studied archaeological evidence for warfare (Arkush and Stanish 2005; Arkush and Tung 2013; Guilaine and Zammit 2005; Keeley et al. 2007; Vencl 1984; Wilcox and Haas 1994). This volume was created with the clear intention of presenting and discussing cases on how people at war has modified their landscapes in different historical and geographical contexts (figure 1.1 and figure 1.2). We invite the reader to compare the presented cases in several ways: y

y

y

14

Old World versus New World developments. Considering that Indigenous societies from the Americas developed almost independently of the Old World since the last glaciation, how do their ways of warfare and their landscapes compare to those of the rest of the world? Different environmental and topographical contexts. How have the characteristics of the terrain and ecology affected how warfare has been waged, and the kinds of defensive strategies favored by communities? How can optimal defensive locations explain reoccupation of the same places (see Scholtus, chapter 8 in this volume)? Degrees of sociopolitical complexity. This volume includes cases of simple and relatively small-scale societies (Borzunov), chiefly/ranked societies (Williams, Chechushkov, Scholtus, Vargas Ruiz, Nichols, Ikehara-Tsukayama

H U G O C . I K EH A R A -T S U K AYA M A A N D J UA N CA R L O S VA RG A S R U I Z

Figure 1.1. Location of the cases presented by authors in this volume: (1) Borzunov; (2) Chechushkov; (3) Williams; (4) Matsugi; (5) Kim and Quick; (6) Earley-Spadoni; (7) Scholtus; (8) Nichols; (9) Vargas Ruiz; (10) Ikehara-Tsukayama; (11) Kohut.

Figure 1.2. Comparison of the time depth of the cases presented by authors in this volume: (1) Borzunov; (2) Chechushkov; (3) Williams; (4) Matsugi; (5) Kim and Quick; (6) EarleySpadoni; (7) Scholtus; (8) Nichols; (9) Vargas Ruiz; (10) Ikehara-Tsukayama; (11) Kohut.

y

and Kohut), and states and empires (Matsugi, Kim and Quick, and EarleySpadoni). Different approaches to the study of landscapes, including differing graphic visualizations. For instance, Borzunov’s maps shows a clear influence from traditional cartography, while other contributors reveal the impact of Geographic Information System (GIS) tools in describing, modeling, and interpreting spatial data.

In the first chapter of this volume, chapter 2, Viktor A. Borzunov presents a unique case: very old (8000–4500 bp) fortified settlements built by nonagricultural groups living in very high latitudes, in the middle of the Taiga zone. Some explanatory models of the origin of war consider the adoption of agriculture and subsequent population increase as causes for competition for lands and catalyzers of violence; however, Borzunov suggests that despite farming life’s incomplete adoption by local populations of northwestern Russia by the end of the Neolithic period, competition for resources arose as the result of the influx of foreign groups. An interesting observation by Borzunov is the shift between a settlement pattern based on fortified hamlets (with several houses each) to a landscape in which the larger houses were specially fortified, occurring in the Neolithic-Chalcolithic transition. Did this shift signal the rise of war leaders during the Chalcolithic? This case has seldom been described outside of Russian academic literature and presents an interesting contrast to the cases in the other chapters of this book. In chapter 3, Igor V. Chechushkov questions the defensive nature of the Sintatshta-Petrovka settlements (4000–3700 bp), proposing that fully understanding these enclosed settlements means considering the multiple environmental challenges people have and had in the Eurasian steppes. Contrary to traditional interpretations that consider the Sintashta-Petrovka settlements forts or defensive villages, Chechushkov argues that enclosed compact settlements may have been a strategy to protect people and their herds from such environmental hazards as freezing winds and river floods. He makes intensive use of GIS tools, including specialized software to analyze wind patterns. Chapter 4, by James T. Williams, focuses on war landscapes formed between 5000 and 4000 bp in Central China, during the Longshan period. If warfare was widespread during this time, Williams inquires, why were only a fraction of the settlements properly fortified with walls, leaving the rest of the population exposed to attacks? By using GIS tools, he tests and disproves several (mostly ecological) hypotheses. He suggests that violence was intricately tied to the formation of political and elite identities in the region. Leaders 16

H U G O C . I K EH A R A -T S U K AYA M A A N D J UA N CA R L O S VA RG A S R U I Z

consolidated their power by providing defense against raids in this tumultuous period. In chapter 5, Takehiko Matsugi analyzes how the rise and fall of fortified landscapes related to the consolidation of political authority in Early Japan (2960–1350 bp). He argues that the popularization and later decline of fortified settlements correspond to a transition between communal identities to a more individualized one. During the Yayoi era, fortified villages were the materialization and symbolic representation of collective identity and power, in a political landscape of competing chiefdoms. The regional domination by a powerful chiefdom first, and a mature state later, was characterized by the consolidation of a political hierarchy. Rulers became the symbol of the whole community. The communal use of labor and resources was reallocated from village defenses to the monumental royal burials of the Kofun era. Comparing this activity to what was occurring in adjacent regions, Matsugi argues that this political process was spatially constrained. In chapter 6, Nam C. Kim and Russel Quick’s introduce a case from the tropical Co Loa polity of Southeast Asia. They relate the emergence of the Co Loa polity to a massive modification of the landscape. A large city, with Indigenous and foreign (Chinese) architectural features, was founded in the Red River Delta (Vietnam). Cosmological elements are also present in the constitution of the city: the elevated terrains resemble the domed shell of a turtle out of legend. In chapter 5, Matsugi argues that Chinese influence in Kofun Japan helped to consolidate power under a religious and legal system, reducing conflict; by contrast, Kim and Quick argue that Chinese influence was felt in northern Vietnam as the influx of refugees and the threat of invasion. These factors, combined with the pressure of local competitors, may have pushed local groups to create a heavily fortified landscape in the polity’s capital. Matsugi’s and Kim and Quick’s chapters are helpful contemporaneous examples of the ways in which Early China imperialism influenced local developments in its periphery and how landscapes can be reconceptualized during times of drastic sociopolitical and ideological transitions. Moving westward, in chapter 7 Tiffany Earley-Spadoni contributes two cases from the Near East: Bronze Age Syria (3950–3750 bp) and Iron Age Assyria (2750–2650 bp). Drawing from multiple historical and archaeological sources, she argues that warfare and communication routes were intimately related, both shaping how landscapes evolved through time. In Syria, landscapes were dominated by fortified city states; fortresses, forts, and towers were built as components of warning systems against attacks. Earley-Spadoni also examines Neo-Assyrian road systems designed to boost imperial expansion L A N D S CA P ES O F P E O P LE AT WA R

17

but permanently altering the landscape and how people interact with others through them. Chapter 8, the last Old World chapter, was written by Lizzie Scholtus, who focuses on fortifications occupied during three intervals: 2480–2351 (Hallstatt D2–La Tène A), 2100–1976 (La Tène D), and finally 1975–1550 (Roman Empire). Scholtus interprets the first two occupations, during the Celtic era, as the results of competition between polities (and their leaders) in a mineralrich zone but also as the loci of important communication routes. These fortifications, with their massive ramparts, seem to have displayed the identity and power of the community in the same way as the Yayoi fortified settlements described by Matsugi. The later Roman reoccupation illustrates a different phenomenon: the formation of frontier territories, where the fortresses of the conquered were converted to a symbol of Roman imperialism but also Roman fear of antagonistic neighboring ethnic groups. The New World section starts with the contribution of Kerry Nichols in chapter 9. He poses the question of what happens when a new weapons technology is introduced to a region, analyzing the impact of the introduction of bow-and-arrow technology in midwestern North American communities during the Late Woodland period (1550–1100 bp). These ancient groups did not form large compact settlements (in contrast with the case in the next chapter), or build fortifications, but instead formed clusters of mutually visible settlements. The threat of a new, deadlier weapon encouraged the cooperative formation of spatially larger social units. In chapter 10, Juan Vargas Ruiz introduces landscapes formed by war between lowland communities between 1000 and 400 bp, in a region with gentle terrain, the Llanos of Casanare of Colombia. These communities’ defensive strategy consisted of forming large compact settlements with no fortifications, separated by empty zones. This spatial distribution of the population had significant economic and ecological consequences for the region: rich areas for farming were underexploited, while agricultural facilities were concentrated around the villages so that they could be readily defended. In chapter 11, Hugo Ikehara-Tsukayama’s chapter focuses on how shifts in defensive systems between 2500 and 1500 bp in coastal Peru reflect changing threats derived from political transitions on local and regional scales. He argues that the presence of a larger and more powerful external enemy encouraged the formation of a defensive alliance first, and of a single polity later. In the 2500–2100 bp period, the Nepeña middle valley was strongly fortified with dozens of fortresses, forts, and walls. This landscape was the result of fear of attacks from other communities of the area but also from a stronger external 18

H U G O C . I K EH A R A -T S U K AYA M A A N D J UA N CA R L O S VA RG A S R U I Z

[128.104.46.206] Project MUSE (2024-03-01 20:28 GMT) UW-Madison Libraries

enemy, maybe the large coetaneous polity emerging in the lower Valley. In the following period (2100–1500) the population decreased dramatically and concentrated in two large hilltop fortified settlements while being exposed to the influence of an expanding cultural and political entity centered in the Virú valley. In our last chapter, chapter 12, Lauren Kohut examines the constitution of political alliances and the social production of landscapes in later periods (1000–500 bp), after the collapse of the Wari and Tiwanaku states. The fact that small and large settlements were both fortified (pukaras) indicates the severity of intergroup conflict during this epoch, which forced highland communities throughout the Andes to move to hilltops far from lands optimal for intensive agriculture. These alliances (like those examined by Nichols in chapter 9 in this volume) manifested via the formation of spatial clusters in which mutual visibility and proximity (for sending help) were given priority. In analysis parallel to that of Earley-Spadoni’s chapter, Kohut suggests that smoke and fire may have been used to pass information between outposts and settlements, rapidly preparing communities for incoming attacks. Even recognizing that this volume has left out important regions where it is known that war was historically important (Africa, Oceania, the American Southwest, the Amazon, and so on), the included chapters present cases from a wide range of geographical contexts, some of them seldom included in comparative volumes. The geographic and temporal scopes of the present volume permit comparison of how communities engaged in war have modified their landscapes in different historical circumstances. From the simple clustering of families in a town to the construction of complex defensive systems with fortress and outposts, landscapes have been marked with violence. These landscapes are inherited by following generations; war not only affects the people who fight them, but, through their permanence in landscapes, are capable of lasting effects even in times of peace. Realizing the magnitude of the transformations and how deeply they affect people’s lives—through the long-term perspective provided by archaeological analysis—may help us to understand how wars waged in a more recent past are still affecting people today. REFERENCES

Allen, Mark W. 2008. “Hillforts and the Cycling of Maori Chiefdoms: Do Good Fences Make Good Neighbors?” In Global Perspectives on the Collapse and Transformation of Complex Systems, edited by Jim A. Railey and Richard Martin Reycraft, 65–81. Albuquerque: Maxwell Museum of Anthropology.

L A N D S CA P ES O F P E O P LE AT WA R

19

Allen, Mark W., and Elizabeth Arkush. 2006. “Introduction: Archaeology and the Study of War.” In The Archaeology of Warfare. Prehistories of Raiding and Conquest, edited by Elizabeth Arkush and Mark W. Allen, 1–19. Gainesville: University Press of Florida. Allentoft, Morten E., Martin Sikora, Karl-Göran Sjögren, Simon Rasmussen, Morten Rasmussen, Jesper Stenderup, and Peter B. Damgaard. 2015. “Population Genomics of Bronze Age Eurasia.” Nature 522 (7555):167–172. Anschuetz, Kurt F., Richard H. Wilshusen, and Cherie L. Scheick. 2001. “An Archaeology of Landscapes: Perspectives and Directions.” Journal of Archaeological Research 9 (2): 157–211. Anthony, David W. 2007. The Horse, the Wheel, and Language: How Bronze-Age Riders from the Eurasian Steppes Shaped the Modern World. Princeton, NJ: Princeton University Press. Arkush, Elizabeth. 2011. Hillforts of the Ancient Andes: Colla Warfare, Society, and Landscape. Gainesville: University Press of Florida. Arkush, Elizabeth. 2014. “ ‘I against my brother’: Conflict and Confederation in the South-Central Andes in Late Prehistory.” In Embattled Bodies, Embattled Places: War in Pre-Columbian Mesoamerica and the Andes, edited by Andrew K. Scherer and John W. Verano, 199–226. Washington, DC: Dumbarton Oaks Research Library and Collection. Arkush, Elizabeth, and Hugo Ikehara. 2019. “Pucarani: Defensive Monumentality and Political Leadership in the Late Pre-Columbian Andes.” Journal of Anthropological Archaeology 53: 66–81. Arkush, Elizabeth, and Charles Stanish. 2005. “Interpreting Conflict in the Ancient Andes.” Current Anthropology 46 (1): 3–28. Arkush, Elizabeth, and Tiffiny A. Tung. 2013. “Patterns of War in the Andes from the Archaic to the Late Horizon: Insights from Settlement Patterns and Cranial Trauma.” Journal of Archaeological Research 21 (4): 307–369. Armit, Ian. 2007. “Hillforts at War: From Maiden Castle to Taniwaha Pā.” Proceedings of the Prehistoric Society 73: 25–38. Armit, Ian. 2011. “Violence and Society in the Deep Human Past.” British Journal of Criminology, Special thematic issue: “Evolutionary and Historical Perspective on Violence,” 51 (3): 499–517. Bauer, Brian S. 2000. El espacio sagrado de los Incas: El sistema de Ceques del Cuzco. Cuzco: Centro de Estudios Regionales Andinos “Bartolomé de las Casas” Blume, Hans-Peter, and Peter Leinweber. 2004. “Plaggen Soils: Landscape History, Properties, and Classification.” Journal of Plant Nutrition and Soil Science 167 (3): 319–327.

20

H U G O C . I K EH A R A -T S U K AYA M A A N D J UA N CA R L O S VA RG A S R U I Z

Bouquet-Appel, Jean-Pierre, and Ofer Bar-Yosef, eds. 2008. The Neolithic Demographic Transition and Its Consequences. New York: Springer. Bradley, Richard. 1998. The Significance of Monuments: On the Shaping of Human Experience in Neolithic and Bronze Age Europe. London: Routledge. Brady, Lisa M. 2012. “Korea’s Green Ribbon of Hope: History, Ecology, and Activism in the DMZ.” Solutions 3 (1): 94–98. Branton, Nicole. 2009. “Landscape Approaches in Historical Archaeology: The Archaeology of Places.” In International Handbook of Historical Archaeology, edited by Teresita Majewski and David Gaimster, 51–66. New York: Springer. Carman, John. 1999. “Beyond the Western Way of War: Ancient Battlefields in Comparative Perspective.” In Ancient Warfare: Archaeological Perspectives, edited by John Carman and Anthony Harding, 39–55. Stroud: Sutton Publishing. Carneiro, Robert L. 1970. “A Theory of the Origin of the State: Traditional Theories of State Origins Are considered and Rejected in Favor of a New Ecological Hypothesis.” Science 169 (3947): 733–738. Carneiro, Robert L. 1998. “What Happened at the Flashpoint? Conjectures on Chiefdom Formation at the Very Moment of Conception.” In Chiefdoms and Chieftaincy in the Americas, edited by Elsa M. Redmond, 18–42. Gainesville: University Press of Florida. Chapman, John 1999. “The Origins of Warfare in the Prehistory of Central and Eastern Europe.” In Ancient Warfare: Archaeological Perspectives, edited by John Carman and Anthony Harding, 101–142. Stroud: Sutton Publishing. Clarke, Joseph. 2010. “The Sacred Names of the Nation’s Dead: War and Remembrance in Revolutionary France.” In Memory, Mourning, Landscape, edited by Elizabeth Anderson, Avril Maddrell, Kate McLoughlin, and Alana M. Vincent, 21–41. Leiden, Netherlands: Brill. Crumley, Carole L. 2007 Historical Ecology: Integrated Thinking at Multiple Temporal and Spatial Scales. In The World System and the Earth System: Global Socio-environmental Change and Sustainability since the Neolithic. Edited by Alf Hornborg and Carole Crumley, 15–28. Walnut Creek CA: Left Coast Press. Crumley, Carole L. 2017. “Historical Ecology and the Study of Landscape.” Landscape Research 42 (Sup. 1): S65-S73. de Barros Damgaard, Peter, et al. 2018. “The First Horse Herders and the Impact of Early Bronze Age Steppe Expansions into Asia.” Science 360 (6396): eaar7711. Dobres, Marcia-Anne, and John E. Robb. 2000. “Agency in Archaeology: Paradigm or Platitude.” In Agency in Archaeology, edited by Marcia-Anne Dobres and John E. Robb, 3–17. London: Routledge.

L A N D S CA P ES O F P E O P LE AT WA R

21

Drennan, Robert D., C. Adam Berrey, and Christian E. Peterson. 2015. Regional Settlement Demography in Archaeology, Principles of Archaeology. New York: Elliot Werner Publications. Dunnell, Robert C. 1992. “The Notion Site.” In Space, Time, and Archaeological Landscapes, edited by Jacqueline Rossignol and LuAnn Wandsnider, 21–41. New York: Plenum. Dunnell, Robert C., and William S. Dancey. 1983. “The Siteless Survey: A Regional Scale Data Collection Strategy.” Advances in Archaeological Method and Theory 6: 267–287. Earle, Timothy, and David E. Doyel. 2008. “The Engineered Landscapes of Irrigation.” In Economies and the Transformation of Landscape, edited by Lisa C. Cliggett and Christopher A. Pool, 19–46. New York: Altamira Press. Earley-Spadoni, Tiffany. 2015. “Landscapes of Warfare: Intervisibility Analysis of Early Iron and Urartian Fire Beacon Stations (Armenia).” Journal of Archaeological Science: Reports 3: 22–30. Fisher, Christopher T., and Tina L. Thurston. 1999. “Special Section Dynamic Landscapes and Socio-political Process: The Topography of Anthropogenic Environments in Global Perspective.” Antiquity 73 (281): 630. Flannery, Kent V. 1999. “Process and Agency in Early State Formation.” Cambridge Archaeological Journal 9 (1): 3–21. Ghezzi, Ivan. 2006. “Religious Warfare at Chankillo.” In Andean Archaeology III, edited by William H. Isbell and Helaine Silverman, 67–84. New York: Springer. Giddens, Anthony. 1984. The Constitution of the Society. Outline of the Theory of Structuration. Berkeley: University of California Press. Guilaine, Jean, and Jean Zammit. 2005. The Origins of War. Malden, MA: Blackwell Publishing. Hayashida, Frances. 2005. “Archaeology, Ecological History, and Conservation.” Annual Review of Anthropology 34: 43–65. Hill, Paul, and Julie Wileman. 2002. Landscapes of War: The Archaeology of Aggression and Defence. Stroud: Tempus Publishing. Hodder, Ian, and Scott Hutson. 2003. Reading the Past. 3rd ed. Cambridge: Cambridge University Press. Hyslop, John. 1984. The Inca Road System. Orlando, FL: Academic Press. Ikehara, Hugo C. 2016. “The Final Formative Period in the North Coast of Peru: Cooperation during Violent Times.” World Archaeology 48 (1): 70–86. Ikehara, Hugo C., and Elizabeth N. Arkush. 2018. “Pucarani: Building a Pukara in the Peruvian Lake Titicaca Basin (ad 1400–1490).” Ñawpa Pacha 38 (2): 157–181.

22

H U G O C . I K EH A R A -T S U K AYA M A A N D J UA N CA R L O S VA RG A S R U I Z

Kaulicke, Peter, Ryujiro Kondo, Tetsuya Kusuda, and Julinho Zapata. 2004. “Agua, ancestros y arqueología del paisaje.” Boletín de Arqueología PUCP No 7: 27–56. Keeley, Lawrence H. 1996. War before Civilization. New York: Oxford University Press. Keeley, Lawrence H., Marisa Fontana, and Russell S. Quick. 2007. “Baffles and Bastions: The Universal Features of Fortifications.” Journal of Archaeological Research 15 (1): 55–95. Kelly, Raymond C. 2000. Warless Societies and the Origin of War. Ann Arbor: University of Michigan Press. Kelly, Raymond C. 2005. “The Evolution of Lethal Intergroup Violence.” Proceedings of the National Academy of Sciences 102 (43): 15294–15298. Kim, Ke Chung. 1999. “Preserving Biodiversity in Korea’s Demilitarized Zone.” Science 278 (5336): 242–243. Kim, Nam C. 2013. “Cultural Landscapes of War and Political Regeneration.” Asian Perspectives 52 (2): 244–267. Knapp, A. Bernard, and Wendy Ashmore. 1999. “Archaeological Landscapes: Constructed, Conceptualized, Ideational.” In Archaeologies of Landscape: Contemporary Perspectives, edited by Wendy Ashmore and A. Bernard Knapp, 1–30. Malden, MA: Blackwell Publishers. Kolb, Michael J., and Boyd Dixon. 2002. “Landscapes of War: Rules and Conventions of Conflict in Ancient Hawai’i (And Elsewhere).” American Antiquity 67 (3): 514–534. Kosiba, Steve. 2015. “Tracing the Inca Past: Ritual Movement and Social Memory in the Inca Imperial Capital.” In Perspectives on the Inca, edited by Monica Barnes, Inés de Castro, Javier Flores Espinoza, Doris Kurella, and Karoline Noack, 178–205. Stuttgart, Germany: Linden-Museum, Sonderband/Tribus. Langlie, BrieAnna S., and Elizabeth N. Arkush. 2016. “Managing Mayhem: Conflict, Environment, and Subsistence in the Andean Late Intermediate Period, Puno, Peru.” In The Archaeology of Warfare and Food: Food Insecurity in Prehistory, edited by Amber VanDerwarker and Wilson Greg D., 259–289. New York: Springer. LeBlanc, Steven A. 2006. “Warfare and the Development of Social Complexity.” In The Archaeology of Warfare: Prehistories of Raiding and Conquest, edited by Elizabeth Arkush and Mark W. Allen, 437–468. Gainesville: University Press of Florida. Lock, Gary. 2011. “Hillforts, Emotional Metaphors, and the Good Life: A Response to Armit.” Proceedings of the Prehistoric Society 77: 355–362. Martin, Paul S., and Christine R. Szuter. 1999. “War Zones and Game Sinks in Lewis and Clark’s West.” Conservation Biology 13 (1): 36–45. Metheny, Karen. 1996. “Landscape Archaeology.” In The Oxford Companion to Archaeology, edited by Brian M. Fagan, 384–385. New York: Oxford University Press.

L A N D S CA P ES O F P E O P LE AT WA R

23

Meyer, William J., and Carole L. Crumley. 2011. “Historical Ecology: Using What Works to Cross the Divide.” In Atlantic Europe in the First Millennium BC: Crossing the Divide, edited by T. Moore and L. Armada, 109–134. Oxford: Oxford University Press. Moore, Jerry D. 2005. Cultural Landscapes in the Ancient Andes: Archaeologies of Place. Gainesville: University Press of Florida. Nielsen, Axel. E. 2009. “Ancestors at War: Meaningful Conflict and Social Process in the South Andes.” In Warfare in Cultural Context Practice, Agency, and the Archaeology of Violence, edited by Axel. E. Nielsen and William H. Walker, 218–243. Tucson: University of Arizona Press. O’Driscoll, James. 2017. “Hillforts in Prehistoric Ireland: A Costly Display of Power?” World Archaeology 49 (4): 506–525. Parsons, Jeffrey R. 1972. “Archaeological Settlement Patterns.” Annual Review of Anthropology 1: 127–150. Peterson, Christian E., and Robert D. Drennan. 2005. “Communities, Settlements, Sites, and Surveys: Regional-Scale Analysis of Prehistoric Human Interaction.” American Antiquity 70 (1): 5–30. Redmond, Elsa M. 1994. “External Warfare and the Internal Politics of Northern South American Tribes and Chiefdoms.” In Factional Competition and Political Development in the New World, edited by Elizabeth M. Brumfiel and John W. Fox, 44–54. Cambridge: Cambridge University Press. Reid, Nicholas, Patrick D. Nunn, and Margaret Sharpe. 2014. “Indigenous Australian Stories and Sea-Level Change.” In Proceedings of the 18th Conference of the Foundation for Endangered Languages, edited by Patrick Heinrich and Nicholas Ostler, 82–87. Berkshire, England: Foundation for Endangered Languages. Roberts, Patrick, Chris Hunt, Manuel Arroyo-Kalin, Damian Evans, and Nicole Boivin. 2017. “The Deep Human Prehistory of Global Tropical Forests and Its Relevance for Modern Conservation.” Nature Plants 3 (8): 17093. Roscoe, Paul. 2013. “War, Collective Action, and the ‘Evolution’ of Human Polities.” In Cooperation and Collective Action: Archaeological Perspectives, edited by David M. Carballo, 57–82. Boulder: University Press of Colorado. Rowlands, Michael. 1993. “The Role of Memory in the Transmission of Culture.” World Archaeology 25 (2): 141–151. Santillana, Julián I. 2012. Paisaje sagrado e ideología inca, estudios andinos. Lima: Fondo Editorial, Pontificia Universidad Católica del Perú. Shanks, Michael. 2008. “Post-processual Archaeology and After.” In Handbook of Archaeological Theories, edited by Christopher Chippindale, R. Alexander Bentley, and Herbert D.G. Marschner, 133–144. Lanham, MD.: Altamira Press.

24

H U G O C . I K EH A R A -T S U K AYA M A A N D J UA N CA R L O S VA RG A S R U I Z

Snead, James. 2009. “War and Place: Landscapes of Conflict and Destruction in Prehistory.” In Bastions and Barbed Wire: Studies in the Archaeology of Conflict, edited by Tony Pollard and Iain Banks, 138–155. Leiden, Netherlands: Brill. Solometo, Julie. 2006. “The Dimensions of War: Conflict and Culture Change in Central Arizona.” In The Archaeology of Warfare: Prehistories of Raiding and Conquest, edited by Elizabeth Arkush and Mark W. Allen, 23–65. Gainesville: University Press of Florida. Spencer, Charles S. 1994. “Factional Ascendance, Dimensions of Leadership, and the Development of Centralized Authority.” In Factional Competition and Political Development in the New World, edited by Elizabeth M. Brumfiel and John W. Fox, 31–43. Cambridge: Cambridge University Press. Spencer, Charles S., and Elsa M. Redmond. 1998. “Prehispanic Causeways and Regional Politics in the Llanos de Barinas, Venezuela.” Latin American Antiquity 9 (2): 95–110. Stanish, Charles. 1997. “Nonmarket Imperialism in the Prehispanic Americas: The Inka Occupation of the Titicaca Basin.” Latin American Antiquity 8 (3): 195–216. Taylor, Gerald. 1987. Ritos y tradiciones de Huarochirí. Lima: Instituto de Estudios Peruanos and Instituto Francés de Estudios Andinos. Tello, Julio C., and Próspero Miranda. 1923. “Wallallo: Ceremonias gentílicas realizadas en la región cisandina del Perú Central (Distrito arqueológico de Casta).” Inca 1 (2): 475–549. Thorpe, I. J. Nick. 2003. “Anthropology, Archaeology, and the Origin of Warfare.” World Archaeology 35 (1): 145–165. Tilley, Christopher. 1994. A Phenomenology of Landscape. Places, Paths and Monuments. Oxford: Berg Publishers. Trigger, Bruce G. 1990. “Monumental Architecture: A Thermodynamic Explanation of Symbolic Behaviour.” World Archaeology 22 (2): 119–132. Vencl, Slavomil. 1984. “War and Warfare in Archaeology.” Journal of Anthropological Archaeology 3 (2): 116–132. Whitridge, Peter. 2004. “Landscapes, Houses, Bodies, Things: Place and the Archaeology of Inuit Imaginaries.” Journal of Archaeological Method and Theory 11 (2): 213–250. Wilcox, David R., and Jonathan Haas. 1994. “The Scream of the Butterfly: Competition and Conflict in the Prehistoric Southwest.” In Themes in Southwest Prehistory, edited by George J. Gumerman, 211–238. Santa Fe: School of American Research. Wilson, David J. 1987. “Reconstructing Patterns of Early Warfare in the Lower Santa Valley: New Data on the Role of Conflict in the Origins of Complex North Coast Society.” In The Origins and Development of the Andean State, edited by Jonathan Haas, Shelia Pozorski, and Thomas Pozorski, 56–69. Cambridge: Cambridge University Press.

L A N D S CA P ES O F P E O P LE AT WA R

25

2 The Neolithic and Chalcolithic Fortified Settlements of the Northernmost Eurasian Taiga Viktor A. Borzunov Translated and adapted by Igor V. Chechushkov

https://doi.org/10.5876/9781646422111.c002

26

It is widely accepted in the Old World archaeology that the origins of fortified settlements and their prototypes—namely, settlements on elevated naturally protected locations, tower houses, and villages surrounded by walls and ditches—are related to the wide adoption of an economy of food production by the collectives of fishermen-hunter-gatherers, a process also known as the “Neolithic Revolution.” The earliest-known fortification systems were developed approximately 10000–8000 bp by early farmers and herders of the eastern Mediterranean (the sites of Jericho, Tell es-Sawwan, Haçilar II, Çatal Hüyük, etc.). Fortified sites gradually appeared in other regions of the Old World: around 7000–6000 bp in western Asia, and some parts of Europe and Central Asia; around circa 6300–5000 bp in the Nile, Indus, and Yellow River valleys, as well as in northeastern Iran and northern Afghanistan; around 5300–3500 bp the simplest forms of fortified settlements existed in most areas of western Europe, as well as in the steppes and forest-steppes of Eurasia. By the Late Bronze Age (ca. 3200–2800 bp) and Early Iron Age (ca. 2700–1700 bp), defensive structures were present almost everywhere on the northern Eurasian landscape. The initial spread of food production economies throughout Eurasia (ca. 8000–5000 bp) is associated with cultural diffusion, intensification of intergroup communication, including exchange. The direct migration of the Near-Eastern and East-Mediterranean

T able 2.1. Summary of cultural sequence for northernmost Eurasia Age/Period Name Stone Age

Bronze Age

Age, BP

Climate Phase

Climate Conditions

Boreal/Atlantic

Dry and cold

8000–5300

Atlantic

Warm and humid

5300–4300

Atlantic/Subboreal

Dry and cold

4300–3200

Subboreal

Dry and cold

Mesolithic

10000–8000

Neolithic Chalcolithic

farmers and herdsmen into adjacent areas and the need to build defensive structures around the agricultural settlements may have been caused by an increasing population pressure on the limited resources and the subsequent conflict over the communal lands and agricultural products. In the nonagricultural areas of northern Eurasia, however, the concept of fortified settlement was adopted and developed by communities of hunter-gatherers as early as circa 8000–4300 bp. This chapter aims to present the archaeological materials from northern Trans-Urals and northwest Siberia to the broad English-speaking audience for the first time, as well as to provide an explanatory model for the presence of fortifications outside the boundaries of Neolithic farming. This chapter focus on the period known as the Neolithic (table 2.1) by Russian archaeologists based on the appearance of polished stone tools and pottery in the archaeological record, while productive economy was not yet adopted. The following periods of Chalcolithic differs by the appearance of evidence for animal husbandry and limited metal production. This period corresponds to the Atlantic phase of the Holocene (9200/8000–5700/5000 bp) which, despite an episode of global cooling around 8200 bp was, broadly speaking, a very warm and humid period, in which the forest boundary was pushed northward, replacing the tundra zone with the forest tundra zone. In western Siberia, the water level of the major interior rivers was higher than today, a feature suggested by the location of the Neolithic and Chalcolithic sites on the high terraces next to what are currently hallow streams and marshy areas. The Atlantic climate moved the northern Taiga boundary toward today’s Arctic, allowing the Neolithic fishermen-hunter-gatherers to settle the far north of western Siberia. In turn, the Chalcolithic (ca. 5300–4300 bp) was a dry and cold period, corresponding to the transition from the Atlantic to the Subboreal climates, when the environmental conditions of the northern taiga deteriorated (Baker et al. 2017).

T H E N E O LI T H I C A N D C H A L CO LI T H I C F O RT I F I ED S E T T LEM EN T S

27

DATING THE EARLIEST FORTIFIED SETTLEMENTS OF NORTHERNMOST WESTERN EURASIA

The Neolithic sites are situated between 56º40´and 64º14´ of the northern latitude within the northern course of the Ob’ River and the forest zone of the Trans-Urals (figure 2.1). There are nine Neolithic fortified sites known in this area: Amnya 1, Kayukovo 2, Bolshaya Umytya 9, Bolshaya Umytya 57, Poludenka 1, Bystryi Kul’egan 66 and Ches-Tyi-Yag fortified settlements and Mikishkino 5, Ust’-Tara XXVIII fortified houses, and two Late Neolithic/ Chalcolithic sites: Imn’egan 2.1 and Nioh-urij 3.3, and so forth. The Stone Age attribution of the settlements is beyond doubt. At first, even the earliest settlements of Amnya 1 and Kayukovo 2 were dated by the end of the Neolithic according to the ceramic typology and few preliminary radiocarbon tests (Morozov and Stefanov 1993; Stefanov et al. 1999, 43–44; Ivas’ko 2002; Ivas’ko and Kardash 2002). More recent radiocarbon dates (Kosintsev et al. 2004), however, forces us to reevaluate that original assumption. According to a series of uncalibrated dates (table 2.2), the northern Neolithic settlements can be dated back to circa 11000–5300 bp. Removing the outliers, the refined range is circa 8000–5300 bp. Date calibrations and Bayesian models were produced in OxCal v.4.3 (Bronk Ramsey 2009) using the IntCal13 Northern Hemisphere atmospheric curve (Reimer et al. 2013).1 Dates were modeled in a sequence to estimate the chronological boundaries of the taiga Neolithic. The outlier model was also applied due to the observation that three calibrated dates from Amnya 1 and Kayukovo 2 appeared to lie far outside the rest of the dates (maybe the effects of old wood). The median points of the period boundaries are 7925 cal. bp and 6250 cal. bp (figure 2.2). This period belongs to the Atlantic climatic phase of the Holocene, while the earliest (Mesolithic) 14C dates correspond to the Boreal phase. The radiocarbon dates demonstrate that the earliest occupations can be dated by the Early Neolithic period in the regional chronological scale (ca. 8000–6550 bp): Amnya 1, Kayukovo 2, and Mikishkino 5. The relative chronological position of House 1 at Bolshaya Umytya 9 can be estimated by the presence of the Neolithic pottery of the Sumpanya type (Pogodin 2010) that can be dated to ca. 6850–6100 bp (Kosintsev et al. 2004). The settlements of Bystryi Kul’egan 66, Ches-Tyi-Yag, and Bolshaya Umytya 57 can be attributed by the Middle and Late Neolithic (ca. 6500–5300 1. Analysis of the radiocarbon dates conducted by Igor V. Chechushkov for the purposes of this chapter.

28

V I K T O R A . B O R Z U N OV

Figure 2.1. Distribution of the Neolithic and Chalcolithic fortified settlements and their prototypes in western Siberia. Neolithic: (1) Amnya I*; (2) Kayukovo 2*; (3) Bystryi Kul’egan (Bystryi Kulyogan) 66*; (4) Bolshaya Umytya 9*; (5) Bolshaya Umytya 57*; (6) Mikishkino 5*; (7) Ust’-Tara XXVIII (object 1)*; (8) Poludenka I*; (9) Ches-Tyi-Yag*. Chalcolithic: (10) Imn’egan (Imnyogan) 2.1; (11) Nioh-urij 3.3; (12) Nivagalskoye 20*; (13) Volvoncha I (lower horizon)*; (14) Teplyi Ruchey; (15) Kayukovo 1. Legend: (a) The Arctic Circle; (b) modern boundaries of environmental zones (I: tundra, II: forest-tundra, III: open woodlands, IV: taiga, V: broad-leaved forest, VI: forest-steppe, VII: steppe, VIII: mountain areas); (c) fortresses; (d) alleged fortress; (e) fortified settlement; (f ) fortified house; (g) island settlement. Note: asterisks mark the sites that were tested archaeologically.

T able 2.2. Radiocarbon measurement yielded by the Neolithic fortified sites Site

C Age, BP

Lab Code

Amnya 1

8760 ± 280

ЛЕ-4974а

Stefanov and Borzunov (2008)

Amnya 1

8630 ± 180

ЛЕ-4974б

Stefanov and Borzunov (2008)

14

Reference

Amnya 1

6900 ± 90

ЛЕ-4973

Stefanov and Borzunov (2008)

Kirip-Vis-Yugan 2, Building 4

6880 ± 50

ЛЕ-6582

Stefanov et al. (2005) Stefanov and Borzunov (2008, 110)

Mikishkino 5

7070 ± 95

СОАН-6230

Ivas’ko (2008)

Mikishkino 5

7060 ± 80

СОАН-6229

Ivas’ko (2008)

Mikishkino 5

6790 ± 110

СОАН-6228

Ivas’ko (2008)

Mikishkino 5

6680 ± 90

СОАН-5720

Ivas’ko (2008)

Mikishkino 5

6590 ± 90

СОАН-5719

Ivas’ko (2008)

Kayukovo 2

6810 ± 40

ЛЕ-6206

Ivas’ko (2008)

Kayukovo 2

6580 ± 35

ЛЕ-6207

Ivas’ko (2008)

Kayukovo 2

6795 ± 65

СОАН-4801

Ivas’ko (2008)

Kayukovo 2

8610 ± 55

СОАН-4800

Ivas’ko (2008)

Kayukovo 2

5495 ± 70

СОАН-4208

Ivas’ko (2008)

Kayukovo 2

5475 ± 85

СОАН-4207

Ivas’ko (2008)

Bystryi Kul’egan 66

6150 ± 210

ЛЕ-5680

Kosinskaya et al. (2002)

Bystryi Kul’egan 66

5930 ± 90

ЛЕ-5335

Kosinskaya et al. (2002)

Bystryi Kul’egan 66

5910 ± 130

ЛЕ-5336

Kosinskaya et al. (2002)

Bystryi Kul’egan 66

5780 ± 130

ЛЕ-5689

Kosinskaya et al. (2002)

Bystryi Kul’egan 66

5725 ± 70

ЛЕ-5337

Kosinskaya et al. (2002)

Bystryi Kul’egan 66

5560 ± 100

ЛЕ-5690

Kosinskaya et al. (2002)

Ches-Tyi-Yag

6455 ± 40

СОАН-2718

Kosintsev et al. (2004)

Ches-Tyi-Yag

6165 ± 35

СОАН-2717

Kosintsev et al. (2004)

Ches-Tyi-Yag

6150 ± 40

СОАН-2721

Kosintsev et al. (2004)

Ches-Tyi-Yag

6095 ± 30

СОАН-2720

Kosintsev et al. (2004)

Ches-Tyi-Yag

5785 ± 40

СОАН-2716

Kosintsev et al. (2004)

Ches-Tyi-Yag

5750 ± 60

СОАН-2719

Kosintsev et al. (2004)

Bolshaya Umytya 57

6030 ± 85

СОАН-7639

Pogodin and Mironov (2009)

Bolshaya Umytya 57

6030 ± 80

СОАН-7640

Pogodin and Mironov (2009)

Figure 2.2. The 14C Outlier Model: Boundaries of the “fortified” Neolithic. Credit: Igor V. Chechushkov.

bp), while Imn’egan 2.1 was used during the Late Neolithic and partially during the Chalcolithic. There are no Bronze and Iron Age materials on most of the sites, and there are only two cases where Chalcolithic dwellings destroyed and covered preceding Neolithic structures (Amnya 1, dwelling 1A and Bolshaya Umytya 57 [figure 2.3, I; figure 2.6, II]). In Bolshaya Umytya 9 and Bolshaya Umytya 57, Chalcolithic, Bronze and Iron Age sherds are scarce and have no spatial association with the Neolithic objects, though there are a cemetery and two Chalcolithic dwellings at Bolshaya Umytya 57. Finally, there are some Chalcolithic sherds at the settlement of Poludenka 1, but they were found outside of circular ditch that surrounds the Neolithic habitational area. For the rest of this chapter, I will present the general characterization of the Eurasian northernmost Neolithic fortified settlements, but an interested reader can find more detailed descriptions in other published papers (Borzunov 2016, 2018). NEOLITHIC FORTIFIED SITES AND THEIR PROTOT YPES IN THE TAIGA OF WESTERN SIBERIA (8000– 5300 bp )

The remains of the early fortified settlements are mostly well preserved, and they are very visible on today’s surface. This good visibility is linked to the slow sediment accumulation process resulting from the leaching regime of the sand terraces covered with coniferous forests. Additionally, the sites are in sparsely populated areas of the taiga, distant from modern economic infrastructure. More important, however, is the respectful attitude of the contemporaneous Indigenous populations to these archaeological remains, which are considered sacred. The combination of these factors produces archaeological sites that look relatively recent for a nonspecialist visitor. The Neolithic fortified sites, as well as their prototypes, ranged in area from 260 m2 to 900 m2, but they can be as large as 1,600–1,900 m2. The number of dwellings varies from one to three on small sites, to five to seven on larger sites. The settlements differ from the preceding Mesolithic sites by the appearance of the fortified parts and by complex architectural patterns. The variation of these Neolithic fortified sites can be classified in the following types: y y

32

Settlements on river capes with a single or doubled line of fortification (Amnya 1, Imn’egan 2.1, Poludenka 1) Settlements on river capes or crests encircled by fortifications (Bolshaya Umytya 9, Bystryi Kul’egan 66, Kayukovo 2)

V I K T O R A . B O R Z U N OV

y y y

Semirectangular three-part fortification on a cape-shaped protrusion (Bolshaya Umytya 57) Small fortified dwellings on capes (House 1 of Amnya 1; Mikishkino 5, House 7 of Imn’egan 2.1) Large fortified dwellings on terraces (Bystryi Kul’egan 66, Ust’-Tara XXVIII)

It is important to mention that unfortified villages have been identified outside the fortified areas in some settlements (Amnya 1, Poludenka 1, Bolshaya Umytya 9, and Bolshaya Umytya 57). The most sophisticated layouts characterize Amnya 1, Kayukovo 2 and Imn’egan 2.1. At Amnya 1 (figure 2.3), the large House 1 was initially built at the tip of the cape and fortified with a ditch on the land side of the cape. At some moment later, the first ditch was backfilled, two additional houses were built on top, and a new ditch and a wooden palisade were added. This second fortification protected an area of 380 m2, but outside it three additional houses and roofed hearths and stone-tool-making areas have been located. House 8 and House 9 were already abandoned by their residents and turned into ruins when the inhabitants of Amnya 1 erected the third line of fortification, destroying the ruins of the previous phase but increasing the protected area to 1,260 m2. Perhaps, the second defensive line was still in use, and the settlement was spatially divided in two parts. This second part of 880 m2 had two new houses, a large fireplace near the outer wall, and an empty central space. In sum, Amnya 1 is the earliest known fortified settlement in northernmost Eurasia (7920–7585 cal. bp at 95.4%), and its existence suggests that these groups regularly faced a hostile social environment and a real military threat. Moreover, the development of this village from a small, stand-alone, fortified dwelling to a citadel with three buildings and an unfortified part, and then to a two-part settlement, suggests an interest in the continuous improvement of the defense system. At Kayukovo 2 (figure 2.4), six houses were arranged in a circle around a central one (House 6) and connected to it by a system of excavated sunken corridors. The houses cover an area of 700 m2, and the total site area is 1,600 m2. The site of Imn’egan 2.1 (figure 2.9) covers an area of 1,900 m2. The large House 7 of Imn’egan 2.1 was located on the tip of the cape and protected by a ditch on the land side of the cape (this site area is about 700 m2). The second habitation space consisted of the two lines of dwellings, separated by a “street.” The three buildings built along the edge of the cape were connected with short corridors. Three other houses were located along the palisade, with two of them connected by a roofed corridor. T H E N E O LI T H I C A N D C H A L CO LI T H I C F O RT I F I ED S E T T LEM EN T S

33

Figure 2.3. Amnya I Neolithic fortress. Located in the lower Ob’ River basin, on the Amnya River, a right tributary of the Kazym River, Beloyar District, Khanty-Mansi Autonomous District. Discovered by Vladimir N. Shirokov in 1987. Topographic survey and excavations by Vyacheslav M. Morozov in 1987–1989; Vladimir I. Stefanov in 1988, 1989, 1993, and 2000; Viktor A. Borzunov in 1993, 2000. (I) site map; (II) map of the Neolithic fortress Amnya I and the Chalcolithic unfortified settlement Amnya II (contour lines at 1 m intervals): (a) forest; (b) trail; (c) ditch; (d) house depression; € excavated area; (f ) house depression; (g) foundational ditch of the palisade under the sand embankment; (h) location of the cross-section; (III) a section of the outer defensive system of the fortress Amnya I (southern wall of cross-section O/26–28): (1) wood bedding and humus; (2) upper black and ash-gray podzol (ancient soil); (3) gray and dark gray buried podzol; (4) rust-brown dense sand; (4е) brown dense sand; (5a) light-yellow sand; (5b) yellow sand; (5c) dark-yellow sand; (6) yellow-gray sand mixed with charcoal; (9) gray-yellow sand mixed with charcoal; (11) dark-gray and black sand mixed with charcoal; (19) pinkish ocherous sand; (23) white “sterile” sand; (23b) egg-yellow layered “sterile” sand; (24) calcined orange-pink sand. Note: figure 2.11 indicates the geographical north.

In other cases, single dwellings were located within the fenced area (House 1 at Bolshaya Umytya 9, Poludenka 1, Mikishkino 5, “Object 1” at the eastern part of Bolshaya Umytya 57). The areas of the fortified parts of 34

V I K T O R A . B O R Z U N OV

Figure 2.4. Kayukovo 2 Neolithic fortress. The Ob’-Irtysh Interfluve, the upper reaches of the Bolshoy Salym River, the vicinity of the Bolshoe Kayukovo Lake, Nefteyugansk District, Khanty-Mansi Autonomous District. Discovered by Georgy P. Vizgalov, Konstantin G. Karacharov, and Gayaz Samigulov in 1991. Surveys and excavation by Lada V. Ivas’ko and Oleg V. Kardash in 2000–2002. (I) site map (primary contour lines at 0.5 m intervals, secondary contours at 0.25 m); (II) site reconstruction; (III) excavation1 (2000–2001), House 3 and cross-section of the ditch; (IV) cross-section of the ditch and House 3. Legend: (a) forest; (b) clearance; (c) old riverbed; (d) bog; € embankment and outer walls of houses 1 to 5; (f ) house depression; (g) excavation 1; (h) charred wood, remains of burnt structures; (i) post-holes; (j) gray charcoal sandy loam; (k) sand with fine charcoal around the hearth; (l) red calcined sand (hearth); (m) sandy loam with ash and calcined bone; (n) sand (removal from modern pits); (o) wood bedding, humus, and upper white podzol; (p) grayyellow sandy loam; (q) light-yellow “sterile” sand.

the Bolshaya Umytya 9 (figure 2.5) and Bolshaya Umytya 57 (figure 2.6) are about 900 m2 and 700 m2, respectively. The exact area of Mikishkino 5 settlement cannot be determined. The large rectangular building of the settlement Bystryi Kul’egan 66 (Kosinskaya and Trufanov 2006) occupied almost the entire area of 360 m2 and is surrounded by an interrupted ditch (figure 2.7). It is most likely that the house had two rooms of 68 m2 and 64 m2. The entire T H E N E O LI T H I C A N D C H A L CO LI T H I C F O RT I F I ED S E T T LEM EN T S

35

Figure 2.5. Bolshaya Umytya 9 Neolithic settlement (a stand-alone fortified house) of the Bolshaya Umytya River, Kondin lowlands, the upper reaches of the Konda River, KhantyMansi Autonomous District. Discovered by Sergey F. Koksharov in 1990; surveyed by Alexey E. Starkov in 1994; Alexander E. Tsemenkov in 2003. Excavations and topographic survey by Andrey A. Pogodin in 2007, 2008. (I) site map (contour lines at 1 m intervals); (II) settlement Bolshaya Umytya 9, excavation area 2 (2008), House 1. Legend: (a) forest; (b) clearance; (c) bushes and meadow; (d) unpaved roads; € breakage; (f ) road slope; (g) house depression; (h) excavation 1 (2007); (i) excavation 2 (2008); (j) surface artifacts; (k) palisade foundation (?); (l) brownish-brown layer with organic materials (hearth); (m) post-holes.

Figure 2.6. Bolshaya Umytya 57 Neolithic fortified settlement. The Bolshaya Umytya River, Kondin lowlands, the upper reaches of the Konda River, Khanty-Mansi Autonomous District. Discovered by Lyudmila A. Pleshkova in 1994, surveyed by Alexander E. Tsemenkov in 2003. Excavations and topographic survey by Andrey A. Pogodin and Pavel V. Mironov in 2007, 2008. (I) site map (primary contour lines at 1 m intervals, secondary contours at 0.5 m); (II) excavation 1, House 1 and surface complexes 1–3. Legend: (a) forest; (b) clearance; (c) wetland floodplain; (d) dirt road; I disturbed surface; (f ) house depression; (g) surface artifacts; (h) test pit; (i) excavations; (j) palisade foundation (?); (k) brownish-brown layer with organic materials (hearth); (l) red-brown ocherous sand; (m) post-holes.

Figure 2.7. Bystryi Kul’egan (Bystryi Kulyogan) 66 Neolithic fortified house. The Bystryi Kul’egan (Bystryi Kulyogan) River, Ob’ basin, Surgut region, Khanty-Mansi Autonomous District. Discovered by Andrey A. Pogodin, Vladimir I. Stefanov, Viktor A. Borzunov, and Alexey S. Sergeev in 1994. Excavations and topographic survey by Lubov L. Kosinskaya in 1997, 1998, 2001. (I) site map (contour lines at 1 m intervals); (II) excavation areas I–III. Houses 2 and 2a, and the bottom of the ditch. Legend: (a) forest; (b) clearance; (c) wetland floodplain; (d) dirt roads; I road escarpment; (f ) power line; (g) house depression; (h) excavated area; (i) pit; (j) ditch; (k) post-hole; (l) hearth; (m) ceramic sherds.

excavated area—including the dwelling, a courtyard, a ditch, and a probable internal defensive log wall—covers 630 m2 (30 m by 22 m). The three dwellings at Poludenka 1 were oriented toward the center of the fortified area of the site also covers most of the site (figure 2.8). At the settlement of Ust’-Tara XXVIII, a ditch surrounded the partially excavated Object 1, which occupied more than 400 m2. These two sites, Bystryi Kul’egan 66 and Ust’-Tara XXVIII, were only prototypes of the large, fortified dwellings that would fully develop in western Siberia during the Bronze Age, as suggested by archaeologists who excavated these two sites (Gorbunova and Tolpeko 2002; Kosinskaya 1999a, 2004;). The defensive walls usually consist of single-lined palisades reinforced with sand and sod at the ground level. These structures are identified after triangular and subtrapezoidal trenches, preserved under reinforcing soil constructions. These trenches vary from 30 cm to 112 cm in width and from 40 cm to 120 cm in depth, usually filled with loamy sand with charcoal (Amnya 1, Bolshaya Umytya 9, Bolshaya Umytya 57, Poludenka 1). One may assume that scaffolds for observation and archery could be attached to the palisades from their inner sides. From the outer side, the palisades could additionally be supported with buttresses, as suggested by grooves near the external part of the palisade at Amnya 1. The slopes of the capes at Amnya 1 and Imn’egan 2.1 could be reinforced with walls made with horizontal logs supported by living or partially felled trees. At the site of Kayukovo 2 (figure 2.4), the external sides of the interconnected houses could play the role of a defensive wall, though the surrounding narrow trench may be the remnants of a palisade as well. The trapezoidalsection ditch usually varies from 1.2 m to 5 m at its top, and from 0.6 m to 1.5 m in depth. The slopes of the ditch were reinforced with wood, and the ditch had a flat bottom. A berm—a flat strip of land bordering a ditch—usually lies along inner edges, protecting sand from falling into the ditch. In general, the shape of these fortifications depends on the location of the site. The fortifications located on terraces are round shaped or rectangular, while those located on capes, are linear. It is worth noting that in all the nine cases evaluated here, the fortifications are always closed and restricted, ultimately protecting the residential space from the outside. Finally, one may assume that the entrances were protected with gates, but there is evidence for elaborate entrances with killing alleys at the sites of Kaykovo 2 and Bolshaya Umytya 9. To wrap up the section on the Neolithic sites, I also need to mention the site of Ches-Tyi-Yag (Vasiliev 1987, 2004), which may represent an additional T H E N E O LI T H I C A N D C H A L CO LI T H I C F O RT I F I ED S E T T LEM EN T S

39

Figure 2.8. Poludenka I Neolithic-Chalcolithic settlement. The mountain-forest zone of the Trans-Urals, the Poludenka River, a tributary of the Tagil River, the Tura River basin, Sverdlovsk region. Discovered and excavated by Оtto N. Bader in 1944–1946 (Bader 1948a, 1948b, 1949); excavated by Nina P. Kiparisova in 1948–1950, and Vera M. Rauschenbach in 1954. (I) site map (contour lines at 1 m intervals); (II) fortified site. Legend: (a) bog; (b) edge of the quarry; (c) settlement boundaries; (d) palisade foundation (?); (e) post-holes; (f ) grid of excavated areas; (g) excavated parts of the houses; (h) plausible house (?).

type of defensive strategy. This larger Neolithic settlement (1,314 m2) has no fortifications, but instead it is located in a triangular island, surrounded by water, bogs, and marshy areas. Out of twenty houses, five excavated vary in area from 60 m2 to 250 m2, and these were contemporary with each other. 40

V I K T O R A . B O R Z U N OV

CHALCOLITHIC FORTIFIED SITES IN WEST SIBERIAN TAIGA (CA. 5300– 4300 bp )

Towards the Chalcolithic (ca. 5300–4300 bp), the global climate had changed, making it a relatively cold and dry period, corresponding to the transition from the Atlantic period to the Subboreal (ca. 5700–2400 bp). Notably, the number of fortified sites decreased, and the number of unfortified villages increased, suggesting that the levels of hostility decreased (Koksharov 2009). Some part of population lived in dugout dwellings, similar to those of the Neolithic, and carried out most of their productive activities in these permanent villages. The seasonal hunting and fishing camps differ from these settlements because their dwellings were only slightly deep, made of small frame-and-pillar structures, and with lighter frame structures such as tents or chums (temporary dwellings used by the nomadic Uralic) (Chemyakin 2008; Koksharov 2009). Fortified villages of 1,900–2,000 m2 were usually located on the capes, some of which were already occupied during the previous period, and are found only in the eastern areas of the Ob’ region, near the modern Russian town of Surgut. The layouts of sites such as Imn’egan 2.1 (figure 2.9), Nioh-urij 3.3 (figure 2.10), and possibly Kayukovo 1 resemble the Neolithic fortified settlement of Amnya 1 (figure 2.3), suggesting some degree of continuity between the two periods. New fortified sites of Volvoncha I (figure 2.11), Nivagalskoye 20 (figure 2.12), and Teplyi Ruchey II also appeared in the region during the Chalcolithic. These sites are novel for the region. Their stand-alone fortified houses had between 200 m2 and 400 m2 in area and were only slightly deepened in the ground or were built above the surface. The structure could have either frame-and-pillar with the shape of a truncated pyramid or timbered prismatic-shaped dwellings with self-supporting vertical walls. Such fortified houses were located only on naturally defensible spots such as capes, cape-like protrusions, and edges of natural terraces, and additionally protected by the ditches and, possibly, palisades. Several of such houses could form a fortified settlement, as can be seen on Nioh-urij 3.3. The sites of Imn’egan 2.1 (1,900 m2) and Nioh-urij 3.3 (2,000 m2) are located on the tips of sandy capes 4 m and 6–7 m, respectively, above the marshy floodplain. Both sites consist of two parts separated by arcuate walls and ditches. These citadels were formed by fortified dwellings with areas of 360 m2 at Imn’egan 2.1 (House 7) and 250 m2 at Nioh-urij 3.3 (House 1). Smaller buildings surrounded by the ditches form the rest of the settlements. The site of Volvoncha I (figure 2.11) is located on the left bank of the Konda River, between its middle and lower reaches, on a large cape-like formation. T H E N E O LI T H I C A N D C H A L CO LI T H I C F O RT I F I ED S E T T LEM EN T S

41

Figure 2.9. Imn’egan (Imnyogan) 2.1 Neolithic-Chalcolithic fortress. The Imn’egan (Imnyogan) River, the Agan River basin, Surgut region, Khanty-Mansi Autonomous District. Discovered and surveyed by Alexander N. Bessmertnykh and Konstantin G. Karacharov in 1995. Site map (primary contour lines at 1 m intervals, secondary contours at 0.5 m). Legend: (a) forest; (b) wetland floodplain; (c) house depression (depths are in meters from the modern surface); (d) ditch; (e) embankment; (f ) test pit.

This site was built in the same place where an earlier unfortified Neolithic site had stood. The transition can be demonstrated by two groups of radiocarbon measurements, acquired from charcoal samples from the floor. The early samples have the uncalibrated radiocarbon age of 5950 ± 80, 5260 ± 40 and 4720 42

V I K T O R A . B O R Z U N OV

Figure 2.10. Nioh-urij 3.3 Neolithic-Chalcolithic fortress. The old bed of Nioh-urij River, the Agan River basin, Surgut region, Khanty-Mansi Autonomous District. Discovered and surveyed by Sergey A. Myznikov in 2004. Site map (primary contour lines at 1 m intervals, secondary contours at 0.5 m). Legend: (a) forest; (b) clearance; (c) wetland floodplain; (d) slope; (e) pits from fallen pines; (f ) ditch; (g) embankment around the house depression; (h) house depression.

± 60 bp (LE-1451, 1452, 1450), and two later samples dated to 4240 ± 40 and 3900 ± 60 bp (LE-1453, 1448) (Koksharov and Stefanova 1993, 63). According to the primary investigators (Koksharov and Stefanova 1993), the Chalcolithic fortified dwelling had the marquee-like appearance, covering an area of 208 T H E N E O LI T H I C A N D C H A L CO LI T H I C F O RT I F I ED S E T T LEM EN T S

43

[128.104.46.206] Project MUSE (2024-03-01 20:28 GMT) UW-Madison Libraries

Figure 2.11. Volvoncha I Chalcolithic settlement (fortified house). The Polymyatskaya anabranch, the lower reaches of the Konda River, Kondin lowland, Khanty-Mansi Autonomous District. Discovered by Alexander V. Malyshkin in 1977. Surveyed and excavated by Nina K. Stefanova in 1979. Legend: (a) water; (b) ditch; (c) palisade foundation under the house wall; (d) house foundation pit, pit; (e) remains of sandy wall; (f ) post-holes; (g) the Early Bronze Age house foundation.

m2 (16 m by 13 m) and a depth of about 1 m below the ground level. The postholes for the supporting pillars are grouped in rows of two or three along the sides of the dugout with a 2–4 m gap between the rows. In addition, the house was surrounded by a 3 m wide ditch (Borzunov 1999). Perhaps the structure 44

V I K T O R A . B O R Z U N OV

Figure 2.12. Nivagalskoye 20 Chalcolithic fortress (fortified house). The Nonk-egan River, the Agan River basin, Surgut region, Khanty-Mansi Autonomous District. Discovered and surveyed by Egor А. Danilov in 2010–2011. Excavations by Pavel S. Bakharev in 2011. (I) site map (primary contour lines at 1 m intervals, secondary contours at 0.5 m); (II) excavated area; (III) western cross-section: (1) sod (wood bedding and humus); (2) white gray podzol; (3) sand with organic material and podzol; (4) yellow-brown sand (rampart); (5) mixed yellow-brown with organic material and podzol sand; (6) gray-brown sand (backfill of illegal excavation); (7) gray-yellow sand with fine embers (foundation of the defensive wall); (8) gray sand; (13) dark-gray sand with organic materials. Legend: (a) forest; (b) bushes; (c) bog; (d) marshy floodplain around the old bed; (e) dirt road; (f ) ditch and walls of the fortified house; (g) house depression; (h) ground raised platform surrounded by pits; (i) the Early Iron Age grave; (j) 2011 excavation.

looked like a sandy hill for the outside viewer and became a prototype for similar Bronze Age houses, known at the nearby settlement of Pashkin Bor I (Stefanova and Koksharov 1988). According to Sergey Koksharov (2009, 234), the dwelling was inhabited by a large group of people who shared productive activities. The site of Nivagalskoye 20 (figure 2.12) locates on a 4–4.5 m high cape. It consists of a solid wooden-earthen construction erected at the ground level. Its remnants look like a semisquare (20–23 m by 20–23 m) platform, protected by an arcuate sandy wall of 5.7 m in width and 1 m in height. Outside the embankment, there is an arc-shaped ditch of 5 m in width, 0.8 m in depth and additional narrow and higher (5 m by 1.3 m) external wall. The difference in elevation between the inner wall and the bottom of the ditch is 1.2 m and between the outer embankment and the ditch is 1.5 m. Additionally, the test pit revealed an inner palisade, located outside the second embankment. In the excavation, it looked like a 0.1–0.2 m wide ditch with the depth of 0.7–1.1 m from the modern surface. The estimated area of the site is 900–950 m2 (35 m by 30 m), and the area of the building is about 400 m2 (Danilov 2011). The fortified dwelling of Teplyi Ruchey II locates on a cape-shaped protrusion of the right bank of the Yem-Yegan River (the basin of the Malaya Sosva River), around 4 m above the floodplain. There is a large (12 m by 11 m) trapezoid depression of 1.7 m in depth, protected by a sandy 4 m wide embankment and a narrow outer ditch. The area of the site is 360 m2 (Borzunov 2016). In all these sites, the artifact assemblages resemble Neolithic technology with stone arrowheads, axes, scrapers, abrasives, various pottery utensils, and clay weights for fishing nets. Thus, there is apparent evidence for the fortification in the Chalcolithic, though the idea of the fortified settlements seems to be less prominent than during the Neolithic. The prototypes of these defensive systems emerged during the Neolithic and then developed further during the Chalcolithic and the following periods. DISCUSSION

There is no agreement between scholars on who built the fortified settlements treated in this chapter. Some believe that the Neolithic villages, including fortified ones, were founded by the herdsmen-farmers who migrated into the Taiga from the Aral Sea region, the Caspian Sea, southeastern Europe, or western Asia. The appearance of a new pottery tradition with flat-bottomed vessels as opposite to the Indigenous Mesolithic tradition of round-bottom 46

V I K T O R A . B O R Z U N OV

vessels supports the above notion (Ivas’ko 2002, 2008; Ivas’ko and Kardash 2002; Kovaleva and Zyryanova 2010; Vasiliev 1987, 1991; Zakh 2003). The closest simultaneous analogs of the northern Neolithic fortifications are fortified cape settlements and large unfortified circular villages of the Cucuteni-Tripolie culture. In the uncalibrated radiocarbon range, these settlements existed circa 7500–6000 bp, or relatively close to the radiocarbon age of the Siberian settlements. Another analog is fortified settlements of the Lengyel (Zymno-Zlota) culture and the Funnel Beaker culture (Zimno, Vinniki), dated back to circa 6000–4500 bp. Some of the Lengyel culture settlements also measure 300 m by 90 m and were protected by concentric ditches (Masson et al. 1982). Another argument in favor of nonlocal origin of the fortifications is that Kayukovo 2 and Mikishkino 5 were ritually burnt after people left them and removed all valuable objects. The principal investigators interpret this ritual as a sign of religious practices that took places behind the erected walls, and this tradition is not considered to be local (Ivas’ko 2002; Ivas’ko and Kardash 2002). However, most of forest settlements from any period have at least some traces of burning, which can be explained by the combination of natural and artificial factors, since the natural fires are also common in Taiga. Finally, Lada Ivas’ko and Oleg Kardash (2002) consider the circular layout of Kayukovo 2 as a direct analog to the Anatolian Neolithic, the Chalcolithic of the Black Sea region and the Bronze Age fortifications of the steppes and, thus, nonlocal. From my point of view, the circular arrangement of dwellings is a universal pattern and cannot be used in favor of the nonlocal hypothesis of the fortification origin. There was no evidence for agriculture on the Neolithic and Chalcolithic sites. Thus, the new pottery styles could have penetrated the Taiga area as a result of cultural diffusion, rather than direct migration. It is hard to believe that early farmers and herders rapidly migrated into the completely unknown environmental zone and abandoned their economic practices, keeping only pottery styles and defensive strategies (fortifications). I agree with my colleagues that it is very tempting to classify the southern sites as the prototypes of the Siberian settlements. For instance, it is possible that some Neolithic groups of hunters and fishermen from the UralKazakhstan steppes, the Aral Sea region, the Caspian Sea, the Volga region, and the Altai have migrated into the Ural-Siberian forest-steppe and then slowly adapted to new climatic conditions and social environment, including intermixing with the aboriginal populations. Furthermore, these heterogeneous groups moved into the Trans-Urals Forest and to the upper T H E N E O LI T H I C A N D C H A L CO LI T H I C F O RT I F I ED S E T T LEM EN T S

47

Ob’ region, interacting with the descendants of the local Mesolithic people (Kosinskaya 1999b, 2002). The decrease in the number of fortifications during the Chalcolithic might be explained as a result of the decrease in the number of immigrants from the neighboring areas of western Siberia and the TransUrals due to worsening climate conditions. However, I think that this similarity is just formal and limited to the sites’ layouts. Economic strategies and material cultures of the Neolithic farmers and herdsmen and the Taiga fishermen-hunter-gatherers differed significantly. While one may assume that the concept of fortification could be borrowed from the distant areas, the mechanisms of information exchange remain unclear. In my view, it is most likely that in western Siberia the earliest fortified settlements and other defense systems were developed independently by the local groups of hunters and fishermen with some influence from outside. Wood-earthen architecture is generally a product of forest groups, and timber strongholds were technical innovations of the West Siberian Taiga people, while in other areas people preferably used stone and adobe houses architecture (Dzhandieri 1981). Wooden palisades were also common for forest areas, where pine, larch, and cedar trees were available. Palisades could also be an independent invention of the Taiga population, derived from the constructions of dams on streams, rivers, and lakes for permanent fishing, as well as residential, economic, and ritual structures. The Taiga fortified villages were multifunctional, serving as communal and tribal year-round centers, outposts on recently populated lands, permanent shelters, and places for productive and religious activities. Nothing convincingly suggests their exclusive specialization on any practice, like religion, but there is evidence for permanent habitation of several large families during a long time. Multifunctionality is a general feature of many early fortifications (Masson 1981). Fences, walls, palisades, and moats protected and marked communal boundaries; prevented wild fauna from coming to the village; and protected inhabitants from floods, winds, landslides, and soil erosion. The ditches served for draining water, for throwing garbage, and, possibly, as fire-protection structures. The latter is not very likely since clearance the area from the forest is more effective in comparison with moats. The embankments were not independent objects, but elements of the fortification that originated after the excavation of the ditches and due to the need to reinforce the palisades. Except for the fortifications themselves, there is no direct evidence for Neolithic/Chalcolithic warfare, but indirect evidence may support it. First, all the settlements were burned down, and some of them several times. While 48

V I K T O R A . B O R Z U N OV

some archaeologists believe this practice was due ritual practices, fires also could be caused by hostilities. Amnya 1 was burned down at least three times, but every time the settlement was rebuilt. Second, seven polished stone arrowheads came from Amnya 1, and they might be related to warfare. The lack of decent raw materials in the north of western Siberia, meant that hunting arrowheads were often made of bone, which is the most affordable material. Until the Middle Ages, combat projectiles were made of nonlocal precious metal, which makes me to believe that these labor-intensive stone artifacts were used mainly in combat. The variability of early fortifications in the Taiga zone represents the experimentation with forms and types of defenses, seeking for the most optimal defense strategy to protect a community in a hostile social environment since fire and dwellings are enough protection from wild fauna such as bears, wolves, and lynxes. Also, settlements Amnya 1, Imn’egan 2.1, Kaykovo 2, and Bolshaya Umytya 57 are manifestations of the strength, high technical level, and complex social organization of their inhabitants as a response to fear. Nevertheless, it should be emphasized that the Neolithic and Chalcolithic settlements were not designed to survive a long siege or an attack of any enlarged squad. Their inability to face a large-scale attack suggests that the possible enemy did not outnumber the local communities, so they were either the aboriginal Mesolithic groups or the groups of migrants, who could penetrate the Taiga area from the south. From the Medieval folk narratives of the Ob’ Ugrians, we learn about numerous military affairs between chiefs who conducted long-distance military campaigns for brides and also about raids of Samudian reindeer herders from the Tundra (Patkanov 1891). After the initial Neolithic/Chalcolithic period, the fortified settlements disappeared in the northern Taiga for more than a thousand years and reappeared again only by the end of the Bronze Age (ca. 3000–2800 bp) until the Iron Age. Stand-alone fortified houses of the total area of 360–400 m2 were prevalent during the Bronze Age (ca. 4300–3500 bp). However, there is no cultural continuity observed between these and the Neolithic/Chalcolithic sites (Borzunov 1999). CONCLUSIONS

The origin of the Old World’s northernmost Stone Age fortifications can be found in the conflict between the local people, sparse groups of newcomers from southern areas, and intermixed groups. Northern Taiga fishing resources and hunting grounds may have attracted people during the climatic optimum T H E N E O LI T H I C A N D C H A L CO LI T H I C F O RT I F I ED S E T T LEM EN T S

49

of the Holocene’s Atlantic period. The competition for these resources may have started intergroup conflict. These fortresses and fortified dwellings may have served also as year-round/permanent habitations, the center for communal and tribal economic and religious activities, and as military outposts in newly occupied territory. There is no reason to believe that fortifications came to the Taiga zone as an imported concept from the southern peoples. The local Taiga groups of fishermen and hunters, unfamiliar to food-production practices, independently developed these defensive strategies and created unique techniques and forms such as the soil-reinforced palisades (usually wrongly interpreted as earthen walls). Fortified settlements and fortified dwellings were absent in the neighboring areas of the Eurasian forest belt and steppes during the Neolithic and Chalcolithic. The sites Amnya 1, Kaykovo 2, Bolshaya Umytya 9, Bolshaya Umytya 57, Mikishkino 5, Imn’egan 2.1, and others were the only fortifications in northern Eurasia, as well as the northernmost fortified settlements in the Old World. ACKNOWLEDGMENTS

The chapter was prepared with the support of the State Assignment of the Ministry of Science and Higher Education of the Russian Federation with the theme “Regional Identity of Russia: Comparative Historical and Philological Studies,” topic no. FEUZ-2020-0056. REFERENCES

Bader, Otto N. 1948a. “Arkheologicheskiye issledovaniya na Urale v 1946” [Archaeological research in the Urals in 1946]. Brief Reports on Presentations and Fieldwork of the Institute of History of Material Culture of the USSR Academy of Sciences (20): 78–81. Bader, Otto N. 1948b. “Neoliticheskaya stoyanka na r. Poludenke bliz g. Tagila” [The Neolithic site on the Poludenka River near the town of Tagil]. First Ural Archaeological Conference, Molotov, USSR. Bader, Otto N. 1949. “Novyy tip neoliticheskogo poseleniya na Urale” [The new type of Neolithic settlement in the Urals]. Soviet Ethnography (2): 144–150. Baker, Jonathan L., Matthew S. Lachniet, Olga Chervyatsova, Yemane Asmerom, and Victor J. Polyak. 2017. “Holocene Warming in Western Continental Eurasia Driven by Glacial Retreat and Greenhouse Forcing.” Nature Geoscience 10 (6): 430–436. 50

V I K T O R A . B O R Z U N OV

Borzunov, Viktor A. 1999. “Novyy areal ukreplennykh zhilishch na severe Yevrazii” [A new area of the fortified Houses in Northern Eurasia]. Russian Archaeology (4): 5–23. Borzunov, Viktor A. 2016. “Ukreplennyye poseleniya eneolita tayezhnogo Priob’ya” [The Chalcolithic fortified Settlement of the Ob’ River Taiga]. Russian Archaeology (3): 34–44. Borzunov, Viktor A. 2018. “Varianty rekonstruktsii i tendentsii razvitiya statsionarnykh zhilishch epokh kamnya i bronzy tayezhnoy zony Zapadnoy Sibiri” [Appearance and development trends of permanent dwellings of the Stone and Bronze Ages in the Taiga Zone of Western Siberia]. Stratum Plus (2): 347–375. Bronk Ramsey, Cristopher 2009. “Bayesian Analysis of Radiocarbon Dates.” Radiocarbon 51 (1): 337–360. Chemyakin, Yuri P. 2008. Barsova Gora: Ocherki arkheologii Surgutskogo Priob’ya. Drevnost [Barsova Gora: Essays on the archeology of the Surgut Ob region. Antiquity]. Omsk, Russia: Omskiy dom pechati. Danilov, Egor A. 2011. “Otchet o nauchno-issledovatel’skoy rabote: Arkheologotopograficheskiye issledovaniya v 2011 godu na Pokachovskom mestorozhdenii nefti v Nizhnevartovskom rayone KhMAO–Yugry” [Research report: Archeological and topographical studies in 2011 at the Pokachyovskoye oil field in the Nizhnevartovsk District of the KMAO—Ugra]. Nefteyugansk, Russia: Archive of the “Northern Archeology-1,” F.1, case 330. Dzhandieri, Maria I. 1981. “Drevneye bashennoye oboronnoye zhilishche” [The ancient tower-dwelling]. Journal of Ancient History 2: 118–155. Gorbunova, Tatyana A., and Irina V. Tolpeko. 2002. “Raskopki v Ust’-Tarskom arkheologicheskom mikrorayone” [Excavations in the Ust-Tarsk archeological micro-region]. In Archaeological Discoveries of 2001, edited by V. V. Sedov and N. V. Lopatin, 405–406. Moscow: Institute of Archeology of the Russian Academy of Sciences. Ivas’ko, Lada V. 2002. “Ukreplennoye poseleniye kamennogo veka Kayukovo 2” [The Stone Age fortified settlement Kayukovo 2]. In Materialy i issledovaniya po istorii Severo-Zapadnoy Sibiri, edited by G. P. Vizgalov, 7–25. Ekaterinburg, Russia: Ural State University. Ivas’ko, Lada V. 2008. “O kayukovskoy arkheologicheskoy kul’ture” [On the Kayukovo archaeological culture]. In Barsova Gora: Drevnosti tayezhnogo Priob’ya, edited by A. Ya. Trufanov, 112–122. Ekaterinburg, Russia: Ural’skoye izdatelsto. Ivas’ko, Lada V., and Oleg Kardash. 2002. “Po sledam ognya: Otkrytiye novogo gorodishcha kamennogo veka” [In the footsteps of fire: The discovery of a New Stone Agesettlement]. Rodina, special issue: 40–42.

T H E N E O LI T H I C A N D C H A L CO LI T H I C F O RT I F I ED S E T T LEM EN T S

51

Koksharov, Sergey F. 2009. Pamyatniki eneolita severa Zapadnoy Sibiri [The Chalcolithic Sites of the north of Western Siberia]. Ekaterinburg, Russia: Volot. Koksharov, Sergey F., and Nina K. Stefanova. 1993. “Poseleniye Volvoncha I na r. Konde” [Volvoncha I settlement on the Konda River]. In Pamyatniki drevney kul’tury Urala i Zapadnoy Sibiri, edited by L. N. Koryakova, 54–67. Ekaterinburg, Russia: Nauka. Kosinskaya, Lubov L. 1999a. “O skhemakh razvitiya severnogo neolita” [On the development of the Northern Neolithic]. In Obskiye ugry, edited by A. V. Golovnev, 38–41. Omsk, Russia: Omskiy gosudarstvennyy pedagogicheskiy universitet. Kosinskaya, Lubov L. 1999b. “Neolit severa Zapadnoy Sibiri: Problema yuzhnykh svyazey” [The Neolithic of the north of western Siberia: On the problem of southern connections]. XIV Ural’skoye arkheologicheskoye soveshchaniye, Chelyabinsk, Russia: Rifei. Kosinskaya, Lubov L. 2002. “Neolit severa Zapadnoy Sibiri” [The Neolithic of the north of Western Siberia]. Tverskoy arkheologicheskiy sbornik, Tver: Tverskoy Gosudarstvennyi Muzei. Kosinskaya, Lubov L. 2004. “Zhilishcha poseleniya Bystryy Kul’yogan 66” [Dwelling at the Settlement of Bystryi Kul’egan 66]. In Zapadnaya Sibir’: Proshloye, nastoyashcheye, budushcheye, edited by T. A. Isaeva, 226–241. Surgut, Russia: Diorit. Kosinskaya, Lubov L., Yuri P. Chemyakin, and Galina I. Zaitseva. 2002. “Radiouglerodnyye daty s arkheologicheskikh pamyatnikov iz okrestnostey Surguta” [Radiocarbon dates of the archaeological sites from the vicinity of Surgut]. In Barsova Gora: 110 let arkheologicheskikh issledovaniy, edited by A. Ya. Trufanov, 141–146. Surgut, Russia: Barsova gora. Kosinskaya, Lubov L., and Alexander Ya. Trufanov, eds. 2006. Poseleniye Bystryi Kul’egan 66: Pamyatnik epokhi neolita Surgutskogo Priob’ya. Kollektivnaya monografiya [Bystryi Kul’egan 66 settlement: A Neolithic site near Surgut. Collective monograph]. Ekaterinburg, Russia: Ural’skoye izdatel’stvo. Kosintsev, Pavel A., Nataliya E. Bobkovskaya, and Evgenyi M. Besprozvannyi. 2004. “Radiouglerodnaya khronologiya arkheologicheskikh pamyatnikov tayezhnoy zony Zapadnoy Sibiri” [A Radiocarbon chronology of archaeological sites of the Taiga Zone of Western Siberia]. In Khanty-Mansiyskiy avtonomnyy okrug v zerkale proshlogo 2, edited by Ya. A. Yakovlev, 17–13. Tomsk, Russia: Tomskiy universitet. Kovaleva, Valentina T., and Svetlana Yu. Zyryanova. 2010. Neolit Srednego Zaural’ya (Boborykinskaya kul’tura) [The Neolithic of the Middle Trans-Urals (the Boborykin Culture)]. Ekaterinburg, Russia: Ural’skiy gosudarstvennyy universitet.

52

V I K T O R A . B O R Z U N OV

Masson, Vadim M. 1981. Altyn-Depe: Raskopki goroda bronzovogo veka v Yuzhnom Turkmenistane [Altyn-Depe: Excavations of the Bronze Age City in southern Turkmenistan]. Leningrad: Nauka. Masson, Vadim M., Nikolay Ya. Merpert, Rauf M. Munchayev, and Evgenyi K. Chernysh. 1982. Arkheologiya SSSR: Eneolit SSSR [Archeology of the USSR: Aeneolithic]. Moscow: Nauka. Morozov, Vyacheslav M., and Vladimir I. Stefanov. 1993. “Amnya I: Drevneysheye gorodishche Severnoy Yevrazii?” [Is Amnya 1 the earliest fortified Settlement in Northern Eurasia?]. Voprosy arkheologii Urala (21): 143–163. Patkanov, Serafim K. 1891. Tip ostyatskogo bogatyrya po ostyatskim bylinam i geroicheskim skazaniyam [The type of the Ostyak hero in Ostyak epics and heroic tales]. St. Petersburg. Zhivaya starina (III—IV): n/p Pogodin, Andrey A. 2010. Poseleniye Bolshaya Umytya 9: Rezul’taty polevykh issledovaniy 2007–2008 gg. v Sovetskom rayone KhMAO–Yugry [Bolshaya Umytya 9 Settlement: The results of the 2007–2008 field studies in the Sovetsky District of KhMAO–Ugra], edited by Ya. A. Yakovlev, Khanty-Mansiysk Autonomous Okrug in the Mirror of the Past 8. Tomsk, Russia: Tomskiy universitet. Pogodin, Andrey A., and Pavel V. Mironov. 2009. Predvaritel’nyye rezul’taty avariynykh raskopok poseleniya Bolshaya Umytya 57 v Sovetskom rayone KhMAO–Yugry (po materialam issledovaniy 2007–2008 gg.) [Preliminary results of the emergency excavations of the Bolshaya Umytya 57 settlement in the Sovetsky District of the KhMAO–Ugra (based on research materials from 2007–2008)], edited by Ya. A. Yakovlev, Khanty-Mansiysk Autonomous Okrug in the Mirror of the Past 7. Tomsk, Russia: Tomskiy universitet. Reimer, Paula J., et al. 2013. “IntCal13 and Marine13 Radiocarbon Age Calibration Curves 0–50,000 years cal. bp.” Radiocarbon 55 (4): 1869–1887. Stefanov, Vladimir I., Viktor A. Borzunov, Andrey A. Pogodin, and Olga N. Korochkova. 1999. “Gorodishche kamennogo veka Amnya I: Novyye dannyye” [The Stone Age fortified Settlement of Amnya I: New data]. In XIV Ural’skoye arkheologicheskoye soveshchaniye: Tezisy dokladov, edited by S. A. Grigoriev, 43–44. Chelyabinsk, Russia: Rifey. Stefanov, Vladimir I., and Viktor A. Borzunov. 2008. “Neoliticheskoye gorodishche Amnya I (po materialam raskopok 1993 i 2000 gg.)” [The Neolithic Site of Amnya 1 (based on 1993 and 2000 excavations)]. In Barsova Gora: Drevnosti tayezhnogo Priob’ya, edited by A. Ya. Trufanov, 93–111. Ekaterinburg, Russia: Ural’skoye izdatel’stvo. Stefanov, Vladomor I., Vyacheslav M. Morozov, and Andrey A. Pogodin. 2005. “Kirip-Vis-Yugan 2: Pamyatnik amninskogo tipa (k voprosu o neolite Prikazym’ya)”

T H E N E O LI T H I C A N D C H A L CO LI T H I C F O RT I F I ED S E T T LEM EN T S

53

[A Kirip-Vis-Yugan 2 site of the Amnya type (on the question of the Neolithic in the Kazym River region)]. In Istochniki po arkheologii Zapadnoy Sibiri, edited by T. N. Glushkova, 19–33. Surgut, Russia: SGPI. Stefanova, Nina K., and Sergey F. Koksharov. 1988. “Poseleniye bronzovogo veka na r. Konde” [The Bronze Age settlement on the Konda River]. Soviet Archaeology (3): 161–174. Vasiliev, Evgenyi A. 1987. “Migratsionnyye protsessy v tayezhnoy polose Zapadnoy Sibiri v eneoliticheskuyu epokhu (prichina i dinamika)” [Migration processes in the Taiga Zone of western Siberia in the Chalcolithic (cause and dynamics)]. In Smeny kul’tur i migratsii v Zapadnoy Sibiri, edited by L. M. Pletneva, 13–14. Tomsk, USSR: TGU. Vasil’yev Eygenyi A. 1991. “K probleme sredneaziatsko-zapadnosibirskikh svyazey v neoliticheskuyu epokhu” [On the Central Asian–West Siberian connections during the Neolithic]. In Problemy khronologii i periodizatsii arkheologicheskikh pamyatnikov Yuzhnoy Sibiri, edited by Yu. F. Kiryushin, 31–33. Barnaul, Russia: Altaiskoi gosudarstvennyy universitet, Institut istorii, filosofii i filologii Sibirskogo otdeleniya Akademii nauk SSSR. Vasiliev, Evgenyi A. 2004. “Raskopki neoliticheskogo poseleniya Ches-Tyi-Yag na Pripolyarnom Urale” [Excavations of the Neolithic settlement Ches-Tyi-Yag in the Subpolar Urals]. Khanty-Mansiyskiy avtonomnyy okrug v zerkale proshlogo (2): 296–301. Zakh, Viktor A. 2003. Epokha neolita i rannego metalla lesostepnogo Prisalair’ya i Priob’ya [The Neolithic and Chalcolithic of the Forest-Steppe Zone of the Salair and Ob’]. Tyumen, Russia: Institut problem osvoyeniya Severa SO RAN.

54

V I K T O R A . B O R Z U N OV

3 The widespread conflict between socially and economically differentiated communities has been seen as a common feature of one of the key periods of Old World prehistory, the Bronze Age (Anthony 2009; Kristiansen and Rowlands 2005). Indeed, the burial of twenty-eight violently killed adult individuals under the Pepkino Mound (4100–3900 cal. bp), the sieged Fortress of Liventsovo (ca. 3750 bp), or the massive battlefield of the Tollense valley (ca. 3250–3150 bp) is each direct evidence of intense hostilities. These cases support the Carneiro’s (1998) view of warfare as a central process for the emergence of political units that transcended local autonomy. However, such vivid examples are rare in Bronze Age archaeology, and scholars studying conflicts have no choice but to rely on indirect evidence. Weaponry is a good indicator, but fortified houses and settlements are on top among such evidence, since they are often interpreted as people’s attempt to protect themselves. One instance of reinforced settlement is the Late Bronze Age Sintashta-Petrovka archaeological phenomenon (4000–3700 cal. bp), situated primarily in the middle part of the Great Belt of the Eurasian steppes within the southern Trans-Ural peneplain (figure 3.1). The area of study lies in the north temperate zone approximately between 50° and 55° north latitude. The present-day environmental conditions of the steppe region developed in the Holocene, approximately 11,700 years ago, and are characterized by cycles

Winter Is Coming Is Bronze Age “Fortification” Always Fortification?

Igor V. Chechushkov

https://doi.org/10.5876/9781646422111.c003

55

Figure 3.1. The map of the Sintashta-Petrovka archaeological sites. Settlements: (101) Stepnoye; (102) Shibaeyvo 1; (103) Chernorechye 3; (104) Bakhta; (105) Paris; (106) Isiney; (107) Kuisak; (108) Ust’ye 1; (109) Rodniki; (110) Konoplyanka; (111) Zhurumbay; (112) Arkaim; (113) Sintashta; (114) Sintashta 2; (115) Kamennyi Ambar; (116) Alandskoye; (117) Chekatay; (118) Selek; (119) Sarym-Sakly; (120) Kamysty; (121) Kizilskoye; (122) Bersuat; (123) Andreyevskoe; (124) Ulak; (125) Streletskoye; (126) Zarechnoye 4; (127) Kamennyi Brod. Cemeteries: (201) Ozernoye 1; (202) Krivoe Ozero; (203) Stepnoye M; (204) Kamennyi Ambar-5; (205) Stepnoye 1; (206) Tsarev Kurgan; (207) Ubagan 2; (208) Solntse 2; (209) Bolshekaraganskyi; (210) Aleksandrovskyi 4; (211) Sintashta; (212) Solonchanka 1a; (213) Knyazhenskyi; (214) Bestamak; (215) Ishkinovka 1; (216) Ishkinovka 2; (217) Novo-Kumakskyi; (218) Zhaman-Kargala 1; (219) Tanabergen 2; (220) NovoPetrovka; (221) Semiozernoye 2; (222) Khalvayi 3. The geographic map is created with the ESRI ArcGIS software package; the source of freely available geographic data is the National Geographic Society. The site locations are provided by the author.

Figure 3.2. Environments of the study area. (1) Climate change in Holocene. A composite curve of δ18O values from the Kinderlinskaya Cave in the southern Urals (Baker et al. 2017) and summed probability distribution for Kamennyi Ambar (intersection of two curves suggest relatively little change in δ18O values since 4000 cal. BP and modern times). (2) A rose wind obtained from the Arkaim meteorological record. (3) A soil profile of the wall at Ust’ye 1. Small particles inside the wall represent alluvium.

of relatively dry and cold climate punctuated by wetter and warmer periods (figure 3.2: 1). The landscape of the Trans-Urals peneplain varies from rolling steppe with some hills on the west to almost flat plain on the east. The Ural Mountains condition the climate by blocking the flow of air from the west W I N T ER I S CO M I N G

57

and providing a tunnel for cold and dry arctic air. In the summer, continental tropical air flows up from Asia, bringing hot weather. The southeastern part of the region is windy with 300–320 windy days annually on average. The climate is characterized as continental with mean temperatures below 0°C in winter and above +10°C during the summer months. The area of study belongs to south steppe region, which is considered as the zone of insufficient humidity (300 mm of precipitation a year) with the largest amount of precipitation during the spring and fall months (75%–78% of annual precipitation), with remaining moisture (22%–25%) coming from snow (Levit 2005). Sintashta-Petrovka represents a rapid shift in the social history of the central Eurasian steppes from small, dispersed mobile groups of herders to nucleated walled communities. At the end of the third millennium bce, at least twenty-five walled settlements—such as Sintashta, Arkaim, Petrovka 2, Ust’ye 1, Alandskoye, and Kamennyi Ambar—emerged between the Ural and Tobol Rivers. The quintessential archaeological evidence of Sintashta-Petrovka communities takes the form of nucleated and enclosed settlements paired with highly recognizable kurgan (burial mound) cemeteries. All settlements were enclosed with 2–3-m high ramparts and 1 m–2 m deep ditches, within which houses were placed immediately adjacent to each other in the manner that neighboring structures shared intermediate walls. The enclosed areas of settlements vary from 0.7 ha to 3.4 ha, where up to sixty houses of 100–200 m2 could be packed, representing some 300–600 people (Koryakova and Epimakhov 2007). It is difficult to distinguish ranked social groups based on households’ architecture because differences in elaboration between the residences are very tenuous. However, the composition and diversity in artifact assemblages allow observing some differences between the houses regarding social status, productive differentiation, and craft specialization (Chechushkov 2018). The most outstanding Sintashta-Petrovka graves are individual male burials accompanied by projectiles and axes, mace heads, craft tools, and a specific set of sacrificed animals (horses, cows, and dogs). Chariots—the most famous and spectacular material component of Sintashta-Petrovka society—are known exclusively from burial contexts. Two-wheeled vehicles represent a complex technology, incorporating crucial innovations, the investment of substantial resources, advanced handcraft, and developed military skills. Burials with chariots probably represent military elites, suggesting that individual leadership played a crucial role in cementing the society and extended into the realm of ideology and general social prestige (Chechushkov and Epimakhov 2018; Earle 2011, 32–33). 58

I G O R V. C H E C H U S H KOV

Bronze Age subsistence relied on a wide variety of resources, among which meat and dairy products played crucial roles: remains from settlements and composition of stable isotope in human bones univocally suggest a key role for livestock products in the diet, with a minor contribution of wild fauna, freshwater fish, and wild plants. Intriguingly, the botanical studies did not reveal any evidence for domesticated plants, suggesting that steppe people neither included domesticates into their diet nor fed animals (Krause and Koryakova 2013; Stobbe et al. 2016). The stable isotope studies indicate that livestock was raised and grazed locally within the same river valley (Kiseleva et al. 2017), while the residential architecture represents fully sedentary year-round residence for at least a part of the population. Summarizing, excavated households represent almost identical architectural patterns, similar levels of wealth and prestige, little productive differentiation, and no evidence of elites amassing substantial wealth through control of craft or subsistence production. However, if the construction of enclosures required planning, organization, and investment of effort, then there had to be a strong reason for such practices. At the same time, the burial records demonstrate social differentiation between elite charioteers and commoners, pointing out the importance of military leadership in social life. It could be hypothesized that the competition between Bronze Age communities resulted in a hostile social environment, forcing people to live in the highly nucleated villages and to praise the most skillful warriors and distinguished leaders (Earle 1997). The role of warfare within Sintashta-Petrovka culture is still under debate; the Sintashta-Petrovka settlements demonstrate no evidence for military catastrophes but the peaceful departure of the inhabitants. In conjunction with this fact, there is no military-related trauma on human remains ( Judd et al. 2018). These facts raise questions about the exclusively military function of these settlements and ask for an alternative economic interpretation. Thus, Nikolai Anisimov (2009), after analyzing the architectural patterns, explained them as comprehensive livestock management systems (inhabited corrals) developed as a social response to the harsh environmental conditions. However, it remains unclear if spatial patterns and settlements’ architecture correlate with the environmental variables of the Trans-Urals steppes that could force people to construct the enclosed shelters. In this chapter, I aim to test if the enclosed settlements could be explained in an environmental rather than in a military sense and to analyze the defensibility of local landscapes. The four well-studied Sintashta-Petrovka settlements are chosen for the analysis of the environmental variables and their relation to defensibility. Kamennyi Ambar and Ust’ye 1 received intense excavations, W I N T ER I S CO M I N G

59

geophysics, and geochemical surveys, while Sarym-Sakly and Konoplyanka were surveyed, briefly tested, and explored through the means of geophysics. Notably, local landscapes and environmental history were also studied in all four cases. The current analysis is focused on bringing together such environmental variables as location altitude, visibility, protective features of landscapes, the risk of flood, the speed of wind, and local temperature variation. ANALYSIS OF LOCAL ENVIRONMENT AND LANDSCAPES

Climatic conditions to a great extent determine how local communities live, so before analysis of past behaviors, the climate between 4000 and 3600 cal. bp should be first understood. Stable oxygen isotope data (δ18O) from two speleothems collected from the Kinderlinskaya Cave, located in the southern Urals, document a long-term trend of warming winter temperatures throughout the Holocene (Baker et al. 2017). The comparison of the composite curve of δ18O values with the summed probability of forty-one radiocarbon measurements from Kamennyi Ambar suggests that the δ18O values are slightly lower in the Late Bronze Age, reflecting slightly wetter and cooler climatic conditions (figure 3.2: 1). Additionally, Jonathan Baker et al. (2017, 434) estimated similar to modern average annual winter temperatures of approximately −10.5°C for 4000 cal. bp. Accordingly, steppe pollen archives suggest that the vegetation patterns are comparable with those of the modern era, leading to the conclusion that at the local scale the climatic conditions of the Late Bronze Age were similar to the present, only slightly wetter (Stobbe et al. 2016, 14). The consideration of independent lines of evidence allows application of the present-day observations of average temperatures, wind speed and directions, and seasonal water accumulation to model the general environmental conditions in the past. ANALYSIS OF WIND PATTERNS

Wind is a natural movement of air of any velocity on a large scale. On a global scale, it is caused by the sun’s radiation heating the atmosphere and the rotation of the Earth. In the Northern Hemisphere, the westerlies (antitrades) are the prevailing winds, blowing from the west and southwest toward the east and northeast. The antitrades play a crucial role in creating local environments and climate conditions of the steppes by carrying precipitation and warm air from the tropical zones and causing weather variation by convergence with cold and denser polar winds that blow along the Ural Mountains (American Meteorological Society 2013). 60

I G O R V. C H E C H U S H KOV

At the local scale, the wind is responsible for the everyday life experience of humans and animals by bringing precipitation, damaging shelters, and moving dust and fire. Wind chill is a part of everyday experience, especially during the wintertime. According to the National Climate Data and Information Archive of Canada, how cold the weather is felt by people can be expressed in the wind chill index, which is derived from a formula that includes temperature and wind velocity values. For example, if the outside temperature is −10°C and the wind chill index is −20°C, a person’s face will feel more or less as cold as it would on a calm day when the temperature is −20°C (Government of Canada 2017). In the steppes, the significance of wind and wind chill is especially high due to the absence of natural barriers and shelters—such as forests, caves, or rock canopies—while the region of study is windy (an annual average of 310 days of wind), and the maximum speed reaches stormy 20–25 m/s (Levit 2005). Consequently, the assumption made here is that in the past people chose to live at calmest possible spots of the local landscapes to prevent themselves and their animals from unpleasant or dangerous conditions. Therefore, modeling the wind conditions at the regional scale might provide a plausible explanation for settlement locations. In order to model the local wind intensity, the day-by-day raw data from the local meteorological station at Arkaim were used. The records for five years of observations were derived from the annual reports for 1998, 2003, 2004, 2007, and 2010, allowing representation of annual fluctuations (Kislenko 2004, 2005, 2008, 2011). The gathered data are the direction of the wind in degrees (an average of three observations per day) and speed of wind in meters per second, registered at the height of 10 m above the ground. For this reason, the average temperatures in degrees Celsius were also gathered from the Arkaim data set. Thus, according to the journal of meteorological observations, on January 24, 1998, the air temperature was −31°C, with the wind blowing at 28.8 km/h. According to the wind chill formula, the index was −47°C, making this day the coldest for the entire winter. The observed mean wind speed is 3.71 ± 0.1 m/s (n = 1,733, 95% CL), including quiet days. The observed averages can be classified as a gentle breeze in the Beaufort scale, though the observed maximum of 12.7 m/s (45.7 km/h) is a strong breeze. The direction of the wind changes several times a day. However, the records for five years (N = 1,538) suggest the southwest direction (225°) dominates in the study area (n = 223) (figure 3.2: 2). The average annual value of wind speed (3.71 m/s) and the dominating wind direction (225°) are the input values for calculating wind speed variability W I N T ER I S CO M I N G

61

depending on the local altitude, thus, they help predict where on the landscape the wind drops its speed and where it increases. The wind speed for each 30 m by 30 m cell of the ASTER GDEM (Fujisada et al. 2005) was computed with WindNinja, a specialized software package to model wind fields (Forthofer et al. 2009). The modeling output is a raster map of predicted wind-speed values assigned to every cell of the digital elevation model (DEM), limited by a buffer with a radius of 1 km around a settlement. Further, the cells of each raster are classified into three groups as low, moderate, and high-speed winds to test if the inhabitants of the walled settlements chose the spots on the landscapes with the lowest speed of wind. For the model, moderate winds are defined as within the range of one standard deviation from the mean speed value observed at each location, while the extreme values are either classified as low- or high-speed winds. This method of classification helps to locate the spots of the local terrains where the chances for the speed decrease and increase are higher (figure 3.4: 5). As to local variation of wind speed, the WindNinja models show that that Kamennyi Ambar is situated in an area of low to moderate winds (3.65 to 3.69 m/s). Both, the settlement of Ust’ye 1 and Sarym-Sakly are in the spots of moderate wind with a modeled speed 3.69 m/s with calmer areas adjacent. The low to moderate speed wind of 3.6–3.76 m/s characterizes the setting of Konoplyanka. In other words, all locations provided the Sintashta-Petrovka communities with relatively quiet spots where the average wind speed is 3.69 m/s (low to moderate), which is below the observed regional average of 3.71 m/s (table 3.1). According to the five-years-long meteorological observation at Arkaim, the average winter temperature is −14.2°C, which suggests that the average wind chill index is −21.2°C. At the same time, the lowest mean temperature recorded at Arkaim in January  24, 1998, was −31°C and the highest speed velocity was 8 m/s, yielding wind chill index of −47°C. Combined with the maximum modeled wind speed of 3.9 m/s in the windiest areas around the settlements, the minimum winter temperature of −14.2°C predicts the wind chill index of −42°C. With the absolute modern winter regional minimum of −50°C, the average wind chill index for all four locations is −65°C, while the absolute minimum hits −71.8°C. Such low temperatures make it virtually impossible for humans and animals to survive without shelter, and an attempt to minimize the impact of chilly wind seems to be a reasonable assumption. Thus, in the especially bad winter days, the locations with the lowest possible wind speed are critically important for survival. In sum, the chosen spots are optimal in the given environment to maintain the highest possible temperatures and prevent heat loss in the cold months. 62

I G O R V. C H E C H U S H KOV

T able 3.1. Wind speed within the 3 km radius buffer zones around the settlements Low Wind (m/s)

Moderate Wind (m/s)

High Wind (m/s)

Speed at the Settlement (m/s ± 1σ)

WCI at the Settlement (−14.2°C)

WCI at Windy Locations (−14.2°C)

Kamennyi Ambar

3.6–3.66

3.66–3.74

3.74–3.9

3.68 ± 0.03

−21.3

−21.5

Konoplyanka

3.5–3.65

3.65–3.76

3.76–3.9

3.68 ± 0.08

−21.3

−21.5

Settlement

Sarym-Sakly

3.5–3.66

3.66–3.75

3.75–3.8

3.71 ± 0.05

−21.3

−21.5

Ust’ye 1

3.6–3.66

3.66–3.75

3.75–3.9

3.69 ± 0.03

−21.3

−21.5

Together with the erected wall and densely packed houses, this strategy allowed the inhabitants of the nucleated settlements to protect themselves and their livestock in the harsh winter conditions. ANALYSIS OF WATER ACCUMULATION

The Sintashta-Petrovka settlements were preferably located in the low and calmest possible spots, and even a brief look at the map suggests their location near the water. The steppe rivers slowly flow from, cutting through the rolling hills of the Trans-Ural peneplain and forming their valleys. In no cases, settlements locate on cliffs, but instead on slightly (1–2 m) elevated first terraces. Nowadays, the south boundary of Kamennyi Ambar is only 20 m from the bank, but there is an old river channel located 260 m to the south (figure 3.3: 2). At the site of Konoplyanka, the modern riverbed is located at 300 m east of the settlement wall. However, the hydrological situation in this locality is complex, since the river changed its channels several times during the Holocene. Based on the sediment study, Svetlana Sharapova et al. (2014) concluded that the modern riverbed was formed during the first millennium ad, while in the Sintashta-Petrovka period the river flowed along the old channel, located 50 m west of the settlement wall. At Sarym-Sakly, the modern riverbed is located only in 30 m from the settlement’s boundary but it is quite possible that in the Bronze Age the river flowed at a distance of 300–400 m. The same can be true for Ust’ye 1 as well, where the old river’s bed is located 200 m from the settlement. The modern environmental record shows that while the summer’s river flow is slow, in the flood period the volume increases sharply, and the level rises as much as 2–2.5 m (figure 3.3: 4). However, the rapid rise of water could occur during the summer or fall, due to extreme rainfalls (Levit 2005). Rapid water W I N T ER I S CO M I N G

63

Figure 3.3. Landscapes of the study area. (1) View from a location near Kamennyi Ambar (marked on the satellite image below). The distance between two marked locations is 8 km. The distance to the forest on the horizon is 2–2.5 km. (2) The area near Kamennyi Ambar as appeared on the Corona Satellite image during the winter of 1963. Both, the modern riverbed and old beds are visible on the image, allowing locating of the settlement in relation to the water stream/ (3) A slowly flowing Karagaily-Ayat River in the midsummer. (4) Spring flood of 2018, with water covering the modern road.

flow is a considerable threat for the dirt-wooden architecture of the SintashtaPetrovka settlements, and the assumption formed here is that inhabitants of the settlements preferred to find less risky locations to protect themselves from occasional floods. For example, during the spring flood of 1998, the water table of the Bolshaya Karaganka River rose for 2.5 m (Kislenko 2004), and Arkaim was fully surrounded by water. The ESRI ArcGIS 10.5 Flow Accumulation tool was used to calculate accumulated flow as the weight of all 30 m by 30 m cells (the highest resolution of ASTER DEM) flowing into each downslope cell in the output raster. The output rasters are classified into three categories, which are low (0 inflowing cells), moderate (0 to 100 inflowing cells), and high (more than 100 inflowing cells) risk of flood. Moreover, the modern and old channels of the river are drawn from the satellite image, and then the value of 100 assigned to each cell within the channels. This value added to the rasters of flow accumulation (Mosaic tool) where they intersect with the river to allow representation of its channels. The result of modeling of water flow is models of the possible seasonal channels and the riverbed (figure 3.4: 5). Their comparison with the settlement locations predicts whether the communities were in danger of being flooded with seasonal waters or rise of rivers. Thus, Kamennyi Ambar is located in the relatively flood-safe area with the seasonal channels about 200 m away in all directions, even though right next to the modern river. Since there is the old bed, it seems to be plausible that the river flowed through this channel in the Bronze Age (figure 3.3: 2). Konoplyanka is surrounded by the river’s channels, but the site itself is located on a relatively safe elevated area. If the river indeed flowed through the channel west from the walls, as suggested by Sharapova et al. (2014), the walled settlement could have experienced some danger of flood. However, the map of microtopography suggests that the site was placed on the highest-possible spot—the difference in elevation between the center part of the settlement and the bank of the old channel is about 2 m—exactly enough to protect the settlement from the most severe flood (figure 3.5: 2). Sarym-Sakly occupies a point bar next to a meander bend of the river, while an oxbow and an old bed are located 400 m southward. According to the model, the exact location is at risk of flood, especially due to closeness to the river (figure 3.6: 5). Perhaps, as in other cases, the settlement was located further from the river on the first terrace and the river flown on the channel, which is nowadays marked with an oxbow lake. Finally, the model predicts that Ust’ye 1 was at constant flood risk, though the site itself is located on the most elevated spot (figure 3.7: 2). A field observation of the seasonal flood in W I N T ER I S CO M I N G

65

Figure 3.4. The settlement of Kamennyi Ambar: (1) Topographic model within the 3 km buffer from the settlement; (2) settlement’s digital elevation model and the profile showing the location in relation to the floodplain; (3) visibility model within the 3 km radius domain (cell size is 1 ha); (4) the wind model within the 1 km radius domain (cell size is 30 ma by 30 m); (5) flow accumulation model within the 1 km radius domain (cell size is 30 m by 30 m).

2013 offers an important insight into the location of Ust’ye: while the flood affects floodplain west and north from the site, the habitational area remains only slightly touched by the streaming water, and the area protected by the 2 m high spoil left by archaeologists remains unaffected. It is self-evident that the inhabitants of the nucleated settlements chose to live very close to the river, since up to 3,000 animals could be kept within the walls during the wintertime. Hence, the demand for water was high, with an adult horse or a cow needing about 50 liters of water per day. Every house within the walls had its well, or few of them. However, this would be first used for the human consumption to avoid contamination, so the location of the barns near the river was the simplest solution to water the animals (Anisimov 2009). Nevertheless, the models of flow accumulation suggest that flood could be a considerable threat to the settlements that had to be dealt with. One possible 66

I G O R V. C H E C H U S H KOV

Figure 3.5. The settlement of Konoplyanka: (1) Topographic model within the 3 km buffer from the settlement; (2) settlement’s digital elevation model and the profile showing the location in relation to the floodplain; (3) visibility model within the 3 km radius domain (cell size is 1 ha); (4) the wind model within the 1 km radius domain (cell size is 30 ma by 30 m); (5) flow accumulation model within the 1 km radius domain (cell size is 30 m by 30 m).

explanation is that the walls and ditches that surrounded the settlements were protective measures against seasonal waters. Several facts can support this notion. Even though the climatic conditions during the Bronze Age were like today’s, a relatively more humid climate predominated between approximately 2400 and 1600 cal. bc (Stobbe et al. 2016, 14), which could have increased the annual amount of precipitation. The stone slab facings of the Kamennyi Ambar and Alandskoye walls can be reasonably explained as erosion protection, including protection from the flood. Finally, according to the sediment study, the earliest construction phase at Ust’ye 1 was conducted in humid conditions: the alluvium layer recorded in situ immediately under the rampart and the wall itself contains redeposited sediment particles (figure 3.2: 3). This sedimentation indicates a frequent occurrence of water level rise before and during the settlement existence. W I N T ER I S CO M I N G

67

Figure 3.6. The settlement of Sarym-Sakly: (1) Topographic model within the 3 km buffer from the settlement; (2) settlement’s digital elevation model and the profile showing the location in relation to the floodplain; (3) visibility model within the 3 km radius domain (cell size is 1 ha); (4) the wind model within the 1 km radius domain (cell size is 30 ma by 30 m); (5) flow accumulation model within the 1 km radius domain (cell size is 30 m by 30 m).

ANALYSIS OF VISIBILIT Y AND LOCAL LANDSCAPES AS DEFENSIVE MEASURES

Since the initial excavations of Sintashta, Petrovka 2, and Arkaim, all sites of that type were interpreted as fortified settlements. The argument is that densely packed houses surrounded by earthen walls and ditches were constructed in this way due to the need for defense (Gening et al. 1992, 23–43). However, defensibility is not limited to the walls and ditches, and protective features of local landscapes are often taken into account (Keeley et al. 2007, 83). One of them is visibility that likely played a role in risk management strategies in dangerous social environments, especially for a community at a constant threat of attack. For this reason, in many other examples around the world, people chose to build their fortresses in spots that allow better visibility and protection or chose to clear up the immediate areas around sites (Llobera 2006). 68

I G O R V. C H E C H U S H KOV

[128.104.46.206] Project MUSE (2024-03-01 20:28 GMT) UW-Madison Libraries

Figure 3.7. The settlement of Ust’ye 1: (1) Topographic model within the 3 km buffer from the settlement; (2) settlement’s digital elevation model and the profile showing the location in relation to the floodplain; (3) visibility model within the e km radius domain (cell size is 1 ha); (4) the wind model within the 1 km radius domain (cell size is 30 ma by 30 m); (5) flow accumulation model within the 1 km radius domain (cell size is 30 m by 30 m).

Thus, it can be expected that for the sake of defense, the Sintashta-Petrovka settlements should be preferably located in the spots with better visibility of the surrounding landscapes to warn communities in case of a military threat. Visibility modeling based on the modern landscape is justified by the relative stability of environmental conditions and landscapes during the Holocene. To test this hypothesis, the analysis of viewshed, or area visible from a particular place, was undertaken by running Viewshed 2 Tool of ESRI ArcGIS 10.5. The visibility models are based on a terrain model with a resolution of 100 m by 100 m. The model extends to a distance of 10 km in each direction from the boundaries of 3 km radius buffer zone around each site, allowing modeling of viewsheds up to 10 km from observation points within the buffers. The visibility limit of 10 km was established empirically during the fieldwork: knowing the local landmarks and distances between them allowed determining how W I N T ER I S CO M I N G

69

T able 3.2. Classification of visibility within the 3 km radius buffer zones around the settlements Settlement

Low Visibility (ha)

Moderate Visibility (ha)

High Visibility (ha)

Kamennyi Ambar

716–2,062

2,062–4,491

4,491–7,975

Konoplyanka

510–2,107

2,107–4,889

4,889–10,547

Sarym-Sakly

30–562

562–2,776

2,776–7,078

Ust’ye 1

31–469

469–2,849

2,849–6,771

far it was possible to see from the most elevated points of the local terrain (figure 3.3: 1). The number of visible cells was calculated for each observation point at the center of each 1 ha cell. Next, the number of visible 1 ha cells was assigned to each viewpoint within 3 km radius domains. These numbers were the z-values of resulting rasters that became the maps of local visibility (figure 3.4: 5). The cells of each raster were classified into three groups, of low, moderate, and high visibility, to investigate whether the inhabitants of the walled settlements had chosen optimal spots on the landscapes for observing large parts of the surrounding territory. The cells classified as moderate visibility are within one standard deviation from the mean visibility value, while the values greater than one standard deviation from the mean are classified as either low or high visibility (table 3.2). According to the modeled maps of visibility, Kamennyi Ambar’s location has moderate visibility (2,196 ha, or 6% of the domain with 10 km radius or area of 34,159 ha), while an observer at Konoplyanka could see 2,889 ha (8% of the domain), and 850 ha are visible from the location of Sarym-Sakly (2%). The most dramatic difference was between the local and the average visibility predicted for Ust’ye 1, since only 516 ha are visible from the site (about 1% of the domain). The modeled values are lower than the mean visibility in all four cases (table 3.3), but they lie within one standard deviation of the mean values for the buffer zones. As one might expect, visibility tends to correlate positively with the elevation—the higher the spot, the better visibility (for example, r = 0.6, 43% of the variance is explained for the Kamennyi Ambar buffer zone). Yet the settlements are in relatively lower parts of landscapes, resulting in low total visibility and generally increased accessibility: Kamennyi Ambar is 19.5 m lower than the mean of 3 km radius domain elevation of 310.5 ± 0.5 m at 95% CL, Konoplyanka lies 13.4 m below the mean of 337.4 ± 0.5 at 95% CL, Ust’ye 1 locates 10.4 m lower than the average 286.9 ± 0.5 m at 95% CL, and Sarym-Sakly lies 17 m lower of than the mean altitude of 383.8 ± 0.2 m at 95% CL. 70

I G O R V. C H E C H U S H KOV

T able 3.3. The visibility values of the settlements’ buffer zones in the Karagaily-Ayat valley (3 km radius from each settlement)

Min (ha)

Max (ha)

Kamennyi Ambar

716

Konoplyanka

Mean (ha)

95.0% CL Range

Visibility from the Site (ha)

Difference between Visibility at the Site and the Mean Value (ha)

σ

ER

7,975 3,277

1,222

23

3,232– 3,323

2,196

−1,081

510

10,547 3,522

1,379

26

3,472– 3,574

2,899

−623

Sarym-Sakly

30

7,078 1,674

1,106

20

1,633– 1,744

850

−824

Ust’ye 1

24

7,137 1,729

1,353

25

1,679– 1,778

516

−1,213

Settlement

If the inhabitants of the walled settlements wished to increase their ability to see around, they could have built up elevated platforms. Archaeological records do not provide data to support this hypothesis. However, the tops of the surrounding walls could have been used for this purpose. The walls could have reached the height of 3 m, which together with the average person’s height of 1.6 m would allow elevating an observer to a height of about 4.5 m. This parameter was applied to the loci of the settlements to remodel the viewsheds and test again whether visibility played a role in settlement location. With this additional height, visibility increased, but in no case did it increase significantly: observing from the wall of Kamennyi Ambar, a watcher could see 2,196 ha of the surrounding area, which is still lower than the mean observing area for the buffer zone (3,277 ha). As a result, a potential enemy approaching Kamennyi Ambar would be spotted only at a distance of 1,000 m north. 1,500 m west, and 1,100 m east. Combined with a very flat landscape and absence of natural obstacles (except the rivers), these distances could be covered in a matter of minutes, providing the advantage of a surprise attack. Ethnographic sources suggest that night raids are the usual form of attack in small-scale societies (Arkush et al. 2005, 8), meaning that an enemy could use the disadvantage of Sintashta-Petrovka wooden constructions to set them on fire in the dark. Finally, a survey of those localities with the highest visibility did not reveal any cultural remains of the Bronze Age (figure 3.3: 1). To summarize, the models indicate that visibility did not influence settlement locations. In all cases, the sites were located in lower spots of the landscape, close to the river, which is typical for other Bronze Age steppe villages. For example, W I N T ER I S CO M I N G

71

the peaks of hills in the Arkaim valley have altitudes of 350 masl and 330 masl, while the site itself is located at the altitude of 319 masl. If the social environment of the Late Bronze Age was hostile and the communities were under threat of attack, the settlements would be in locations with better visibility and decreased accessibility—on the peaks of surrounding hills, as predicted by the models. DISCUSSION

The above analysis supports Nikolai Anisimov’s (2009) notion that the Sintashta “fortified towns” were well-thought-out artificial ecological niches, designed to protect communities of herders from the harsh environmental conditions, including extreme temperatures, heavy wind, and seasonal floods. A more traditional view, though, interprets the surrounding embankments and ditches as defensive measures. The ultimate defensive measure is a fort or a system of fortifications, that usually consists of ditches, walls, entrances and gates, towers, fortified households, and other secondary defensive features (Arkush 2011, 62). Ditches are additional fortification features, since the earthen material for walls was commonly obtained through excavation. A ditch is usually located outside of a fort following an outer wall, becoming the first line of defense that increases height and inaccessibility of an embankment. Typically, the depth and width of a ditch have the same dimensions and constructive features all along the length of a wall, though some segments could be reinforced with optional defensive measures. Statistical analysis indicates that an average ditch width (at Ust’ye and Kamennyi Ambar, Sintashta) is 3.6 ± 0.4 m (at 95% CL). However, the measurements range from 1 m to 8.8 m, resulting in a very large standard deviation (σ = 1.4, n = 60). Notably, significant differences in widths can be seen along the ditch of the same settlement, as in the cases of Sintashta and Kamennyi Ambar. Further, the excavated ditches are rarely deeper than 2 m (x̄ = 1.6 ± 0.2 m at 95% CL) and sometimes as shallow as 1.2 m, and these inconsistences in the depths would undermine the military function. Moreover, none of the studied cases are characterized by V-shaped profiles, considered to be the most effective for defense, as it is the most difficult to cross (Keeley et al. 2007, 60). On the contrary, the profiles of the excavated moats are usually trapezoidal or rounded, with ledges on both sides. At Kamennyi Ambar, the deepest sections showed evidence of still water, suggesting that the ditch functioned as drainage (Berseneva et al. 2015, 8). A considerable exception is the ditch at the settlement of Alandskoe, which exceeds 3 m in both dimensions at least in 72

I G O R V. C H E C H U S H KOV

some segments (Zdanovich 2011, 51). Perhaps, variations in depths and design may be due to the difference in specific geographic conditions, with this site located on the spit of land between two rivers with a high risk of flooding, and slab facing of the walls would help them resist flood waters especially well. In sum, the studied cases of walled villages show a high degree of heterogeneity and lead to the conclusion that no universal design for a ditch existed. Defensive walls, whose ultimate purpose is the protection of the inhabitants of a settlement, are usually complemented by natural barriers such as cliffs, deep chasms, rivers, and swamps, when available (Keeley et al. 2007, 57–80). Sintashta-Petrovka sites, however, were placed in locations without such barriers. High places are advantageous for reasons of better visibility and longer effective range of projectile weapons. However, none of the settlements is located in such positions, what would condition the total enclosure of the settlement (Zdanovich and Batanina 2007). The relatively elevated position of Konoplyanka allows for flood protection but unlikely increases defensibility, since the elevation of the terrain does not exceed 2 m and the increase in the height is gradual (figure 3.5: 2). In some cases, multiple lines of walls could be constructed to enhance a fort’s defensive capacity, combining an outer wall with an inner wall that serves to protect more vulnerable or important parts of a community (Arkush 2011, 68). Indeed, just such patterns found at Sintashta and Arkaim make archaeologists think of them as effective forts (Anthony 2007, 393–395). However, the rest of the walled villages have only one line of walls each. Other archaeologists have noted that if walls do not exceed 2–3 m in height, they could be barriers for livestock to prevent straying and to protect them from nonhuman predators (Keeley et al. 2007, 81). This idea fits with the estimated height of the SintashtaPetrovka walls: 3–3.5 m for Sintashta, 1.7–2.6 m for Ust’ye, 2–3 m for SarymSakly, and 2–3.2 m for Kamennyi Ambar (Chechushkov et al. 2018). Combined with the ditch, though, the wall heights would be 3–4 m. The plausible explanation is that the walls were erected following ditch constructions that would leave a large volume of excavated dirt, which was used to construct additional measures against seasonal floods and as barriers to block chilly winter wind. The principal investigators had created detailed artistic reconstructions of the enclosures for Sintashta and Arkaim, which were subsequently borrowed by the authors of the secondary literature on the Eurasian Bronze Age (Anthony 2007). In these reconstructions, a reader would have a view of the walled settlements as developed strongholds, comparable to Early Medieval castles. However, a detailed examination of the initial Sintashta field drawings does not support this view. First, bastions are depicted, but the only W I N T ER I S CO M I N G

73

archaeological evidence for their existence is the irregular outline of the inner boundary of the ditch (Gening et al. 1992, 29). However, to be effective, such bastions must be built in a regular pattern, with overlapping shooting lines. Otherwise, their defensive value drops sharply. On the plans, we see the irregular outlines of the ditch, which can be reasonably explained as the result of taphonomy. Second, the authors reconstruct buttresses that were used as observation towers and rifle niches (Gening et al. 1992, 32). The general plan of the excavation shows that the grooves under the so-called foundations of buttresses cut through the body of the main wall, and do not adhere to it, which contradicts the reconstruction drawing and the function the wall implies for this architectural element. Thus, the existence of such buttresses is doubtful, and the unevenly distributed grooves can be explained as drains. Third, questions arise about the reconstruction of the Sintashta gates. It remains unclear why the entrances had such distinctive designs: complex structures of moats were recorded at the southern entrance to the settlement facing the river, and nothing comparable was found on the northern side, which would have been the first point to be attacked. The southern entrance, most likely, was regularly flooded, so there was a need to drain the rising water into the moat and then discharge it with the help of drainage moats. In sum, the reconstruction of the Sintashta fortified settlement seems to overcomplicate the real picture and cannot be used as bases for discussing the function of the fortifications. To summarize, the fortification features of the settlements were not universal even at the same site. Walls were not strengthened by bastions or towers, and the ditches did not represent a very difficult obstacle for attacking infantry. Also, the settlements are in the lower parts of the terrain, allowing a potential enemy to covertly approach them, especially in the dark, when the settlement was illuminated by bonfires. At the same time, such features of the walled settlements as heat loss prevention and flood protection were vital to maintaining the population of domestic animals. The Sintashta-Petrovka herd consisted of approximately 52 percent cows, 42 percent sheep, and 6 percent horses (Stobbe et al. 2016), whose calving season is the late winter and early spring—the most volatile seasons regarding temperature fluctuations and flood risk. Preventing of newborns loss would be the most critical annual problem to solve for local communities, while the threat of attack could not be constant. CONCLUSION

The Sintashta-Petrovka case study indicates that at least in some cases the strategy of settlement design and location can be explained as creating artificial 74

I G O R V. C H E C H U S H KOV

ecological niches that provide communities with safety and economic stability in the harsh climate conditions. Even though military affairs, indeed, could have had a place in the Late Bronze Age, there are not many facts supporting intense warfare—human remains do not demonstrate conflict-related trauma, and there is no evidence for assaults on the settlements. In contrast, in the following Alakul’ period (3700–3500 bp), settlements are not reinforced but decapitated human remains and abandoned weapons are vocal evidence for military catastrophes (Chemyakin 2015). The climate change in this period toward aridization explains well the decrease in the construction of moats and walls (Baker et al. 2017). Presumably, if intergroup military conflict took place during the Sintashta-Petrovka period, it was limited in scale elite-style warfare (chariots and mace-heads support this notion) and did not influence the local communities. Robert Carneiro’s (1998) question of why individuals choose to integrate and give up their independence can be answered as some combination of two necessities: to persist as a larger community in the ecological niche of the newly settled region, and to protect herds. There is general agreement among researchers that the Sintashta phenomenon had no local roots and originated with the migration of pastoral communities from eastern Europe to the marginal area of the southern Urals. This process forced families to stay together and fueled the necessity of the walled villages for ensuring the reproduction of herds in the extreme climatic conditions of the southern Urals, which are colder and dryer than the eastern Black Sea region from which the Sintashta populations are thought to have migrated (Anthony 2007). At the same time, the herds needed protection from the cold, as well as from animal and human predators. Those who were talented at managing the construction of closely packed villages surrounded by ditches and walls and at protecting people and livestock, and who otherwise served the community in the newly colonized zone became the most prominent members of society. These people, who usually are associated with chariots, formed the core of the Sintashta-Petrovka communities but were not able to accumulate much personal wealth or create significantly different lifestyles, acquiring instead high social prestige. REFERENCES

American Meteorological Society. 2013. Glossary of Meteorology. AMS. https://www .ametsoc.org/index.cfm/ams/publications/glossary-of-meteorology/.

W I N T ER I S CO M I N G

75

Anisimov, Nikolai P. 2009. “Arkaim strana kard: Empirica prostranstva zaural’skoyi sredy” [Arkaim and the country of barns: Environments of the Trans-Urals]. Gradostroitel’stvo (2): 16–21. Anthony, David W. 2007. The Horse, the Wheel, and Language: How Bronze-Age Riders from the Eurasian Steppes Shaped the Modern World. Princeton, NJ: Princeton University Press. Anthony, David W. 2009. “The Sintashta Genesis: The Roles of Climate Change, Warfare and Long-Distance Trade.” In Social Complexity in Prehistoric Eurasia: Monuments, Metals, and Mobility, edited by Bryan K. Hanks and Katheryn M. Linduff, 47–73. Cambridge: Cambridge University Press. Arkush, Elizabeth, Charles Stanish, Christine A. Hastorf, Axel E. Nielsen, Theresa Topic, and John W. Verano. 2005. “Interpreting Conflict in the Ancient Andes: Implications for the Archaeology of Warfare.” Current Anthropology 46 (1): 3–28. Arkush, Elizabeth N. 2011. Hillforts of the Ancient Andes. Gainesville: University Press of Florida. Baker, Jonathan L., Matthew S. Lachniet, Olga Chervyatsova, Yemane Asmerom, and Victor J. Polyak. 2017. “Holocene Warming in Western Continental Eurasia Driven by Glacial Retreat and Greenhouse Forcing.” Nature Geoscience 10 (6): 430–436. Berseneva, Nataliya A., Andrej V. Epimakhov, Vladislav V. Noskevich, and Nataliya V. Fedorova. 2015. “Vozmozhnosti sinteza geofizicheskoy i arkheologicheskoy informatsii pri interpretatsii rezul’tatov raskopok (na primere poseleniya bronzovogo veka Kamennyy Ambar)” [Towards synthesis of geophysical and archaeological information in interpreting the results of excavations (the example of the Bronze Age settlement of Kamenny Ambar)]. Vestnik arkheologii, antropologii i etnografii [Bulletin of Archaeology, Anthropology and Ethnology] 1 (28): 4–14. Carneiro, Robert L. 1998. “What Happened at Flashpoint? Conjectures on Chiefdom Formation at the Very Moment of Conception.” In Chiefdoms and Chieftaincy in the Americas, edited by Elsa M. Redman, 18–42. Gainesville: University Press of Florida. Chechushkov, Igor V. 2018. “Bronze Age Human Communities in the Southern Urals Steppe: Sintashta-Petrovka Social and Subsistence Organization.” PhD diss., University of Pittsburgh. Chechushkov, Igor V., and Andrei V. Epimakhov. 2018. “Eurasian Steppe Chariots and Social Complexity during the Bronze Age.” Journal of World Prehistory 31 (4): 435–483. Chechushkov, Igor V., Artyom S. Yakimov, Olga P. Bachura, Yan Chuen Ng, and Evgeniya N. Goncharova. 2018. “Social Organization of the Sintashta-Petrovka Groups of the Late Bronze Age and a Cause for Origin of Social Elites (Based on Materials of the Settlement of Kamenny Ambar).” Stratum plus (2): 149–166.

76

I G O R V. C H E C H U S H KOV

Chemyakin, Yurii P. 2015. “Sledy voyennykh konfliktov na alakul’skikh poseleniyakh” [The evidence for military affairs on Alakul’ settlements]. Etnicheskiye vzaimodeystviya na Yuzhnom Urale. VI Vserossiyskaya nauchnaya konferentsiya, Chelyabinsk, Russia: Chelyabinsk, CGKM. Earle, Timothy K. 1997. How Chiefs Come to Power: The Political Economy in Prehistory. Stanford, CA: Stanford University Press. Earle, Timothy K. 2011. “Chiefs, Chieftaincies, Chiefdoms, and Chiefly Confederacies: Power in the Evolution of Political Systems.” Social Evolution and History 10 (1): 27–54. Forthofer, Jason, Kyle Shannon, and Bret Butler. 2009. “Simulating Diurnally Driven Slope Winds with WindNinja.” Proceedings of the 8th Symposium on Fire and Forest Meteorological Society, Missoula, MT. Fujisada, H., G. B. Bailey, Glen G. Kelly, S. Hara, and M. J. Abrams. 2005. “ASTER DEM Performance.” IEEE Transactions on Geoscience and Remote Sensing 43 (12): 2707–2713. https://doi.org/10.1109/TGRS.2005.847924. Gening, Vladimir F., Gennadiy B. Zdanovich, and Vladimir V. Gening. 1992. Sintashta: arkheologicheskie pamiatniki ariiskikh plemen Uralo-Kazakhstanskikh stepei [Sintashta: The archaeological evidence of Aryans in the Ural-Kazakhstan Steppes]. Chelyabinsk, Russia: Uzhno-Ural’skoe knizhnoe izd-vo. Government of Canada. 2017. “Canadian Climate Normals 1971–2000.” http:// climate.weather.gc.ca/climate_normals/index_e.html. Judd, Margaret A., Jessica L. Walker, Alicia Ventresca Miller, Dmitry Razhev, Andrey V. Epimakhov, and Bryan K Hanks. 2018. “Life in the Fast Lane: Settled Pastoralism in the Central Eurasian Steppe during the Middle Bronze Age.” American Journal of Human Biology 30 (4): e23129. Keeley, Lawrence H., Marisa Fontana, and Russell Quick. 2007. “Baffles and Bastions: The Universal Features of Fortifications.” Journal of Archaeological Research 15 (1): 55–95. https://doi.org/10.1007/s10814-006-9009-0. Kiseleva, Dariya V., Ludmila N. Koryakova, Svetlana V. Sharapova, Pavel A. Kosintsev, Mariya V. Zaytseva, Mariya V. Streletskaya, Nadezda V. Cherednichenko, A. K. Fokina, and Evgeniy S. Shagalov. 2017. “Opredeleniye mikroelementnogo i izotopnogo sostava Sr v obraztsakh kostnoy i zubnoy tkani cheloveka i zhivotnykh iz arkheologicheskikh pamyatnikov Yuzhnogo Urala epokhi bronzy” [Determination of the microelement and isotopic composition of Sr in bone and dental tissue samples of humans and animals from the Bronze Age archaeological sites in the Southern Urals]. In Geoarchaeology and Archaeological Mineralogy—2017: Proceedings of IV Scientific Youth School, 41–45. Miass, Russia: Institute of Mineralogy UB RAS.

W I N T ER I S CO M I N G

77

Kislenko, Aleksandr I. 2004. Letopis’ prirody 1998–2003 [Nature’s observations 1998–2003]. Chelyabinsk, Russia: Scientific Archive of Arkaim Reserve. Kislenko, Aleksandr I. 2005. Letopis’ prirody 2004 [Nature’s observations 2004]. Chelyabinsk, Russia: The Scientific Archive of Arkaim Reserve. Kislenko, Aleksandr I. 2008. Letopis’ prirody 2007 [Nature’s observations 2007]. Chelyabinsk, Russia: Scientific Archive of Arkaim Reserve. Kislenko, Aleksandr I. 2011. Letopis’ prirody 2010 [Nature’s observations 2010]. Chelyabinsk, Russia: Scientific Archive of Arkaim Reserve. Koryakova, Ludmila N. and Andrej V. Epimakhov. 2007. The Urals and Western Siberia in the Bronze and Iron Ages. Cambridge: Cambridge University Press Krause, Rüdiger, and Ludmila N. Koryakova. 2013. Multidisciplinary Investigations of the Bronze Age Settlements in the Southern Trans-Urals (Russia). Frankfurt am Main: Habelt. Kristiansen, Kristian, and Michael Rowlands. 2005. Social Transformations in Archaeology: Global and Local Perspectives. Abingdon: Routledge. Levit, Aleksandr I. 2005. Yuzhnyy Ural: Geografiya, ekologiya i prirodopol’zovanie [The Southern Urals: Geography, ecology, and environment management]. Chelyabinsk, Russia: Yuzhno-Ural’skoe knizhnoe izdatel’stvo. Llobera, Marcos. 2006. “What You See Is What You Get? Visualscapes, Visual Genesis and Hierarchy.” In Digital Archaeology: Bringing Method and Theory, edited by Thomas L. Evans and Patrick Daly, 148–168. London: Routledge. Sharapova, Svetlana V., Rüdiger Krause, Ivan V. Molchanov, Astrid Stobbe, and Nikolay V. Soldatkin. 2014. “Mezhdistsiplinarnyye issledovaniya poseleniya Konoplyanka v Yuzhnom Zaural’ye: Predvaritel’nyye rezul’taty” [Multidisciplinary investigations of Konoplyanka Settlement in the Southern Trans-Urals: Preliminary results]. Vestnik Novosibirskogo Gosudarstvennogo Universiteta [Bulletin of Novosibirsk State University] 13 (3): 101–109. Stobbe, Astrid, Maren Gumnior, Lisa Rühl, and Heike Schneider. 2016. “Bronze Age Human–Landscape Interactions in the Southern Transural Steppe, Russia: Evidence from High-Resolution Palaeobotanical Studies.” Holocene 26 (10): 1–19. Zdanovich, Gennadiy B. 2011. “«Ukrepleniya», «krepost’» v arkheologii sintashtinskoarkaimskoy kul’tury i v poeticheskoy traditsii «Rigvedy»” [«Fortifications», «fortress» in the archeology of the Sintashta-Arkaim culture and in the poetic tradition of the «Rig-veda»]. Voprosy arkheologii Urala: Sbornik nauchnykh trudov (26): 40–63. Zdanovich, Gennadiy B., and Iya M. Batanina. 2007. Arkaim—Strana gorodov: Prostranstvo i obrazy [Arkaim and the country of towns: Spatial dimension and images]. Chelyabinsk, Russia: Krokus.

78

I G O R V. C H E C H U S H KOV

4 The construction of fortification systems is a community affair: it requires the aggregate labor of many families, and when ready, the fortifications provide protection for the whole group. For archaeologists, they serve as an indicator of threat(s) of violence from other communities but also as an approximation of otherness as perceived and defined in the past. This chapter aims to explore the nature of violence in Central China, Henan Province, and how that violence relates to increasing social and political complexity, which, in turn, can help to characterize the nature of communities within the study area. This chapter will study how warfare, or group-togroup violence, relates to the regional landscape of the Longshan archaeological culture (5000–4000 bp). This study will address the particularity of locations and landscapes of the fortified sites in comparison to the other settlements in the region. I analyze environmental and sociopolitical factors, at regional and local scales, as potential catalysts of war. Moreover, by exploring the uniqueness of the fortified sites within this region, I want to contribute to the search of generalizable principles of the warfare’s emergence in other regions. This analysis will provide better understanding of how warfare is integrated into social and political organizations by asking the following questions: How did violent conflict shape the development of communities at local, regional, and supraregional scales? Why do we see a massive investment in fortifications in some areas

Communities, Violence, and Fortification A Study of Longshan Landscapes

James T. Williams

https://doi.org/10.5876/9781646422111.c004

79

but not in others? Where can we find the sources of threat and how it affected settlement location? POLITICS, WAR, AND GROUP VIOLENCE

In Raymond Kelly (2000), Julie Solometo (2006), and D. Webster (2000), the authors define warfare by somewhat-related criteria. Warfare not only denotes violence at the group level or scale (Solometo 2006, 26) but also has implications in power relations (Webster 2000, 72). Thorpe additionally notes the political nature of the violence related to warfare (Thorpe 2003, 146). In part, the analysis presented in this chapter will help to understand the political aspects of warfare as they relate to indications of violence. Construction labor, location, and placement of fortifications are evidence of critical decisions for maintaining and establishing geographic power; studying them helps identify the political motivation for warfare if it exists. Functionally, if the elite is the prime mover behind the construction of the fortification, every time an attack is thwarted by the defensive structure that elite is able to reinforce his or her status. The fortification serves as the embodiment of the protection commoners enjoy. Symbols of war used to reinforce power dynamics have been examined in Denmark (Earle 1997), Italy (Robb 1997), and other parts of Europe (Hanks 2008, 268). The proliferation of bronze artifacts in the context of elites or aggrandizers further emphasizes their symbolic role in reinforcing the power dynamics. In the Chinese Erligang period (3600–3400 bp), we see some of the earliest bronzes are “incorporated with non local styling.” which “reinforces the transmission of symbols . . . which results in reinforcement of elites” (Barnes 1999, 123). Furthermore, the earliest bronze vessel from the Erlitou period (4000–3600 bp) was found in context with jade ceremonial axes, further solidifying the connection between violence, symbols of prestige, and novel technology (Underhill and Habu 2006, 134). Understood symbolically, fortification can be understood in the same manner as artifacts. Fortification occupies a location on the landscape that indicates ownership and reinforces identities and inequality (Earle 2001). Competition over landscapes for political gain is what one would expect in a peer-polity conflict situation. If the construction of fortification is primarily the result of peer-polity conflict or competition, either symbolic or violent or both (Liu 2004), then we should see a strong relationship between political community boundaries and the locations of fortifications. At minimum, the presence of fortifications remains in this region suggest the existence of violent interaction in the past, and the motivations for 80

JA M ES T. W I LLI A M S

violence can be understood by analyzing the kind of boundaries and variables to which these fortifications aligned. For example, if the fortification was built to defend against competing polities there should be a correlation between the boundaries of the polities and the locations of fortification. BOUNDARIES

The study of boundaries is a key component to the study of social complexity and its relationship to warfare. How frontiers and boundaries form as a result of conflict and competition is important in understanding settlement patterning. The potential causal agents for population nucleation are as numerous as the potential causal agents for warfare. Warfare can cause population nucleation and form the creation of buffer zones or boundaries (Hill and Wileman 2002; LeBlanc 2006). This increased nucleation causes communities to become more delimited. Nucleation can be then measured through analyses such as k-means clustering, nearest neighbor statistics, and intervisibility (Arkush 2009). Political and territorial boundaries are an element of the landscape that have been explored through the use of Thiessen polygons (Banning 2002, 78; Renfrew 1976). The relationship between Longshan polities, the boundaries of these polities, and locations of fortification has not been investigated but rather assumed (Liu 2004; Liu and Chen 2006, 157). CENTRAL CHINA’S TRAJECTORY

The periods and archaeological cultures addressed in this chapter are the Yangshao, Longshan, Qujialing, and Dawenkou. the Yangshao and Dawenkou are the earlier of the two archaeological cultures, dating from 7000 to 5000 bp and 6300 to 4600 bp respectively, both overlapping with the Longshan period of 5000–4000 bp and the Qujialing of 5500–4600 bp. The Yangshao period is characterized as being relatively egalitarian, with little evidence for status differentiation in burial patterning or in household architectural evidence in the first millennium of the period. During the second half of the period, incipient social inequality seems to appear. It is in this period that the first evidence for fortification is identified. The sites of Banpo, Jiangzhai, and Xishan all have a moat and Xishan includes a wall. Xishan, which is associated with the late Yangshao period, as well as Banpo and Jiangzhai, are seen as the precursors to the type of fortification synonymous with the Longshan archaeological culture and period (Wiesheu 1997, 103). CO M M U N I T I ES, V I O LEN C E , A N D F O RT I F I CAT I O N

81

The Dawenkou period predates the Longshan, but as with its overlap the Yangshao, there it overlaps into the Longshan period. Dawenkou pottery has been found in the same context as Longshan-style pottery. This archaeological culture is concentrated in areas east of the study area, especially Shandong Province. The Dawenkou archaeological culture has evidence of inegalitarian social organization. Evidence from burial data indicates that during this period, there are status differentiation as well as gendering of economic activities (Lui 2004). South of the Longshan region the Qujialing overlaps with the Longshan chronologically and spatially. The Qujialing archaeological culture and period are characterized by particular pottery styles and characteristically designed spindle whirls (Chang 1986, 226). The Longshan period is often mentioned when scholars address the emergence of warfare (Barnes 1999, 117; Chang 1986). This period has been described as having indicators of violent conflict between competing polities (Shao 2000), or conflict between nonlocal archaeological cultures (Liu 2004). Other scholars cite the Longshan as a period characterized by increasing levels of violent conflict but do not indicate the scale at which this takes place. Liu and Chen indicate that conflict may have been over resources (Liu and Chen 2006; Underhill and Habu 2006, 132). Walburga Wiesheu notes the increase in violence but poses the question of whether this is a result of “internal or external conflict.” The indicators of warfare that scholars have cited for this period include an increase in formal weaponry and violent trauma. In addition to this evidence, the most important line of evidence for the purposes of this chapter is the increase in the construction of fortifications to which we will return in the “The Survey Area and Data Source” section of this chapter. NONFORTIFICATION EVIDENCE RELATED TO WARFARE

Claudio Cioffi-Revilla (2000) cites six lines of evidence that can be used to indicate warfare: forensic, locational, structural, artifactual, iconographic, and epigraphic lines of evidence have been used as a means of understanding the existence and nature of warfare. For this study, I will concentrate on the structural and locational lines of evidence that can be examined through the available data but briefly deal with epigraphic, artifactual, and forensic evidence. Walled sites have been noted as particularly important in the study of increasing social complexity in China, as they represent a milestone in urban planning. This is especially true of square or rectangle fortification as it represents the prototype of the early dynastic city layouts. The character cheng (城), meaning “city,” was first used in its current form in the second century 82

JA M ES T. W I LLI A M S

bc. Prior to this, the Shang (3550–2996 bp) oracle character for cheng clearly indicates a wall, and the left portion of the character has been translated as guo, meaning city wall (figure 4.1). The definition of city, urbanization, and in some regards civilization implies some notion of fortification. Cheng also uses the radical tu Figure 4.1. Chinese character (±) meaning earth, which is a reference to the from the Shang Dynasty. tamped earth construction used to build a fortification in China from the Longshan period until the nineteenth century. Many of the fortifications from historic dynasties were built with tamped earth substructures and brick facades. Traditionally, the other lines of evidence have been included in an interpretation of warfare during the Longshan period. Liu cites increasing numbers of projectile points and human sacrifice as forensic and artifactual evidence of warfare. The distribution of the artifactual evidence has not been systematically collected and would only serve to indicate where modern archaeological excavations have been focused, not the distribution of violence in the past. The sample size of the forensic evidence is not large enough to make accurate conclusions about the Longshan population as a whole, and the published data includes three sites outside of the survey area. Furthermore, the lack of systematic sampling makes the data unsuitable for a study of regional social and political complexity. It does, however, serve as useful evidence about the variety of violent behaviors during the Longshan period. The settlement of Wangchenggang provides good evidence for violent behaviors, including decapitated individuals interred in a well, and sacrificed individuals possibly related to a groundbreaking ceremony of the wall, as well as three examples of scalping (Chang 1986, 271). This bioarchaeological evidence from this settlement indicates the taking of trophies, which could be used to perpetuate the memory of violence in the same way that the permanence of fortification can also create persistent memory. The main agricultural products of the Longshan are millet and foxtail millet (Underhill 1997), which both require well-drained soils to prosper (Oelke et al. 2000). Cattle, sheep, and goats are the main domesticates in the Longshan archaeological period (Underhill 1997). These would have required grasslands for grazing in the summer and are generally foddered with agricultural products for the winter. Domesticated animals were presumably utilized for their secondary products, since much of the protein diet is from hunted animals especially deer (Liu 2004). CO M M U N I T I ES, V I O LEN C E , A N D F O RT I F I CAT I O N

83

THE SURVEY AREA AND DATA SOURCE

The study area, located in central Henan Province, consists of a rectangle approximately 300 km east-west by 240 km north-south (figure 4.2). The data are originally from the National Cultural Resource Inventory. The data may have biases toward larger sites. and sites smaller than 1 ha may have been potentially overlooked. For the purposes of this study, which focuses on the locations and contexts of centers and fortified sites, the bias should not affect the results or interpretation. The Yellow River forms an approximate northern boundary to the study area and the Song Mountain range the western boundary. This area is no stranger to warfare and violence; for example, the site of the famous Shaolin Temple is at the foot of Song Mountain. The earliest evidence of defensive structures within the study area are dated to a much earlier epoch. Xishan is a walled settlement that dates to the late Yangshao period. The settlement is estimated at 25 ha of which 6,385 m2 has been excavated. The wall is partially preserved and is circular with two gates within the preserved section; the northern gate incorporates a structure that could be used as a guardhouse or tower. The construction of the fortification would have taken approximately 3,530 person-hours according to estimates from Erasmus (1965). The study area was chosen due to several factors. First, the area has settlements that represent a number of periods and archaeological cultures. Second, relatively high-resolution data for soils are available. Last, the study requires fairly comprehensive survey data with relatively high spatial resolution. The raster files of the soils data have a resolution of 1 km cells. The settlement raster data result in a resolution of 500 m cells. The data from Liu (2004) are primarily for settlements; through mathematical manipulation, the settlement data can be transformed into a landscape of economic resources, local and foreign boundaries that can be examined as they change over time. The four fortified sites in the study area vary in size and architectural composition. Some have been excavated, while others are only understood through surface remains. The architecture of the site can be used to understand in part the nature of the violent conflict. Guchengzhai is the largest of the four fortifications in the study area, with a 17 ha enclosure. Within this area is a palatial compound with fortifications for further protection. The exterior walls do not have preserved gatehouses or baffle gates, and the western wall may have been destroyed due to the Zhen River changing course. The walls exterior are impressive and could have supported individuals with defending with projectiles atop. Using estimates from Charles Erasmus (1965), we can theorize that the construction of the exterior walls would have taken approximately 600,000 person-hours to construct. The 84

JA M ES T. W I LLI A M S

Figure 4.2. Location of the research zone.

palace incorporates additional defensive architecture, what Lawrence Keeley characterizes as baffle gates, either labyrinthine or bent axis (Keeley 2007, 63). The architecture of the site indicates long-term investment in fortification, and the perception of violent threat against the central building is prevalent. Pingliangtai is a fortified village site measuring approximately 5 ha and 3.5 ha of the site is contained within a tamped earth enclosure. The walls would have taken 37,000 person-hours to construct. The main gate to the site includes two guardhouses on either side of the opening. Keeley (2007) characterizes this type of fortification as a balance between “military defense, peaceful passage, and regulation of traffic through heavily traveled gates.” The size and permanence of the architecture again indicate community effort in its construction. The site of Haojiatai is a walled site, enclosing 3.4 ha. The fortification is approximately 250 m on a side and 3 m high, requiring the effort of 10,000 person-hours. This site is the lowest of Longshan fortifications in this study, but the entire site is on a 2 m platform, raising the estimates to 225,000 person-hours. While the site report gives no indication of the type of gate or bafflement, it does mention that the fortification is similar to those found at Pingliangtai and Wangchenggang (HENAN 1992). CO M M U N I T I ES, V I O LEN C E , A N D F O RT I F I CAT I O N

85

Wangchenggang is estimated at 1 ha in size, but there has been significant damage to the site due to fluvial erosion, so the site may actually include a second enclosure that would have added an additional hectare to the site as well as 400 more m of tamped earth fortification. As it exists today, there are 290 m of the wall remaining. The existing wall would have taken 62,600 personhours to construct. If we consider the additional destroyed fortification, the figure rises to 149,000 person-hours. This site has evidence of human sacrifice, decapitation as discussed earlier. The most interesting aspect about this information is evidence of a well. This presence indicates that the settlement could withstand long-term sieges and attacks, and it provides more evidence for long-term investment and planning in the settlement and the landscape. The measures of construction hours in this section serve two purposes. First, they show the community investment in the landscape. Second, they serve to show that the fortifications in the Longshan period are much more substantial features than in the previous period. The change from the Yangshao period to the Longshan is not a slight increase in size, but one that is one or two orders of magnitude e larger. While the architecture of the fortification speaks to the community that completed the construction and the nature of the attack the fortifications are constructed to defend, the main focus of this study will be to examine the landscape they inhabit as a means of testing the proposed hypotheses and assess what communities were in conflict with each other. Data Interpolation, Manipulation, and Classification

Site maps (figure 4.3 and figure 4.4) were produced by digitizing the maps from the Chinese Neolithic (Liu 2004, 6.16 and 6.18). These maps were subdivided to create several layers that could be examined independently and recombined when necessary. These layers are divided by chronological period, archaeological culture, and size. Size is indicated by Liu as major, minor, and small. Typological classifications are fortified and unfortified. The maps of major and minor centers were combined to create a “Longshan centers” map. The “Longshan centers” (figure 4.3) and the “Longshan walled” maps are the two maps used to establish if there are differences between settlements that are fortified and those that are not. The Food and Agricultural Organization (FAO) of the UN has compiled and freely distributed GIS soils data. Soils maps from the FAO were the base for understanding the relationship between the economic potential of an area and its potential to be related to the erection of fortification. There are thirtyfour soil types in the study area according to the FAO, and these include six 86

JA M ES T. W I LLI A M S

Figure 4.3. Longshan sites in the central Henan cluster. Modified from Liu (2004, 6.16).

that for various reasons are classified as no soil, including bodies of water and urban areas. These are treated as missing data, and a data mask was deployed so that the missing soils do not affect the regression analyses. The remaining twenty-eight soil types were combined by using both the CLUSTER module in IDRISI and some manual reclassification. This combination of soils was performed for two main reasons. First, estimating and ranking twenty-eight soil types is beyond the scope of this study. Second, much of the classification in the original FAO data set is related to the elevation of the soil type or underlying bedrock and not the quality of the soil. The CLUSTER module is commonly used as a means of creating a hard classification for vegetation, biomass, or soil characteristics (Eastman 2006, 36) and is useful here for classifying similar soils. After the data were processed, an additional manual reclassification was preformed, in which modern Anthrosols were reclassified as Fluvisols. The Anthrosols in the data set characterize a very small proportion of the survey area. This soil type consists of human-disturbed sediments and could include CO M M U N I T I ES, V I O LEN C E , A N D F O RT I F I CAT I O N

87

Figure 4.4. Sites with Dawenkou and Qujialing cultural elements on the Central Plains. Modified from Liu (2004, 6.18).

human-made levees, mines, or urban garbage. The surrounding material is either a body of water or Fluvisol. The following is a description of the eight soil types (figure 4.5) and is derived from the UN working group on soil classification (IUSS 2007). They are in order of area within the study region, and table 4.1 indicates the ranks used for this study, 1 being most productive and 8 least productive. Calcaric Fluvisols are river-deposited soils that contain high amounts of calcium and are less well drained. With water management and irrigation, these soils can be very agriculturally productive but only through more intensive farming. Calcaric Cambisols, relatively young upland soils, are well drained and do not contain high amounts of clays. These provide excellent soils for cultivation of grass crops, especially foxtail. Eutric Planosols are usually waterlogged soils and do not support crops very well other than rice paddy agriculture. These soils generally support grasses and make for excellent grazing. Calcic Gleysols are soils that generally need draining to support most crops but are also usable as pasture or for grazing but not as well as other soils. 88

JA M ES T. W I LLI A M S

Figure 4.5. Map with the most economically important soils in the region. T able 4.1. Soil resource potential Ranked Measures of Productivity or Abundance of Resources Soil Type

Agricultural

Pastoral

Forest

Calcaric Fluvisols

3

6

7

Calcaric Cambisols

1

4

8

Eutric Plansols

7

1

3

Calcic Gleysols

5

3

4

Gleyic Luvisols

2

5

6

Eutric Cambisols

4

7

1

Eutric Vertisols

8

8

2

Haplic Luvisols

6

2

5

Gleyic Luvisols are soils with a period of longer development and are higher in clays and silts. They still well drained and can support a variety of crops. Eutric Cambisols are somewhat shallow rocky soils with some soil development, commonly known as forest brown soils in many parts of Europe. Eutric Vertisols are rocky soils that have a great deal of vertical movement of material. This quality makes them less desirable for farming; the unstable surface can only support plant species with deeper root systems, making this soil more CO M M U N I T I ES, V I O LEN C E , A N D F O RT I F I CAT I O N

89

Figure 4.6. Site clusters model.

common in forested areas. Haplic Luvisols are very common in steppe environments and generally support grassland environments. The maps that define boundaries are created by using the DISTANCE module, which assigns higher values to areas farther from the target. In this case the settlements are of a particular archaeological culture (figure 4.6). The distance maps of all the site types and sizes, when combined, create a boundary map, indicating where the most and least interaction would take place (figure 4.7). The map that results from this is similar to a Thiessen polygon map but takes into account communities that are formed through small site clustering. About sixteen clusters of population can be seen in figure 4.7. This is generally a good correspondence to the polities interpreted by Liu through Thiessen polygons. The methodology could be improved with more precise measures of population by putting a finer point on the scale of interaction within sites and between them. This method also takes into account permeable and fluid nature of boundaries. Buffer zones or boundaries are created by isolating areas where there is little settlement (i.e., higher values in distance 90

JA M ES T. W I LLI A M S

Figure 4.7. Boundary distance model.

maps). Then interpolating increasing values from the boundaries or buffer zones creates a permeable area between areas of heavy activity (figure 4.7). Correlating the locations of buffer zones and settlements (both fortified and unfortified) can then be quantified by using a spatial linear regression that would indicate a strong positive correlation, if the fortified sites were located close to the buffer zones, and weaker as the settlement moved further from the buffer zone. This linear regression analysis would take into account the values of cells within two raster images and result in a best-fit line: the higher the r-squared value the more that the two images correlate with each other. The linear regression can also be used for correlating the locations of settlements with soil productivity. Measures of Soil Productivity

This set of variables is introduced into the study as a means of quantifying potential agricultural and pastoral productivity. In addition, the analysis CO M M U N I T I ES, V I O LEN C E , A N D F O RT I F I CAT I O N

91

quantifies forest resources of fortified sites in comparison to nonfortified sites. Do the catchment and environment of fortified sites differ at all from nonfortified ones? The goal is not to produce absolute measures of productivity but relative measures. Understanding the relationship between staple resources and settlement patterning addresses the potential of staple resources and economies as a base or source of power (Earle 1993, 204) The results also provide fodder for the proposition that this is an example of Robert Carneiro’s (1970) circumscription model (Wiesheu 1997, 103). This analysis addresses as well the rationale put forward by Li Liu and Xingcan Chen (2006, 162) that conflict was over resources. Last, the soil productivity analysis can also speak to the nature of the buffer zones. Do the buffer zones actually relate to conflict, or do they form as a result of less productive land not being utilized in favor of more productive locales? This analysis is not about resource scarcity but the idea that if there were locales where food could be produced easier and more productively, people would move as opposed to engaging in conflict. The analysis of productivity is assessed with regard to both agricultural and pastoral potential. Last, a measurement is used to assess the proximity to forest resources, which include deer, a large part of the protein diet in the Longshan. Walled settlements are slightly closer to areas where agriculture resources would be most abundant than other Longshan centers. However, the strength of correlations of both walled and unfortified settlements is generally low, and there is not a good correspondence between the better soils and locations of Longshan sites of any size. Longshan sites tend to occupy soils ranging from the very best to the very worst in terms of agricultural productivity. The results for the relationship between grassland environments best suited for a pastoral economy and locations of Longshan settlements are similar to the measures of agricultural productivity, indicating that settlements of all sizes and types are located on grasslands. Finally, the results of the forest resources regression indicate that as with the other resource types, the settlement distribution and patterning do not relate to the proximity of forest resources. The examination of the correlations between settlement distribution and potential economic resources shows that conflict is not a result of staple resource scarcity or that the settlement pattern has any relation to ecological zones or economic products within them. Furthermore, a lack of any correlation between the buffer zones and soil productivity indicates that these buffer zones are not wastelands of unproductive areas. It shows that the centripetal social forces of centers overpower environmental ones (table 4.2). 92

JA M ES T. W I LLI A M S

T able 4.2. Linear regression r² strength of correlation values Walled Settlement

Unfortified Settlement

Agriculturally ranked soils

Variable

4.36%

0.00%

Pasture-ranked soils

0.05%

3.84%

Forested-ranked soils All rivers

5.14%

0.60%

34.34%

20.12%

Yellow River

0.06%

−1.74%

Class 2 rivers

33.97%

48.09%

Class 3 rivers

12.43%

13.99%

Class 4 rivers

2.57%

3.47%

Cluster/buffer

47.65%

65.53%

Macro cluster/buffer

4.38%

4.32%

Interaction areas

1.52%

3.84%

Distance to River

An assessment of the proximity to water is incorporated into this study for three reasons. First, it serves as an additional measure for agricultural potential. As understood in the previous section, particular soils could be cultivated but only through intensive irrigation; therefore, access to water could help explain the locations of sites. Second, distance to rivers is also incorporated to understand the function of the fortification and the amount of time that could be spent under siege. Third, this assessment examines the possibility that the fortifications are for water diversion purposes (Bar-Yosef 1986; Liu 2004; Wiesheu 1997, 102). Distance to any river is a very good predictor of site location, and the location of the sites of all sizes seem to correspond well to river distance. Overall, when all rivers are taken into consideration, there is a slight difference between fortified sites and unfortified centers. The analysis further divided the size of the rivers into four categories to explain what rivers in particular impacted settlement location and whether fortified sites were located closer to larger rivers that would be more prone to flooding and would potentially require protection. The classes of rivers are as follows, from largest to smallest: Yellow River, Class 2 river, Class 3 river, and Class 4 river (figure 4.8). The distance to the Yellow River is a very small a factor in the location of settlements of any size and type. Distance to a Class 2 river is a better predictor of site location as well as distance to a Class 3 river. Distance to a Class 4 river has very little to do with site location or type. The regressions CO M M U N I T I ES, V I O LEN C E , A N D F O RT I F I CAT I O N

93

[128.104.46.206] Project MUSE (2024-03-01 20:28 GMT) UW-Madison Libraries

Figure 4.8. River classification.

from this analysis show that locations of Longshan sites correlate strongly with distances to Class 2 and 3 rivers, but since there is no difference between the fortified sites and the other sites, fortification has very little to do with location. BOUNDARIES WITH OTHER GROUPS

Human communities and clustering can be seen at different scales. The next section will investigate three scales at which we might find boundaries between clusters of sites. Local Political Boundaries

If Longshan communities created fortification because of threats from other Longshan communities, (i.e., interpolity conflict), then we would expect to see that fortification would occupy areas that are farthest from buffer zones or 94

JA M ES T. W I LLI A M S

located closest to them. The logic is that fortifications are either protecting the boundaries of the community or they are centrally located, with the maximum distance from the boundary, serving as a last line of defense. Either way, these distances will be indicated as a strong negative or positive correlation, or the correlation for walled sites should be significantly different from the unfortified sites. Longshan fortification is located slightly closer to the local buffer zones than to other major and minor centers, with r-squares for 65.5 percent to 47.6 percent for fortified sites; this does not indicate the type of periphery or boundary protection expected. The strong r-squared values indicate that there is a trend toward major settlements being centrally located within clusters, not an earth-shattering revelation, as it demonstrates that there is little difference between fortified and unfortified centers. Larger Community Boundaries

At a spatial scale larger than the political communities previously delimited, structures of other spatially recognizable communities emerge. Five large clusters of Longshan population emerge. These larger community structures show very little correlation to the location of fortifications, as well as little spatial correlation to the unfortified centers. The regression results are nearly equivalent for both fortified and unfortified sites and explain very little about the locations of the settlement distribution of any type. Longshan Boundaries with Dawenkou and Qujialing

During this period, there is evidence of three archaeological cultures within the survey area. In order to test if warfare is related to the influx of “Other” groups, the patterning of centers and fortified sites are correlated against the distribution of “foreign activity.” The distribution of “foreign activity” is interpolated by combining the distance maps of Dawenkou and Qujialing with the distance maps from the Longshan (figure 4.9). This map shows the areas where these archaeological cultures would have been in contact most frequently. Much like the local political boundary analysis, either a strong positive correlation or negative correlation indicates that the creation of fortification is related to the proximity to Dawenkou and Qujialing sites. If the result is positive, then it means purposely created fortification in areas of high cultural exchange and that potentially this exchange could be characterized as conflict or a least a perception of conflict. Negative results would indicate fortification in areas of low cultural exchange and, again, could be characterized as CO M M U N I T I ES, V I O LEN C E , A N D F O RT I F I CAT I O N

95

Figure 4.9. Interaction model map.

conflict. The results of the correlation, however, indicate very little patterning of Longshan settlements, fortified or unfortified, with respect to Dawenkou or Qujialing cultures. CONCLUSIONS

The role of warfare and/or conflict in relation to Longshan landscapes has been tested with a number of scales. The nature of the warfare has been tested in terms of scale, staple economies, and political implications. To reiterate the hypotheses—how violent conflict shaped the development of communities at local, regional, and supraregional scales; why we see a massive investment in fortifications in some areas but not in others; and where we can find the sources of threat and how it affected settlement location—tested in this chapter. First, the fortification is due to increasing interpolity conflict (Shao 2000, 204). Second, conflict is over resources. (Liu and Chen 2006, 162; Underhill and Habu 2006, 132) Finally, the emergence fortification in this area has been attributed to “population movement from other cultures” (Liu 2004, 189). All 96

JA M ES T. W I LLI A M S

three of these hypotheses have been disproven in the settlement patterning and landscapes of the Longshan within this study area. Given the results of this study, we continue to lack an explanation with a foundation in archaeological material as to why fortification emerged as an element of settlement structure during the Longshan period. Even without a definitive explanation for the emergence of fortification, disproving three of the potential causal agents allows for more productive study of the region. Furthermore, the results of this study indicate the spatial scale of investigation that will be most productive. As with most of the Longshan centers, the fortified settlements investigated in this study are centers of craft production and specialization (Liu 2004, 103–107; Underhill 1991, 23). These centers are in the production of specialized goods that could be part of wealth economies and relate to the power base of the elites within the polity (Earle 1997). The facets of nested Longshan identity that are exposed through this study are the identities of Longshan people, who all use similar pottery. Additionally, identities were related to political affiliation, which largely determine settlement patterning. Within these polities, in some situations, local identities emerge that result in expressions of violence. Scalping of individuals sends a more powerful message if those being scalped are locally known. More intensive survey and excavation at and around one or more of these fortifications could provide more evidence about the scale and nature of the violence. The distribution of weapons within the polity, for example, could inform where violence was occurring most frequently. Finally, I offer an additional hypothesis about the nature and scale of violence and the emergence of Longshan fortification. In the Longshan case, the relationship between political organization and violence is complex. The political boundaries do not correspond to the locations of fortifications and do not indicate the type of competition between neighboring polities resulting in further aggrandizement of individuals as a peer polity model might suggest. The scale of the conflict is occurring within the polity but carried out at high intensity. This situation can be described as high-intensity and small-scale warfare. Violence of this nature could be characterized as persistent raiding and fits with the interpretation of specialized production of pottery being one of the sources (Underhill 1991) of elite power. It also explains why the fortifications were built: to accommodate a higher volume of traffic to promote the production for trade and exchange of specialized craft goods. Asymmetrical violence and raiding may not result in political gain on the part of the attackers. However, there is a political consequence of violence in the Longshan on the part of the raided. The fortifications bolster elite ties to the landscape and serve to protect the wealth CO M M U N I T I ES, V I O LEN C E , A N D F O RT I F I CAT I O N

97

economy from local pilferers. Violence on the periphery and in the boundaries and the frontiers is widely discussed as being part of the dynamic process of culture change (Barfield 1989; Ferguson and Whitehead 1992; Lightfoot and Martinez 1995). This discussion should not obscure the potential for understanding the Othering of local groups and understanding the manifestations of violence at smaller scales within archaeological cultures and within polities. The communities of violence in the Longshan left lasting impressions on the landscape but serve a reminder of variability within the Longshan archaeological culture and between different political communities. REFERENCES

Arkush, Elizabeth. 2009. “Warfare, Space, and Identity in the South-Central Andes: Constraints and Choices.” In Warfare in Cultural Context: Practice, Agency, and the Archaeology of Violence, edited by Axel E. Nielsen and William H. Walker, 190–217. Tucson: University of Arizona Press. Banning, Edward. 2002. Archaeological Survey. New York: Plenum Publishers. Barfield, Thomas. 1989. The Perilous Frontier: Nomadic Empires and China. Cambridge: Blackwell. Barnes, Gina. 1999. The Rise of Civilization in East Asia: The Archaeology of China, Korea and Japan. London: Thames and Hudson. Bar-Yosef, Ofer. 1986. “The Walls of Jericho: An Alternative Interpretation.” Current Anthropology 27 (2): 157–162. Carneiro, Robert. 1970. “A Theory of the Origin of the State: Traditional Theories of State Origins Are Considered and Rejected in Favor of a New Ecological Hypothesis.” Science 169 (3947): 733–738. Chang, Kwang-chih. 1986. The Archaeology of Ancient China. New Haven, CT: Yale University Press. Cioffi-Revilla, Claudio. 2000. “Ancient Warfare.” In Handbook of War Vol II, edited by Manus I. Midlarsky, 59–89. Ann Arbor: University of Michigan Press. Earle, Timothy K. 1993. Chiefdoms: Power, Economy, and Ideology. Cambridge: Cambridge University Press. Earle, Timothy K. 1997. How Chiefs Come to Power: The Political Economy in Prehistory. Stanford, CA: Stanford University Press. Earle, Timothy K. 2001. “Institutionalization of Chiefdoms: Why Landscapes Are Built.” In Leaders to Rulers, edited by Jonathan Haas, 105–124. New York: Kluwer Academic / Plenum Publishers.

98

JA M ES T. W I LLI A M S

Eastman, J. Ronald 2006. IDRISI 15.0: The Andes Edition (software). Worcester, MA: Clark University. Erasmus, Charles. 1965. “Monument Building: Some Field Experiments.” Southwestern Journal of Anthropology 21 (4): 277–301. Ferguson, Brian, and Whitehead, Neil, eds. 1992. War in the Tribal Zone: Expanding States and Indigenous Warfare. Santa Fe: School of American Research Press. Hanks, Bryan. 2008. “The Past in Later Prehistory.” In Prehistoric Europe: Theory and Practice, edited by Andrew Jones, 255–284. Chichester, UK: Wiley-Blackwell. HENAN. 1992. “河南省文物研究所郾城县许慎纪念馆,郾城郝家台遗址的发 掘” [Henan Provincial investigations of Haojiatai site], 华夏考古 1992 年第三期 62–91。[Huaxia Archaeology 1992, 3: 62–91]. Hill, Paul, and Julie Wileman. 2002. Landscapes of War: The Archaeology of Aggression and Defense. Stroud, UK: Tempus Publishers Limited. IUSS Working Group. 2007. World Reference Base for Soil Resources 2006, first update 2007. World Soil Resources Reports No. 103. FAO, Rome. Kelly, Raymond. 2000. Warless Societies and the Origin of War. Ann Arbor: University of Michigan Press. LeBlanc, Stephen. 2006. “Warfare and the Development of Social Complexity: Some Demographic and Environmental Factors.” In The Archaeology of Warfare, edited by Elizabeth Arkush and Mark W. Allen, 437–468. Gainesville: University of Florida Press. Lightfoot, Kent, and Antoinette Martinez. 1995. “Frontiers and Boundaries in Archaeological Perspective.” Annual Review of Anthropology 24: 471–492. Liu, Li. 2004. The Chinese Neolithic: Trajectories to Early States. Cambridge: Cambridge University Press. Liu, Li, and Xingcan Chen. 2006. “Sociopolitical Change from Neolithic to Bronze Age China.” In Archaeology of Asia, edited by Miriam Stark, 149–176. Malden, MA: Blackwell. Oelke, Ervin A., Edward S. Oplinger, Dan H. Putnam, Beverly R. Durgan, J. D. Doll, and Dan J. Undersander. 2000. Millets. http://www.hort.purdue.edu/new crop/afcm/millet.html. Last updated January 11, 2000. Renfrew, Colin. 1976. Before Civilization: The Radiocarbon Revolution and Prehistoric Europe. Harmondsworth, UK: Penguin. Robb, John. 1997. “Violence and Gender in Early Italy.” In Troubled Times: Violence and Warfare in the Past. Vol 4m edited by Debra L. Martin and David W. Frayer, 111–144. London: Psychology Press, 1997. Shao, Wangping. 2000. “The Longshan Period and Incipient Chinese Civilization.” Journal of East Asian Archaeology. 2 ( January): 195–226. CO M M U N I T I ES, V I O LEN C E , A N D F O RT I F I CAT I O N

99

Solometo, Julie. 2006. “The Dimensions of War: Conflict and Culture Change in Central Arizona.” In The Archaeology of Warfare, edited by Elizabeth Arkush and Mark W. Allen, 23–65. Gainesville: University of Florida Press. Thorpe, Ian. 2003. “Anthropology, Archaeology, and the Origin of Warfare.” World Archaeology 35 (1): 145–165. Underhill, Anne. 1991. “Pottery Production in Chiefdoms: The Longshan Period in Northern China.” World Archaeology 23 (1): 12–27. Underhill, Anne. 1997. “Current Issues in Chinese Neolithic Archaeology.” Journal of World Prehistory 11 ( June): 103–160. Underhill, Anne, and J. Habu. 2006. “Early Communities in East Asia: Economic and Sociopolitical Organization at the Local and Regional Levels.” In Archaeology of Asia, edited by Miriam Stark, 121–138. Malden, MA: Blackwell. Webster, David. 2000. “The Not So Peaceful Civilization: A Review of Maya War.” Journal of World Prehistory 14 (1): 65–119. Wiesheu, Walburga. 1997 “China’s First Cities.” In Emergence and Change in Early Urban Societies, edited by Linda Manzanilla, 87–105. Boston: Springer.

100

JA M ES T. W I LLI A M S

5 This chapter aims to reconstruct the changing relationship between warfare and sociopolitical processes as manifested in the landscapes of protohistoric Japan, more specifically during the transition between the chiefdoms of the Yayoi period (2950–1700 bp) and an early state of the Kofun era (1700–1350 bp). In this chapter I review the archaeological evidence of warfare—including weapons, fortifications, human remains, warrior’s burials, and military rituals—through time to reconstruct the process by which warfare emerged and developed, how it was manifested in landscapes, and how it related to regional processes of social transformation.

Process of Warfare and Its Landscape in Protohistoric Japan Takehiko Matsugi

CULTURAL HISTORY

The prehistoric and protohistoric times in the main areas of the Japanese archipelago (figure 5.1) are divided in four stages according to traditional archaeological sequences: the Paleolithic, Jomon, Yayoi, and Kofun periods. The earliest dates for the Jomon period, characterized by the appearance of the first pottery in the archipelago, place its origin around 16500 bp, during the transition between the Last Glacial Era and the Holocene, when the global warming process led to the formation of ecologically rich ecosystems capable of sustaining sedentary complex foragers. These Jomon groups subsisted mainly through an economy based on gathering, fishing, and gathering during nearly 14,000 years and with limited horticulture toward the end of the period.

https://doi.org/10.5876/9781646422111.c005

101

Figure 5.1. Areas of the Japanese archipelago.

By 3000 bp, intensive agriculture based in wet-rice cultivation was imported from the Korean Peninsula to the archipelago, marking the beginning of the Yayoi culture, the first agricultural society in Japan. It is considered that hundreds of people migrated from the Korean Peninsula, introducing many warfare practices including the use of weapons and fortified settlements as related later in detail. By 2350 bp (fourth century bc), metal tools—bronze weapons and mirrors as well as smaller iron instruments such as chisels and adzes—were introduced from the peninsula, triggering processes of rapid social stratification and the emergence of local chiefdoms (Tsude 1967). Chinese written documents recorded Nu, a chiefdom believed to have been in Kyushu, the westernmost 102

TA K EH I KO M AT S U G I

Figure 5.2. Status system represented in the scale and form of burial mounds. After Tsude 1989.

main island of the archipelago of Wo ( Japan) in ad 57 (1893 bp) and Shuai Sheng, the king of Wo in ad 107 (1843 bp), indicating that the local chiefdoms were integrated into a kingship over the main areas of the archipelago during the first and second centuries (between 1950 and 1750 bp). According to radiocarbon dating, by 1700 bp (in the middle of the third century ad), a series of kingly zenpō-kōen-fun (keyhole-shaped burial mounds with round rear mound) started to be constructed in the central Kinki region (mainly the modern Nara, Osaka, and Kyoto Prefectures). The Kinki region is known for having been historically the seat of the Japanese capital from these early times until the early modern era. Smaller copies of the kingly mounds known as zenpō-kōhō-fun (keyhole-shaped burial mounds with square rear mound), as well as simple round or square burial mounds that belonged to various ranks of local elites, were erected throughout the main area of the archipelago, marking the beginning of the Kofun period. This hierarchical formulation of mounds construction (figure 5.2) is regarded as a manifestation of status system of the early state reigned by Daiō, who was the paramount chief or the king of the Yamato kingship and who succeeded in extending his power by controlling the local elites. The Yamato kingship reached its peak by the fifth century bc (1550–1450 bp) and declined through the sixth century bc (until 1350 bp) as reflected in the scale of the burial mounds. It is interpreted P RO C ES S O F WA R FA R E A N D I T S L A N D S CA P E I N P RO T O H I S T O R I C JA PA N

103

that this early kingship, dependent on the material display of its power and prestige, was gradually transformed into a mature state, whose government was based on legal codes and world religion (Buddhism) practiced across ethnic borders, imported from the Chinese culture. YAYOI PERIOD: BEGINNING OF WARFARE AND ITS LANDSCAPE

Overall, the archaeological evidence for warfare for the Jomon period is very scarce. Few wounded bodies from graves have been identified, but these kinds of discoveries are, in general, very rare (Nakao et al. 2016). Given that there were neither interpersonal weapons nor fortified settlements, Jomon is regarded as an egalitarian foraging society in which internal social organization prevented the concentration of individual power and the development of intergroup violence (Matsumoto 2018). Furthermore, it is considered that competition for food resources was lessened by the adoption of a highly diversified subsistence economy that provided abundant food year round without the need of practicing intensive agriculture. The first clear evidence of warfare belongs to the beginnings of the Yayoi period. Interpersonal weapons—such as polished stone daggers and arrowheads—and the design of fortified settlements were introduced by the Korean immigrants together with wet-rice cultivation from the peninsula to the northern coastal areas of Kyushu, marking the start of the Yayoi period as early as 2950 bp (late tenth century bc). A clear example of the rise of conflict along with the adoption of agriculture was found in Shin’machi (northern Kyushu), where a middle-aged man skeleton presented a thighbone penetrated by a polished stone arrowhead. However, the overall evidence for warfare for initial Yayoi is still scarcer when compared with later Yayoi. Even settlement fortifications, such as circular moats and ramparts, were so modest in size that archaeologists tend to consider them more in terms of their symbolic value than of their real defensive benefits. After 2350 bp (the beginning of fourth century bc), when the first metal implements—including bronze weapons—were introduced, wounded bodies rapidly increased in frequency in the northern Kyushu region (Hashiguchi 1987). Initially, by 2350 bp, people were settled mainly in the alluvial plains. This changed after 2350 bp, when some people started moving into the low hills that separate these plains, suggesting that the intraregional conflict developed into an interregional conflict (Fujio 1996; Nakahashi 1991) (figure 5.3 and 5.4). It is considered that the transformation of conflict described here reflects regional-wide processes of social integration and stratification, 104

TA K EH I KO M AT S U G I

Figure 5.3. Distribution of injured bodies before 2350 BP in the northern Kyushu region. Black skeletons represent casualties with bones deadly damaged by weapons, without skulls or skulls only. Gray skeletons represent bodies with signs of violence. After Shin’ichiro Fujio (1996) (partially modified).

Figure 5.4. Distribution of injured bodies after 2350 BP in the northern Kyushu region. Black skeletons represent casualties with bones mortally damaged by weapons, bones without skulls, or skulls only. Gray skeletons represent bodies with signs of violence. After Fujio (1996) (partially modified).

Figure 5.5. Location of weapons in the Japanese archipelago between 2150 and 1950 BP (first and second centuries BC). After Matsugi (1995) (partially modified).

including the consolidation of multiple local chiefdoms under the rule of a few powerful chiefs. Some of these chiefs were buried in clay coffins at Sugu-Okamoto and Mikumo-Minamishoji sites around 2050–1950 bp (to first century ad), and accompanied by abundant grave goods including dozens of bronze mirrors and several glass discs (gifts from Chinese dynasties) as well as indigenous bronze weapons specialized for ritual use. In contrast with the substantial evidence of conflict in the form of injured bodies, fortified landscapes seldom developed in the central zone of northern Kyushu. In peripheral regions, such as the coastal areas of Ariake Sea, a few large moated settlements, such as the Yoshinogari site, were founded and developed by the later Yayoi period. In the eastern islands of Honshu and Shikoku, where wet-rice cultivation arrived 100–700  years later than in northern Kyushu, local types of stone daggers and arrowheads for interpersonal use appeared by 2150 bp (second century bc) (figure 5.5). Tipped stone daggers and arrowheads were most notably developed in western Honshu including the Kinki region, though the frequency of discoveries of injured bodies is much lower than in northern Kyushu. Furthermore, the degree of political integration and social stratification found in northern Kyushu is not observed in the eastern regions, where egalitarian communal burials without exceptional tombs were the prevailing pattern until the transition to the Kofun period. This observation suggests that conflict in these areas did not promote, but instead may have inhibited, social 106

TA K EH I KO M AT S U G I

Figure 5.6. Plan of Ikegami-Sone site. After Inui 1996 (partially modified).

integration and stratification (Matsugi 1995; Teramae 2018) as is often the case with “pandemic war” in tribal societies (Allen 2008; Arkush 2011). In these eastern regions, the defensive landscapes developed more than in northern Kyushu. For instance, in western Honshu, many of those large settlements considered as regional centers—such as Ikegami-Sone (figure 5.6) and Karako-Kagi (both in Kinki)—were protected by multiple concentric circular moats and probably ramparts between them around 2150 bp. The impressive visual effect of these defenses may have dominated the landscape, evoking communal defensive efforts. At the same time, hilltop settlements emerged along the coast of Seto Inland Sea in Shikoku and western Honshu, which are thought to have functioned as forts (Kobayashi and Sahara 1964; Ono 1953), lookout places over trade routes, and fire beacon platforms in time of emergency (Tsude 1979; Earley-Spadoni, chapter 7 in this volume), as well as landmarks of regional groups (Shibata 2006). In any case, people from P RO C ES S O F WA R FA R E A N D I T S L A N D S CA P E I N P RO T O H I S T O R I C JA PA N

107

Figure 5.7. Topographical map of the Tawayama site. After Matsue-shi Kyoiku-iinnkai (Matsue Municipal Education Board) 2005 (partially modified).

this region of Japan produced defensive landscapes that incorporated moated settlements. Also, in eastern Honshu, especially in the southern Kanto region (the present location of Tokyo), many settlements with a single circular moat appeared on alluvial plateaus during the 2150–1950 bp lapse (first and second centuries bc). One of the most striking defensive sites of the Yayoi period is Tawayama, in San’in region and facing the Japan Sea in western Honshu. This site is located on a hill 60 masland supposedly had a shrine and an observation tower within a hilltop enclosure and surrounded by triple circular moats and ramparts constructed on the steep slope (figure 5.7). Of interest, there were dozens of pit houses and buildings outside the moats and ramparts, suggesting that the site was built as a symbolic and monumental focus of defensive landscape by the regional group(s). 108

TA K EH I KO M AT S U G I

KOFUN PERIOD: INDIVIDUALIZING OF WARFARE LANDSCAPE

Large settlements with single or multiple circular moats decreased between 1950 and 1750 bp in many areas and finally disappeared by 1650 bp from the archipelago. Hilltop settlements increased suddenly in numbers after 1850 bp until finally being abandoned by 1650 bp, too. The disintegration of these fortified landscapes coincides in time with the rise of new forms of monumentality and power. For example, in the Nara basin, in the center of the Kinki region, Hashihaka, the initial kingly keyhole-shaped burial mound of the Yamato kingship, was constructed around 1700 bp (in the middle of the third century bc), just after the sudden decline of Karako-Kagi, a large settlement with multiple circular moats, which was during long time the political center of the basin. Also in the Hokuriku region, facing the Japan Sea in eastern Honshu, some hilltop settlements were abandoned and keyhole-shaped burial mounds of local chiefs were built above them, such as Furutsu-Hachiman’yama. The end of the Yayoi period was the end of all the moated and hilltop settlements in the archipelago. After the beginning of the Kofun period, the only moated structures were the kings’ and chiefs’ rectangular residences. Not only were these moats of little defensive function, but also these residences were located in the low plains. This shift suggests that the defensive landscape and the authority linked to war, prevalent during the Yayoi period, were swept away by other forms of authority that materialized in the construction of massive burial mounds at the beginning of the Kofun period. These burial mounds of the successive kings and subordinate chiefs in central Kinki had, however, many grave goods centered on weapons. One of the largest tombs ever excavated in Japan is Mesuriyama, a keyhole-shaped burial mound 240 m in length and built in the Nara basin around 1600 bp. Archaeologists discovered 212 iron spearheads and 238 bronze arrowheads placed in a stone-offering chamber. Kurohimeyama is a keyhole-shaped burial mound 114 m in length that contained twenty-four iron armor suits in a stone chamber for offerings. Kurohimeyama was built in the Osaka plain, an area where numerous king’s tombs were built between 1550 and 1450 bp (ad fifth century). Other extraordinary discoveries of iron weaponry were found in Nonaka (11 armor suits, 153 swords, and around 740 arrowheads) and Ariyama (77 swords and 1,542 arrowheads). Both Nonaka and Ariyama were small square mounds specially built for royal funerary rituals. In the 1550–1450 bp period. The abundance of weapons as offering were not restricted to kings’ mounds. Burial mounds of local chiefs throughout the archipelago had considerable amounts of weapons offerings—such as armors, swords, daggers, and P RO C ES S O F WA R FA R E A N D I T S L A N D S CA P E I N P RO T O H I S T O R I C JA PA N

109

Figure 5.8. Offerings of iron armors, weapons, and other instruments in Nonaka Burial mound. After Kitano 1976.

spears—inside and outside of coffins sometimes placed within stone chambers (coffins of local chiefs often lacked stone chambers covering them; see figure 5.8). These offerings seem to have been smaller copies of those ritually buried in the kingly burial mounds, and many of them are revealed to have been delivered by the Yamato rulers not only as armaments but also as prestige goods. In a clear contrast with the increasing number and quality of weapons deposited as offerings in graves, the frequency and ratio of injured bodies 110

TA K EH I KO M AT S U G I

decreased during the Kofun period (Fujiwara 2018; Nakao et al 2016). The fact that fortifications are seldom found for the Kofun period is used as an indicator of the absence of violent intergroup conflict within the archipelago during this time. The large amount of armor and of other weapons is assumed to have been produced for the display of military prestige of chiefs and kings. It means that the warfare landscape as material manifestation converged with individual elites through the burial mound construction including the rituals focused on offerings of military goods. After 1450 bp (sixth century ad), the practice of depositing large caches of armor and weapons was replaced by the offering of one or two ornamented swords in the burials of the elite. These ornamented swords are regarded as status symbols in a new political structure headed by kings. The formation of a mature archaic state sanctioned by legal codes and Buddhism brought from the continent during the seventh century ad was determinant for the disintegration of the warfare landscapes and the burial mound tradition. Warfare landscape faded along with the decline of the burial mound construction undermined by the formation of a mature archaic state based on legal codes and Buddhism in seventh century from the continent. The landscape associate with the newly born state of Japan was novel in that cities were not fortified or walled. The only exception was the construction of a dozen hill forts by late seventh century in the northern Kyushu and central Kinki, in the western part of the archipelago, which may be a response to the threat of invasion from Silla (Korea) and Tang (China). DISCUSSION

In this chapter I have described the spatial distribution of the archaeological evidence of warfare, contrasting the different realities of northern Kyushu and Kinki regions during the Yayoi period, and explaining the transition between the Yayoi and Kofun periods (table 5.1). In short, moated and hilltop settlements were prevalent on Shikoku and Honshu Islands including Kinki, while in northern Kyushu region they seldom developed despite the higher frequencies of wounded bodies. These moated and hilltop settlements in Honshu and Shikoku disappeared in conjunction with the beginning of the tradition of burying the elite in mounds and accompanied with large caches of military offerings (Kofun). It is considered that the political consolidation by powerful chiefdoms such as Kyushu’s Nu during the Yayoi period and by the much larger Yamato kingdom during the Kofun period brought intergroup conflict to an end, made P RO C ES S O F WA R FA R E A N D I T S L A N D S CA P E I N P RO T O H I S T O R I C JA PA N

111

T able 5.1. Distribution of evidence of warfare during the Jomon period to the Nara (the ancient state) period Evidence of warfare

Jomon

Yayoi

Kofun

×

〇

◎

〇

Fortifications

×

◎

△

△（periphery）

Injured bodies

△

◎

△

×

Interpersonal weapons

Nara

Offered weapons in burials

×

〇

◎

△（periphery）

Worship of weapons

×

〇

△

×

Artistic expressions of warfare

×

×

〇

×

◎ ample; 〇 often; △ rare; × none

fortifications unnecessary, and reduced the number of direct victims of these internal struggles that plagued the archipelago for centuries. The large quantity of military offerings in the Kofun period are considered to have been served to compete against oversea polities such as Koguryo, Silla, Paekche kingdoms, and Kaya chiefdoms in the Korean Peninsula. The evidence of war victims, however, did not disappear completely, even after political consolidation and pacification carried out by the powerful chiefdoms in Kyushu. A dozen of bodies injured by weapons have been found in underground horizontal tombs of the Kofun period in Southern Kyushu (Fujiwara 2018). If this discovery is not a consequence of differential preservation of human remains throughout the archipelago, it suggests this region witnessed higher levels of violence than did other regions of Japan. Additionally, some evidence from the Korean Peninsula indicates that the Japanese may have been involved in warfare overseas during this time. Due to the variety in forms of evidence, the Hobbesian assumption that the difference between and transformation of the archaeological evidence of conflicts were caused by political integration with the change in the quality of warfare—from internal conflicts to overseas competition—does not seem a full explanation. In terms of landscape, the material evidence can be interpreted not only as traces of human behavior but also as the manifestation of their cognition including sense, thoughts, knowledge, habits, beliefs, technology, and worldview. “Landscape” is defined as a built environment that makes and is made by human action. From this perspective, the changes in the spatial distribution of the archaeological evidence related to warfare from the Yayoi to the Kofun periods (table 5.1) can be considered a shift in the ideological role of warfare in the people’s worldview. 112

TA K EH I KO M AT S U G I

The material culture of the Yayoi period reflects a collective ethos. Despite the fact that no single large settlement has been fully excavated, many of them, including hilltop and moated ones, were large enough to contain substantial numbers of pit houses and other buildings, sometimes into the few hundreds, surrounding ritual places such as shrines. These settlements, where many kin groups were living together in a single large community, are regarded as characteristic of the Yayoi period (Okubo 2011). Yayoi period burials also reflect this communal ethos. Tens of hundreds of individual coffins—made of timber, slate, or clay—were arranged in an apparent lack of order in large aggregations. These cemeteries, together with the settlements, produced a collective landscape, where interpersonal relationships were apparently manifested (Mizoguchi 2001). In the same way, religious beliefs seem to have emphasized the collective character of this society. Bronze instruments such as dohoko (spearheads), mainly found in Kyushu, and dotaku (bells), mainly found in Kinki, which were of important ritual use, are often found in isolation at a short distance from the settlements and graveyards. This discovery suggests that objects belonged to the whole community and were not the property of specific individuals or families. By contrast, the beginning of the Kofun period also marked the development of an individualist ethos. The large communities, formerly aggregated in large villages, split into smaller villages of several dozens of pit houses and buildings, large enough to be occupied by only few kin groups. Although several large residential areas such as Makimuku (Nara basin) appeared occasionally, their pit houses and buildings were placed in small groups with no moat surrounding them. Chiefs and kings were living in rectangular enclosed residences segregated from the rest of the settlement and monopolizing the defenses of the community (Tsude 1979). The large communal graves of the Yayoi period were replaced by small groups of burial mounds each containing one or a few inhumations in them. These are thought to be for an individual person and/or her/ his siblings according to the analysis for estimation of kinship based on toothcrown measurements of skeletal remains from these burial mounds (Tanaka 1995). Metal artifacts started to be deposited ritually in kofun (individual burial mounds), especially in keyhole-shaped mounded tombs. These burial mounds were built as a manifestation of the power and prestige of the kings in Kinki and other lesser chiefs throughout the archipelago. The change in the distribution of archaeological evidence of warfare from the Yayoi to Kofun periods can be explained in similar terms. Defensive constructions in the Yayoi period, such as moated and hilltop settlements, are considered to have served as a material manifestation and a visual representation of P RO C ES S O F WA R FA R E A N D I T S L A N D S CA P E I N P RO T O H I S T O R I C JA PA N

113

communal identity. The hilltop settlements in the Seto Inland Sea, in western Honshu, may have served as landmarks of community identity against sailors crossing their coasts (Shibata 2006). Their identity was represented in the form of collective self-defense and sharing a common goal in their worldview. During the Kofun period, by contrast, community members’ identity was constructed and manifested through their chiefs, in the form of their military authority and heroic qualities. Plenty of weapons were thought to have been offered to the burial mounds of a chief or a king to show his/her militaristic and heroic power to protect and lead the community. In the core zone of northern Kyushu, where defensive construction did not develop during the Yayoi period, the communal representation of identity is considered to have been undermined by the rise of powerful chiefs buried with many bronze weapons as early as 1950 bp (first century ad). CONCLUSION

I have demonstrated that the differential spatial distribution of the evidence of warfare during the Yayoi period, and its temporal change from the Yayoi to Kofun periods, stemmed from the material manifestation of the transformation and divergence of worldviews. Recently, the same change in the pattern of material culture, including burials, is observed in surrounding areas such as Korean Peninsula (Matsugi 2019). Following the disorganization created by the decline of Han Dynasty, the powerful ancient empire of China, many ethnic groups rose politically and economically, competing with one another in a campaign to dominate each region. In such volatile and conflictive circumstances, those societies, grouped around powerful chiefs or kings, may have been more resilient than collective organized ones when confronting rapid political fluctuation, such as the fall of the Han Dynasty. In part, these different capacities to cope with external change are the result of conservative tendencies of collective organizations versus the dynamism injected by ambitious leaders. The described transition between the Yayoi and Kofun periods in the Japanese archipelago, including how warfare was accommodated and landscapes transformed, may correspond to local impacts of this continentalwide political change. REFERENCES

Allen, Mark W. 2008. “Hillforts and the Cycling of Maori Chiefdoms: Do Good Fences Make Good Neighbors?” In Global Perspectives on the Collapse and 114

TA K EH I KO M AT S U G I

Transformation of Complex Systems, edited by Jim A. Railey and Richard Martin Reycraft, 65–81. Albuquerque: Maxwell Museum of Anthropology. Arkush, Elizabeth. 2011. Hillforts of the Ancient Andes: Colla Warfare, Society, and Landscape. Gainesville: University Press of Florida. Fujio, Shin’ichiro. 1996. “Wakokuran ni sakidatsu tatakai” [Conflicts before Wakokuran]. In Wakoku Midaru: Himoko no tojo made [Wo state fell into disorder: Until Himiko appeared], edited by Makoto Sahara and Shin’ichiro Fujio Tokyo: Asahi Newspaper. Fujiwara, Satoshi. 2018. Nihon-retto ni okeru senso to kokka no kigen [Origins of warfare and state in the Japanese archipelago]. Tokyo: Doseisha. Hashiguchi, Tatsuya. 1987. “Shuraku ricchi no hensen to tochi-kaihatsu” [Transition of settlement location and land development]. In Higash-Ajia no kouko to rekishi: Okazaki Takashi sensei taikan-kinen-ronshu [Archaeology and history of eastern Asia: Retirement memory essays for Professor Takashi Okazaki]. Kyoto: Dohosha. Inui, Tetsuya. 1996. “Kinai Daikibo-shuraku no Kozo” [Structure of large settlement in central Kinki region]. In Yayoi no Kango-toshi to Kodai-shinden [Moated city and prehistoric shrine in the Yayoi period], edited by Ikegami-Sone-iseki Shisekishitei 20 Shunen-kinen-jigyou Jikko-iinnkai [The Executive Committee for the Project of the Commemorative Events of the 20th Anniversary of the Designation as a National Historic Site]. Izumi. Kitano, Kohei. 1976. “Kawachi Nonaka-kofun no Kenkyu” [Study of Nonaka Burial Mound in Kawachi District]. Osaka-daigaku Bunngaku-bu Kokushi-kenkyushitsu [Department of Japanese History, Faculty of Letters, Osaka University]. Kobayashi, Yukio, and Makoto Sahara. 1964. Shiude: Kagawa-ken Mitoyo-gun Takuma-cho Shiudeyama-Yayoi-iseki no kenkyu [Shiude: Study on Shiudeyama Yayoi site at Takuma town, Mitoyo county, Kagawa prefecture]. Takumacho bunkazai-hogo-iinnkai [The committee for cultural properties protection of Takuma town]. Matsue-shi Kyoiku-iinnkai (Matsue Municipal Education Board). 2005. Tawayamaiseki-gun Hakkutsu-chosa-houkoku [Excavation report of Tawayama sites]. Matsue. Matsugi, Takehiko. 1995. “Yayoi-jidai no senso to Nihon-retto shakai no hattenkatei” [Warfare in the Yayoi period and social development process of prehistoric Japanese archipelago]. Kokogaku kenkyu [Quarterly of Archaeological Study], 42 (3): 33–47. Matsugi, Takehiko. 2019. “Yayoi-jidai kara Kofun-jidai e: Higashi-Ajia-funbo-bunka no Teishou” [From the Yayoi to the Kofun Periods: Advocacy of ‘East Asian Burial Mound Culture’]. In Saiko Jomon to Yayoi (Rethinking the Jomon and Yayoi), edited by Shin‘ichiro Fujio, Tokyo: Yoshikawa Kobun.

P RO C ES S O F WA R FA R E A N D I T S L A N D S CA P E I N P RO T O H I S T O R I C JA PA N

115

Matsumoto, Naoko. 2018. “Shuryo-saishu-shakai ni okeru senso: Shudankan no bouryoku wo sokushin/yokusei suru youin ni tsuite kangaeru” [Warfare among Hunter-gatherer societies: Considering promoting/preventive factors of intergroup violence]. Kokogaku kenkyu (Quarterly of archaeological study) 13 (3): 22–36. Mizoguchi, Koji. 2001. “Yayoi-jidai no shakai” [Society in the Yayoi period]. In Sonraku to shakai no kokogaku: Gendai no kokogaku 6 [Archaeology of villages and societies: Contemporary archaeology 6], edited by Ryuzaburo Takahashi, 135–160. Tokyo: Asakura Shoten. Nakahashi, Takahiro. 1991. “Haka no kazu kara shiru jinko-bakuhatsu” [Population Explosion Reconstructed by Counting the Numbers of Graves]. In Gen-Nihonjin: Yayoi-jin to Jomon-jin no nazo [Indigenous Japanese people: Puzzle of the Jomon people and the Yayoi people]. Tokyo: Asahi Newspaper Publisher. Nakao, Hisashi, Kohei Tamura, Yui. Arimatsu, Tomomi Nakagawa, Naoko Matsumoto, and Matsugi Takehiko. 2016. “Violence in the Prehistoric Period of Japan: The Spatiotemporal Pattern of Skeletal Evidence for Violence in the Jomon Period.” Biology Letters 12 20160028. Okubo, Tetsuya. 2011. “Kyodai-kango-shuraku no seicho to sore wo sasaeta shisutemu” [Development of gigantic moated settlements and the system supported them]. In Tayouka suru Yayoi-bunka: Yayoi-jidai no kokogaku 3 [Diversified Yayoi culture: Archaeology of the Yayoi period], edited by Shin’ichiro Fujio, Hiromi Shitara, and Takehiko Matsugi, 71–87. Tokyo: Doseisha. Ono, Tadahiro, ed. 1953. “Shimatagawa: Suo Shimatagawa-ryuiki no iseki-chousa-kenkyu-hokoku” [River Shimata: Survey and research report of archaeological sites on River Shimata, Suo district]. Yamaguchi: Yamaguchidaigaku Shimatagawa gakujutu-chosadan. Shibata, Shoji. 2006. “Chu-seibu-Setouchi no kouchisei-shuraku to yamazumi no shuraku: tokuni Hiuchinada, Iyonada, Akinada Engan’iki wo chushin to shite” [Hilltop settlements and mountain-living settlements in the middle and western Seto Island Sea Area: With special attention to Hiuchi, Iyo and Aki Sea areas], Kodai Bunka [Paleology] 58 (2): 69–81. Tanaka, Yoshiyuki. 1995. Kofun-jidai shinzoku-kouzou no kenkyu: jinkotsu ga kataru kodai shakai [Study of kin structure in Kofun Period: Ancient society reconstructed with human bones]. Tokyo: Kashiwa Shobo. Teramae, Naoto. 2018. Bunmei ni koshita Yayoi no hitobito [The Yayoi people against civilization]. Tokyo: Yoshikawa-kobunkan. Tsude, Hiroshi. 1967. “Nogu tekkika no futatsu no kakki” [Two epochs in spread of iron farming equipment], Kokogaku Kenkyu [Quarterly of Archaeological Study] 13 (3): 36–51.

116

TA K EH I KO M AT S U G I

Tsude, Hiroshi 1979. “Zenpo-koen-fun shutsugen-ki no shakai” [Societies in the emerging period of keyhole-shaped burial mounds], Kokogaku Kenkyu [Quarterly of Archaeological Study] 25 (4): 4–6. Tsude, Hiroshi, ed. 1989. “Kofun-jidai no Ou to Minshu”: Kodaishi-fukugen 6 [Kings and the people: Reconstructing Ancient History 6]. Tokyo: Kodansha.

P RO C ES S O F WA R FA R E A N D I T S L A N D S CA P E I N P RO T O H I S T O R I C JA PA N

117

6 Inscribing Power on a Landscape The Case of Co Loa in Vietnam

Nam C. Kim and Russell S. Quick

https://doi.org/10.5876/9781646422111.c006

118

Over the last forty years, archaeological research has increasingly demonstrated that warfare has been a major factor in the lifeways and cultural aspects of many societies in prehistory. The significance of warfare is not restricted to any time period, world region, or type of society. Decades ago, most researchers suggested that only state-level societies had the capacity to engage in warfare, thus downplaying the existence or significance of warfare among smaller-scale, nonstate societies (see Allen 2015; Keeley 1996; and Kim and Kissel 2018 for an overview). Today, however, the preponderance of evidence indicates that organized violence (i.e., “war”) was conducted by nonstate societies. Nonstate warfare should not, therefore, be conceptually separated from that practiced by more politically centralized communities, it is simply harder to detect, especially in the archaeological record. As a result, researchers have developed methodological frameworks for recognizing warfare across time and space. These frameworks pertain to categories such as settlement data, defensive works, specialized weapons, iconography, buffer zones, and skeletal trauma (see Arkush and Stanish 2005; Estabrook 2014; Golitko 2015; Haas 2001; Keeley 1996; Keeley et al. 2007; Kim et al. 2015; Lambert 2007; LeBlanc 2014; Martin et al. 2012). According to archaeologist Lawrence Keeley (1996, 36), the two most direct and unequivocal archaeological signatures of armed conflict are human skeletons exhibiting weapons trauma and ditch-and-palisade

[128.104.46.206] Project MUSE (2024-03-01 20:28 GMT) UW-Madison Libraries

fortifications. Traces of warfare—embedded projectile points, wounds from blunt and bladed weapons, scalping marks, skull fractures, and injuries to the back and forearms—are often found on the skeletal remains of its victims (Quick 2010, 48–50). Just as violence and warfare leave traces on bodies, landscapes can be viewed as “artifacts” upon which clues about their inhabitants are inscribed, illuminating past behaviors, beliefs, and cultural practices (Kim 2013, 245); including warfare. This marking is particularly true when landscapes have been heavily modified, as is common during times of expected or actual warfare. A case in point is the construction of fortified settlements, which require large-scale efforts that significantly alter and modify natural landscapes. These physical alterations are often accompanied by transformations in the meanings of these landscapes, thus leaving material traces indicating the social importance of violence, coercive power, or warfare. In some instances, the perceptions and meanings of landscapes altered by warfare persist through subsequent generations, as later societies continue to occupy and utilize these geographic spaces. In this chapter, we present a case study from northern Vietnam’s Red River valley that highlights this theme. Specifically, we discuss the late prehistoric settlement known as Co Loa located on the outskirts of present-day Hanoi (see figure 6.1). The available archaeological evidence from the wider region demonstrates that a mix of variables led to critical changes throughout the first millennium bce, culminating in the formation of the Co Loa Polity and the founding of its capital settlement, Co Loa. This case demonstrates how specific historical trajectories, environmental conditions, and modes of social interaction can foster changes in sociopolitical organization. In terms of archaeological evidence, the material categories of fortification, specialized weaponry, and iconographic depictions are most pertinent for the Co Loa case. Recent field investigations and analysis of the settlement’s monumental system of ramparts, embankments, moats, ditches, and waterways demonstrate not only a healthy concern for defense, but also the capacity to marshal significant amounts of labor and resources to put that concern into practice. Co Loa’s defensive system was put into place by a prehistoric, state-level polity, whose power and authority were essentially inscribed into the landscape through large-scale permanent modification of the local terrain. The case illustrates the complex relationships between warfare, settlement choices, leadership strategies, and social change. Ultimately, Co Loa underscores the variable uses of landscapes, demonstrating changing meanings and functions of the settlement and its fortification system over centuries and millennia. The construction of the defenses, involving thousands of laborers, likely I N S C R I B I N G P OW ER O N A L A N D S CA P E

119

Figure 6.1. Location of Co Loa in the Red River Delta. ArcView 3.x Data & Maps. 2002. ESRI (Environmental Systems Research Institute), Redlands, CA.

communicated powerful messages for both local and extralocal communities. For the former, the modified landscapes probably served to promote social bonds and the coalescence of communal identity, whereas for the latter, the formidable constructions likely served as a deterrent and clear reminder of the polity’s power and capabilities. BACKGROUND: THE RED RIVER VALLEY AND ITS PREHISTORY

The Co Loa settlement is located in the Bac Bo region of Vietnam within the Red River valley and delta (see figure 6.1). This watershed contains three terrain zones: the delta, the midlands, and a mountainous zone to the north and northwest. Most of the site lies within the Co Loa Commune of the Dong Anh District, though portions of the ramparts lie within surrounding communes. Dong Anh’s topography, like most of the midland zone within which it lies, consists primarily of very flat floodplains. Elevations range between 5.5 and 26 m AMSL. Historically, Dong Anh has relied on its watercourses, many no longer extant, but whose locations have been determined using a combination of ground survey, satellite, and Laser Detection and Ranging (LiDAR) data. These infilled features include streambeds, oxbow and “finger” lakes, and 120

N A M C . K I M A N D R U S S ELL S. Q U I C K

irrigation channels. Such elements would have been extremely important both for the initial selection of the settlement site and the ways in which the settlement developed over time. The Red River Delta (RRD), formed by the 1,200 km long Red River and its numerous tributaries, is the largest watershed in Vietnam and one of the largest in Southeast Asia. The Red River originates in the mountainous Yunnan Province of China and the uplands of the Truong Son Cordillera, and flows southeastward (Li et al. 2006, 5) joining its two main tributaries—the northern Lo (Clear) River and the southern Hac Giang (Black) River near the city of Việt Trì—before emptying into the Gulf of Bac Bo near Nam Định. The sedimentary basin, approximately 500 km long and 50–60 km wide, is filled with Neogene and Quaternary sediments with a thickness of more than 3 km (Li et al. 2006, 6). The drainage area (160,000 km2) consists primarily of subtropical lowlands with a few mountainous areas. Large, highly seasonal water (120 cubic km/yr.) and sediment (100–130  million t/yr.) discharges make it one of the major fluvial suppliers in Southeast Asia (Hanebuth et al. 2006, 121). Apart from the coastal strip, the alluvial plains are surrounded almost completely by mountainous terrain (Nishimura 2005, 99). Due to its shape, topography, and location along mainland Southeast Asia’s eastern edge, Vietnam experiences several climatic regimes. The RRD is characterized by a tropical monsoonal climate with a pronounced maritime influence (Li et al. 2006, 6). There are two main seasons—a cool, dry winter, and a warm, wet summer; these seasons are punctuated by short transitional periods (Sterling et al. 2006, 7). Annual rainfall averages 1,300–1,800 mm, 85 percent of which falls during the summer rainy season (April–October), with the heaviest rainfall deposited by August and September monsoons (Li et al. 2006, 6). Monsoonal rainfall has long been a factor in the social development of the region, as both past and present populations have relied on the control of annual flood waters for agricultural production (Nishimura 2005, 99). The climate and agriculturally fertile soils make the region extremely productive; several crops can be grown each year. Its climate is less extreme than most of low-lying mainland Southeast Asia, largely because the dry season is tempered by moist winds that move across the Gulf of Bac Bo (Higham 2002, 170). The RRD has long been one of the great “rice bowls” of the world, literally and figuratively. The lower Red River valley and delta comprise one of the most extensive tracts of flat agricultural land in Southeast Asia (Higham 2002, 278). Although the Delta’s diverse climate makes it suitable for the cultivation of tropical and subtropical crops, it is the production of rice—both in modern times and historically—that dominates the land use of the area. I N S C R I B I N G P OW ER O N A L A N D S CA P E

121

Rice agriculture accounts for 90 percent of the cultivated land and 80 percent of the total value of food crops in the Delta (Le and Le 2001, 6). Wet-rice agriculture was first introduced to the Delta lowlands during the Neolithic sometime after 4500 bp (Bellwood 2005, 131–133). Prior to this, late Neolithic communities had occupied the upland area to the west. During the late second and early first millennia bce, as bronze working began to spread and the Delta plain increased through alluviation, populations moved from the highlands and began to form settled communities in the lowlands and coastal plain. Palynological sequences obtained from core samples across the region reveal an increase in large Gramineae granis after 3340 cal. bp (Li et al. 2006, 25). These findings suggest that the increase in wet-rice cultivation coincides with the initial construction of settlements in the lowlands, especially after 4000 bp when the use of bronze tools became widespread. The geography of the riverine plain and the locations of prehistoric settlements would have been deeply influenced by changes in sea level during the Holocene (Nishimura 2005, 99). As farming and animal domestication became more widespread, agricultural production and fuel requirements for pottery production resulted in greater degrees of forest clearance (Sterling et al. 2006, 26). Whereas the coastline was probably quite close to presentday Hanoi (and Co Loa) near the end of the Neolithic (ca. 3500 bp), the accelerated process of erosion and alluvial deposition associated with intensive agricultural practices lengthened and broadened the Delta region. Over time, as riziculture became more prominent and the fertile lowlands extended seaward; corresponding shifts in settlement locations also occurred. As the population increased, previously unused land on the fringes of the plain became incorporated into the irrigated systems of the lowlands (Nishimura 2005, 99), further extending the area of riziculture. By the mid-first millennium bce, Bac Bo was home to the Dongson Culture (ca. 2600–1800 bp). The late prehistoric Dongson Culture is renowned for the production of ceremonial bronze artifacts, especially large drums (Pham 2004). The Dongson Culture is marked by intensified agricultural production, major habitation sites, and increased military and ritual practices. The rising disparity of grave furnishings in Dongson burials indicates an increased social differentiation of rank, status, and wealth compared to preceding archaeological sequences in the region (Murowchick 2001, 175). More than 100 Dongson sites have been discovered in varying environmental and geographic regimes including river deltas, coastal areas, and mountains (Pham 2004) (figure 6.2). The distribution of sites across differing topographical and ecological situations suggests close interaction between communities. While pre-Dongson 122

N A M C . K I M A N D R U S S ELL S. Q U I C K

Figure 6.2. Extent of Dongson Culture materials. The heaviest concentration of sites is within the Bac Bo (northern) region of Vietnam and the Red River valley.

political economies hint at some degree of craft specialization and social ranking, Dongson phase sites offer much stronger evidence for a high degree of metallurgical expertise, craft specialization, agricultural intensification, and differential status (Kim 2015). Bronze plows first appear in Vietnam at Dongson sites, with some 200 known specimens (Pham 2004, 199). Combined with the use of water buffalo for power, such agricultural innovations caused tremendous economic and political changes in Dongson communities. Aside from its agricultural importance, the RRD conferred other benefits upon its inhabitants. Its geographic location facilitated access to key transportation routes; providing opportunities for interaction and trade (see Calo 2009; Yang 2004; Yao 2016). Links to resources, materials, ideas, and innovations enabled leaders to develop strategies to obtain greater wealth, higher status, and political legitimacy. During the first millennium bce, long-standing exchange patterns allowed certain groups to gain advantages in economic competition related to metal industries. Moreover, political turmoil associated with Warring States China just to the north would have spurred the movement of peoples within and between southern China and northern Vietnam (Higham 2014, 198). As a result, various military innovations and technologies likely also found their way into the RRD. The manufacture and circulation of highly specialized crafts and prestige goods also intensified; access to various raw materials and attached craft specialists became restricted to certain segments of societies (Calo 2009; Murowchick 2001; Nguyen 2005). Differential access to resources, ideas, and technologies had a profound effect on social life within the RRD, contributing to an increase in competition between communities and the perceived threat of I N S C R I B I N G P OW ER O N A L A N D S CA P E

123

Figure 6.3. Co Loa’s wall system (derived from 2017 LiDAR data) overlaid on satellite imagery (USGS 1967).

actual conflict. It is within this wider social context that the massive fortification system encompassing the Co Loa settlement was constructed. During the Dongson Culture period, Co Loa was the largest settlement in Bac Bo. Its total area is estimated to be between 400 and 600 ha, depending on how the site is delineated (Higham 2014), making it one of the largest prehistoric settlements anywhere in Southeast Asia during the first millennium bce. The extant remnants of the fortification system consist primarily of three earthen rampart circuits (Outer, Middle, and Inner; see figure 6.3), each in varying states of disrepair. Recent investigations, discussed below, suggest the presence of additional defensive features. The settlement was likely the capital and seat of power for what is referred to as the Co Loa Polity (Kim 2015). Previous publications have explored the 124

N A M C . K I M A N D R U S S ELL S. Q U I C K

nature of the polity (ca. 2300–2100 bp) and the settlement, as well as the various factors for the emergence of sociopolitical complexity and urbanism (Kim et al. 2010; Kim 2013, 2015, 2017; Lai 2014). Arguably, one of the most significant factors was the use of physical or coercive power for militarism and warfare. There are two main sources of data that suggest the presence of competition and conflict within Bac Bo during the mid-first millennium bce and later, just when the Co Loa Polity emerges. These consist of textual traditions and the archaeological record, both discussed below. GENERAL EVIDENCE FOR VIOLENCE IN THE RED RIVER VALLEY

For the purposes of this discussion, warfare is defined as “organized violence between two independent political units in pursuit of social, economic, or political gain.” Archaeological evidence suggests that warfare and militaristic competition were important elements of Dongson culture. This evidence consists primarily of specialized weaponry and iconographic depictions. Aside from Co Loa evidence for Dongson defensive features and fortifications is limited by the available datasets. Most evidence for the Dongson culture comes from mortuary settings, far less from actual settlements. Agricultural intensification, rising populations, bronze working, and the growth of a prestige ideology surrounding bronze artifacts contributed to increased militarism in Bac Bo. As bronze working becoming more sophisticated during the mid-first millennium bce, certain agents and social segments had new means to effect changes in political relationships within and between communities. This period saw increased forms of competition and differentiation in social status, with competition and warfare likely becoming tied to political interactions (Kim 2017). Mortuary data, including the large number of bronze weapons recovered from burials, indicate that bronze was important for social and military power. Bronze weaponry accounts for 50  percent of recovered Dongson bronze implements (Hoang and Bui 1980, 64). This weaponry can be divided into three main categories: projectile weapons (javelins, socketed spearheads, socketed and tanged arrowheads), proximity or shock weapons (swords, fighting axes, daggers, and ge halberds), and shields/armor (Pham 2004, 199). Of these, daggers were extremely common; more than 230 were recovered from Dongson burial contexts. Depictions of a warrior class on Dongson bronzes reinforce the notion of militaristic practices. The iconography on bronze drums and other products often includes depictions of warriors. The lavish decorations also depict ritual and ceremonial activities, along with battle scenes and war canoes with plumed I N S C R I B I N G P OW ER O N A L A N D S CA P E

125

warriors (Higham 2014). Warriors are shown standing on firing platforms on boats and in some cases with captives. The depiction of warriors on boats is reminiscent of iconographic evidence from Late Bronze Age Europe, whose archaeological record is similarly marked by a proliferation of weaponry and armor, as well as signs of habitation in defensive locations (Hill and Wileman 2002, 35). The European depictions of warriors on boats suggest that raiding by boat was frequent in Scandinavia (Hill and Wileman 2002, 36). Dongson societies likely also utilized boats for raiding; their potential targets included the settlements along Bac Bo’s numerous rivers. As suggested by Calo (2009, 2), the distribution of bronze drums throughout the region may have been tied to exchange systems related to strategies of power and alliance building, and competition and conflict could have fueled by such exchange networks. Overall, artifacts and iconography point to the role of coercive strategies, along with associated tactics of physical intimidation, raiding, and conquest. Elites in the RRD were probably able to accumulate considerable status, wealth, and power through the exchange of bronze goods within an interregional network that connected southern and southwestern China, northern Vietnam, and parts of Southeast Asia through overland, riverine, and coastal routes. Increased use of bronze (and iron during the middle and later Dongson periods) furnished new advantages for military weaponry and agricultural intensification. Military power and coercive force, especially after the introduction of metal weaponry, presented novel avenues for consolidating power. Notably, the Warring States period of China was drawing to a close during this period as well, as the Qin Empire consolidated its power over China’s Central Plains region. It is quite likely that turmoil caused by warfare pushed communities away from threats and into areas of southern China and northern Vietnam. This may be one of the primary reasons for the sudden appearance of fortification features on the scale of those at Co Loa, which was absent elsewhere within the Dongson cultural sphere. Whereas local warfare may have been marked by raiding tactics between smaller-scale societies, the installations at Co Loa indicate considerable concern with threats posed by much larger forces using mass-produced and powerful technologies such as the crossbow. Co Loa represents the best prehistoric case of fortified settlement in the region. CO LOA’S FORTIFICATION SYSTEM: MONUMENTAL LANDSCAPE MODIFICATION

Today, most people in Vietnam view Bac Bo as the nucleus of Vietnamese cultural identity, and the Delta is home to semilegendary figures associated 126

N A M C . K I M A N D R U S S ELL S. Q U I C K

with early Vietnamese history and civilization during the late prehistoric period. According to Vietnamese texts, oral traditions, and folktales, Co Loa was founded by the Au Lac Kingdom at an ascribed date of 258 bce (Lai 2015; O’Harrow 1979; Taylor 2013; Trinh 2015). It is viewed as the first capital of an emerging Vietnamese civilization (Taylor 2013). According to textual sources, the founding figure of the Au Lac Kingdom, a man named An Duong Vuong (also known as Thuc Phan), was responsible for overthrowing the previous kingdom and establishing Co Loa as his capital and seat of power. Romanticized tales describe An Duong Vuong as being aided by a magic turtle that offered advice for constructing its defenses, also giving the king one of its claws to be used as the trigger mechanism for a mystic crossbow. Legendary accounts about An Duong Vuong were officially recorded in Chinese and Vietnamese texts during the Common Era. Many of the Vietnamese sources that purport to describe these events were written well after the fact, dating to the early second millennium ce. This long interval, combined with the supernatural aspects of the folktales, has led to much scholarly debate about the actual chronology and cultural history of the settlement. Compounding the issue is the annexation of Bac Bo by the Han Empire of ancient China at an ascribed date of 111 bce, with Sinitic domination lasting until the tenth century ce. Han chroniclers make little mention of Co Loa, and they essentially deny the presence of indigenous (“barbarian”) social complexity before the arrival of the empire. In terms of archaeological evidence, Co Loa consists of a monumental system of earthen ramparts, moats, and other constructions. Many of the modifications made to the historic landscape are still visible today. The system consists of three main earthen rampart enclosures, and river-fed moats and ditches, along with artificially constructed mounds, bastions, and towers. Modern ground and aerial surveys confirm that large sections of the Outer, Middle, and Inner walls are extant, though they have suffered from historic, and modern, alterations. The Inner Wall is roughly rectangular in shape, punctuated at regular intervals by bastions and measures 1.65 km in perimeter (Nguyen and Vu 2007). In places, the walls stand approximately 5 m in height, and some are 20–30 m wide at the base. The Middle (figure 6.4) and Outer walls form an irregularly shaped enclosures measuring 6.5 km and 8 km in circumference, respectively, and range from 3 to 10 m in height and up to 30 m in base width (Nguyen and Vu 2007). As discussed below, the irregular shapes stem from the natural topography, with the interfluves between water courses being intentionally connected to form the enclosures. Excavations of the rampart system (2007 through 2014) have provided stratigraphic data, artifacts, and radiocarbon determinations suggesting that I N S C R I B I N G P OW ER O N A L A N D S CA P E

127

Figure 6.4. View of the Middle Wall rampart’s interior face prior to excavation (2007).

much of the rampart system was built concurrently during the Co Loa Polity period (ca. 2300–2100 bp), with evidence of later refurbishment or amplification phases (see Kim 2013, 2015; Kim et al. 2010). Although these data do not completely substantiate the Vietnamese folktales, they do support the notion that a local, politically centralized society was present and responsible for the Co Loa settlement’s fortifications during the third and second centuries bce, well before Han annexation. The archaeological data also indicate that the primary impetus for the system’s construction was defense and concerns over security, though this certainly does not preclude secondary functions (water control, transportation, and others). This “defensive” interpretation is bolstered by the contemporaneous military weapons and weapon production facilities discovered across Co Loa. Other excavations uncovered a centralized facility, within the Inner Wall’s so-called Citadel, that produced standardized bronze crossbow bolts during the mid-third century bce (see Lai 2014). Its presence in such a restricted location suggests that a highly centralized polity held control over the production of implements of war. Elsewhere, researchers analyzing the crossbow technology of China’s imperial Qin period (ca. 300  bce) 128

N A M C . K I M A N D R U S S ELL S. Q U I C K

Figure 6.5. Smaller-scale defensive features. The smaller-scale clay wall is visible in the foreground, with the platform and earth structure beyond.

determined that it was not likely to be common or widespread at the time and that associated knowledge, skills, and materials would have been restricted (Li et al. 2014). Li and colleagues (2014) also maintain that the introduction of trigger-fired crossbows during the Warring States period revolutionized military tactics. It is plausible that the introduction of new military technologies and associated tactics into Bac Bo, such as the crossbow, affected levels of militarism and influenced decisions in fortification styles. One of the intriguing finds made during excavation of the Middle Wall in 2007 was the remains of a set of smaller-scale, defensive features found buried in situ within the larger rampart (figure 6.5). This set consisted of a much smaller wall made of soil and clay, a platform mound and structure also made of soil and clay, and ditches associated with them (Kim et al. 2010). The slumped appearance of the reddish clay suggests the wall was larger when first constructed and likely suffered a period of disrepair before eventually being buried and preserved beneath the rampart. Associated with this wall was an earthen mound at the same depth, and sitting atop the mound was a roughly I N S C R I B I N G P OW ER O N A L A N D S CA P E

129

rectangular earthen structure with the remains of a wall still present. Given their characteristics, the features may have served a military function (see Kim et al. 2010). Found in association with these suspected defensive features were Dongson potsherds and charcoal, and radiocarbon dates provide a window of circa 2500–2300 bp. The features clearly predate the monumental rampart and were likely constructed by a community culturally and politically distinct from the Co Loa Polity. There would be no structural or architectural purpose for building the smaller features if construction of the larger rampart were to follow shortly thereafter. The scale of construction is vastly different, by several orders of magnitude, signaling a different set of available labor and technological capacities, along with a different scale of perceived threats. It is thus interesting to consider changing needs, strategies, and tactics in terms of warfare and security (especially in the same location). It should be noted that Co Loa is unique within Bac Bo in terms of the scale and nature of its rampart system. What accounts for the Co Loa phenomenon? Regionally, Co Loa is not alone as an example of a moated site. The second half of the first millennium bce witnessed an emergence of numerous moated settlements across mainland Southeast Asia (Moore 1988, 1992). Hundreds of them have been found south and west of Co Loa, in Vietnam and modern-day Cambodia, Myanmar, and Thailand. These sites range in size, from a few to hundreds of hectares, typically enclosed by circular, oval, or irregular earthworks and moats (Dega 1999; Fletcher et al. 2008; Higham 2014; McGrath and Boyd 2001; Moore 1988, 1992; Moore and Win 2007; O’Reilly 2008). In every case, the earthworks and moats were constructed by communities; many may have functioned as fledgling, protourban settlements. Although the function for these earthwork and moat features varied depending on time and cultural context, some features likely served defensive purposes at some point in their existence. Despite similarities with other Southeast Asian moated settlements, Co Loa’s areal extent and the scale of its earthworks and moats were far greater than those of its contemporaries. In terms of scale, the best analogs of Co Loa lie within the emergent Chinese civilization to the north, where the size of urban settlements and their wall systems approach—and even surpass—Co Loa. The use of rammed or stamped earth in the construction of Co Loa’s ramparts, along with the presence of ceramic roof tiles (which are absent anywhere else in Vietnam until centuries later), is another similarity with sites to the north. These material indicators combine to suggest local awareness and knowledge of Sinitic forms of urbanism, elite symbols of authority, and military technologies (Kim 2015). A major difference, however, is the rectilinear wall shapes enclosing 130

N A M C . K I M A N D R U S S ELL S. Q U I C K

Figure 6.6. A 50 cm LiDAR DEM (main image) compared with 10 m DEM based on the GDEMv2 (inset).

Sinitic cities (see von Falkenhausen 2008), which stand in stark contrast to the irregular shapes of the Middle and Outer walls of Co Loa. What accounts for the seemingly amorphous shape of Co Loa’s defenses, and why was the settlement constructed in its particular location? As discussed above, the 2007–2014 field investigations focused on gathering data relevant to the construction of the rampart system, specifically in an effort to determine chronological information and building methods. In recent years, new methods (magnetometry, aerial photo analysis, LiDAR data, and others) have been employed to complement the excavation data by gaining a better understanding of the landscape around Co Loa and how its builders used the natural terrain. Particularly helpful has been a 25.4 km2 LiDAR data set, originally collected by the Vietnam Natural Resources and Environment Corporation, a division of the Ministry of Natural Resources and Environment, in Hanoi. The digital elevation model (DEM) derived from the LiDAR data forms the basis for reconstructing not only the site’s defenses, but also the ancient landscape and environment (see figure 6.6). It has been particularly useful for mapping the historic waterways of Dong Anh, many of which are no longer extant. Analysis of the DEM shows the locations of several kinds of I N S C R I B I N G P OW ER O N A L A N D S CA P E

131

infilled waterway including streambeds, oxbow and “finger” lakes, and irrigation channels—many of which, as noted above, influenced decisions regarding the settlement and its fortifications. The Dong Anh District is bounded by the Red River to the southwest, the Ca Lo River to the north, and the Duong River/Canal to the south. The Hoang Giang River lies on the western and southern edges of the Co Loa site, and it flows within a remnant meander loop channel around the site until it reaches its southeast corner. From this point, the river has been diverted into a new channel that empties into the Ngu Huyen Khe River (which has itself been converted into an east-west canal) approximately 1 km south of Co Loa. Historically, the Hoang Giang River followed an easterly route after passing Co Loa. The original channel took the river east-northeast to eventually join the Cau River—a major north-to-south-flowing river—near the city of Bac Ninh, some 22.6 km away. The historic route of the Hoang Giang River is important because of its potential implications for the transportation of both goods and armed forces into and out of the area of Co Loa. Prior to our access to LiDAR data, the precise historic route was difficult to determine due to the lack of high-resolution terrain data. As noted earlier, the region is extremely flat, and small variations in topography were absent from the satellite acquired GDEMv2 data set. Analysis of the processed LiDAR data conducted in 2017–2018 clearly shows the river’s route to Bac Ninh. While this knowledge is useful for reconstructing the historic landscape, contemporary mapping or documents and soil cores are still required to confirm which river channels were actually active at approximately 2300 bp, when the Co Loa Polity period commenced. Researchers have long suspected that the irregular shape of the outer ramparts was due to efforts to incorporate elements of the existing terrain, such as waterways, ridgelines, and hills. These landscape features were connected with purely anthropogenic constructions to create a monumental system of earthworks, bastions, moats, ditches, and other fortification features on a scale unprecedented for the RRD. The new LiDAR information substantiates this notion while providing additional insights. As indicated by analyses presented elsewhere (Quick 2019), Co Loa is situated on an enormous “island” formed thousands of years ago by the wanderings of the Red River and its tributaries. Over millennia, multiple meandering events deposited enough material to create what is essentially a dome-shaped mound of earth between two meander loops. The channels surrounding Co Loa eventually filled in, leaving a mounded landform that is approximately 4 km (2.5 mi) in diameter, as shown in figure 6.7. Intriguingly, this landform is reminiscent of descriptions from folktales associated with the capital city; it was constructed on a landform 132

N A M C . K I M A N D R U S S ELL S. Q U I C K

Figure 6.7. Outline of Co Loa overlaid on a 150 m DEM showing the domed landform on which it was constructed. NASA / METI / AIST / Japan Spacesystems, and US/Japan ASTER Science Team. ASTER Global Digital Elevation Model V003. 2019, distributed by NASA EOSDIS Land Processes DAAC, https://doi.org/10.5067/ASTER/ASTGTM .003. Accessed January 1, 2020.

that resembles a turtle’s domed shell. In other versions of the tale, Co Loa’s walls were built in a way to mimic the windings of a snail shell (Taylor 2013). The now defunct river channels surrounding the settlement may have been either inspiration for these tales or approximations of actual events and plans associated with the city’s emergence. In either case, it is clear that the location of Co Loa allowed its political leaders and residents to effectively capitalize on unique natural landscape features to build a massive system of defense in a systematic, expedient, and rapid fashion. Architectural energetic analyses suggest the system could have been built within a range of a few years or decades, likely utilizing thousands of laborers to move the 1–2 million m3 of earthen materials that were used (Kim 2013). Co Loa’s leadership would, therefore, have required the vision, resources, political will, and military strength to select this location and transform the landscape. As previously mentioned, the growing threats posed by imperial China at the end of the Warring States period could have been a significant motivating factor for the massive nature of the fortification system and its relatively rapid I N S C R I B I N G P OW ER O N A L A N D S CA P E

133

construction. Recent and past surveys of the ramparts indicate the presence of bastions, especially around the Inner Citadel. The dimensions, scale, and spacing of the bastions reflect the potential uses of crossbow firing technologies as well as the nature of perceived threats. The extant Citadel bastions are 135 m apart, almost the exact range of a wood and sinew crossbow (~137 m), meaning that each wall section was covered by fire from at least two bastions. Given the presence of competition and warfare, both locally (Dongson Culture) and extraregionally (i.e., Warring States / Imperial China), it is not surprising to see political and physical power being inscribed in such a way at Co Loa. CONCLUDING REMARKS: INSCRIBING POWER ON LANDSCAPES

The massive scale and morphological characteristics of Co Loa’s fortifications suggest that militarism and defense were important parts of social lifeways during the mid- to late first millennium bce. Military use does not preclude, of course, other possible functions for these monumental features, which may have included social demarcation of space, ritual practices, flood control, and waterway transport. Symbolically, the impressive scale and extent of the ramparts would have communicated tremendous wealth and power. The ramparts, along with other associated features and constructions of the city, would have conveyed considerable control over resources, labor, and physical power, thus complementing physical functions with ideological ones. Owing to issues of accessibility, much of the interior space within the enclosures has not been fully explored yet, leaving many open questions regarding internal settlement patterns. However, we suspect that much of the interior would have been used for agricultural production given the environmental conditions. In that sense, the fortified settlement could have easily functioned as a refuge during times of trouble (see Quick 2010 for a discussion). The system of ramparts and moats was designed to both harness water and fortify its inhabitants against potential, large-scale threats. Undoubtedly, the geographic setting and natural features contributed to broad social and cultural changes. The landscapes, natural resources, water courses, climate, and agriculturally productive potential of the area surrounding Co Loa were vital components in allowing societies to establish far-ranging trade contacts, grow in population size, consolidate political power, and compete through violence. The Sinitic domination periods of Bac Bo ended during the tenth century through warfare, and the first independent Vietnamese dynasty to take power was that of a man named Ngo Quyen (Taylor 2013). Textual accounts state that he made Co Loa the capital of his dynasty. The fledgling Ngo Quyen polity 134

N A M C . K I M A N D R U S S ELL S. Q U I C K

would have had legitimate concerns over future Chinese invasion attempts, making his selection of Co Loa sensible from a military standpoint. Arguably, the use of Co Loa would also have been symbolically potent, with the new dynasty claiming links to a distant, pre-Sinitic past. That event, combined with the decision of various Vietnamese regimes in succeeding centuries to refurbish and amplify Co Loa’s rampart system, speaks to the ongoing cultural and political significance of the locale. Essentially, power, war, and social memory had been inscribed into the landscape for future generations to continue reading. In the end, the Co Loa case offers an insightful glimpse into how local trajectories of change—as related to competition, coercive power, and warfare—can shape and influence lifeways, settlement decisions, and human-environment interactions. The case is even more interesting when one considers the larger, interregional circumstances, with Bac Bo societies being located on the margins of a burgeoning imperial power to the north. ACKNOWLEDGMENTS

The authors thank Juan Carlos Vargas Ruiz and Hugo C. Ikehara-Tsukayama for the invitation to contribute this chapter. We also wish to thank our Vietnamese colleagues for their efforts during our field investigations, specifically members of the Vietnam Institute of Archaeology, the Thang Long–Hanoi Heritage Conservation Center, and the people of the Co Loa Commune. REFERENCES

Allen, Mark. 2014. “Hunter-Gatherer Conflict: The Last Bastion of the Pacified Past?” In Violence and Warfare among Hunter-Gatherers, edited by Mark Allen and Terry Jones, 15–25. Walnut Creek, CA: Left Coast Press. Arkush, Elizabeth, and Charles Stanish. 2005. “Interpreting Conflict in the Ancient Andes.” Current Anthropology 46 (1): 3–28. Bellwood, Peter. 2005. The First Farmers: The Origins of Agricultural Societies. Malden, MA: Blackwell Publishers. Calo, Ambra. 2009. The Distribution of Bronze Drums in Early Southeast Asia: Trade Routes and Cultural Spheres. BAR International Series 1913. Oxford: Archaeopress. Dega, Michael. 1999. “Circular Settlements within Eastern Cambodia.” Indo-Pacific Prehistory Association Bulletin 18 (2): 181–190. Estabrook, Virginia. 2014. “Violence and Warfare in the European Mesolithic and Paleolithic.” In Violence and Warfare among Hunter-Gatherers, edited by Mark Allen and Terry Jones, 49–69. Walnut Creek, CA: Left Coast Press.

I N S C R I B I N G P OW ER O N A L A N D S CA P E

135

Fletcher, Roland, Dan Penny, Damian Evans, Christophe Pottier, M. Barbetti, Matti Kummu, Terry Lustig, and Authority of the Protection and Management of Angkor and the Region of Siem Reap (APSARA) Department of Monuments and Archaeology Team. 2008. “The Water Management Network of Angkor, Cambodia.” Antiquity 82 (317): 658–70. Golitko, Mark. 2015. LBK Realpolitik: An Archaeometric Study of Conflict and Social Structure in the Belgian Early Neolithic. Archaeopress Archaeology Series. Oxford: Archaeopress. Haas, Jonathan. 2001. “Warfare and the Evolution of Culture.” In Archaeology at the Millennium: A Sourcebook, edited by Gary Feinman and T. Douglas Price, 329–350. New York: Kluwer/Plenum. Hanebuth, Till J.J., Yoshiki Saito, Susumu Tanabe, Quang Lan Vu, and Quang Toan Ngo. 2006. “Sea Levels during Late Marine Isotope Stage 3 (or Older?) Reported from the Red River Delta (northern Vietnam) and Adjacent Regions.” Quaternary International 145–146: 119–134. Higham, Charles. 2002. Early Cultures of Mainland Southeast Asia. Bangkok: River Books. Higham, Charles. 2014. Early Mainland Southeast Asia: From First Humans to Angkor. Bangkok: River Books. Hill, Paul, and Jill Wileman. 2002. Landscapes of War: The Archaeology of Aggression and Defence. Charleston: Tempus Publishing. Hoang, Xuan Chinh, and Van Tien Bui. 1980. “The Dongson Culture and Cultural Centers in the Metal Age in Vietnam.” Asian Perspectives 23 (1): 55–65. Keeley, Lawrence. 1996. War before Civilization. Oxford: Oxford University Press. Keeley, Lawrence, Marisa Fontana, and Russell Quick. 2007. “Baffles and Bastions: The Universal Features of Fortifications.” Journal of Archaeological Research 15 (1): 55–95. Kim, Nam C. 2013. “Cultural Landscapes of War and Political Regeneration.” Asian Perspectives 52 (2): 244–267. Kim, Nam C. 2015. The Origins of Ancient Vietnam. Oxford: Oxford University Press. Kim, Nam C. 2017. “Coercive Power and State Formation in Northern Vietnam.” In Feast, Famine or Fighting? Multiple Pathways to Social Complexity, edited by Richard Chacon and Ruben Mendoza, 165–196. New York: Springer. Kim, Nam C., and Marc Kissel. 2018. Emergent Warfare in Our Evolutionary Past. New York: Routledge. Kim, Nam C., Chapurukha Kusimba, and Lawrence Keeley. 2015. “Coercion and Warfare in the Rise of State Societies in Southern Zambezia.” African Archaeological Review 32 (1): 1–34.

136

N A M C . K I M A N D R U S S ELL S. Q U I C K

Kim, Nam C., Lai Van Toi, and Trinh Hoang Hiep. 2010. “Co Loa: An Investigation of Vietnam’s Ancient Capital.” Antiquity 84 (236): 1011–1027. Lai, Van Toi. 2014. Đền Thượng Cổ Loa: Và Những Bí Ẩn Trong Lòng Đất [Thuong Temple of Co Loa: Mysteries in the Ground]. Hanoi: Nha Xuat Ban Chinh Tri Quoc Gia. (In Vietnamese). Lai, Van Toi. 2015. “Co Loa: the Capital of the Au Lạc Kingdom in the 3rd and 2nd Centuries bce.” In Perspectives on the Archaeology of Vietnam, edited by Andreas Reinecke, 129–156. Bonn: German Archaeological Institute. Lambert, Patricia. 2007. “Ethnographic and Linguistic Evidence for the Origins of Human Trophy-Taking in California.” In The Taking and Displaying of Human Trophies by Amerindians, edited by Richard J. Chacon and David H. Dye, 65–89. New York: Springer. Le, Van Tiem, and Le Quoc Doanh. 2001. “Soil Properties of the Red River Delta.” In Long Climate Change and the Environment Change of the Lower Red River Delta, edited by Shigeko Haruyama, Jun Matsumoto, Yumio Sakurai, Le Quoc Doanh, Le Van Tiem, and Le Khanh Phon, 5–11. Hanoi: Agriculture Publishing House. LeBlanc, Steven. 2014. “Forager Warfare and Our Evolutionary Past.” In Violence and Warfare Among Hunter-Gatherers, edited by Mark Allen and Terry Jones, 26–46. Walnut Creek, CA: Left Coast Press. Li, Xiuzhen Janice, Andrew Bevan, Marcos Martinon-Torres, Thilo Rehren, Wei Cao, Yin Xia, and Kun Zhao. 2014. “Crossbows and Imperial Craft Organization: The Bronze Triggers of China’s Terracotta Army.” Antiquity 88 (339): 126–140. Li, Zhen, Yoshiki Saito, Eiji Matsumoto, Yongji Wang, Susumu Tanabe, and Quang Lan Vu. 2006. “Climate Change and Human Impact on the Song Hong (Red River) Delta, Vietnam, during the Holocene.” Quaternary International 144: 4–28. Martin, Debra, Ryan Harrod, and Ventura Perez. 2012. “Introduction.” In The Bioarchaeology of Violence, edited by Debra Martin, Ryan Harrod, and Ventura Perez, 1–10. Gainesville: University Press of Florida. McGrath, Roger, and William Boyd. 2001. “The Chronology of the Iron Age ‘Moats’ of Northeast Thailand.” Antiquity 75 (80): 349–360. Moore, Elizabeth. 1988. Moated Sites in Early North East Thailand. British Archaeological Reports International Series 400. Oxford: Archaeopress. Moore, Elizabeth. 1992. “Water Enclosed Sites: Links between Ban Takhong, Northeast Thailand and Cambodia.” In The Gift of Water: Water Management, Cosmology and the State in South East Asia, edited by J. Rigg, 26–46. London: School of Oriental and African Studies. Moore, Elizabeth, and San Win. 2007. “The Gold Coast: Suvannabhumi? Lower Myanmar Walled Sites of the First Millennium A.D.” Asian Perspectives 46 (1): 202–232.

I N S C R I B I N G P OW ER O N A L A N D S CA P E

137

Murowchick, Robert. 2001. “The Political and Ritual Significance of Bronze Production and Use in Ancient Yunnan.” Journal of East Asian Archaeology 3(1/2): 133–92. Nguyen, Giang Hai. 2005. “Ancient Metallurgy in Vietnam: An Ethno-Archaeological Investigation.” Bulletin of the Indo-Pacific Prehistory Association 25: 121–123. Nguyen, Quang Ngoc, and Van Quan Vu. 2007. Dia Chi Co Loa [The Co Loa Site]. Hanoi: Nha Xuat Ban Ha Noi. Nishimura, Masanari. 2005. “Settlement Patterns on the Red River Plain from the Late Prehistoric Period to the 10th century ad.” Bulletin of the Indo-Pacific Prehistory Association 25: 99–107. O’Harrow, Stephen. 1979. “From Co-loa to the Trung Sisters’ Revolt: Vietnam as the Chinese Found It.” Asian Perspectives 22 (2): 140–163. O’Reilly, Dougald. 2008. “Multivallate Sites and Socio-economic Change: Thailand and Britain in Their Iron Ages.” Antiquity 82 (316): 377–389. Pham, Minh Huyen. 2004. “Northern Vietnam from the Neolithic to the Han Period—Part II: The Metal Age in the North of Vietnam.” In Southeast Asia: From Prehistory to History, edited by Ian Glover and Peter Bellwood, 189–201. New York: Routledge. Quick, Russell. 2010. “Refuge Fortifications in the Hesbaye Region of Liege Province, Belgium: An Explanation for Linienbandkeramik Site ‘Clustering.’” PhD diss., University of Illinois at Chicago. Quick, Russell. 2019. “2017 Archaeological Investigations of the Bronze Age Fortification and Citadel at Co Loa, Dong Anh, Ha Noi, Viet Nam: Ground Truthing, LiDAR, GIS, and Geophysics.” Unpublished report. Sterling, Eleanor, Martha Hurley, and Le Duc Minh. 2006. Vietnam: A Natural History. New Haven, CT: Yale University Press. Taylor, Keith. 2013. A History of the Vietnamese. Cambridge: Cambridge University Press. Trinh, Sinh. 2015. “The First States in North Vietnam.” In Perspectives on the Archaeology of Vietnam, edited by Andreas Reinecke, 71–84. Bonn: German Archaeological Institute. von Falkenhausen, Lothar. 2008. “Stages in the Development of ‘Cities’ in Preimperial China.” In The Ancient City: New Perspectives on Urbanism in the Old and New World, edited by J. Marcus and J. Sabloff, 209–228. Santa Fe: School for Advanced Research Press. Yang, Bin. 2004. “Horses, Silver, and Cowries: Yunnan in Global Perspective.” Journal of World History 15 (3): 281–322. Yao, Alice. 2016. The Ancient Highlands of Southwest China: From the Bronze Age to the Han Empire. Oxford: Oxford University Press.

138

N A M C . K I M A N D R U S S ELL S. Q U I C K

7 Now the true account of the road in question is the following: royal stations exist along its whole length, and excellent caravanserais; and throughout, it traverses an inhabited tract, and is free from danger. In Lydia and Phrygia there are twenty stations within a distance of 94 1/2 parasangs. On leaving Phrygia the Halys has to be crossed; and here are gates through which you must needs pass ere you can traverse the stream. A strong force guards this post.

Fortified Roads as Communication and Defense Networks in the Ancient Near East Tiffany Earley-Spadoni

Herodotus, H ISTORIES , Book V

Much has been made, historically speaking, of the famed Royal Road network in Achaemenid Persia, mostly thanks to the promotion of its many attributes by Herodotus. According to the various sources, we learn that the king’s roads boasted a series of small forts to ensure safe passage and also served as a kind of pony express, linking up the empire with integrated networks of message relays (Briant 2002, 2012; Graf 1994; Kuhrt 1995; Wiesehöfer 2001). A system of fire-beacon-signaling stations is another marvel of the Persian road described by Herodotus, and these claims are supported by archaeological evidence at Anatolian sites (Dusinberre 2013). Additional details from the Persepolis Fortification Tablets supplement Herodotus’s account by describing officials charged with duties such as surveilling the road or managing “express mail” services (Briant 2012; Hallock 1969). Yet, the success and overall structure of the Persian road

https://doi.org/10.5876/9781646422111.c007

139

network owe much to older ancient Near Eastern models of road and fort construction, a fact which is rarely appreciated. While historical studies frequently treat the role of warfare in ancient Near Eastern state development, landscape studies of the region have preferred to focus on economic or ecological explanations for change in the past. With the intention of disrupting this narrative, I will focus on two case studies, one from Middle Bronze Syria and the second from Iron Age Assyria. The cases presented in this chapter demonstrate that the exigencies of warfare and strategic communication shaped subsequent landscape formation. To achieve this goal, I incorporate archaeological data with textual documentation, primarily from epistolary corpuses. In addition to communication schemes such as fire beacons and smoke signals, fortress networks were powerful platforms across which to gather and distribute information such as espionage reports and letters across vast imperial hinterlands. In the context of this volume, examples from the ancient Near East offer the opportunity to situate archaeologically attested landscapes of warfare within known historical frameworks. THE LANDSCAPE CONTEXT OF WARFARE IN THE ANCIENT NEAR EAST

A regional approach is critical for situating defensive structures within their meaningful, socially constructed contexts (Fachard 2016). Historical accounts around the world describe systems of fortresses that serve as defensive, communication, and mobility networks; therefore, the structures cannot be studied in isolation (Earley-Spadoni 2015b; Swanson 2003). Furthermore, an approach that focuses upon the landscape context of defensive structures is more likely to yield information about nonelite people in the past, revealing the more ephemeral traces of past habitation and land use (Greene and Lindsay 2012; Hammer 2014). Last, studying nonsite level features is the only way to understand rural, dispersed phenomena (Franklin and Babajanyan 2018). In sum, a comprehensive study of patterns of conflict requires a large-scale, regional perspective, in addition to site-level and off-site studies (Arkush 2011; Arkush and Allen 2006). Structures such as fortresses are best studied within their landscape contexts because the defensive origins of certain features can shape subsequent regional development. Tells, large mounds formed from millennia of occupational debris, are the most characteristic of ancient Near Eastern settlement types and are, arguably, the product of landscapes of pervasive warfare. The defense preoccupations represented by the construction of outer defensive walls are among the social practices that led to tell development in the ancient 140

T I F FA N Y E A R LE Y- S PA D O N I

Near East. Wilkinson hints at the defensive origin of these features, noting that one of a variety of important cultural factors in the formation of tell mounds is the presence of an outer defensive wall that inhibits the erosional expansion of the sediments, thereby damming them up (2003, 108). In order for tells to develop, there must be a compelling reason for people to stay in one place, such as the need to stay within a defensive wall. Defensive structures alter their surrounding landscapes in important ways. Elevated places such as tells and fortresses, besides providing impressive views out onto a landscape, are also highly visible from the landscape itself. Prominent structures act as focal anchor points when moving through landscapes (Ristvet 2014). Their presence over time creates a socially constructed place that is symbolically embedded within its surroundings and with which individuals and communities create strong ties and enduring relationships (Bachhuber 2014). Visual presence and cultural cachet are some of the special qualities of prominent places such as tells and fortresses. A long history of establishment and the connection to ancestors implied in constructions such as tells would have been a source of social power that priests and kings alike could have drawn upon (Wilkinson 2003, 108). Moreover, places that evoke the social memory of violence, such as fortresses, may also be complicit in reproducing violence and beget future outbreaks of conflict (Rowe 2007). Accordingly, archaeologists have grappled with the implications of material agency in recent years (Smith 2015; Tilley 2004), reflections that should be applied to investigating regional phenomena. Defensive concerns and the significance of place are not mutually exclusive categories, but are, instead, mutually reinforcing. Besides obvious utility in defending ancient cities from threat, city walls came to be highly symbolic of urbanism, kingship, and place more generally (Creekmore 2014; Ristvet 2007). It is clear from the opening lines of the Mesopotamian Epic of Gilgamesh that the great walls of Uruk were a symbol of civic pride and were associated with the figure of the king (George 2003, 539). In Gilgamesh, the city walls are Uruk. These structures became complex and mutually reinforcing symbols of defense, authority, and urbanism more generally. Yet, the practical needs of defense remain an important part of the social origins of city walls. FORT AND FIRE BEACON NETWORKS IN THE MIDDLE BRONZE AGE

This chapter will discuss two different time periods and geographical settings in the ancient Near East. The first of these is the early second millennium bce F O RT I F I ED ROA D S A S CO M M U N I CAT I O N A N D D EF EN S E N E T WO R K S

141

in northern Syria (ca. 2000–1800 bce) during the Middle Bronze Age (MBA). In addition to the archaeological remains known from the region, epistolary archives from the palace at Mari document the affairs of the rulers of the state and their associated functionaries.1 Many of the letters date to the reigns of Shamshi-Adad and Zimri-Lim, rulers who presided over the territory in the time preceding the conquest of the territory by none other than Hammurabi. The landscape of northern Syria in the second millennium bce was dominated by highly fortified city-states with associated hinterlands that were subjected to dominion by a constantly shifting cast of rulers (Charpin and Ziegler 2003). The predominant anthropogenic landscape features in the undulating northern Syrian landscape are tells. The walled cities of the MBA were reconstructed on tells after a widespread urban collapse at the end of the third millennium. Architecture overwhelmingly emphasized practical notions of defense, which included the fortification of towns and cities with mud-brick enclosure walls built upon stone foundations, towers, glacis structures, and ramparts (Akkermans and Schwartz 2003, 321; Burke 2008; Rousset et al. 2017). The historical sources indicate that siege warfare was central to the military activities of the city-states in question (Earley-Spadoni n.d.), a reality the helps to explain the high level of investment in city defenses. Although diplomacy and economic exchange were important modes of interaction, relations among the various states in northern Syria in the early second millennium bce can be characterized as persistent and nearly constant warfare (Schwartz 2013). The environmental and social consequences of the constant warfare in the region were, arguably, dire as evidenced by a deurbanization of the region and an overall reduction of the resident population that attended the end of the period (2013, 8). Ideologies also contributed to the warfare of the period and the experience of places. An extraordinary value was attached to a view of leadership that was inextricably linked to prowess in battle and the performance of a militarized masculinity that is readily apparent in the art of the period (Bahrani 2008; Winter 1996). Old Babylonian divinatory texts from the northern Mesopotamian city-state of Tigunānum relating to the practice of teratomancy also reveal the striking use of androcentric and martial imagery 1. Mari, modern Tell Hariri, was an important state in the middle Euphrates region in the early second millennium (for a summary and additional bibliography, see Akkermans and Schwartz 2003, 313–317). The Mari Letters have been published as the result of decades of research, primarily under the auspices of the French project, Archives Royal de Mari (ARM).

142

T I F FA N Y E A R LE Y- S PA D O N I

(De  Zorzi 2017).2 While the texts are clearly inspired by southern models, the Tigunānum versions rework the originals to emphasize agonistic masculinity. For example, the enemy Other is feminized, is dehumanized through the comparison to animals, and must be conquered through coercion, often sexual. The forts, fortresses, and towers of the MBA constituted fire-beaconsignaling networks. In particular, Georges Dossin (1938) assembled the extensive evidence for fort- and fortress-based fire beacons at Mari from dozens of letters pertaining to them discovered in the palace archive. In one letter, an official named Banum writes his king (probably Zimri-Lim) a message stating that while traveling along the road north from Mari, he saw the fire beacons being lit town by town by the Yaminites in the district of Terqa (Dossin 1938, 178). Banum is unsure why the fire beacons have been lit, but plans to write back with more details just as soon as he has them. In the meantime, he recommends that the Mari city defenses be reinforced, indicating fire beacons were being used as an early warning system to signal other stations situated along the road about incoming threats. Although the ancient writer does not know why the beacons have been lit, there seems to be a mutual understanding between Banum and the intended recipient that these signals are meant to indicate danger. Additional letters are relevant to the discussion of communications among forts. In the first of these, a functionary apologizes if the lighting of the beacons worried his lord and explains that the Yaminite cities remain in a state of revolt. The king’s loyal servant, the letter reassures, lit the fires to summon reinforcement troops, and his majesty should not be overly concerned (Dossin 1938, 181). Yet another letter informs Zimri-Lim that an attack near Terqa is imminent, and his general Sammetar has a force assembled in the general area to meet the attack. When the fire beacon is lit at the location of attack, unknown at the moment of writing, Sammetar reassures the king that he will come to the rescue (182–183), indicating that there was previous agreement about what the meaning of a lit fire would be. In a third letter, a certain Zindria responds to a complaint from the king. Zindria writes that in order to avoid future confusion, he will muster the troops and signal to other stations only when he sees two fires lit rather than one (183). The described letter indicates that fire communication was at times an imperfect system, that the meaning of signals could be contextual and agreed upon in advance, and that a mustering of troops at stations situated along the roads was a common response to 2. Teratomancy is the observation of malformed or atypical human or animal births. F O RT I F I ED ROA D S A S CO M M U N I CAT I O N A N D D EF EN S E N E T WO R K S

143

[128.104.46.206] Project MUSE (2024-03-01 20:28 GMT) UW-Madison Libraries

the lighting of the beacons. While fire beacon signaling was certainly fast, it left something to be desired in terms of its capability for message complexity. A final Mari letter shows the importance of state-run systems of fortresses (ARM 5 65). Ašqudum, a diviner and obsequious official (Heimpel 2003, 529), writes the king to inform on his colleague who has not completed the required extispicy rituals, the cultic evaluation of sheep entrails. The snitch adds an additional assurance that the necessary protocols have been followed at fortresses under his own jurisdiction, revealing that the continued safety of the fortresses was of such vital importance to the state that they would entreaty the gods to understand future risks. Organized systems of forts and fire beacon stations situated along roads were not an innovation of the later Neo-Assyrian or Persian empire, as is sometimes claimed in the scholarly literature, since such systems were in use by the state of Mari and its neighbors by the early second millennium bce. Letters describe the presence of waystations situated along roads (Heimpel 2003, 116), and archaeological investigations have revealed evidence of intentionally constructed fortress networks. The Arid Margins of Northern Syria project (Rousset et al. 2017) has cataloged more than one hundred Middle Bronze sites, and one of its central research goals is to investigate the observed uptick in defensive network construction during the period. Team members note a multitiered regional defense network comprised of forteresses, fortins, tors, petites tours, and villes fortifiées (fortresses, forts, large towers, small towers, and fortified settlements, respectively). Regular spacing and intervisibility were important factors for the placement of structures in the landscape, an observation that is consistent with the ample textual documentation from the MBA. Not only would the regular spacing of structures (ca. 20 km) have ensured excellent visibility, such spacing also ensures safety by providing regular rest stations where travelers can overnight. The team also observes that surveillance and protection of transportation corridors were together an important function of the intentionally constructed regional networks, and it notes that the placement of ancient towers sometimes coincides with the placement of modern telecom emplacements, which I have noted in my own investigation of Iron Age forts and fortresses in Armenia. Such placement suggests redundancy and intervisibility of emplacements, which is the case in modern telecommunications networks. In sum, a number of salient traits of defensive networks emerge from MBA Syria. The forts and fortresses played a vital role in regional communication, as evidenced by their role in fire beacon communications. Furthermore, the sites served as garrisons from which soldiers could be mustered in times of 144

T I F FA N Y E A R LE Y- S PA D O N I

peril. The fire beacon signals in the Mari Letters have only one recorded meaning—danger—so their role in regional defense is clear. Moreover, fortified structures along transportation corridors, and roads ensured the efficient functioning of imperial communication as well as provided safety along the road by functioning as waystations. The archaeological survey data show that the kingdoms of MBA Syria sought to address the risk of violence by a heavy investment in defensive and communication infrastructure. Finally, the defensive structures were part of larger, intentionally constructed, multitier networks. FORTRESS NETWORKS IN THE NEOASSYRIAN EMPIRE AND URARTU

The second case to be discussed is from the Neo-Assyrian empire, specifically the eighth century bce, with a particular emphasis upon its interactions with, Urartu, a powerful highland state centered upon fortresses and a formidable enemy. Urartu was located in the mountains of modern-day eastern Turkey, Armenia, and northwest Iran, while the Neo-Assyrian state encompassed much of the remainder of the ancient Near East. Warfare was integral to the Neo-Assyrian state identity, and constant warfare was the operational environment of the state. Assyrian military tactics are well documented in textual sources and represented in elaborate pictorial scenes in contexts such as palaces (Bahrani 2008). While the Neo-Assyrians are famous for their large standing armies and siege warfare tactics, strategies of psychological terror served as both state aesthetic and propaganda. Scenes depicting atrocities such as the flaying of live victims or the counting of the heads of fallen enemies are but two well-known examples of the violence that dominates the state-sponsored art of the period (Bagg 2016; Dolce 2016). Much like the Middle Bronze case, the Assyrians practiced an agonistic masculinity predicated upon feminizing and dehumanizing enemies with violence (De Zorzi 2017, 144–46; Kessler Guinan 1997). The Assyrians used military aggression and the threat of violence as ways to quell a large, multiethnic empire that featured large, urban centers (Melville 2016; Radner 2011). Rather than engage in more direct military tactics such as pitched battle on the open field, the Assyrians would approach enemy territories en masse and attack villages and cities that presented easy, lucrative targets, committing atrocities upon the populations in question if they did not immediately surrender (Van de Mieroop 2007, 248). The Neo-Assyrian case, like Middle Bronze Syria, can also be described as existing in a state of nearconstant warfare (Earley-Spadoni 2015b; Ristvet 2018). F O RT I F I ED ROA D S A S CO M M U N I CAT I O N A N D D EF EN S E N E T WO R K S

145

Epistolary corpora are central to contextualizing the role of fortresses in the administration of the Neo-Assyrian state and its defense. In particular, letters from the reign of Sargon II (late eighth century bce) discovered in archives at Nineveh reveal the polyvalent functioning of forts and fortresses (Fuchs and Parpola 2001; Lanfranchi and Parpola 1990; Parpola 1987). From the letters, we learn of the existence of bēt mardēti, Assyrian forts that served as waystations along roads (e.g., SAA 1 177, SAA 10 361).3 Messages exchanged among Assyrian officials could be delivered either by a relayed letter (e.g., ABL 1021) or personally, by envoy, the preferred method in preceding eras (Favaro 2007).4 The former method was considerably faster, however, though it would have required more organization to achieve. While fort-lined roads that served as communication conduits were not new, the concept of kalliu (message relays) may have been an innovation of the Neo-Assyrian empire (Kessler 1997; Parpola 1987, xiv). Assyrian roads were highly organized systems complete with specific infrastructure and staffed with a variety of specialized officials. Official state correspondence (ABL 414, 1021) indicates that bēt mardēti lined the major corridors (Graf 1994, 171; Parpola 1987, xiv). The waystations were spaced at regular intervals along the roads (e.g., ADD 1096).5 The letters, furthermore, describe a variety of officials charged with ensuring communication along the state roads of Assyria, including mār šipri (messengers or envoys), kallȗ (express messengers), kallāp šipirti (military messengers), and raksū (escort riders). Royal letters were delivered by members of the ša-qurbūti (imperial guard), an elite corps consisting of eunuchs, the unsullied of the Assyrian world. Rab kallie (officials charged with ensuring the delivery of mail) were assigned to forts. Some forts had a residential function meant for housing soldiers and other officials, and forts could also be positioned along rivers that were, along with roads, important transportation conduits (Parker 1997). Information could only be moved from one part of the empire to another through a network of roads (Parpola 1987, xiii). The most informative text for reconstructing the communications infrastructure in the empire is a letter that 3. The State Archives of Assyria (SAA) is a series of translations and critical editions of Neo-Assyrian texts published by the Neo-Assyrian Text Corpus Project at the University of Finland. 4. The abbreviation ABL refers to Assyrian and Babylonian Letters, which were excavated at the site of Nineveh. The letters in question were originally published in the early twentieth century by the University of Chicago as a series of volumes. 5. The abbreviation ADD refers to Assyrian Deeds and Documents, a series of volumes originally published in the early twentieth century by Cambridge.

146

T I F FA N Y E A R LE Y- S PA D O N I

describes fortified relays spaced along a road, in northern Syria (SAA 1 177), and the letter proposes the organization of another relay. The waystation at Hēsa was administered by a rab kallie, a postmaster, illustrating the link between forts and strategic communication. An intriguing point about Hēsa related in the letter is that the village needed thirty new families to ensure proper staffing at the waystation. One implication is that the requested workers may have also been responsible for the construction and maintenance of road infrastructure. Regional insecurity and warfare shaped the imperial road system. The military ties to the forts situated along the road is logical in this part of NeoAssyrian empire, where numerous problems vis-à-vis troublesome desert people, insurgents, and fugitives would have made an official Assyrian presence desirable in the region (Kessler 1997; Richardson 2016). Daniele Morandi Bonacossi interprets the formation of landscapes in the Syrian Jezireh as linked to the intense traffic created by the systematic warfare of Assyria, arguing that the continuous stream of soldiers, merchants, prisoners of war, booty from military campaigns, taxes, and tribute would have had a dramatic effect on the local economy and the social formation of landscapes (2000). In other words, the Neo-Assyrian roads do not simply pass through places but are embedded within and influenced by a rich social matrix. Roads shape and are shaped by social processes, which are related to the exigencies of expansionary states and systematic warfare. The Assyrians maintained a network of forts along the roads, rivers, and frontiers of their empire, and defense and security were the primary motivations for investment in this infrastructure. Forts within the empire would have served the purpose of keeping a watchful eye not just on external invaders but also upon the Assyrian subjects themselves. In fact, the letters in the Sargonid corpus are the products of an elaborate, tiered scheme of espionage gathering that featured various levels of spy “middle managers” whose job it was to compile reports for the king and his functionaries (Dubovský 2006). A formulaic element of many letters is a reassurance that all of the Assyrian forts are well; that is, the forts have not been attacked or captured. Such knowledge implies that the Neo-Assyrian informants were well placed to observe the comings and goings within the empire—in some cases from within the forts themselves. In one letter, the crown prince of Assyria writes the king that all of the guards at the Assyrian forts along the border have sent him the same account regarding an Urartian defeat (SAA 1 31), suggesting that the Assyrian forts, located along roads, constituted a network of surveillance. Moreover, the described report implies that the regional network of spy administrators collected information directly from fort personnel. F O RT I F I ED ROA D S A S CO M M U N I CAT I O N A N D D EF EN S E N E T WO R K S

147

A systematic study of the landscape correlates of Neo-Assyrian warfare is desperately needed at the regional level. While historical and art historical studies of Mesopotamian warfare have been frequent in recent years (Bagg 2016; Bahrani 2008; Fales 2010; Melville 2016; Radner 2012), landscape studies focusing upon the traces of ancient warfare are rare. Nonetheless, the systems of regional and imperial road networks (i.e., harrān šarri, hūl šarri) known from textual documentation (Fales 1990) may or may not be archaeologically discernable. An analysis of “hollow ways”6 associated with Neo-Assyrian sites in satellite imagery suggests a hierarchical road network (Wilkinson et al. 2005), although this interpretation is disputed. Specifically, hollow ways cannot be securely dated to the Neo-Assyrian period since they are negative features without associated material culture (Ur 2012, 527, 2017). Meanwhile, scholars of the neighboring and contemporaneous highland empire of Urartu have studied the archaeological evidence for evidence of regional networks of defense and communication (Earley-Spadoni 2015b; Zimansky 1985). A multiregional study of hundreds of Early Iron (EI) and Urartian archaeological sites in northwest Iran has revealed a multitiered system of fortresses, forts, fortified settlements, and towers, mostly situated along roads (Earley-Spadoni 2015a). As in the MB Syrian case, there was an effort to space forts and fortresses along the roads for a variety of reasons, particularly to surveil roads and to ensure the safety of travelers (figure 7.1). In the Urartian case, critical intersections of roads were prioritized (Earley-Spadoni 2015a). Forts that function as waystations are well known from the Urartian context; the Urartian fort at Getap-1 (figure 7.2) in Vayots Dzor, Armenia, features a central beacon-signaling platform as well as structures that would have been ideal for stabling animals such as horses (Melkonyan et al. 2010). While archaeological approaches that utilize social network theory have become more frequent in recent years (Brughmans 2010; Collar et al. 2015), the scholarly literature has only begun to explore the placement of landscape features to create robust communication networks (Earley-Spadoni 2015b). One particular Neo-Assyrian source merits special attention here, since the text relates important information about defensive communication in the Urartian state. The Neo-Assyrian text, known alternately as “The Eighth Campaign of Sargon II” or the “Letter to Aššur,”7 is a detailed account describing an 6. Hollow ways are shallow, relatively straight valleys that radiate out from Mesopotamian tell sites. While less visible from the ground, they are quite visible using aerial reconnaissance techniques such as the analysis of declassified CORONA imagery (Ur 2012; Wilkinson 2003, 62). 7. See Foster (2005), Luckenbill (1927), and Thureau-Dangin (1912) for various editions.

148

T I F FA N Y E A R LE Y- S PA D O N I

Figure 7.1. Visibility channeled around roads on Urmia plain, Urartian period (fortresses large hexagons, forts small hexagons, and settlements points).

extensive military campaign by Sargon II to Iranian Azerbaijan in the late eighth century bce. The “Letter” is addressed to the chief god of Assur, the other gods of his temple, and the people of the city, and the text styles Sargon as the writer. In one scene, the letter describes the Urartians lighting an elaborate system of fire beacons as a sign of their dread at the arrival of Sargon’s army. Archaeological correlates of the fire beacon communication described in the Assyrian sources are discernable in the archaeological record. A dense network of intervisible sites can be observed south of Lake Sevan in Armenia and has been the object of systematic GIS and social network analysis studies (Earley-Spadoni 2015b). The dense interconnections of sites during the EI (figure 7.3) and subsequent Urartian annexation (figure 7.4) mirror levels of redundancy found in modern telecommunications networks. Statistical

F O RT I F I ED ROA D S A S CO M M U N I CAT I O N A N D D EF EN S E N E T WO R K S

149

Figure 7.2. Urartian waystation of Getap-1.

verification methods applied in the study suggest that the fort and fortress networks were not the result of the organic development of the state. Instead, the sites were constructed as an integrated communications system requiring an extraordinary level of cooperation among the ancient inhabitants of the region. Some sites, for example, function as relays, or intermediate sites that forward messages from other sites in the network, a strategy that has parallels to the Assyrian system of relayed letters. DISCUSSION

A renewed interest in the study of ancient road networks suggests that much can be learned by exploring the multifaceted societal and cultural context of these landscape features (Alcock et al. 2012, 1–5; Babajanyan and Franklin 2018), but the relationship of roads to warfare and systems of regional defense has been underappreciated. Complex cultural constructions such as the royal roads of Assyria are unintelligible when divorced from their social matrix, which is inextricably linked to the war machine that its existence served. In 150

T I F FA N Y E A R LE Y- S PA D O N I

Figure 7.3. Intervisibility of Early Iron fortified sites.

Figure 7.4. Intervisibility of Urartian fortified sites.

fact, harrānu, the Assyrian word for road, is also the word for military campaign.8 Even at the lexical level, the concepts of road and war were associated. Improved roads and rapid communication would have served the purpose of facilitating the Neo-Assyrian empire’s systematic warfare. One of the primary purposes of Assyrian roads was to facilitate its systematic warfare and expansionist aims. The use of forts and fortresses along roads to conduct the strategic communication needed to run empires has heretofore been underappreciated in landscape studies and the archaeology of empire alike. The evidence suggests that fortresses play a valuable role as foci of communication, acting as hubs along which information is moved across landscapes. However, Urartu and Assyria’s networks of surveillance, at times, may have made each of these states more vulnerable to attacks from the outside since their respective systems of information collection could also provide information to the rival state by way of treacherous double agents. The Sargonid letters give the impression that corruption and double agents were widespread. The case of Hu-Teshub, the king of the Anatolian buffer state of Šubria, is a particularly relevant example since the sources disclose that he compiled a detailed report on the Assyrians for the Urartian king while on other occasions he supplied the Assyrians with information on the Urartians (SAA V 44, SAA V 45, SAA V 35). In other words, elaborate schemes of surveillance and information collection may contribute to the fragility of states in unanticipated ways. Several common traits emerged from the ancient Near Eastern case studies evaluated above. Road networks tend to be organized by states on a regional level as multitiered systems. Forts served as waystations that guaranteed safety along the road and were also important hubs in communications networks. Specialized officials presided over the administration of the roads as well as the transmission of a variety of state and private communications. Rapid but imprecise fire beacon signaling complemented express delivery services. In the ancient Near East, all known cases of fire beacon communication are in context of danger. The forts provided an excellent way for the ancient states to surveil the comings and goings along the roads. What emerges is a template of fortified roads that was remarkably durable over time (Elizabeth Arkush, chapter 13 in this volume). Models of road construction developed in the context of expansionary ancient Near Eastern states were undoubtedly influential in the later development of imperial road networks in the Persian, Hellenistic, Roman, and Byzantine spheres. All of the salient traits of the famed Persian Royal 8. Chicago Assyrian Dictionary: “h”; Akkadisches Handwörterbuch I: 4.

152

T I F FA N Y E A R LE Y- S PA D O N I

Road—such as fire beacons, waystations, and message relays—have predecessors in the ancient Near East. Based upon the various cases discussed from the ancient Near East, a number of common traits emerge in the discussed road networks. While it is tempting to conceive of a set of more-or-less universal principles that defined secure roads given the coherence of the discussed Syrian, Neo-Assyrian, Urartian, and Persian examples, future research may certainly elucidate historical particulars that are unique to certain places and times. ACKNOWLEDGMENTS

I would like to express my gratitude to the editors of this volume for their organization of the original Society for American Archaeology session in Washington, DC, and for their subsequent efforts to produce this volume. I also thank Elizabeth Arkush and Glenn Schwartz, who supplied helpful comments on my manuscript. I thank Artur Petrosyan and Teagan Wolter for supporting drone photography at Getap-1. REFERENCES

Akkermans, Peter M.M.G., and Glenn M. Schwartz. 2003. The Archaeology of Syria: From Complex Hunter-Gatherers to Early Urban Societies (ca. 16,000–300 BC). Cambridge: Cambridge University Press. Alcock, Susan, John Bodel, and Richard Talbert, eds. 2012. Highways, Byways, and Road Systems in the Pre-modern World. Malden, MA: John Wiley and Sons. Arkush, Elizabeth. 2011. Hillforts of the Ancient Andes: Colla Warfare, Society and Landscape. Gainesville: University Press of Florida. Arkush, Elizabeth, and Mark Allen, eds. 2006. The Archaeology of Warfare: Prehistories of Raiding and Conquest. Gainesville: University Press of Florida. Babajanyan, Astghik, and Kathryn Franklin. 2018. “Everyday Life on the Medieval Silk Road: VDSRS Excavations at Arpa, Armenia.” ARAMAZD 12 (1): 155–183. Bachhuber, Cristoph. 2014. “Citadels in Spectacle-Scapes in Bronze Age Anatolia.” In Approaching Monumentality in Archaeology, edited by James F. Osborne, 291–310. Albany: State University of New York Press. Bagg, Ariel. 2016. “Where Is the Public? A New Look at the Brutality Scenes in Neo-Assyrian Royal Inscriptions and Art.” In Making of Pictures of War: Realia et Imaginaria in the Iconography of the Ancient Near East, edited by Laura Battini, 57–82. Oxford: Archaeopress Publishing.

F O RT I F I ED ROA D S A S CO M M U N I CAT I O N A N D D EF EN S E N E T WO R K S

153

Bahrani, Zainab. 2008. Rituals of War: The Body and Violence in Mesopotamia. New York: Zone Books. Briant, Pierre. 2002. From Cyrus to Alexander: A History of the Persian Empire. Winona Lake, IN: Eisenbrauns. Briant, Pierre. 2012. “From the Indus to the Mediterranean: The Administrative Organization and Logistics of Great Roads of the Achaemenid Empire.” In Highways, Byways and Road Systems in the Pre-modern World, edited by Susan E Alcock, John Bodel, and Richard Talbert, 185–201. New York: John Wiley and Sons. Brughmans, Tom. 2010. “Connecting the Dots: Towards Archeological Network Analysis.” Oxford Journal of Archaeology 29 (3): 277–303. Burke, Aaron. 2008. Walled up to Heaven: The Evolution of Middle Bronze Age Fortification Strategies in the Levant. Winona Lake, IN: Eisenbrauns. Charpin, Dominique, and N. Ziegler. 2003. Mari et Le Proche-Orient à l’époque Amorrite: Essai d’histoire politique. Paris: Société pour l’étude du Proche-Orient Ancien. Collar, Anna, Fiona Coward, Tom Brughmans, and Barbara J. Mills. 2015. “Networks in Archaeology: Phenomena, Abstraction, Representation.” Journal of Archaeological Method and Theory 22 (1): 1–32. Creekmore, Andrew. 2014. “The Social Production of Urban Space in Third Millennium Cities of Upper Mesopotamia.” In Making Ancient Cities: Space and Place in Early Urban Societies, edited by Andrew Creekmore and Kevin D. Fisher, 32–60. New York: Cambridge University Press. De Zorzi, Nicla De. 2017. “Teratomancy at Tigunānum: Structure, Hermeneutics, and Weltanschauung of a Northern Mesopotamian Omen Corpus.” Journal of Cuneiform Studies 69: 125–150. Dolce, Rita. 2016. “ ‘Losing One’s Head’: Some Hints on the Procedures and Meanings of Decapitation in the Ancient Near East.” In Making of Pictures of War: Realia et Imaginaria in the Iconography of the Ancient Near East, edited by Laura Battini, 45–56. Oxford: Archaeopress Publishing. Dossin, G. 1938. “Signaux lumineux au Pays de Mari.” Revue d’Assyriologie et d’archéologie Orientale 35 (3/4): 174–186. Dubovský, Peter. 2006. Hezekiah and the Assyrian Spies: Reconstruction of the NeoAssyrian Intelligence Services and Its Significance for 2 Kings 18–19. Biblica et Orientalia 49. Rome: Pontifical Biblical Institute. Dusinberre, Elspeth R.M. 2013. Empire, Authority and Autonomy in Achaemenid Anatolia. New York: Cambridge University Press. Earley-Spadoni, Tiffany. n.d. Landscapes of Warfare in the Ancient Near East. Submitted for initial review. Boulder: University Press of Colorado.

154

T I F FA N Y E A R LE Y- S PA D O N I

Earley-Spadoni, Tiffany. 2015a. “Envisioning Landscapes of Warfare: A Multiregional Analysis of Early Iron Fortress-States and Biainili-Urartu.” PhD diss. Baltimore, MD: Johns Hopkins University. Earley-Spadoni, Tiffany. 2015b. “Landscapes of Warfare: Intervisibility Analysis of Early Iron and Urartian Fire Beacon Stations (Armenia).” Journal of Archaeological Science: Reports 3: 22–30. Fachard, Sylvian. 2016. “Regional Fortifications and the Fortification of Regions.” In Focus on Fortifications: New Research on Fortifications in the Ancient Mediterranean and the Near East, edited by Silke Frederiksen, Peter Schneider, and Mike Schnelle, 413–416. Oxford: Oxbow Books. Fales, Frederick Mario. 1990. “The Rural Landscape of the Neo-Assyrian Empire: A Survey.” State Archives of Assyria Bulletin 4: 81–142. Fales, Frederick Mario. 2010. Guerre et paix en Assyrie: Religion et impérialisme. Paris: Éds. du Cerf. Favaro, Sabrina. 2007. Voyages et voyageurs à l’époque Néo-Assyrienne. State Archives of Assyria 18. Helsinki: Neo-Assyrian Text Corpus Project. Foster, B. R. 2005. Before the Muses: An Anthology of Akkadian Literature. 3rd ed. Bethesda: University Press of Maryland. Franklin, Kathryn, and Astghik Babajanyan. 2018. “Approaching Landscapes of Infrastructure: Methods and Results of the Vayots Dzor Silk Road Survey.” In Landscape Archaeology in Southern Caucasia: Finding Common Ground in Diverse Environments. Proceedings of the Workshop Held at 10th ICAANE in Vienna, April 2016, edited by William Anderson, Kristen Hopper, and Abby Robinson, 131–143. Vienna: Austrian Academy of Sciences. Fuchs, Andreas, and Simo Parpola. 2001. The Correspondence of Sargon II, Part III. State Archives of Assyria 15. Helsinki: Helsinki University Press. George, Andrew R. 2003. The Babylonian Gilgamesh Epic: Introduction, Critical Edition, and Cuneiform Texts. Vol. 1. Oxford: Oxford University Press. Graf, David F. 1994. “The Persian Royal Road System.” Achaemenid History 8: 167–189. Greene, Alan, and Ian Lindsay. 2012. “Mobility, Territorial Commitments, and Political Organization among Late Bronze Age Polities in Southern Caucasia.” Archaeological Papers of the American Anthropological Association 22 (4): 54–71. Hallock, Richard. 1969. Persepolis Fortification Tablets. Chicago: University of Chicago Press. Hammer, Emily. 2014. “Highland Fortress-Polities and Their Settlement Systems in the Southern Caucasus.” Antiquity 88 (341): 757–774. Heimpel, W. 2003. Letters to the King of Mari: A New Translation with Historical Introduction and Commentary. Winona Lake, IN: Eisenbrauns.

F O RT I F I ED ROA D S A S CO M M U N I CAT I O N A N D D EF EN S E N E T WO R K S

155

Herodotus. 2021. “The History of Herodotus.” October 26, 2021, http://classics.mit .edu/Herodotus/history.mb.txt. Kessler, Karlheinz. 1997. “Royal Roads and Other Questions of the Neo-Assyrian Communication System.” In Proceedings of the 10th Anniversary Symposium of the Neo-Assyrian Text Corpus Project, September 7–11, 1995, edited by Simo Parpola and R. M. Whiting. Helsinki: Neo-Assyrian Text Corpus Project. Kessler Guinan, Ann. 1997. “Auguries of Hegemony: The Sex Omens of Mesopotamia.” Gender and History 9 (3): 462–479. Kuhrt, A. 1995. The Ancient Near East c. 3000–330 BC. London: Routledge. Lanfranchi, G. B., and Simo Parpola. 1990. The Correspondence of Sargon II, Part II: Letters from the Northern and Northeastern Provinces. Edited by G. B. Lanfrachi and S. Parpola. Helsinki: Helsinki University Press. Luckenbill, Daniel David. 1927. Ancient Records of Assyria and Babylonia, Vol. I. No. 279–280. Chicago: University of Chicago Press. Melkonyan, H., I. Karapetyan, and N. Yengibaryan. 2010. “The Excavations of the Newly Found Urartian Fortress in Getap.” ARAMAZD no. 5 (2): 90–98. Melville, Sarah C. 2016. “Insurgency and Counterinsurgency in the Assyrian Empire during the Late Eighth Century.” In Brill’s Companion to Insurgency and Terrorism in the Ancient Mediterranean, edited by Timothy Howe and Lee L. Brice, 62–92. Leiden, Netherlands: Brill. Morandi Bonacossi, Daniele. 2000. “The Syrian Jezireh in the Late Assyrian Period: A View from the Countryside.” In Essays on Syria in the Iron Age: Ancient Near Eastern Studies Supp. 7, edited by Guy Bunnens, 349–396. Louvain, Belgium: Peeters. Parker, Bradley J. 1997. “Garrisoning the Empire: Aspects of the Construction and Maintenance of Forts on the Assyrian Frontier.” Iraq 59: 77–87. Parpola, Simo. 1987. The Correspondence of Sargon II, Part I. State Archives of Assyria. Vol. 1. Helsinki: Helsinki University Press. Pecorella, P., and Mirjo Salvini. 1984. Tra Lo Zagros e l’Urmia: Ricerche storiche ed archeologiche nell’Azerbaigian Iraniano. Rome: Edizioni dell’Ateneo. Radner, Karen. 2011. “Assyrians and Urartians.” In The Oxford Handbook of Ancient Anatolia: (10,000–323 BCE), edited by Sharon R. Steadman and Gregory McMahon, 734–751. Oxford: Oxford University Press. Radner, Karen. 2012. “Between a Rock and a Hard Place: Musasir, Kumme, Ukku and Subria—the Buffer States between Assyria and Urartu.” In Biainili Urartu: The Proceedings of the Symposium Held in Munich 12–14 October 2007, edited by Stephan Kroll, C. Gruber, U. Hellwag, Michael Roaf, and Paul Zimansky, 243–64. Leuven, Belgium: Peeters. 156

T I F FA N Y E A R LE Y- S PA D O N I

Richardson, Seth. 2016. “Insurgency and Terror in Mesopotamia.” In Brill’s Companion to Insurgency and Terrorism in the Ancient Mediterranean, edited by Timothy Howe and Lee L. Brice, 31–61. Leiden, Netherlands: Brill. Ristvet, Lauren. 2007. “The Third Millennium City Wall at Tell Leilan, Syria: Identity, Authority, and Urbanism.” In Power and Architecture: Monumental Public Architecture in the Bronze Age Near East and Aegean, Peeters, Leuven, edited by J. Bretschneider, J. Driessen, and K. Van Lerberghe, 183–211. Leuven, Belgium: Peeters. Ristvet, Lauren. 2014. Ritual, Performance and Politics in the Ancient Near East. Cambridge: Cambridge University Press. Ristvet, Lauren. 2018. “Negotiation, Violence, and Resistance: Urartu’s Frontiers in the Iron Age Caucasus.” In The Archaeology of Imperial Landscapes: A Comparative Study of Empires in the Ancient Near East and Mediterranean World, edited by Bleda Düring and Tesse Stek, 175–205. Cambridge: Cambridge University Press. Rousset, M. O., B. Geyer, N. Awad, and S. Shabo. 2017. “Un réseau défensif de l’âge Du Bronze Moyen dans Les Marges Arides de Syrie Du Nord.” Paléorient 43 (2): 115–63. Rowe, William T. 2007. Crimson Rain: Seven Centuries of Violence in a Chinese County. Stanford, CA: Stanford University Press. Schwartz, Glenn M. 2013. “An Amorite Global Village: Syrian-Mesopotamian Relations in the Second Millennium B.C.” In Cultures in Contact: From Mesopotamia to the Mediterranean in the Second Millennium BC, edited by J. Aruz, S. Graff, and Y. Rakic, 1–11. New York: Metropolitan Museum of Art. Smith, Adam T. 2015. The Political Machine: Assembling Sovereignty in the Bronze Age Caucasus. Princeton, NJ: Princeton University Press. Swanson, Steve. 2003. “Documenting Prehistoric Communication Networks: A Case Study in the Paquime Polity.” American Antiquity 68 (4): 753–767. Thureau-Dangin, François. 1912. Une relation de La Huitième Campagne de Sargon (714 Av. J-C). Paris: Librairie Paul Geuthner. Tilley, Cristopher. 2004. The Materiality of Stone: Explorations in Landscape Phenomenology. Oxford: Berg Publishers. Ur, Jason Alik. 2012. “Landscapes of Movement in the Ancient Near East.” In Proceedings of the 7th International Congress on the Archaeology of the Ancient Near East, edited by Roger Matthews, John Curtis, Michael Seymour, Alexandra Fletcher, Alison Gascoigne, Claudia Glatz, St. John Simpson, Helen Taylor, Jonathan Tubb, and Rupert Chapman, 12: 521–38. Washington, DC: Harrassowitz. Ur, Jason Alik. 2017. “Physical and Cultural Landscapes of Assyria.” In The Blackwell Companion to Assyria, edited by Eckart Frahm, 13–35. Oxford: Wiley-Blackwell.

F O RT I F I ED ROA D S A S CO M M U N I CAT I O N A N D D EF EN S E N E T WO R K S

157

Van de Mieroop, Marc. 2007. A History of the Ancient Near East, ca. 3000–323 BC. Malden, MA: Blackwell Melden. Wiesehöfer, Josef. 2001. Ancient Persia from 550 BC to 650 AD. London: I. B. Taurus Publishers. Wilkinson, Tony James. 2003. Archeological Landscapes of the Ancient Near East. Tucson: University of Arizona Press. Wilkinson, Tony James, Eleanor Barbanes Wilkinson, Jason Alik Ur, and Mark Altaweel. 2005. “Landscape and Settlement in the Neo-Assyrian Empire.” Bulletin of the American Schools of Oriental Research, no. 340 (November): 23–56. Winter, Irene J. 1996. “Sex, Rhetoric, and the Public Monument: The Alluring Body of Naram-Sîn of Agade.” In Sexuality in Ancient Art: Near East, Greece and Italy, edited by Natalie Boymen Kampen, 11–26. New York: Cambridge University Press. Zimansky, Paul. 1985. Ecology and Empire: The Structure of the Urartian State. Studies in Ancient Oriental Civilization 41. Chicago: Oriental Institute.

158

T I F FA N Y E A R LE Y- S PA D O N I

8 This chapter presents the case of several fortified settlements in eastern France occupied during two different moments (figure 8.1; table 8.1). The first existed at the end of the Second Iron Age, during which Gallia was divided into several different political entities, sometimes in competition with each other. During the second fortification phase, the entire territory was under control of the Roman Empire. How can these fortifications be explained for each period? One particular characteristic of these sites is that they are located very close to each other, which raises additional questions. Why so many sites in such a limited space? Why were these sites fortified? What were they protecting? What were the relationships between them?

The Fortified Settlements in the Saint-Dié-desVosges Basin during the Iron Age and the Roman Empire, France Lizzie Scholtus

CONTEXT

The area of the Saint-Dié-des-Vosges basin, located in the Vosges Mountains (northeastern France, figure 8.1), is a microregion dotted with several fortified sites of similar size and dating. Therefore, it can be considered as a good case for the study of fortified landscapes. The region is separated from the rest of the massif and possesses abundant metal resources, such as iron and silver, which has been exploited at least since the Middle Age (1550–550 bp). A large valley is formed by the Meurthe River between the contemporary regions of Alsace and Lorraine. The area, thus, is an important link through the Vosges Mountains by its access to the important

https://doi.org/10.5876/9781646422111.c008

159

Figure 8.1. Location map of the Saint-Dié-desVosges basin. Made by CROLL | OpenStreetMap, OpenTopoMap, GWC-CIGAL.

passes of Donon, Saales, Sainte-Marie-aux-Mines, and Bonhomme (Devel 1999, 10; Féliu 2008, 18). Because of the topography, several settlements were able to set up on the spurs on the summits of the mountains surrounding Saint-Dié-des-Vosges valley. Five of these hilltops were inhabited during La Tène D (2100–1976 bp) and are known through excavations or surveys. They all present the same features: their invariable high situation overhanging the valley and its roads; their protection by a fortification; their small surface area; a later reoccupation between 1850 and 1650 bp (second or the third centuries ce), and their location in a range of 12 km from one another and, therefore, within sight of each other (Scholtus 2016a). These sites are located at the eastern end of the civitas of Leuci, a Gallic tribe that was in contact with similar civitates of the Rauraci and the Mediomatrici (Devel 1999, 10; Fichtl 2012). Indeed, at the time of the Roman conquest, Caesar explained that the various Gallic peoples were organized into civitates; however, their social, political, or even military structure is still difficult for archaeologists to understand today 160

LI Z Z I E S C H O LT U S

T able 8.1. Chart of the chronological systems used in the chapter Archaeological Culture Period Celtic Period

Phase First Iron Age

Second Iron Age

BC/AD

BP

Hallstatt C

C1

800–726 bc

2750–2676

C2

725–621 bc

2675–2571

Hallstatt D

D1

620–531 bc

2570–2481

La Tène A

D2

530–501 bc

2480–2451

D3

500–461 bc

2450–2411

A1

460–431 bc

2410–2381

A2

430–401 bc

2380–2351

La Tène B

B1

400–321 bc

2350–2271

B2

320–261 bc

2270–2211

La Tène C

C1

260–201 bc

2210–2151

C2

200–151 bc

2150–2101

La Tène D

D1

150–76 bc

2100–2026

D2 Roman Empire

Dates

Subphase

Augustan period

75–26 bc

2025–1976

25 bc–ad 36

1975–1914

Early Empire

ad 21–260

1929–1690

Late Antiquity

ad 260–450

1690–1500

(Fichtl 2012, 5). It can be deduced, though, that Gallia included all peoples located to the west of the Rhine. It was a Roman construction to distinguish them from the other Celtic peoples, who occupied the rest of Europe at the time. The first traces of occupation in this area date from the Neolithic (8950–5250 bp). Some artefacts have been found on several sites, but no real settlements are known. The population grew during the First Iron Age (2750–2411 bp) with the foundation of two hilltop settlements: La Pierre d’Appel (figure 8.2: 1) and Varrinchâtel (figure 8.2: 2) in Étival-Clairefontaine. During this period, the basin’s resources also began to be exploited, as shown by the example of Les Fossottes millstone quarry (figure 8.2: 12) in La Salle (Michler 2004, 55). It was during the Second Iron Age (2410–1976 bp), however, when the occupation of Saint-Dié-des-Vosges basin intensified. Indeed, five fortified hilltops are known for this period: La Pierre d’Appel (figure 8.2: 1) and Varrinchâtel (figure 8.2: 2), which were reoccupied at that time; La Corre (figure 8.2:), Le Chastel (figure 8.2: 7), and La Bure (figure 8.2: 6). T H E F O RT I F I ED S E T T LEM EN T S I N T H E S A I N T- D I É - D ES - VO S G ES BA S I N

161

Figure 8.2. Location map of sites and discoveries in the Saint-Dié-des-Vosges basin (CAD L. Scholtus, base map Géoportail).

Other locations are also likely to reveal occupation for this period, but they have never been investigated and, consequently, have never been documented. Among those we can mention the Camp Romain in Combrimont (figure 8.2: 8), which has a land embankment that can be interpreted as a rampart, or La Corne de Lesse in La Bourgonce (figure 8.2: 4), which is acknowledged 162

LI Z Z I E S C H O LT U S

to be a fortification from the Second Iron Age (Scholtus 2016a). In addition, the discovery of a wooden bridge at the foot of the fortified site of La Pierre d’Appel (figure 8.2: 1), dating from the first century bc, clearly illustrates the presence of roads that allow the Vosges Mountains to be crossed by this basin (Deyber 1978). During the Gallo-Roman period (1975–1500 bp), the Saint-Dié-des-Vosges basin still appeared as a privileged passage area. Indeed, fortuitous discoveries show an antique occupation. Of particular note is the presence of several funeral monuments, a well, three monetary treasures, and numerous sculptural fragments. There is also some evidence of a Gallo-Roman occupation in the city of Saint-Dié-des-Vosges, where a fortification and artefacts from 1950 to 1650 bp have been found. Moreover, some of the fortified settlements known during the Second Iron Age were reoccupied as early as 1850 bp (Scholtus 2016a). I will now present each of the fortified settlements in this area, and in particular their defense systems, so that I can then come back to their similarities and differences. After, I will reflect on the reasons behind the organization of this fortified landscape for each period observed. LA BURE, SAINT- DIÉ- DES- VOSGES

One of the best-known archaeological sites in this area is La Bure in the city of Saint-Dié-des-Vosges (figure 8.2: 6). It was first discovered at the end of the nineteenth century and then excavated for almost thirty years, from 1964 to 1993 by Georges Tronquart (1989). It is located on a plateau 350 m long and 120 m wide for a total area of almost 3 ha. It culminates at an elevation of 575 masl. Its south, west, and north sides are bordered by steep slopes, making it a naturally defensible spur closed by a rampart on the east side. Given its geographical position, La Bure dominates the outlet of the roads leading down from the Saales or Donon passes. It is therefore a privileged location to observe the Meurthe valley and the surrounding hills (Scholtus 2016c). Despite that this hill fort was occupied from La Tène C2 (2150–2101 bp) to the end of the fourth century ce (1650–1550 bp), the occupation was especially important during two short periods. During La Tène D1 (2100–2026 bp), the first fortification wall was built with a bend midcourse and closed the site by linking two rocky peaks on either side, to the north and south of the spur. It is a Murus Gallicus (figure 8.3. and figure 8.4) with a base of large blocks topped by a dry-stone wall mixed with sand and internal horizontal oak beaming, visible on the facing and fixed with iron pins (Scholtus 2016b). T H E F O RT I F I ED S E T T LEM EN T S I N T H E S A I N T- D I É - D ES - VO S G ES BA S I N

163

Figure 8.3. La Bure, plan of the rampart sector. CAD L. Scholtus from Tronquart (1989) and Boulanger (1997).

Figure 8.4. Facing of the Murus Gallicus (Tronquart 1976) with white representations of the spaces where the beams were located and an iron pin.

It is preserved over a length of 14 m and measures between 6.50 m and 7.50 m wide in the northwestern part, while the north-south section is still preserved over a length of 24 m and a width of 6.50 m. This type of rampart is known almost only on the west side of the Celtic world (which means today’s France, Belgium, Luxembourg, Switzerland, and the west part of Germany territories) and only at La Tène D (2100–1976 bp), usually on large size sites but also sometimes on smaller fortifications such as La Bure (Fichtl 2005a, 52). For this type of rampart, the strategic function exists, but it is not the most important. The ostentatious function is more predominant, as illustrated by the use of construction techniques not necessary for the infrastructure—metal nails, for example (Fichtl 2005b). Then, at an unknown moment in the past, a new section straightens the route of this rampart. It consists of a wall adorned on the east façade and a ramp made of stones and sand on the west side that encloses the previous Murus Gallicus into its volume (Scholtus 2016c, 43). Few artefacts dated to the transition from the Gallic period to the Roman era (1975–1929 bp) have been found in the site, suggesting that the population may have partially abandoned the settlement. The occupation seems to have gradually resumed as early as 1850 bp. The second moment of intensive occupation occurred during the first part of the fourth century ce (1650–1550 bp). This period was also marked by the reconstruction of the barrier rampart in front of the previous system (figure 8.3). It consists of an agger 3 m to 4 m high on which a base of large blocks of cut sandstone and headstone were placed. Preceding this wall, a 3–5 m wide and 1.5–3 m deep ditch was carved directly into the bedrock (Scholtus 2016b). At the same time, this fortification system was complemented by a surrounding 2.2 m wide wall built with an exterior and interior dry stone facing filled with earth, sand, and rubble. It also contained fragments of reused steles, and in particular a fragment of one representing a blacksmith also discovered in the Gallo-Roman rampart, which therefore seems to confirm that these two buildings are coetaneous. Although it is close to the cliffs, this wall does not exactly follow its course. Indeed, it is rather as straight as possible. Moreover, because of these cliffs, it has no real defensive function and seems to have a more ostentatious objective (Scholtus 2014, 96). The site was abandoned around 1550 bp. Only a few sporadic artefacts indicate later human presence on the site until its discovery in the mid-nineteenth century. The excavations of the rampart did not revealed signs of intentional destruction during the abandonment events, which therefore do not seem to be linked to war events. By contrast with the overwhelming presence of T H E F O RT I F I ED S E T T LEM EN T S I N T H E S A I N T- D I É - D ES - VO S G ES BA S I N

165

fortifications, very few weapons have been discovered on the site. Only two javelins, one spearhead and two arrowheads dated from 2100–1975 bp (Reich 2011, 2016), a spur tip, three pila tips, a bolt of catapult, and a spearhead for the Roman period (1950–1550 bp), and fourteen other undated weapons have been recorded (Scholtus 2014, 254). In addition to these metal artefacts, Tronquart also notes the presence of 4,500 pebbles, which he interprets as projectiles (1989, 28). This function seems unlikely, however, given their distribution throughout the site and not only near the wall, especially since these pebbles had to be mounted from the plain while the site is established on accessible rock. (Scholtus 2014, 217). This relative lack of weapons may be explained due to the fact that during these periods, most of the warring equipment was made of wood, which has disintegrated through time (Reich 2011). In any case, twenty-five metal weapons do not seem to be enough for a defensive settlement occupied for nearly four centuries. Moreover, only two of them have been discovered in the ramparts area. The others are concentrated in the center of the settlement. It is therefore questionable whether these pieces were actually used in confrontation, or whether they were relics of a manufacturing workshop. There is indeed much evidence of workshop endeavors on this site such as all the blacksmith’s tools, a headstone representing a blacksmith, and some artefacts in-process. They could also be deposits within a sanctuary. The presence of other elements common in Celtic sanctuaries such as large quantities of coins (over 1,300) and jewelry (almost 500 artefacts) also reinforces this last hypothesis, meaning that these weapons were not used in confrontations (Bigoni 2016; Manisse 2016; Scholtus 2014). LA PIERRE D’APPEL, ÉTIVAL- CLAIREFONTAINE

La Pierre d’Appel (figure 8.2: 1) is the secondmost-well-known site of the area. It was excavated from 1967 to 1981 by Alain Deyber (Deyber et al. 1984). Once again, it is a rocky promontory of about 2.5 ha, bounded on three sides by cliffs whose height sometimes exceeds 10 m. It culminates at an elevation of 492,3 masl and dominates a crossroads of the main roads through the Vosges. The settlement was protected by a curved rampart with a ditch on its west side, complemented on both sides by a smaller peripheral rampart. The archaeological excavations established five phases of fortifications between La Tène C (2210–2101 bp) and the beginning of the third century (1750–1650 bp). According to Deyber (et al. 1984), the first three buildings belong to the first fortification system, dated from 2210 to 1975 bp (figure 8.5). The first defensive system was made of a mixed wall (5–8 m wide and 5–8 m high) with an earth 166

LI Z Z I E S C H O LT U S

Figure 8.5. La Pierre d’Appel, plan of the Celtic ramparts, CAD L. Scholtus from Deyber (1984).

bank and a wooden beam grid, covered with a facing of stones and wood. A ditch of 8 m to 10 m wide was also carved into the bedrock. This construction was then transformed into a 4–5 wide stone-core wall with faces on both sides, known as Rostbau type (Fichtl 2005a). Finally, the third phase consists of a reconstruction and widening of the previous wall. At the end of this period, the site was abandoned, and the rampart was left to ruin. At the beginning of the third century ce (1750–1650 bp) the site was occupied again, and the rampart was rebuilt reusing materials from previous phases. A palisade was placed directly into the rock surrounding the site (Deyber et al. 1984). More recent research, however, indicates that these different phases cannot be clearly differentiated (Féliu 2008, 113). The structures unearthed inside these fortifications correspond to the remains of buildings, some of them used as metallurgical workshops. Despite the variety of artefacts found in the place, it is not possible to characterize the occupation of the settlement beyond those discoveries. According to Deyber, the organization of the site was designed to respond to a major threat, in an emergency, by reconciling technical constraints and tactical imperatives. He also reports signs of fire, and a brutal destruction (layers of ash, remains of burnt palisades, broken artefacts) linked to fighting at strata corresponding to 2100–1975 bp. He states also that in each of these layers of destruction projectiles were found (Deyber 2009, 380). These weapons consist of two single barbed arrowheads and a fragment of a sword from 2100–2025 bp, a spearhead with a wooden casing from 2025–2000 bp and a javelin iron from 2000–1975 bp. Thus, they all correspond to the first major phase of the site’s occupation. However, their contexts were not clearly specified, being the only information that they were found behind the rampart (Reich 2011). For Deyber, these traces of fighting are to be put in relation with the political instability of the tribe system in Gallia for the period (2009). T H E F O RT I F I ED S E T T LEM EN T S I N T H E S A I N T- D I É - D ES - VO S G ES BA S I N

167

The idea that this rampart was erected as a response to an imminent threat does not seem to be the only explanation. Indeed, the surrounding wall cannot be explained with that hypothesis. Like the one at La Bure, it is not necessary for the immediate defense of the site, which is already protected naturally by cliffs. It does certainly highlight the military function of the fortifications, but other explanations are possible such as the manifestation of financial power, prestige, or independence, or the expression of political power, for example. According to Deyber (2003, 2009; et al. 1984), despite its small size, the site of La Pierre d’Appel would correspond to an aristocratic settlement with ostentatious fortifications. This status would explain the importance of the fortifications, and their regular renovation, but also their construction techniques generally reserved for the oppida, that is, for much larger sites. This hypothesis is based in part on the presence of a coin workshop on the site, evidenced by casting waste, crucibles remains, weights, and scales (Deyber 2003, 2009, 338). It should be noted, however, that only thirteen coins have been discovered on the site. In addition, the question of the aristocratic status of these sites will also be discussed in the synthesis. LA CORRE, HOUSSERAS

The information for the site of La Corre in Housseras (figure 8.2: 5) comes from two small digs in the rampart, one in 1962 by D. Claude and one in 2003 by O. Caumont and T. Le Saint-Quinio. It is therefore difficult to date it, but according to the shape of the rampart, the occupation must date from the first century bc (2050–1950 bp). This site, located at the western limit of the sandstone massif of the Deodatien basin, is found along a penetration route from the massif through the Saint-Dié basin and to the Alsace plain. Covering 6 ha, it occupies the western end of a small hill at an altitude of 420 masl, which was closed by a rectilinear rampart. Unlike the other sites presented in this chapter, La Corre is not prominent in the landscape. It does not occupy a dominant position but is bordered to the north and south by two higher hills. However, it has all the characteristics of a defensive site: it is naturally protected by rocky peaks to the southwest and steep slopes to the west and north. The southeast side is protected by a 266 m rampart and the south by a natural constriction and a steep slope (figure 8.6). In addition, there is no topographical indication of a lateral defense system or a peripheral enclosure wall (Caumont and Le Saint-Quinio 2003). This absence can be explained by high soil erosion but also perhaps by the apparent lack of a late reoccupation of the site. 168

LI Z Z I E S C H O LT U S

[128.104.46.206] Project MUSE (2024-03-01 20:28 GMT) UW-Madison Libraries

Figure 8.6. La Corre, map of the site with localization of the rampart, CAD L. Scholtus from Caumont and Le Saint-Quinio (2003).

As explained above, the only information we have on this site comes from the rampart built on the southeastern side of the site. It remains 1.5 m high from the interior side and 6.5 m high from the bottom of the adjacent ditch, and 7.5 m to 8.5 m wide according to surface remains. The excavations revealed facing made of cut sandstone blocks, interrupted every 130 cm by vertical spaces of 30 cm wide corresponding to the location of vertical beams on the façade of a Kelheim type. The internal volume seems to have been composed of sand and pebbles (Fichtl 2005a). The entrance, located in the center of the rampart, was a corridor of 7 m wide of a reentrant wings type (Caumont and Le SaintQuinio 2003). The interior of the site itself has never been explored, but the excavation of the many uprooted trees, following the 1999 storm, revealed only two artefacts dated from 2100–1975 bp and no remains of structures (Caumont and Le Saint-Quinio 2003). LE CHASTEL, TAINTRUX

This site (figure 8.2: 7) is known by some surveys conducted between 1975 and 2006. It is located at the top of a mountain at an altitude of 657 masl and it is visible from other hilltop sites. While its elevation may have permitted its T H E F O RT I F I ED S E T T LEM EN T S I N T H E S A I N T- D I É - D ES - VO S G ES BA S I N

169

Figure 8.7. Le Chastel and Varrinchâtel, maps of the site, CAD L. Scholtus from Triboulot (2006).

occupants a good view of the valley, it was only partial, due to its remoteness from the Saint-Dié basin (Féliu 2008, 241). Two consecutive enclosures spaced 30 m apart delimit a space of 2 ha (figure 8.7). The first surveys, carried out by Deyber in 1975 and Tronquart in 1979, established that the first enclosure included a simple earth mound. Its entry system consisted of a simple interruption. The second enclosure was made of a rampart with facing and a beam structure between 5 and 10 m wide and still 1 to 3 m high. The entrance was through a chicane door, which creates a series of curves making it easier to defend. During these early excavations not many artefacts were found, only a few pottery fragments, and a millstone from the Les Fossottes quarry (figure 8.2: 12), but there was enough evidence to date the occupation to La Tène D1 phase (2100–2026 bp) with a possible later reoccupation in the third century ce (Féliu 2008, 122; Triboulot 2006). A new survey campaign of the uprooted trees in 2004 carried out by Bertrand Triboulot uncovered 979 sherds of ceramics dated to Hallstatt D2–D3 (2550–2400 bp) and La Tène A (2400–2250 bp), which pushed back the occupation of the site by nearly four centuries (Triboulot 2006). A few slingshot bullets were also discovered during these surveys (Reich 2011). However, these 170

LI Z Z I E S C H O LT U S

artefacts seem to have been uncovered under the rampart. This would mean that they date from a previous occupation of the site before the establishment of the fortification (Féliu 2008, 122). VARRINCHÂTEL, ÉTIVAL- CLAIREFONTAINE

This site (figure 8.2: 2) has never been excavated, and little information comes from Triboulot’s survey of uprooted trees. This site is very similar to Le Chastel: it was also established on a hilltop at 516 masl and protected by two perimeter walls spaced 30 m apart, enclosing an area of 4 ha (figure 8.7). The enclosure at the summit consists of a 0.5 to 1 m high earth mound with an interruption at the west that can be interpreted as the entrance. The second enclosure, on a peripheral terrace, complements the access device with an entrance through a chicane door to the southwest. A collapse of this second enclosure led to the discovery of artefacts dated from 2550–2400 bp as well as fragments of millstones reused in the construction. This dating is confirmed by a radiocarbon dates of charcoal from a fire layer under the rampart. Therefore, this site has the oldest rampart known in the region. Artefacts from 2400–2250 bp were also present in the layers above the rampart. Finally, ceramic sherds dated to 2100–1975 bp and the GalloRoman period were also discovered during pedestrian surveys, and some slingshot bullets were found (Reich 2011). It is also interesting to note that a large number of mines have been identified all around this site (Triboulot and Michler 2006, 15). DISCUSSION

The study of these fortified sites, even if not all of them have been excavated, allows us to establish a possible chronology of the formation of the fortified landscape in the Saint-Dié-des-Vosges basin. The first phase of occupation from the Hallstatt D2–D3 (2480–2411 bp) to La Tène A (2410–2351 bp) is found on the sites of La Pierre d’Appel, Varrinchâtel and Le Chastel. However, apart from the artefacts, no structures could be identified for this period, except for the rampart on Varrinchâtel. These sites were abandoned between 2350 and 2101 bp, and then reoccupied again during La Tène D (2100–1976 bp). It is also during this second occupation that most of the fortification systems were built. The posterior abandonment of these sites was not accompanied with clear evidence of intentional destruction. Finally, a third phase of occupation extends from the third century ce to the end of the fourth century ce (1750–1550 bp) T H E F O RT I F I ED S E T T LEM EN T S I N T H E S A I N T- D I É - D ES - VO S G ES BA S I N

171

and is marked by the reconstruction of most of the ramparts of the previous period, and the addition of peripheral walls with no clear defensive function. Very little information is available about the occupation of the valley itself. No traces of settlements could be found for the Celtic period in the entire valley. The only known elements are limited to ancient roads and the bridge mentioned previously. The only known site in the valley is Les Fossottes in the commune of La Salle (figure 8.2: 12). It is a rhyolite quarry used to make millstones of different sizes between 3150 bp and 1550 bp. This is an important resource for the sector since the export of these products is documented on non-Vosges sites and in aristocratic contexts such as the settlement of Vix in Bourgogne more than 200 km away (Farget 2007; Michler 2004, 54). There are several indications of a Roman city in the commune of Saint-Dié-des-Vosges, a burial site in the Mortagne valley at Les Lions, and a third-century ce settlement in Bruyères on the Mont Avison. In addition to these few sites, occasional discoveries suggest a more considerable occupation, but the human presence in the area remains poorly known for the Roman period (Michler 2004). This lack of information is mainly due to the limited archaeological research carried out in the area. Overall, these sites have similar characteristics: They were built in elevated terrain; they overlooked the valley and roads (except for La Corre); they were protected by fortifications; they were small in area; and for the most part they were occupied at the same time. In addition, they are all located within a 12 km radius and are therefore within sight of each other (Walter 2015). As mentioned at the beginning of this chapter, all these elements raise several questions: Why are there so many sites in such a limited space? Why were these sites fortified? What were they protecting? What were the relationships between them? One possible interpretation of this concentration of fortifications would be the protection of their inhabitants. Moreover, some authors interpret these places as evidence of battlefields. It should be noted, however, that all these sites are small, the largest reaching only 4 ha. Despite that excavations are limited in these sites, no dwellings have been identified, except for La Pierre d’Appel. These sites could therefore not contain large populations. The inhabitants of the region had to live elsewhere, or it was very sparsely populated, which seems unlikely given the resources used to build the fortifications. Second, very few weapons were discovered on all these sites, given that there are about ten sites, but this can be explained by the lack of preservation of wooden weapons. Guillaume Reich reports a possible concentration of weapons for the period of 2100–1975 bp around the Meurthe River bed (Reich 2011). Nonetheless, the evidence of actual military actions justifying these interpretations is tenuous. 172

LI Z Z I E S C H O LT U S

Finally, no destruction events have been identified on the excavated sites allowing us to support the hypothesis of armed conflicts related to the site abandonment or destruction. Most of the structures were just abandoned and left to ruin. A fire stratum has been identified at the Varrinchâtel site, but it predates the construction of the rampart (Triboulot 2006). Moreover, it is common to find small quantities of weapons or militaria on the fortified sites of the Leuci territory. Under these conditions, this presence would not be linked to warfare but rather suggest, for the Celtic period and the beginning of the Roman era, an important role for these sites, to be linked to transport routes and in particular watercourses according to Thierry Dechezleprêtre (2008, 101). Despite the other uses and the lack of use for direct military conflict, there is a real defensive use of these sites. The rebuilding of ramparts in subsequent phases of occupation suggests this impression. It is very likely that privileged position, at high terrains, have permitted its users to exercise control on roads crossing the basin but also to provide protection to people living in the surroundings and maybe control the exploitation and trade or the mineral resources of the area. I have already mentioned the importance of the stone trade in this sector for the entire period studied, with the Les Fossottes quarry. Sandstone, which forms the basin’s bedrock, was also exploited. There are indeed numerous traces of extraction of this stone on the fortified sites and in particular on the site of La Bure (Boulanger 2016). Many fragments of stelae, sculptures, and inscriptions from the Roman period, sometimes in the process of being carved, have been discovered in the Les Lions site in the commune of Housseras (Michler 2004, 195). The exploitation of ores, especially for iron metallurgy, has been also documented in this region. As already mentioned, many mines have been identified on the mountain of Varrinchâtel, but also on the one of La Bure (figure 8.8) (Scholtus 2015), with uninterrupted use from the Early Iron Age (around 3150 bp) to the Medieval period (1550–550 bp). Many metalworking tools found on La Bure attest to this activity at least for the late La Tène and the Roman period (Scholtus 2014, 288). Wood may have been another economically important resource from this area, since it was the preferred building material. Wood remains did not preserve well in the region, but tools discovered in La Bure are clear evidence of this activity (Scholtus 2014, 292). The defensive needs sure were behind these fortifications when they were first erected, but their builders always kept in mind the ostentatious role of those constructions. The first ramparts were indeed necessary for blocking the access to these naturally protected locations, but the care taken in their T H E F O RT I F I ED S E T T LEM EN T S I N T H E S A I N T- D I É - D ES - VO S G ES BA S I N

173

Figure 8.8. Map of the mines (on red) around the site of La Bure (Scholtus 2015).

construction seems more important than simply closing off access to the site. They would have been expensive to build considering the amount of wood and metal used, as well as the use of carved stones. Moreover, the need of the surrounding walls is not that clear. These walls were built only during the Roman reoccupation, using different techniques, and they were less massive than the ramparts, but they would have been seen from the roads in the valley, showing the wealth of the sites (figure 8.9). In addition, throughout the Saint-Dié-des-Vosges basin, there is a great diversity of architectural solutions adopted in the construction of these fortification systems, over a relatively short period of time and in a limited area (Caumont and Le Saint-Quinio 2003). Thus, La Bure has an internal horizontal wall with horizontal beams of the Murus Gallicus type with its large iron nails, La Pierre d’Appel has a vertical wall with vertical beams on the façade and an internal wall with two facing for its first state, and La Corre has a wall with vertical beams on the façade and cut blocks. The presence of a ditch is attested at La Pierre d’Appel and La Corre but not at La Bure for 2100–1975 bp period. Finally, access was via side gates to La Pierre d’Appel and La Bure, 174

LI Z Z I E S C H O LT U S

Figure 8.9. View of the valley from La Bure.

a central gate to La Corre, and chicane gates to Le Chastel and Varrinchâtel. So, it seems that we are facing a set of organized settlements that wish to display their power and wealth through imposing and multiple fortification systems. The construction also shows a common understanding of the rampart as a prestigious element that testifies to the wealth and power of the authority that erected it (Féliu 2014, 235). This wealth certainly came from the trade of raw materials presented above. It is also perceptible through artefacts that come from long-distance exchanges. This is the case in particular for eastern Europe and Bohemia for the Celtic period (Pierrevelcin 2012), but artefacts from Egypt have also been found in La Bure (Scholtus 2014, 262). None of these sites corresponds to the model of fortified sites known for the 2100–1975 bp period in the Celtic world. They are generally aristocratic fortifications of the oppidum type, dispersed within the land they control. The concentration of small fortifications in the Saint-Dié region suggests that this control is divided between different competing powers. These sites are also very original because of the artisanal and commercial functions that develop there, while not having any of the proper aristocratic markers for the period (Kaurin and Marion 2016). However, the reoccupation of the heights and their fortification is a common pattern for the Roman period and in particular the end of the third century ce. This phenomenon is often linked to a period of insecurity associated T H E F O RT I F I ED S E T T LEM EN T S I N T H E S A I N T- D I É - D ES - VO S G ES BA S I N

175

with the movement of the “Great Barbarian Invasions” at the end of the fourth century, which preceded the fall of the western Roman Empire. This interpretation is now being challenged, and these “invasions” are rather interpreted as migrations. CONCLUSION

The settlements of the Saint-Dié basin are a special case in the Leuci territory for the Celtic period. They form a series of small contemporary fortifications that must be part of the political organization of the civitas, by controlling borders and access (Féliu 2014, 237). Moreover, they are also part of a developed economic network, as shown by economic markers such as Mediterranean imports and currencies but also by the diffusion of raw materials such as, among others, the rhyolite of the Les Fossottes quarry (Féliu 2014, 243). Thus, the various fortification systems of the sector seem to have a more ostentatious role in showing the wealth and power of these sites located at the entrance to the territory of the civitas to which they belong. Even if their defensive function cannot be completely rejected, it does not seem to be the original purpose of these constructions. Moreover, recent research has shown the relative fragility of this type of construction, which often had to be redesigned without even having been subjected to fighting (Gruel and Buchsenschutz 2015, 310). For the Roman period the role of this fortified landscape seems to change. The occupation or reoccupation of this type of small fortified settlements of heights seems much more widespread in the whole area of the Vosges Mountains (Gentner and Walter 2017, 133). Unlike during the Celtic period, these fortifications are related to a period of insecurity. However, as we have demonstrated, they retain an ostentatious aspect that seems to indicate the permanence of the power of these sites between the two periods of occupation. REFERENCES

Bigoni, Rose. 2016. “Les monnaies gauloises du site de hauteur fortifié de La Bure (Saint-Dié-Des-Vosges).” Mémoire des Vosges, Le «Camp Celtique» de La Bure à Saint-Dié, 1964–2015, special edition no. 7: 72–78. Boulanger, Karine. 1997. “Rapport de synthèse sur les fouilles du site de Saint-Dié–La Bure (Vosges), Étude documentaire préalable au projet de mise en valeur du site de La Bure.” Overview Report. Metz: SRA Metz. 176

LI Z Z I E S C H O LT U S

Boulanger, Karine. 2016. “L’extraction et l’utilisation du Grès Sur le site de La Bure dans l’Antiquité.” Mémoire des Vosges, Le «Camp Celtique» de La Bure à Saint-Dié, 1964–2015, special edition 7: 62–71. Caumont, Olivier, and Thomas Le Saint-Quinio. 2003. “Un site de hauteur du Massif Gréseux Vosgien: ‘La Corre’ à Housseras (Vosges).” Archaeologia Mosellana, no. 5: 107–122. Dechezleprêtre, Thierry. 2008. “Présence de militaria sur quelques oppida de l’Est de la Gaule.” In Sur les traces de César: Militaria tardo-républicains en contexte gaulois, Actes de La Table Ronde de Bibracte, Octobre 2002, edited by Mathieu Poux, 93–101. Bibracte 14. Glux-en-Glenn: Bibracte, Centre Archéologique Européen. Devel, Pierre. 1999. “Le Bassin de Saint-Dié à La Tène Finale.” Master’s thesis, Strasbourg: Université Marc Bloch, Strasbourg II. Deyber, Alain. 1978. “Le pont celtique d’Étival-Clairefontaine (Vosges): Points de connaissances, problèmes et directions de recherches.” Revue Archéologique de l’Est et Du Centre-Est, no. 29: 105–116. Deyber, Alain. 2003. “Les monnaies gauloises de l’oppidum de La Pierre d’Appel à Étival-Clairefontaine (Vosges).” Archaeologia Mosellana, no. 5: 123–136. Deyber, Alain. 2009. Les Gaulois en guerre: Stratégies, tactiques et techniques. Essai d’histoire militaire, IIe–Ier siècles av. J.-C. Paris: Editions Errance. Deyber, Alain, Marc Dalaut, Edmée Ladier, and André Weisrock. 1984. “L’habitat fortifié laténien de la «Pierre d’Appel» à Etival-Clairefontaine (Vosges).” Gallia 42 (1): 175–217. https://doi.org/10.3406/galia.1984.1915. Farget, Virginie. 2007. “Les Carrières de Meules au Lieu-Dit ‘Les Fossottes’ (La Salle—88) : La Carrière N°11.” Excavation Report. Metz, France: SRA Metz. Féliu, Clément. 2008. “Leuques et Médiomatriques à La Tène moyenne et finale. Organisation sociale et territoriale de l’habitat dans deux cités du nord-est de la Gaule du IIIe au Ier siècle avant notre ère.” PhD diss., Strasbourg: Université Marc Bloch, Strasbourg II. Féliu, Clément. 2014. “Structures politiques, sociales et économiques dans deux cités du nord-est de la Gaule (Leuques et Médiomatriques) à La Tène Finale.” In Produktion—Distribution—Ökonomie: Siedlungs–und Wirtschaftsmuster der Latènezeit; Akten des internationalen Kolloquiums in Otzenhausen, 28.–30. Oktober 2011, 231–245. Universitätsforschungen zur prähistorischen Archäologie Aus dem Institut für Vor- und Frühgeschichte der Universität Mainz. Bonn: Habelt. Fichtl, Stephan. 2005a. La ville celtique: Les “oppida” de 150 av. J.-C. à 15 ap. J.-C. Paris: Editions Errance. Fichtl, Stephan. 2005b. “Murus et pomerium: Réflexions sur la fonction des remparts protohistoriques.” Revue archéologique du Centre de la France [En ligne], no. Tome 44. http://journals.openedition.org/racf/515.

T H E F O RT I F I ED S E T T LEM EN T S I N T H E S A I N T- D I É - D ES - VO S G ES BA S I N

177

Fichtl, Stephan. 2012. Les peuples gaulois: IIIe–Ier siècle av. J.-C. Paris: Errance. Gentner, Steeve, and Maxime Walter. 2017. “L’éperon barré du Schieferberg à Oberhaslach (67): Prospection thématique avec sondages. Rapport 2017.” Prospecting Report. Strasbourg, France: SRA Grand-Est–Strasbourg. Gruel, Katherine, and Olivier Buchsenschutz. 2015. “Un urbanisme perché.” In L’Europe celtique à l’âge du Fer (VIIIe–Ier siècles), edited by Olivier Buchsenschutz, 307–315. Paris: Presses universitaires de France. Kaurin, Jenny, and Stéphane Marion. 2016. “Le site de hauteur fortifié de ‘la Tête du villé’ sur le massif de La Bure, a la recherche de la ville gauloise.” Mémoire Des Vosges, Le «Camp Celtique» de La Bure à Saint-Dié, 1964–2015, special series no. 7: 54–61. Manisse, Pierre-Damien. 2016. “Des monnaies romaines Du Camp de la Bure (88): Récolement et distribution spatiale.” Mémoire Des Vosges, Le «Camp Celtique» de La Bure à Saint-Dié, 1964–2015, special series no. 7: 79–85. Michler, Matthieu. 2004. Les Vosges: Carte archéologique de la Gaule, dressée sous la dir. de Michel Provost; 88. Paris: Ed. Académie des Inscriptions et Belles-Lettres. Pierrevelcin, Gilles. 2012. “Les relations entre la Bohême et la Gaule du IVe au Ier siècle avant J.-C.” Dissertationes archaeologicae Brunenses / Pragensesque 12. Praha, Czech Republic: Univ. Karlova, Filozof. Fak. Reich, Guillaume. 2011. “L’armement laténien en Alsace et en Lorraine: Objets, pratiques et contextes de découverte.” Master’s thesis, Strasbourg: Université de Strasbourg. Reich, Guillaume. 2016. “L’armement gaulois du site de La Bure.” Mémoire Des Vosges, Le «Camp Celtique» de La Bure à Saint-Dié, 1964–2015, special series no. 7: 93–97. Scholtus, Lizzie. 2014. “La Bure: Relecture des données.” Master’s thesis, Strasbourg: Université de Strasbourg. Scholtus, Lizzie. 2015. “La Bure, Saint-Dié-Des-Vosges (Vosges): Prospection inventaire, rapport 2015.” Prospecting Report. Metz, France: SRA Metz. Scholtus, Lizzie. 2016a. “Histoire de la recherche archéologique des sites fortifiés dans Le Bassin de Saint-Dié-Des-Vosges.” Archimède [En Ligne], no. 3: 8–19. Scholtus, Lizzie. 2016b. “La Bure: Relecture des données.” Mémoire Des Vosges, Le «Camp Celtique» de La Bure à Saint-Dié, 1964–2015, no. 7: 98–107. Scholtus, Lizzie. 2016c. “Le Camp Celtique de La Bure.” In Actes des Journées d’Études Vosgiennes. Saint-Dié-des-Vosges: Fédération des Sociétés Savantes des Vosges; Société Philomatique Vosgienne. Triboulot, Bertrand. 2006. “Aristocratie Celtique sur les habitats fortifiés d’Étival-Clairefontaine / Saint-Benoît-La-Chipotte, ‘Varrinchâtel’ et de Taintrux,

178

LI Z Z I E S C H O LT U S

‘Chastel.’” Nouvelles Archéologies, Prospection et Découvertes dans La Moyenne Vallée de La Meurthe, special series no. 3: 21–30. Triboulot, Bertrand, and Matthieu Michler. 2006. “Nouvelles traces d’extraction de minerai dans La Moyenne Vallée de La Meurthe (88): Première campagne de prospection inventaire.” Nouvelles Archéologies, Prospection et Découvertes Dans La Moyenne Vallée de La Meurthe, special series no. 3: 9–20. Tronquart, Georges. 1976. “Fouilles de La Bure.” Excavation Report 13. Saint-Dié, France: SRA Metz. Tronquart, Georges. 1989. Le “Camp celtique” de la Bure (Saint-Dié): Un castellum du massif vosgien. Saint-Dié, France: Éd. le Chardon. Walter, Maxime. 2015. “Les sites fortifiés de ˙auteur du Massif Vosgiens: Actualisation des données et modalités d’implantation.” Master’s thesis, Strasbourg: Université de Strasbourg.

T H E F O RT I F I ED S E T T LEM EN T S I N T H E S A I N T- D I É - D ES - VO S G ES BA S I N

179

9 Late Woodland Cultural Adaptations in the Lower Missouri River Valley Archery, Warfare, and the Rise of Complexity

Kerry Nichols

https://doi.org/10.5876/9781646422111.c009

180

Beginning in Mesolithic times and probably much earlier, the development of archery was an innovation that revolutionized both the procurement of food and the conduct of warfare. This was an innovation that enabled a hunter or warrior to carry out a potentially lethal strike on animals or people at distances greater than a hundred yards or so using a standoff projectileweapons technology. Prior to this, other projectile technologies—such as the spear, sling, or atlatl (spear thrower)—were mostly effective at much closer ranges and had relative disadvantages for their use. including a lack of flexibility in casting position and greatly decreasing accuracy as range increased. The Late Woodland Period in midwestern North America was unusual in that widespread societies lacking archery were thrust into contact with an advancing and very dangerous archery technology that promised major changes for their societies. This was not a change that took place in a uniform or regular pattern but was instead a very halting and long-term process that culminated with the arrival of formal, organized warfare that was endemic among the later Mississippian cultures. The process may be best described as the adaptation of new technology that fundamentally altered the fabric of society by enabling increased social cooperation, greater behavioral sophistication, and information exchange on ever-increasing scales each time it occurred (Bingham and Souza 2009; Morris 2014). The primary enterprise in generating large amounts of

highly varied culturally transmitted information to be disseminated and diffused is found in the rise of non-kin-based coalitions that met in societies throughout human history (Bingham and Souza 2009, 313). The commonality that defines the rise and increasing sophistication of human non-kin-based coalitions is access to potentially coercive weapons technology applied to the area of human relations to help manage conflicts of interest (Bingham and Souza 2009). The following research addresses the impact of the introduction of the bow and arrow on southeast societies living along the Missouri River. I suggest that the introduction of the bow and arrow into prehistoric Missouri during the Late Woodland period changed the entire Middle Woodland social dynamic and settlement pattern arrangement such that there was a major increase in social cooperation between settlements tied closely to defensive settlement strategies. I seek to employ a series of spatial-statistical tests to demonstrate that at least for the lower Missouri River valley, changes in settlement patterns from Late Woodland to Early Mississippian periods strongly correlate with the introduction of the bow and that these changes also spurred increases in social complexity. I argue here that conditions created by the arrival of the bow called for increased coordination and increased communications among settlement locations such that farmer-collector and band-level social groups could coordinate defensive settlement strategy. Such relationships may denote incipient alliance formation or a possible effort at collective defense of closely interacting settlements. I further argue that the first steps toward increased complexity in the study area are primarily due to defensive actions in response to long-range weapons technology. This study is based on the application of Geographic Information System (GIS) approaches to look at viewsheds, site distances, and the degree of intervisibility among settlements, in combination with other archaeological information about subsistence and settlement structure patterns. I ultimately suggest that the introduction of the bow and arrow may have caused settlements to move to locations that provided both excellent viewsheds relative to previous settlements and intervisibility among settlements, both of which help decrease the likelihood of surprise attacks and increase the speed with which aid and reinforcements could be delivered to attacked settlements. This defensive structure was then replaced by larger settlements, where people came together to provide defense through numbers. Settlement locations also changed, perhaps reflecting the formation of somewhat stable boundary zones along natural geographic breaks.

L AT E WO O D L A N D C U LT U R A L A DA P TAT I O N S

181

CULTURE HISTORY OF THE LOWER MISSOURI RIVER VALLEY

For much of Missouri prehistory, humans lived primarily in areas where settlements were not far from Missouri’s major river systems and associated drainages. A major locus of cultural development in Missouri’s river systems was in the lower Missouri River valley, which has consequently been the focus of studies by Joseph Harl (1991) and Michael O’Brien and W. Raymond Wood (1998). The lower Missouri River valley is located within the River Hills ecoregion, a transitional zone between the Central Plains and Ozark Highlands ecoregions that is characterized by wooded, loess-covered hills and karst features (Chapman et al. 2002). As an area bordering two very different physiographic regions to the north and south, it would have been a major crossroads for human interactions moving from different directions. It is also an area of high landform variability with upland bluffs, ridges, hills, and lowland terraces and floodplains, providing a wide variety of environments and exploitable resources for humans (figure 9.1). The chronology of prehistoric Missouri relevant to this study generally conforms to the following temporal ranges as indicated by O’Brien and Wood (1998, 4): Middle Woodland period (2250–1550 bp), Late Woodland period (1550–1100 bp), and Early Mississippian period (1100–800 bp). Middle Woodland Period (2250– 1550 bp)

For both the Middle and Late Woodland periods, the dominant subsistence pattern was that of a village-based mixed collector-farming strategy that utilized native cultigens. A simple hunting-gathering existence also persisted for people in many regions (Railey 2010, 3). The Middle Woodland period in Missouri was characterized by a time of vast panregional networks of trade and contact whereby exotic goods traveled great distances and mostly had points of origin in circumscribed areas of Illinois, Ohio, Wisconsin, and Louisiana (O’Brien and Wood 1998, 217). Burial treatments in Middle Woodland Missouri also somewhat reflected the presence of these exotic goods and perhaps differential status in a variety of burial types. The consensus among scholars seems to be that high-status individuals were almost always adult males who achieved their standing through abilities valued by society and not through ascribed status and the trappings of permanent social hierarchy. The entire panregional system was a highly dynamic apparatus set up as a hedge against resource unpredictability and as a means of transmitting information. In addition to trade, it enabled what were probably tribal-level groups to occasionally come together as larger aggregates 182

K ER RY N I C H O L S

Figure 9.1. Study region map of the lower Missouri River valley.

for purposes of exchanging information and new blood (O’Brien and Wood 1998, 218–219). Late Woodland Period (1550– 1100 bp)

Around 1800 bp, the system of panregional exchange changed in that there was increased decorative standardization and stylistic similarity between localities, which is suggestive of decreased social distance between groups. Objects representing individual prestige were less common than objects focused on increasing group unity (Harl 2000, 272). The flow of exotic goods was no longer necessary to maintain far-flung social networks, which is reflected in the widespread standardization in pottery style that comes about in the Late Woodland period. The similarities in pottery styles from so many different parts of Missouri during the Late Woodland period indicate either direct contact between groups or widespread diffusion of ideas (Braun 1977; O’Brien and Wood 1998, 221–224). The Late Woodland period is interesting not only for evidence for increased social contacts, but also for the appearance L AT E WO O D L A N D C U LT U R A L A DA P TAT I O N S

183

of transformative technology and the revolutionary drive toward increased social cooperation and political complexity. Early Mississippian Period (1100– 800 bp)

In Early Mississippian times after about 1100 bp, the rapid shift toward increased political complexity culminates in chiefdom-level political units with depopulated hinterlands and large numbers of people living in nucleated settlements within the larger river valleys. The large Mississippian towns of the American Midwest represent kin-based, hierarchical chiefdoms with ascribed status, intensive agriculture, craft specialization, and concomitant mechanisms of wealth redistribution and the rise of organized, endemic warfare. SETTLEMENT PATTERNS AND THE CONDUCT OF WAR

Gordon Willey (1953) once described settlement pattern studies as a starting point for the functional interpretation of archaeological cultures that represented not just their relationship with the natural environment, but technology, and mechanisms of social interaction and control. The conventional view is that where warfare is endemic, settlements are quite often built on elevated ground or on the bends of rivers where walls or palisades may be more easily constructed ( Jones 2004; Moss and Erlandson 1992, 74; Schaepe 2006, 674; Trigger 1968, 66). Where warfare is not a constant threat, it is likely that sites are located more in deference to an available nearby resource, rather than a location where defensive considerations are more prevalent. However, there is also the possibility of a redoubt where if warning were sufficient, the settlement’s occupants could go to avoid a marauding enemy (Maschner and Reedy-Maschner 1998, 33). Settlement location may sometimes be suggestive of a concern for transport costs and efforts aimed at minimizing movement or making it easier (Hodder and Orton 1976, 58; Johnson 1977, 485). Hodder and Orton (1976, 85) state that clustering of settlements can result from several factors, including the presence of localized resources, the presence of large towns or religious shrines, availability of services, and protection or defense. Very often, regular site spacing is associated with central place theory models that stipulate competition between market centers and control over their peripheries. A concern with maintaining/defending borders or frontiers is another factor favoring regular site spacing. Sites may be regularly spaced in a cooperative manner to facilitate communication between settlements in a common defensive strategy. 184

K ER RY N I C H O L S

Long- Range Weapons and Settlement Considerations

When the bow and arrow were first introduced, it would have made raiding and warfare vastly more efficient and productive from the standpoint of killing enemies and capturing resources. Relative to earlier spear or atlatl (spear thrower) weapon systems, the bow conferred a major advantage in that it had a vastly greater range (Cotterell and Kamminga 1992, 166–182). Raiders with bows could attack from much greater distances, from behind heavy cover, and at very little risk to themselves. Prior to this time, effective warfare often required close contact using spears and shock weapons (axes, clubs). The atlatl was effective as a distance weapon, but lacked the range, accuracy, and stealth of the bow. The arrival of the bow and arrow at least doubled and may have tripled the ability of a hunter or warrior to kill animal or human targets (Bettinger 2013, 119). Its utilization in all likelihood led to the intensification of prehistoric warfare in much the same manner as the introduction of firearms led to the violent transformation of societies throughout the protohistoric and historic periods (Blitz 1988, 136). The techniques used for hunting with advanced projectile weapons such as stealth, concealment, surprise and pursuit/stalking also would have been useful for military applications and the progression from hunting to warfare would be a natural adaptation of the new technology. It has been argued that the decisive factor in the balance of power between intruders and territorial defenders is detection over weaponry (Kelly 2005). It follows then that asymmetrical detection abilities might outweigh superiority in weapons and numbers and might help determine the outcome of border incursions by outside groups. A possible solution to the problem of defense against long-range weapons might be found in increased social cooperation between settlements, which would increase the likelihood of detecting raids before they occur and the ability to extract costs (through capture or killing) raiders either before or after a raid was conducted. At the very least, alliances among geographically close settlements would decrease the likelihood of a massacre through the application of overwhelming enemy force, given that potential victims could flee to allied villages and/or reinforcements could be summoned to help villages under attack. The formation of such defensive alliances would then be reflected in changes in settlement patterns, which would still necessarily reflect other requirements such as access to necessary resources. Site clustering or nucleation then can be a defensive strategy for collective defense in that attackers are at far greater risk of being killed or injured due to the increased ability for those being attacked to quickly generate a counterattack. It can also be a useful strategy for defense against attack in geographic locations with high resource potential but poor defensive potential. L AT E WO O D L A N D C U LT U R A L A DA P TAT I O N S

185

With increasing cooperation and population aggregation as a means of selfpreservation, the presence of localized resources in considering settlement location is more important than movement of site locations to higher altitudes or obviously defensive topography. The ability to signal other sites is also a major advantage, since it could help compensate for the lack of defensive topography and increase a settlement’s defensive capabilities. Changes in settlement patterns can therefore reflect changing resource exploitation or a survival response to increased warfare, or both. With the introduction of the bow and arrow, outlying sites located where they were not part of larger site clusters or associations might see increased efforts to locate in defensive or in highly concealable locations, or to construct defensive features such as earthworks or palisades. These site locations could also be “redoubts” for use on occasion as the need arises (Maschner and Reedy-Maschner 1998, 33). Movement to locations that are defensible but with suboptimal access to food resources could potentially be offset by the greater hunting abilities offered by the long-range capabilities of the bow. Faunal resources could now be exploited to a much greater degree than with older, less efficient projectile technology due to far greater range, accuracy, and stealth (Bettinger 2013). THE ARRIVAL OF THE BOW AND ARROW IN MISSOURI

There is certainly some debate about how early the bow makes an appearance in the North American Midwest, but it is generally considered to be the beginning of the Late Woodland period around 1500–1300 bp (Blitz 1988, 131; Shott 1993, 425). It is no accident that concomitant with the rise of increasingly complex Mississippian societies at the end of the Late Woodland period, evidence of arrow-related trauma in humans increases in many parts of North America (Blitz 1988, 136; Jones 2004). Direct skeletal evidence of arrow wounds associated with the initial introduction of the bow and arrow is lacking in Missouri, probably in part because of the lack of relevant skeletal analysis and in part because arrow wounds often produce soft tissue trauma that does not leave evidence of arrow strikes (Emerson 2007, 138–139). Although there is evidence of bow and arrow adoption in Late Woodland period Missouri by the presence of definite arrowheads in site assemblages, the only known example of a projectile point embedded in human remains from a burial in the lower Missouri River valley is a Late Archaic dart-point tip embedded in the spine of a subadult burial in St. Charles County, Missouri (Harl 2000, 272). Although much removed in time from the Late Woodland 186

K ER RY N I C H O L S

period, this lone example of death by projectile weapon probably reflects a much more common event that is underrepresented in the archaeological record. Although the skeletal evidence from Missouri is scant, there is abundant archaeological evidence for endemic warfare in the Late Woodland and Early Mississippian periods in other areas of the Northern Midcontinent and Mississippi River valley of the United States (Milner et al. 1990; Turner 1986, 132). Unequivocal signs of warfare include human skeletal traumas; the construction of fortified or defensible settlements, which increases dramatically outside of Missouri at about 1200–1000 bp; and analysis of Late Woodland mortuary data, indicating the definite use of the bow and arrow in warfare as indicated by injuries from arrow strikes (Emerson 2007, 130). ARCHERY AND THE RISE OF COMPLEXIT Y

Paul Bingham and Joanne Souza (2009) suggest archery fundamentally transformed prehistoric cultures throughout the continent by providing a coercive means of allowing the formation of larger political units. Others, such as Robert Bettinger (2013), suggest that the introduction of the bow led to a reduction in levels of sociopolitical complexity and reduced the drive toward site aggregation due to the increased economic independence and protection offered by the bow. Still others believe that the superiority of the bow against such earlier projectile weapons as the atlatl is vastly overstated and that it did not offer a significant economic or coercive advantage prior to the emergence of complexity (Shott 1993; Walde 2013). The impact of archery does in fact seem to vary depending on local culture history and environmental contexts. Certainly, other factors contributed to the emergence of complexity where it occurs after the appearance of the bow, and there are situations, such as in the US Southwest, where the emergence of complexity substantially predates the appearance of the bow (Van Pool and O’Brien 2013, 112). Available evidence from the Southwest suggests that “intensely lethal” warfare involving the bow and arrow was at least partly responsible for migrations, site aggregation, nucleation, and definite inter- and intracommunity social differentiation, even after the first appearance of complexity (Van Pool and O’Brien 2013, 114). RESEARCH DESIGN

I expect that the introduction of the bow and arrow into prehistoric Missouri during the Late Woodland period changed the entire Middle Woodland L AT E WO O D L A N D C U LT U R A L A DA P TAT I O N S

187

social dynamic and settlement pattern arrangement such that there was a major increase in social cooperation among settlements tied closely to defensive settlement strategies. I seek to employ a series of spatial-statistical tests to evaluate whether changes in the lower Missouri River valley settlement patterns from Late Woodland to Early Mississippian periods strongly correlate with the introduction of the bow and whether these changes also spurred increases in social complexity. Specifically, I consider four main lines of evidence: (1) site viewshed, (2) site intervisibility, (3) settlement clustering and nucleation, and (4) settlement landform. I anticipate that these lines of evidence will reflect the following general trends in the study area over previous time periods: 1. Sites in the study area should demonstrate increased settlement clustering with a bimodal pattern of clustering in high resource areas and single sites in more defensible or obscured locations. By “high resource” areas, I mean bottomland areas of the lower Missouri River valley that are highly conducive to resource procurement as documented by Harl (1991, 26–35). 2. Sites or site clusters in the study area should exhibit increased line-of-sight visibility to each other relative to periods without bow and arrow technology. 3. Sites or site clusters in the study area should exhibit increased viewsheds at possible avenues of approach, or conversely increased concealment as ways of dealing with territorial incursions. 4. Site area in the study area should increase in size as population nucleation increases over time.

A concern for settlement viewshed has been documented elsewhere and has produced multiple patterns based on the characteristics of the perceived threat and local geography (Maschner and Reedy-Maschner 1998, 32). Diachronic settlement studies such as those by Herbert Maschner and Katherine ReedyMaschner (1998) have been used to demonstrate that the presence of warfare is often accompanied by settlements situated with concern for defensive viewshed arrangements or locations that are hidden or difficult to access. In studies from other geographic areas, defensive settlement patterns associated with the bow and arrow allow excellent surrounding views that provide for maximum early warning of an attack and allow the use of projectile weaponry such that projectiles may be “rained down” on the attackers from a distance (Lambert 2002, 215; Maschner and Reedy-Maschner 1998, 32–33). The two key factors when considering defensive uses of viewshed in settlement choice are thus early warning of attack and maximizing the effect of projectile weaponry against attack. 188

K ER RY N I C H O L S

In considering interaction among archaeological sites of the Late Woodland period, we begin with an initial premise that interaction decreases as the distance between sites increases. This is not always the case as with the high levels of long-distance interaction in the preceding Middle Woodland period but is more generally true due to the increased costs of long-distance movement. We also typically assume that similarities in artifact assemblages between sites relate to the degree and nature of interaction between those sites ( Johnson 1977, 481). If the focus of a defensive network is to allow pooled bodies of defending/ attacking warriors and/or rapid retreat to safety, there must be time for allied settlements to respond quickly to the point of attack. Settlements separated by several kilometers might not offer the ability to quickly assist each other, especially if information must be transmitted in person. However, defensive settlement locations may also hinge on signaling ability in addition to actual geographic distance, which may allow reinforcements from a distance to arrive in cases where they otherwise wouldn’t have time and/or for settlements not yet under attack to prepare their defenses. Of course, sites might be organized to allow intervisibility for reasons aside from defense, instead reflecting factors such as hunting or preventing crop predation. However, signaling and site intervisibility would be central to a defensive settlement system. To be sure, intervisibility between sites can occur by mere chance placement of sites over time, and there is no absolute means of indicating whether sites with intervisibility are contemporaneous (Swanson 2003, 760). However, if a site is part of a larger site settlement pattern, with intervisibility relationships in a physiographically variable environment, the question of a chance occurrence becomes less likely. Due to the fact that landforms in the study area are highly variable, a useful approach to test for significance of site intervisibility is to compare sites of multiple time periods, especially those where archery technology is present and those where it is not and also to generate and test the same number of completely random site locations by time period. If two or more sites are located such that they were able to signal each other, or spaced to avoid intruding on other site catchments, the argument for contemporaneity is strengthened. This argument does not hold everywhere, and there are known situations in the US Southwest where communities moved a short distance and rebuilt, creating a similar pattern (Van Pool 2014, personal communication). Still, by examining the multiple lines of evidence considered here, it should be possible to evaluate the propositions I outlined above. Based on the work of H.  M. Wobst (1972), Grahame Clark (1975), and R. Dennell (1983), Stephen Savage (2006, 331) proposes the hypothesis that foraging societies divided their social territories based on minimum group L AT E WO O D L A N D C U LT U R A L A DA P TAT I O N S

189

subsistence (resource collection) bands and maximum group socialization bands. The minimum group subsistence bands would be oriented around the material needs and requirements of day-to-day life. The larger, maximum group band would be oriented to facilitate contacts with other maximum group bands, trade, acquisition of mates, and the defense of commonly held territory and to maintain harmonious coexistence within maximal band territory between the smaller subsistence-level bands. It is possible that maximum band relationships are visible in the presence of line-of-sight relationships between sites so that the entire band could be mustered in the event of some emergency such as an attack, but even if just portions of the band are intervisible with any given location, it would be expected that defensive networks would use some means of communicating quickly by line of sight. Given the much greater range of the bow and arrow over earlier weapons, the ideal site pattern might be a cooperative network of site clustering such that defensively arrayed outer settlements could warn of an attack to quickly mobilize inner settlements to stem the incursion, creating a pattern like a spiderweb in which outer settlements act as tripwires that ultimately protect inner settlements from surprise attacks. Defenses at the point of contact would have good line-of-sight communications with other sites in the network such that they could quickly signal for reinforcements from inner settlements exactly when and where they were needed from warriors in other parts of the defensive network, creating a system of defense in layers. This would make the need for permanent defensive structures or palisades, which would be difficult to build for short-term settlements, much less necessary as any raid on the connected sites would trigger a rapid response with overwhelming defensive force at the point of attack. Such an arrangement would also allow for sufficient mobility to follow a seasonal round of resource procurement. Such a “circling the wagons” strategy with good interior routes of movement and communication might allow site clusters in geographic areas of high resource potential but less than ideal defensive potential. Regular site spacing or site clustering with good line-of-sight communications could thus reflect an overwhelming concern with territoriality, collective response against attackers, and the desire not to be “overrun.” Defense against attackers has also been proposed as an important factor in population nucleation into site clusters or one large site, since increasing compactness of a community offers several advantages for defense against raiders. In placing settlements closer together or moving people into fewer larger settlements, people can act more effectively as a group. It also makes it easier to construct features such as palisades or fortifications to defend larger numbers of people 190

K ER RY N I C H O L S

(Drennan 1986, 284–285). I therefore propose that increased clustering of sites, regular spacing of sites, or movement to ever larger sites on the landscape is in many cases a direct response to the emergence and spread of long-range projectile weapons technology. These patterns could stand in contrast to sites distributed evenly or clustered in resource-rich zones, which would be expected in the absence of defensive considerations. A consideration of viewsheds, site intervisibility, and local ecology will consequently allow me to differentiate among clustering caused by mapping onto resources and the patterns I expect to result from the introduction of archery, especially when comparing settlement patterns before and after the adoption of the bow. ANALYSIS AND RESULTS

The basic data set used is found in the Missouri State Historic Preservation Office (SHPO) Archaeological Sites and Surveys Geodatabase. The research also utilized the geospatial statistical package found in ESRI ArcGIS 10 software and two different Laser Detection and Ranging (LiDAR) databases consisting of both mosaic and individual tile data, maintained on servers at the Missouri Spatial Data Information Service (MSDIS). These databases allow the creation of high-resolution digital elevation models (DEMs) for accurately modeling the effects of landforms and terrain for analysis. The analysis indicates that line-of-sight relationships between settlements are far more important considerations in the Late Woodland period, when archery was widely adopted, than in the Middle Woodland period: 78 percent of Late Woodland settlements are linked by line of sight, whereas only 50 percent of Middle Woodland settlements are. In the Early Mississippian period, this figure drops to 28  percent. In the Middle Woodland period, there are only twenty-one discrete clusters of sites linked by line of sight. In the Late Woodland, this number jumps to sixty-four clusters before a massive falloff in numbers of sites in the Early Mississippian period. As a percentage of total site numbers by time period, non–visually linked, isolated sites are far more prevalent both before and after the Late Woodland period. Accompanying increased settlement intervisibility, there is a dramatic increase in upland settlement in the Late Woodland period followed by movement to less intervisible lowland settings in the Early Mississippian period (table 9.1, figure 9.2). To compare these results with the same numbers of randomly generated points by time period in the study area, only 16 percent of the random Middle Woodland points were linked by line of sight and only 31 percent of the random Late Woodland points were linked by line of sight. None of the randomly L AT E WO O D L A N D C U LT U R A L A DA P TAT I O N S

191

T able 9.1. Summary of line-of-sight analysis # Sites Linked by Line of Sight

# Sites Linked by Line of Sight (Upland)

# Sites Linked by Line of Sight (Bottomland)

Total Number of Sites

# Isolated Sites

Line-ofSight Linked Clusters

Middle Woodland

121

60

21

61

41

20

Late Woodland

289

61

64

228

177

51

14

10

2

4

1

3

Period

Early Mississippian

Figure 9.2. Middle-Late Woodland and Early Mississippian line-of-sight relationships.

generated points for Early Mississippian were linked by line of sight. These numbers reflect that the observed line-of-sight relationships for all time periods are not a chance occurrence. All time periods under consideration display similar patterns of viewshed arrangements. Late Woodland settlements seem to manifest an increased 192

K ER RY N I C H O L S

Floodplain

Bluff top

Slope

Ridge

Hill

Knoll

River / Stream Terrace

Hillside/ Bench

T able 9.2. Percentage of sites by landform

Middle Woodland Sites (121 Sites)