Europe's Early Fieldscapes: Archaeologies of Prehistoric Land Allotment 3030716511, 9783030716516

Table of contents :
Contents
About the Editors
Contributors
1: Europe’s Early Land Allotment: Questions of Time, Scale and Stewardship
1.1 Fieldscapes: Global and European Issues
1.2 Methods
1.3 Ambitions
1.4 A Future for Field system Research
1.5 Chapter Summaries
1.5.1 Mappings Fieldscapes
1.5.2 New Methods
1.5.3 In-Depth Archaeological Investigations of Field Systems
1.5.4 The Economy of Fieldscapes
1.6 Final Remarks
References
2: Enclosing and Dividing Land: The Neolithic and Bronze Age Field Systems of Shetland
2.1 The Archaeology of Prehistoric Shetland: Houses, Fields and Function
2.2 Mapping the Neolithic and Bronze Age Field Systems of Shetland
2.3 Homestead Enclosures
2.4 Field Systems
2.5 Enclosing and Dividing Land: The Neolithic and Bronze Age Field Systems of Shetland
2.6 Conclusions
References
3: Formation, Use and Chronology of Celtic Fields: New Perspectives from the Groningen Celtic Field Research Programme
3.1 Later Prehistoric Agricultural Landscapes in the Low Countries
3.2 Prelude: A Long-term Perspective on Fields and Subsistence – And Its (in)Visibility
3.3 Celtic Fields and Where to Find Them
3.4 Celtic Fields and How to Date Them
3.5 How Celtic Field Banks Came to Be
3.6 Celtic Field Economy
3.7 A Cultural-Landscape Approach to Celtic Fields: The Before and During
3.8 Fields of Opportunity: What Is Left to Learn?
References
4: A Large-Area Prehistoric Cultural Landscape in the Sachsenwald Forest Near Hamburg
4.1 Introduction
4.2 Sachsenwald Celtic Fields
4.3 Celtic Field Land Forms
4.4 Implications
References
5: A Sheep’s Eye View: Land Division, Livestock and People in Later Prehistoric Somerset, UK
5.1 Introduction
5.2 Background
5.3 Livestock in the Landscape – Understanding the Needs of the Animals
5.4 Managing Grazing – Understanding the Contribution of the Land
5.5 Fields as Infrastructure
5.6 Model Systems
5.7 Applying the Models – Cadbury Castle and its Environs, Somerset, UK
5.8 Time and Place – Discussion
5.9 Conclusion
References
6: Terraced Crop Fields in the Eastern Pyrenean Mountains (France): The View from Pedoarchaeology
6.1 Cultivated Fields, Soils and Palaeosols in Terraced Mountains
6.1.1 The Dating of Terraces
6.1.2 Bronze Age Fields in the Mediterranean Mountains
6.1.3 Bronze Age Agricultural Terraces
6.1.4 Archaeology and Soil Science
6.2 A mountain Only for Shepherds?
6.3 Under the Grassland, the Crop Fields
6.4 The Spread of Terraced Field Systems
6.5 The Significance of Mountainous Crop Fields in Terms of Social and Territorial Organization
6.5.1 Soil Improvement at High Altitude and Agricultural Know-How
6.5.2 Settlement and Land-Use Patterns from Bottom-Up and Top-Down
References
7: Fields and Farming-Systems in Bronze Age Scotland
7.1 Introduction
7.2 The Extent of the Bronze Age Landscape
7.3 The Character of Settlement
7.4 The Landscape Context of Burnt Mounds
7.5 The Character of Fields and Enclosures
7.6 A Landscape System in Action
7.7 Conclusion
References
8: Terraced Fields, Farming, and Farmers at the Settlements of Kalamianos and Stiri, Greece
8.1 Introduction
8.2 Terraces in the Bronze Age Aegean
8.3 Field Methodology and Results
8.4 Cultivation at Kalamianos and Stiri
8.5 Changed Land, Changed Labour
8.6 Women’s Farm Work
8.7 Terrace Construction as Agricultural Labour
8.8 Conclusions and More Questions
References
9: The Changing Fieldscapes of Loughcrew: New Insights from Airborne Lidar
9.1 Introduction
9.2 The Loughcrew Landscape
9.3 The Loughcrew Landscape Project
9.4 Linear Features and Boundaries
9.4.1 Categorising Boundaries
9.4.2 Irregular Banks
9.4.3 Rectilinear Boundaries
9.4.4 Boundary Changes
9.5 The Effect of the more Recent Land Use
9.6 Conclusions
References
10: My Home Is My Castle! Field Systems and Farms: Rhythm and Land Appropriation During the Bronze Age in North-West France (2300–800 BCE)
10.1 Introduction
10.2 The Third Millennium: Twenty-First – Eighteenth Century BCE
10.2.1 A Historical Interpretation
10.3 The Seventeenth–Sixteenth Centuries BCE
10.3.1 A Historical Interpretation
10.4 The Fifteenth–Thirteenth Centuries BCE
10.4.1 A Historical Interpretation
10.5 Thirteenth–Twelfth Centuries BCE
10.5.1 A Historical Interpretation
10.6 Conclusion
References
11: Tracing the Remains of a Late Bronze Age Field System in Central Mainland Greece
11.1 Glas and the Late Bronze Age Drainage of the Kopais
11.2 Geophysical Results
11.2.1 Connecting Glas with the Drainage System
11.2.2 Evidence of an Irrigated Agricultural Field System
11.3 ‘Ground-Truthing’ the Character and Nature of the Anomalies
11.4 Age of the Features
11.5 Other AROURA Investigations
11.6 Reconstruction of the Hydraulic Systems, and Bronze Age Comparanda
11.6.1 Hypothetical Reconstruction
11.7 Cases for Comparison
11.8 Why the Hydraulic Engineering?
References
12: Reconstructing Enclosed and Parcelled Out Landscapes from the First Millennium BC in Himmerland, Denmark: Arable Fields, Grazing Land and Settlement Patterns Examined in Three Micro-regions
12.1 Introduction
12.2 Sources and Methods
12.3 The Vegetation History of Western Himmerland
12.4 Cultivated land and Sediment Deposition
12.5 The Topography and the Settlement History of Skørbæk Hede
12.6 Land Use at Skørbæk Hede
12.7 The Topography and Settlement History Around Gundersted
12.8 Land Use at Gundersted
12.9 The Topography and Settlement History at Store Binderup
12.10 Land Use at Store Binderup
12.11 Land Use in the Micro-regions: A Summary
12.12 Settlements in Himmerland During the Bronze Age and Early Iron Age and Their Relation to Field Systems
12.13 Conclusion
References
13: Understanding the Chronologies of England’s Field Systems
13.1 Introduction
13.2 Project Context and Aims
13.3 Methodology
13.3.1 Sources
13.3.2 Database Structure
13.3.3 Field System Categorisation
13.3.4 Dating Methods
13.3.5 Biases
13.4 Dating Methods and Materials
13.5 Dating Strategies
13.6 Geographic Variation
13.7 Chronological Variation by Region
13.8 Chronological Variation by Type
13.8.1 Aggregated Fields
13.8.2 Coaxial Fields
13.8.3 Linear Land Division (Including Pit Alignments)
13.8.4 Uncategorised Boundaries and Field Systems
13.9 Discussion and Conclusions
13.9.1 Historical Narratives
13.9.2 Methodologies and Data Standards
Appendix: Table of Sites
References
14: The Prehistoric Agrarian Management Pattern in the Volcanic Landscape of the Campanian Plain, Southern Italy
14.1 Introduction
14.2 Campanian Plain: A Geo-Archaeological Overview
14.3 Prehistoric Settlement and Agrarian Evidence in the Campanian Plain
14.3.1 The Territory of Gricignano d’Aversa, Caserta Province
14.3.2 The Prehistoric Evidence Between the Rivers Clanis/Regi Lagni and Sebeto
14.3.3 Naples Underground Lines and Fuorigrotta
14.4 Discussion
14.5 Conclusions
References
Index

Citation preview

Themes in Contemporary Archaeology

Stijn Arnoldussen Robert Johnston Mette Løvschal Editors

Europe’s Early Fieldscapes

Archaeologies of Prehistoric Land Allotment

Themes in Contemporary Archaeology Series Editors Peter Attema, Groningen Institute of Archaeology, University of Groningen, Groningen,  The Netherlands Agathe Reingruber, Institut Prähistorische Archäologie, Freie Universität Berlin, Berlin, Germany Robin Skeates, Department of Archaeology, Durham University, Durham, UK

The Themes in Contemporary Archaeology series provides cutting edge perspectives on key areas of debate in current archaeological enquiry, with a particular emphasis on European archaeology. The series has a broad coverage, encompassing all archaeological periods and all approaches. Examples of topics welcome in the series include, but are not limited to: • • • •

from theoretical debate to archaeological practice landscape studies bioarchaeology issues of cultural heritage

The volumes are based on research presented at the Annual Meetings of the European Association of Archaeologists. They include proceedings of individual sessions, which can be enlarged if necessary to provide coherence for publication. Each volume undergoes a strict peer-review process, ensuring volumes of high quality that capture current debates in the field. SERIES EDITORS The Series editors are Peter Attema, Agathe Reingruber and Robin Skeates ([email protected]). DISCOUNT FOR MEMBERS All EAA members can buy print and e-copies of all THEMES volumes at 30% discount please email [email protected] who will assist you.

More information about this series at http://www.springer.com/series/16659

Stijn Arnoldussen  •  Robert Johnston Mette Løvschal Editors

Europe’s Early Fieldscapes Archaeologies of Prehistoric Land Allotment

Editors Stijn Arnoldussen Groningen Institute of Archaeology Groningen University Groningen, The Netherlands

Robert Johnston Department of Archaeology University of Sheffield Sheffield, UK

Mette Løvschal Department of Archaeology and Heritage Studies Aarhus University and Moesgaard Museum Højbjerg, Denmark

ISSN 2730-7441     ISSN 2730-745X (electronic) Themes in Contemporary Archaeology ISBN 978-3-030-71651-6    ISBN 978-3-030-71652-3 (eBook) https://doi.org/10.1007/978-3-030-71652-3 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Contents

1 Europe’s Early Land Allotment: Questions of Time, Scale and Stewardship���������������������������������������������������������������������������������������������    1 Mette Løvschal, Stijn Arnoldussen, and Robert Johnston 2 Enclosing and Dividing Land: The Neolithic and Bronze Age Field Systems of Shetland�������������������������������������������������������������   13 Claire Christie 3 Formation, Use and Chronology of Celtic Fields: New Perspectives from the Groningen Celtic Field Research Programme���������������������������������������   27 S. Arnoldussen 4 A Large-Area Prehistoric Cultural Landscape in the Sachsenwald Forest Near Hamburg�������������������������������������������������������������   43 Volker Arnold 5 A Sheep’s Eye View: Land Division, Livestock and People in Later Prehistoric Somerset, UK �������������������������������������������������������������������������   53 Clare Randall 6 Terraced Crop Fields in the Eastern Pyrenean Mountains (France): The View from Pedoarchaeology�����������������������������������������������������������������������������   71 Romana Harfouche and Pierre Poupet 7 Fields and Farming-Systems in Bronze Age Scotland�������������������������������������������   87 Stratford Halliday 8 Terraced Fields, Farming, and Farmers at the Settlements of Kalamianos and Stiri, Greece�����������������������������������������������������������������������������  101 Lynne A. Kvapil 9 The Changing Fieldscapes of Loughcrew: New Insights from Airborne Lidar�������������������������������������������������������������������������  115 Corinne Roughley, Elizabeth Shee Twohig, Colin Shell, and Gillian Swanton 10 My Home Is My Castle! Field Systems and Farms: Rhythm and Land Appropriation During the Bronze Age in North-West France (2300–800 BCE) �����������������������������������������������������������������  131 Cyril Marcigny and Rebecca Peake 11 Tracing the Remains of a Late Bronze Age Field System in Central Mainland Greece������������������������������������������������������������������������������������  143 Michael F. Lane and Vassilis L. Aravantinos

v

vi

12 Reconstructing Enclosed and Parcelled Out Landscapes from the First Millennium BC in Himmerland, Denmark: Arable Fields, Grazing Land and Settlement Patterns Examined in Three Micro-regions���������������������������������������������������������������������������  167 Michael Vinter 13 Understanding the Chronologies of England’s Field Systems �����������������������������  185 Robert Johnston, Rowan May, and David McOmish 14 The Prehistoric Agrarian Management Pattern in the Volcanic Landscape of the Campanian Plain, Southern Italy�������������������  209 Fabio Saccoccio Index�����������������������������������������������������������������������������������������������������������������������������������  225

Contents

About the Editors

Stijn Arnoldussen is an Associate Professor in Later Prehistory at Groningen University. He is fascinated by the long-term development of (agri)cultural landscapes – particularly field systems and the interplay of settlements, funerary sites and deposition between the Neolithic and the Roman Iron Age. He has recently excavated various later prehistoric field systems (known as Celtic fields or raatakkers in Dutch) in the Netherlands.  

Robert Johnston is a Senior Lecturer in Landscape Archaeology at the University of Sheffield. He recently published ‘Bronze Age Worlds’ (2021), in which he considers the ways that kin relations were fundamental in forming the social life and landscapes of Bronze Age Britain and Ireland. He is currently researching landscape transformations in the upland and coastal landscapes of western Britain.  

Mette Løvschal is an Associate Professor at Aarhus University’s Department of Archaeology and Heritage Studies and Moesgaard Museum. She is the Principal Investigator on the ERC project ‘Anthropogenic Heathlands: The Social Organization of Super-Resilient Past Human Ecosystems (ANTHEA) 2020–2025’. Her research engages with a range of archaeological, social anthropological and philosophical debates pertaining to spatial ontology and deep time trajectories, land tenure changes, disturbance ecologies, spatial perception and the becoming of biosocial entanglements.  

vii

Contributors

Vassilis L. Aravantinos  9th Ephorate of Prehistoric and Classical Antiquities, Thebes, Greece Stijn  Arnoldussen  Groningen Institute of Archaeology, Groningen University, Groningen, The Netherlands Volker Arnold  Berliner Strasse 61, Heide, Deutschland Claire Christie  Headland Archaeology, Edinburgh, Scotland Stratford Halliday  Rath Manach, Peeblesshire, UK Romana  Harfouche CNRS et Université Panthéon-Sorbonne, UMR 7041 ArScAn (Archéologie et Sciences de l’Antiquité), Robiac-Rochessadoule, France Robert Johnston  Department of Archaeology, University of Sheffield, Sheffield, UK Lynne A. Kvapil  Butler University, Indianapolis, IN, USA Michael  F.  Lane Ancient Studies Department, University of Maryland, Baltimore County, Baltimore, MD, USA Mette  Løvschal  Department of Archaeology and Heritage Studies, Aarhus University and Moesgaard Museum, Højbjerg, Denmark Cyril Marcigny  Inrap, UMR 6566-CReeAAH, Institut national de recherches archéologiques préventives, Centre archéologique Inrap de Normandie, Bourguébus, France Rowan May  ArcHeritage, Sheffield, South Yorkshire, UK David McOmish  National Strategy Team, Historic England, London, UK Rebecca Peake  Inrap, UMR 6298-ArTeHiS, Institut national de recherches archéologiques préventives, Centre archéologique Inrap de Passy, Passy, France Pierre Poupet  Soil Science, CNRS, Robiac-Rochessadoule, France Clare  Randall Department of Archaeology and Anthropology, Bournemouth University, Bournemouth, Dorset, UK Corinne Roughley  Hughes Hall, University of Cambridge, Cambridge, UK Fabio  Saccoccio Department of Classics and Archaeology, University of Nottingham, Nottingham, UK Colin  Shell McDonald Institute for Archaeological Research, University of Cambridge, Cambridge, UK Gillian Swanton  Kynet Consultancy, Marlborough, Wiltshire, UK Elizabeth  Shee  Twohig  Department of Archaeology, University College Cork, Cork City, Ireland Michael Vinter  Moesgaard Museum, Højbjerg, Denmark ix

1

Europe’s Early Land Allotment: Questions of Time, Scale and Stewardship Mette Løvschal , Stijn Arnoldussen , and Robert Johnston

Abstract

Early field systems, in some cases several millennia old, are tangible relicts of past large-scale processes of landscaping and land allotment in many regions across Europe. Yet our cultural landscapes, created in both the near and distant past, are disappearing fast and often irreversibly so, showing how the management and preservation of fieldscape heritage is a real and urgent issue for landscape conservation policy and practice. In this introductory chapter, we provide a conceptual framework for the study of later prehistoric land allotment in Europe, including a discussion of major methodological approaches, findings and future research priorities. We outline the main scope of this volume, followed by a series of summaries of the individual chapters and describe the cross-chapter themes and approaches. Moreover, we use this introductory chapter as an opportunity to critically evaluate the research field of prehistoric fieldscapes, its current state in European archaeology as well as its future challenges and perspectives.

M. Løvschal (*) Department of Archaeology and Heritage Studies, Aarhus University and Moesgaard Museum, Højbjerg, Denmark e-mail: [email protected] S. Arnoldussen Groningen Institute of Archaeology, Groningen University, Groningen, The Netherlands e-mail: [email protected] R. Johnston Department of Archaeology, University of Sheffield, Minalloy House, Sheffield, UK e-mail: [email protected]

Keywords

Land allotment · Land enclosure · Land-use change · Anthropocene · Fieldscape heritage · Agriculture · Pastoralism

1.1

 ieldscapes: Global and European F Issues

Field systems of immense time-depth weave historical textures into the fabric of Europe’s present-day landscapes. In many regions, field systems dating back to the 2nd millennium BC are tangible relicts in contemporary landscapes. Often, the organisation of rural landscapes embodied in banks, ditches, hedges or terraces echoes much older historic and prehistoric forerunners (e.g. Müller 2013), illustrated by the long-term influence of large-scale landscape structures such as Roman period centuriation (e.g. Caravello and Michieletto 1999; Clavel-Lévêque and Orejas 2002; Upex 2002; Bonnie 2010). These historically constituted, deep-set fieldscapes  represent the largest cultural ‘artefacts’ in Europe. In England, historic fieldscapes covered at least 70% of the land (Johnston, et  al., this volume); for parts of Denmark, prehistoric  landscape parcelling may have even covered more than 75% of all available space (Vinter 2011) – leaving little room for alternative landscape forms. Critically, however, fieldscape heritage is unlike an archaeological artefact hidden beneath the surface or displayed in a museum. Fieldscape heritage is emergent and influential in the contemporary landscape, actively co-shaping the cultural identities and characters of regions (cf. Gruffudd 1995; Pinto-Correia 2000; Antrop 2006: 188; Allen 2014: 165). This is not a phenomenon exclusive to landscapes of the deep past. But despite their vast extent and influence, the relics of prehistoric field systems remain vulnerable to neglect, erosion and economic developments across Europe, as are ancient landscapes and rural ecologies globally.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 S. Arnoldussen et al. (eds.), Europe’s Early Fieldscapes, Themes in Contemporary Archaeology, https://doi.org/10.1007/978-3-030-71652-3_1

1

2

Across the globe, the fabrics of cultural landscapes created by generations of predecessors are disappearing fast and often irreversibly so. As we prepare this book for publication, international companies, military forces and small-­ scale landholders are clearing many thousands of hectares of Amazonian forest for soya production and livestock grazing (Stabile et al. 2020). Similar large-scale human-driven landscape transformations play out across vast areas of the planet, including the forests in Indonesia and Malaysia (Teheripour et al. 2019), Southeast Asia and Africa (Vijay et al. 2016). Concurrently, land apportionment and fencing are appearing in formerly open, unfenced pastoral landscapes as part of large-scale land confiscation and privatization  processes (Galvin et al. 2008). This has recently been the situation in Inner Mongolia (Taylor 2006) and East Africa (Said et  al. 2016; Løvschal et al. 2017; Løvschal and Gravesen 2021), reminding us of  the rural histories of  Australia and the American Great Plains in the late nineteenth century (Netz 2004; Greer 2012). Such fast-paced land-use changes are leading to deforestation and desiccation at unprecedented scales  – contributing to the ‘great acceleration’ of the Anthropocene (Steffen et al. 2015). The systematic destruction of forest and grasslands and enclosure of landscapes is partly driven by international geopolitics and the globalised economy. Its destruction will have unpredictable climate-, species-, human-, cultural- and historical consequences on a local through to  planetary scale (e.g. Ogden et  al. 2013; Steffen et  al.  2015). And as new fieldscapes are created in responses to current global demands for food and fuel, the historic fabric and connections with land are often also lost. Hence, the conservation and governance  of cultural landscapes, including those of past field systems, is a real and urgent issue in heritage management policies across Europe (e.g. Chaps. 9, 12 and 13 this volume). Moreover fast-paced landscape transformations are not merely, and frequently negatively, eroding the cultural heritage of rural landscapes. They also impair researchers’ efforts to study how past communities sustained long- and short-term modes of tenure and land use, often over centuries. As rich, living archives of past pastoral and agricultural livelihoods and economies, Europe’s ancient field systems provide foundations for recognising European rural histories. Research on ancient fieldscapes can shape solutions that ease the tensions between heavy carbon-footprint agriculture and ecologically and socially resilient polyculture in foodeconomies. Moreover, given our current environmental crisis, questions pertaining to large scale land allotment and the social and economic significance of fieldscapes have never been as actual or as urgent. In this volume, we explore a series of case-studies across Europe to address their potential as sources of information on past relations between humans, animals and their environments. The existence of field systems in pre- and protohis-

M. Løvschal et al.

toric Europe has revealed fascinating insights into the deep history of pastoral and agricultural economies (e.g. Van Giffen 1918; Curwen and Curwen 1923; Hatt 1949; Bradford 1957; cf. Rackham and Moody 1992; Franceschelli 2015). Field systems mark the ongoing investment, improvement, structuring and restructuring of human and other-than-­ human relationships within landscapes. Therefore, by studying the allotment and long-term  use of fieldscapes, as represented in this book, we get an insight into the complex organisation of cultural and economic relations that shaped the European countryside.

1.2

Methods

The methodological toolkit for investigating Europe’s vast prehistoric fieldscapes has expanded significantly over the last five decades. For much of the twentieth century, surveys and excavations were constrained in extent and ambition by the available archaeological technologies (e.g. Van Giffen 1918; Curwen and Curwen 1923; Hatt 1949, cf. Brongers 1973; Klamm 1993: 5–16). Since the 1970s, prehistoric field systems have more often, more deliberately, and more extensively been targeted for open area excavations, and the means for their identification and mapping have significantly improved through ambitious aerial mapping projects and advances in remote sensing (cf. Chap. 7). Amongst the plethora of new methods and data, we need to remember the merits of intensive analytical survey. Both for mountainous terrace landscapes and lowland landscapes, where field systems generally comprise banks, lynchets or ditches, detailed ‘feet on the ground’ surveys have proven vital to accurately identify, map and better understand field systems (e.g. Chaps. 5, 6, 8 and 11). Yet due to their spatial extent, it is often not feasible to employ analytical ground survey on extensive scale. In these cases, the availability and refinement of mapping from LiDAR (Light Detection and Ranging) datasets has proved crucial in identifying earthwork field and terrace systems (e.g. cf. Humme et al. 2006; Kooistra and Maas 2008; Clemmensen 2010; Hesse 2010). The high resolution of the topographic models created from LiDAR data, from which the spatial fabric of past field systems can be mapped, should not be considered an end-point of the investigative process. However accurately they may be mapped, archaeological features present themselves as a deceptively atemporal snapshot of centuries of field system making and use, and masking details and diachronic changes in agricultural specifics, layouts and regimes (Nielsen and Dalsgaard 2017: 416). A new generation of investigations have emerged in which pedological and palaeo-ecological analyses are integral to  research design (e.g. Spek et  al. 2003; Nielsen and

1  Europe’s Early Land Allotment: Questions of Time, Scale and Stewardship

Dalsgaard 2017; Nielsen et  al. 2019). This integration has enabled the reconstruction of the pastoraland  agricultural practices, beings and processes that operated within and helped to shape the field systems. For example, details of manuring strategies involving plaggen, muck and midden materials have come to light in case-studies in Denmark, Scotland, England, the Netherlands, the Pyrenees as well as in Greece (Chaps. 6–8, 12 and 13). Barely tangible aspects, such as the frequencies and durations of fallow periods, can be recognised by integrating palynology, anthracology and pedology. Incorporation of the study of algae and non-pollen palynomorphs (NPPs) have helped to identify manuring, non-local soil admixture and to create more robust reconstructions of local vegetation (cf. Arnoldussen and Linden 2017; Enevold et al. 2019). These details are needed if we want to assess the intensities and rhythms of land-use within ancient field systems. For example, Behre (2008: 115) proposes that no more than 10% of the Flögeln Celtic field plots were in use simultaneously for crop-cultivation. For most prehistoric field systems, however, we still lack robust data on fallow duration, derived from percentages of biennials, and nutrient conditions (e.g. depletion risks and manuring requirements). The methods, such as environmental aDNA and those listed above, are available to address these gaps in the future. Moreover, high resolution chronologies are required in order to effectively archaeological and environmental histories. There is an increased awareness of the limitations of dating prehistoric field systems by morphology (cf. Chap. 13), cultural material (cf. Chap. 3) or single radiocarbon dates (cf. Chap. 4). In systems where re-use of older sites was common or manuring involved an admixture of settlement waste the reliability of dating agricultural phases using pottery, charcoal or other artefacts should be evaluated ­ critically. There are two evident routes towards more robust chronologies for Europe’s field systems. The first involves critically evaluating aggregated dates for particular types of field system phenomena (Løvschal 2014, 2020; Chap. 13). For example, basing themselves on a corpus of 323 dates for 120 sites, Johnston and colleagues (Chap. 13) identify an initial phase of large-scale apportionment of England’s field systems in the early centuries of the second millennium BC, which became more widespread after 1700 BC. They also use the dataset to argue for a waning in the construction and maintenance of field systems during 1000 BC  and 600  BC.  The second approach can be characterised by the realisation that efforts to date specific field system elements by just a single date, or single type of date, are intrinsically flawed for building detailed diachronic narratives (Chaps. 3 and 13). Foremost, the application of OSL and TL as direct dating strategies (e.g. Chaps. 6 and 11) or combined application of OSL and

3

AMS dating will aid in unravelling the internal developmental trajectories of aggregate complexes of field systems. Moreover, for the north-west European field systems consisting of earthen banks, stratified AMS- and OSLdates have already shown how perimeter banks can act as long-term chronostratigraphic repositories of agricultural use, not just preserving specifics of agricultural usage, but forcefully illustrating the resilience of such systems (Arnoldussen 2018; Chaps. 3 and 13). As a consequence, there is good reason to be optimistic about the extensive and still-growing methodological toolkit for making new discoveries about Europe’s prehistoric fieldscapes.

1.3

Ambitions

This book’s focus is on the development of fieldscapes through time and space and in their wider landscape context. The chapters address established topics relating to past land management regimes, including manuring, water, land and livestock and  crop management, and technologies such as slash-andburn, ard and plough. Several research themes and questions link the chapters: (a) the mapping and understanding of field systems at various scales; (b) interpreting social processes from field system morphologies; (c) the relations between field systems and cultural and natural features of their environments; (d) the time-depths and temporalities of usage; and (e) the specifics of the underlying land tenure systems, with special attention to matters of resilience and changing practices (Table 1.1). The chapters take a largely anthropocentric perspective, which focuses on the human shaping and inhabitation of landscapes over time. This reflects the volume’s attention to the physical, human-driven allotment of landscapes through various uses and manipulations. This approach is deeply embedded in the research history of prehistoric fieldscapes. For example, the early and continuing importance of remote sensing of field systems through aerial imagery has so a large degree disconnected researchers from the land and its processes, and made humans remote from these landscapes (cf. Kostyrko et al. 2016). There are current theoretical positions that decentre humans and recognise alternative ontologies for relating human and nonhuman histories. They offer other logics than those of the westernized, sedentary and human-­ oriented, which more explicitly take into account the role of multispecies assemblages in the resilience of allotted landscapes (e.g. Deleuze and Guattari 1980; Haraway 2008; Latour 2004; Povinelli 2016; Tsing 2015). For the moment, few scholars researching fieldscapes have pursued such paths. It is our anticipation – and hope – that this will change in coming years.

Table 1.1  Coverage of main research themes by chapter. The number of crosses indicate the relative importance given to that specific theme in a paper. For details see relevant chapters

Ch. 2 3 4 5 6 7 8 9 10 11 12 13 14

Author(s) Christie Arnoldussen Arnold Randall Harfouche & Poupet Halliday Kvapil Roughly, et al. Marcigny & Peake Lane & Aravantinos Vinter Johnston, et al. Saccoccio

Mapping and understanding field system morphologies at various scales +++ +++ +++ ++ ++ ++ +++ +++ ++ +++ ++ + +

The extraction of information on social processes from field system morphologies +++ +++ + +++ ++ ++ +++ ++ +++ ++ + + ++

The relations between field systems and cultural and natural features of their environment, +++ ++ +++ +++ +++ +++ ++ ++ +++ +++ +++ + +++

Time-depths and temporalities of usage + ++++ ++ + + + + ++ ++ +++ ++ ++++ +++

Specifics of the underlying agricultural systems + +++ + ++++ +++ ++ +++ + + +++ ++++ + +

Ch. 2 systems of shetland

Ch. 7 Fields and farming systems in Bronze Age Scotland

Ch. 12 The enclosed agrarian landscape of Himmerland

Ch. 13 Understanding the chronologies

Ch. 9

Ch. 4 Prehistoric Cultural landscape in the Sachsenwald Forest

of Loughcrew

Ch. 3 New perspectives from the

Ch. 5 Land division, livestock and people in later prehistoric Somerset

Ch. 10 Field systems and farms in Bronze Age north-west France

Ch. 6 eastern Pyrrenean mountains

Ch. 14 The Bronze Age field systems of the Campanian Plain

Ch. 11 in central mainland Greece

Ch. 8 at Kalamianos and Stiri

Fig. 1.1  Distribution and characterisation of case-studies targeting European field systems in this volume

1  Europe’s Early Land Allotment: Questions of Time, Scale and Stewardship

While accepting this limitation, this volume addresses the above five research themes in the first comprehensive insight into prehistoric fieldscape research across Europe. We wish to raise a broader awareness of some of the main scientific questions that are addressed by scholars working in various fieldscapes across Europe (Fig.  1.1). Their questions raise fundamental issues with the interpretation of European fieldscapes and their long-term histories, and their case-studies exemplify the established and novel methodologies that can progress our knowledge. In addition, the volume offers inspiration and guidance for the conservation management of fieldscape heritage, which we hope will stimulate the strategic guidance that will frame and support improved protection of Europe’s fragile landscape heritage. A series of important research questions are raised and answered across the chapters. Several papers examine distinct field system morphologies within their culture-­ historical contexts. Christie (Chap. 1) and Halliday (Chap. 7) discuss the landscapes of Shetland and Scotland more broadly in which organic, small scale enclosures and field boundaries regulated agricultural use, without evidence for extensive co-­ axial structures. Roughly and colleagues’ case-study of Loughcrew, Ireland (Chap. 9), shows how detailed map-­regression can reveal the deep temporal trajectories of patterns of irregular banks and rectilinear boundaries. For England, Johnston and colleagues (Chap. 13) review the scientific dates available for coaxial/cohesive, aggregated (i.e. regular and irregularly accreted), formal terraced, open, enclosed and parliamentary fields. The case-studies by Randall (Chap. 5) and Marcigny and Peake (Chap. 10) are excellent examples of how systems of ditches reflected (changing) agricultural strategies  – and may have been tailored as much to livestock rearing as crop-cultivation. In the low-altitude regions around the North Sea coast, Vinter (Chap. 12; Denmark), Arnold (Chap. 4; Germany) and Arnoldussen (Chap. 3; the Netherlands) discuss systems of sub-rectangular embanked fields that extend across hundreds of hectares. The case studies from higher-altitude zones, such as the French Pyrenees (Chap. 7) and Corinthia (Chap. 8), illustrate that terracing is best understood in tandem with the study of the settlement systems and subsistence economies that required their construction (Chaps. 6 and 8). The chapters also demonstrate the variation in the boundary architecture of field systems that existed across Europe: ditches (Chaps. 5, 10, 13 and 14), earth and stone banks (Chaps. 2, 3, 4, 7 and 12), lynchets (Chaps. 7, 12 and 13), terracing (Chaps. 6, 8 and 11) and canals (Chap. 11). It is important to realise that unenclosed fields too may have formed part and parcel of prehistoric agricultural strategies. Bjerre Enge in Thy, northern Denmark, is a well-documented case, where a series of irregular fields from c. 900 to 700 BC were separated only by unploughed wetter areas (Bech and Mikkelsen 2018; cf. Nielsen 1993). Another buried culti-

5

vated field was relatively recently discovered at Swifterbant S4, the Netherlands, where the mixing of anthropogenic materials such as carbonized plant remains and burnt bone in the soil is interpreted as the result of tillage (Huisman et al. 2009). Some features, such as hedges, fences and pathways, are often difficult to detect archaeologically. Yet, Chaps. 3, 5, 10 and 13 offer evidence for landscape structuring using fences or hedges. Pathways and roads that opened-up field systems to maintenance, manuring and harvesting, are similarly infrequently documented but discussed in Chaps. 4, 8, 10 and 14. We hope that such descriptions of rarer components of field systems may aid their conceptualisation and recognition elsewhere in the future. A comparable observation can be made with respect to the practical life of fields: hoe-based horticulture (Chaps. 3 and 14), gardening and polyculture, manuring strategies, ard-marks as proxies for sowbed-preparation or clearance (Chaps. 6, 7 and 14; Arnoldussen 2018: 315), the roles of clearance through burning and fallow cycles (Chaps. 3, 6, 7 and 12; Dev 2018), crop-rotation (Chaps. 5 and 8) and the dynamics of managing livestock and crop-plots (Chaps. 5 and 7). While the chapters touch on many of these topics, they also illustrate the need for further sustained research beyond the case-studies. Other vital research questions and topics are underrepresented in the chapters. For instance, how might variation in social organization correspond with different forms of land management, and agricultural and tenurial regimes? We still understand relatively little about the social and labour organisation of agricultural and pastoral regimes in prehistory, and how they influenced and were related to other spheres of life. The volume also has chronological gaps. The chapters generally present the earliest manifestations of fieldscapes in later prehistory (c. 2000–500 BC; Table 1.2). However, other studies from later periods have shown the potential of relating late prehistoric and early historic land allotment processes through a combination of archaeological and historical evidence, as has been carried out fruitfully by Susan Oosthuizen (2013) and Peter Fowler (2000). Finally, there remain unaddressed questions about how prehistoric field systems can have relevance for present-day communities throughout Europe by connecting people with land, contributing to sustainability education, and shaping the identities of localities. We hope the book can offer a catalyst for research on the public archaeology or heritage of fieldscapes.

1.4

A Future for Field system Research

Notwithstanding the rich and detailed narratives offered by the case-studies in this volume, we judge that there remains significant work to be done. A first issue to raise concerns

Ch. 2 3 4 5 6 7 8 9 10 11 12 13 14

Author(s) Christie Arnoldussen Arnold Randall Harfouche & Poupet Halliday Kvapil Roughly, et al. Marcigny & Peake Lane & Aravantinos Vinter Johnson, et al. Saccoccio . .

.

.

3500-3000 . .

.

.

4000-3500 . .

.

.

. .

X

3000-2500 x

.

X x . . x

2500-2000 X

X x X x . X X . x X

2000-1500 X X

X X X? X X x X x

X? x x X x

x

X

X

X X

1000-500

1500-1000

.

x X X

500-0

.

X X

.

.

. x

AD 0-500 AD 500-1000

Table 1.2  Overview of main periods of field system activity for each case-study. Dots indicate older or later activity, small crosses minor and big crosses major activities. For details see relevant chapters

6 M. Løvschal et al.

1  Europe’s Early Land Allotment: Questions of Time, Scale and Stewardship

chronologies, tempos and temporalities. It is still a rather open question when and how fast the semi-permanent field systems emerged in the different areas of Northern Europe. Do we know enough about continuity and potential periods of abandonment or shifting land-use patterns? And which rhythms of labour and movement were embedded in the fieldscapes? Investigating such issues, including the speed and phasing of land allotment processes, requires a chronological resolution often not yet available. This renders difficult narratives about how boundaries may have changed from notional and invisible to tangible and visible  (Løvschal and Gravesen 2021), or from one boundary type to another, such as from fence to bank, and bank to ditch. Imprecise chronologies impact in our knowledge of the earliest phases of the uselife of fieldscapes. The earliest formation of fieldscapes marked the instantiation of particular landscape logics that, combined with landscape affordances, structured the long-term trajectories along which landscapes developed (Løvschal 2020). Much more knowledge is needed about the chronologies involved in such trajectories, in order to investigate causalities and  correspondences with other ecological and cultural trajectories. Second, research on the underpinning forms of social organisation and traditions of tenure are underrepresented, particularly outside Britain (cf. Wickstead 2007). The study of field systems is still a research field that is dominated by positivist notions of economic rationality, maximization and individual (human) choice, and too frequently studied as a domain separate from other aspects of human and nonhuman life. To counter overly functionalist and economic approaches, we need an increased focus on the social, organisational and more-than-human dimensions of field systems. These concern, but are not limited to, the study of the organisation of labour and which forms of biosocial organisation, rules, and regulations of access produced, managed and manipulated the fieldscapes. New studies of manuring (e.g. Lauer et  al. 2014; Nielsen and Kristiansen 2014; Dev 2018; Nielsen et al. 2019) target the composition and intensity of manuring on the prehistoric field systems. Unfortunately, for proxies such as coprostenoles, stable bile acids and coprophilous spore bodies, it remains difficult to argue whether they entered the fields ‘on the hoof’ as droppings or as part of a manure-mixture prepared in settlements. This ambiguity feeds into wider questions pertaining to the relationship between pastoralism and agriculture (Chaps. 5, 6 and 8; cf. Holst and Rasmussen 2013), and may call for a reopening of the debate from the 1970s on the relationship between pastoralism and land enclosure (e.g. Fleming 1987, 2007;

7

Fowler 1983), as well as between pastoralist and agricultural landscapes. Many field systems are situated in landscapes typically regarded as ‘marginal’, such as the many Celtic fields situated in former heathlands, suggesting that grazing and herding could have formed a key part of their management and even initial allotment. In turn, the question of how particular landscapes and ways of organising landscapes shaped interaction and multispecies relationships should also be addressed. Particularly the ways and (social) scales on which usage and use-rights were coordinated to arrive at the uniform outlook that various field systems pose, hinges on the availability of studies that target the social rather than physical aspects of field systems. Third, issues related to the causality of fieldscape  change remain critically absent. Particularly the possible social, cultural, demographic,  climatic and spatiotemporal factors in the emerging and reconfiguring of field systems need to be addressed and disentangled. Whereas first-millennium BC climatic changes have been identified as important agents of change (Chaps. 4, 6, 7, 10 and 11; cf. Groenman-van Waateringe and Geel 2017), they are by no means to be seen as universal and unicausal drivers of change. If anything, the case-studies in this volume illustrate the extent that local factors all affect the resilience of local field systems, including factors such as water proximity and soil-nutrient content, the role and proximity of (older) monuments and influences of social organisation and forms of land governance. Fourth, field systems should, in future, be studied from more diachronic or biographical perspectives, in which ‘before’ and ‘afterlives’ of fields are addressed alongside their use histories (Johnston 2005). Earlier monuments and indications of vegetation change due to prior use or settlement can have influenced the situation and development of field systems, yet too few are excavated with such questions in mind. When not in tabula rasa landscapes, what types of prior uses were seen as unproblematic or even favoured? Several case-studies in this book have highlighted how use and reuse have created rich – but difficult to disentangle – palimpsests of landscape structures (Chaps. 4, 7, 8, 9 and 14). This speaks volumes on the long-term consequences of land allotment. Over what durations did field systems have structuring effects on people’s practices and ways of organising and negotiating  landscapes (Løvschal 2020)? A stronger landscape-biographical approach – sensitive to its palimpsest character – can help identify ways of reinterpretation that showcase the sustained influence of previous structures and features related to past fieldscapes.

8

1.5

M. Løvschal et al.

Chapter Summaries

We have grouped the chapters by four major approaches and themes: ‘Mapping fieldscapes’, ‘In-depth archaeological investigations of fieldsystems’; ‘New methods’ and ‘The economy of fieldscapes: pasture and agriculture’. We of course recommend a close reading of the entire volume, however, what follows are short introductions to each of the case-studies to guide readers towards their specific interests.

1.5.1 Mappings Fieldscapes Volker Arnold (Chap. 4) demonstrates the value of LiDAR for the identification and mapping of fieldscapes in landscapes shrouded in woodland. Arnold’s case study is the Sachsenwald Forest, east of Hamburg, northern Germany. Once the digital terrain models were generated, the unmistakable patterns of field banks appeared beneath 15 km2 of the woodland. The largest undisturbed and continuously enclosed area covers 5 km2. The field systems vary in their regularity and sizes. Their origins post-date the Late Bronze Age/Early Iron Age funerary barrows and predate the thirteenth century AD, when the woodland was established. Scientific dates from relict field systems elsewhere in the region mainly fall within the first century BC to the first century AD. These field systems are not as extensive, complete or morphological variable as the fields in Sachsenwald Forest. The Sachsenwald fields have considerable potential for extending the chronology and better understanding the evolution of early land apportionment in northern Germany. Along similar lines, Michael Vinter’s paper (Chap. 12) seeks to reconstruct three fieldscapes from the first millennium BC in Himmerland, northern Jutland, Denmark. Through in-depth analysis and a combination of aerial photographs, pollen data, archaeological data, topographical and historical maps, Vinter is able to reconstruct and critically evaluate the potential extent of the Celtic field systems. He estimates the original extent of these field systems in prehistory, pointing to the fact that existing distribution maps, such as those based on LiDAR (Arnold, this volume), are most often highly biased by historical land-use and modern infrastructural expansion. Moreover, he points to a potential, marked change in the use-life of these field systems in the middle of the Pre-Roman Iron Age, which corresponds with settlement nucleation, more sedentary lifeways, and a peak in heathland vegetation. Cyril Marcigny and Rebecca Peake’s paper (Chap. 10) adds a largely over-looked region in field system research within western Europe, by presenting an overview of the Neolithic to Late Bronze Age field systems and settlements

of northern France. The fields show remarkable similarities to the extensive allotted landscapes described by Arnoldussen (this volume), Vinter (this volume) and Arnold (this volume) in that there are no clear demarcations of their actual limits, and they are present as a mixture of both coaxial and more irregular fields. Marcigny and Peake see the emergence of these field systems as related to a marked population expansion and increased population density, corresponding with a demand for an increase in agricultural production. Corinne Roughley and colleagues spent two decades mapping and interrogating the multiperiod fieldscapes of Loughcrew, County Meath, eastern Ireland (Chap. 9). Their work began with one of the earliest LiDAR surveys undertaken for archaeological purposes, to which they added historic and recent aerial photography along with published and archival maps. These multiple sources are essential for unravelling the complexities amongst 150  km of linear earthworks in their study area. Following map regression, Roughley and colleagues used the morphology of the earthworks and their stratigraphic relationships to propose a sequence of transformations of Loughcrew’s fieldscapes, which may have begun in later prehistory. The Loughcrew area remains a changing landscape in this and the last centuries, as a dispersed pattern of new-build bungalows takes the place of earlier farms. The fragile remains of the early fieldscapes are especially vulnerable as they are extensive and difficult to protect. Loughcrew exemplifies the widespread challenge of conserving the character and archaeological potential of ancient fields for the future’s living landscapes.

1.5.2 New Methods Robert Johnston, Rowan May and David McOmish’s paper (Chap. 13) adds an important chronological dimension to the understanding of large-scale fieldscapes such as those just mentioned. Their results are based on a project that sought to collate and review the existing evidence of scientific dating of field systems as well as other types of boundaries such as larger landscape demarcations in England. With 393 scientific dates from 120 sites from the early Bronze Age to the post-­medieval period, they are able to point out the earliest dates for the emergence of large-scale landscape allotment to the early 2nd millennium BC as well as periods of interruption and reemergence. Stijn Arnoldussen’s paper (Chap. 3) focuses on the Dutch Celtic fields assumed to date from the Late Bronze Age-­ Early Iron Age. The fields belong to an extensive but highly heterogeneous group of enclosed field systems, concentrated on the sandy soils, spanning an area from Belgium in the southwest to Estonia and Poland in the east. The study uses

1  Europe’s Early Land Allotment: Questions of Time, Scale and Stewardship

state of the art excavation and sampling methods in both the investigation of dating of bank formation as well as use-­ practices pertaining to their growth. He shows that the banks consist of a mixture of soil, plants from wetlands, and household refuse, suggesting the use of manuring from byre-­ bedding in houses, and indicating a significant agricultural use for the fields.

9

1.5.4 The Economy of Fieldscapes

Two papers are particularly focused on questions of pastoralism associated with fieldscapes. Clare Randall’s paper (Chap. 5) considers the extensive, regular field systems emerging in southern Britain around the sixteenth century BC.  They are not unlike the Celtic fields described by Arnoldussen (Chap. 3), however, there appears to be a chronological discrepancy of several hundred years. She critically 1.5.3 In-Depth Archaeological Investigations engages with former notions linking enclosed field systems of Field Systems to arable agriculture and an ignorance of the importance of pastoral husbandry in fieldscapes and late prehistoric landClaire Christie’s paper (Chap. 2) focuses on the early farm- scapes in general. She uses morphological analyses as a way ing landscapes of Neolithic and Bronze Age Shetland. The of re-engaging with such notions. The study is focused on study deploys an extensive suite of mapping methods, Somerset, where the relationship between hillfort occupaincluding high-resolution aerial photographs combined tion, landscape and livestock holding can be studied, by intewith field surveys to detect the character and distribution of grating archaeological investigation with faunal and plant these very early fieldscapes surrounded by stone walls and macrofossil data from within that landscape. banks. She shows that the field systems were much more Stratford Halliday (Chap. 7) reviews the evidence for later extensive than hitherto expected as well as their connection prehistoric agricultural systems in Scotland. Roundhouses with a series of associated structures and features such as are widely recorded and excavated across many of Scotland’s stone-­ built houses, lynchets, clearance cairns and kerb lowland and upland landscapes. Despite their ubiquity, they cairns. were rarely accompanied by bounded fields. Instead, Michael F.  Lane and Vassilis L.  Aravantinos’s paper Halliday recognises evidence for what he terms ‘churning’: (Chap. 11) focuses on field systems ascribed to the Late relatively short-lived occupations of roundhouses with mulBronze Age on Crete and central Greece. They deploy a radi- tiple rebuilds, and scattered clearance cairns and fragmencally different approach to the other papers. They make use tary banks left by intermittent agricultural activity. The lack of the texts in the Linear B script together with ethno-­ of stockproof boundaries leads Halliday to suggest that cularchaeological data to set up a hypothetical model for the tivated plots and grazing areas were kept apart until hardimensions and organisation of the land plots. This model vested ground could be manured ‘on the hoof’. The system served as a basis for conducting subsequent fieldwork in an was sustained through relatively frequent movements of ancient polder landscape in central Greece, including mag- fields and houses, which periodically returned to abandoned netometry, dating (OSL and thermoluminescence) and soil plots and stances. Land and settlements were managed with profiling. The investigations resulted in the documentation of a dynamism that negated a need for or the means to create a complex network of drained and irrigated fields, demar- permanent field systems. Halliday concludes by asking if cated by low levees, which in scales corresponded to the this same dynamism may also have operated amongst the expected model. They see this type of landscape manage- more regimented and extensive field systems of southern ment as important for the population expansion and expand- Britain. ing land claims of the Late Bronze Age. Two other papers interpret the field systems in the light Fabio Saccoccio synthesises the evidence for agricul- of agrarian production and cultivation techniques. Romana tural landscapes dating to the fourth to second millennia Harfouche and Pierre Poupet’s paper (Chap. 6) adresses the BC in the Campanian Plain, southern Italy (Chap. 14). long-term dynamics of terraced fields in the Eastern Tephra layers, laid down by eruptions of Mount Vesuvius Pyrenees – with particular focus on the Late Neolithic and and the Campi Flegrei caldera, provide stratigraphic mark- the Bronze Age. They make use of a rich palette of multiers and a chronological framework for the archaeological disciplinary approaches, including archaeological methods deposits. Evidence for agricultural activity primarily con- and excavations and pedoarchaeological surveys and soil sists of plough-marks left by the tip of the ard cutting into science. They consider long-term dimensions in the develthe sediment. The paper adds an important case study and opment of fieldscapes, including the expansion and abanits socio-economic implications to an otherwise quite over- donment of cultivated lands, and their relationship with looked region, and sees it in the wider context of prehistoric human-induced environmental changes as well as biophysiagrarian regimes in northern European contexts that can be cal factors. Unlike the field systems investigated by Clare used to interpret underlying socio-economic implications Randall, the paper develops a strong focus on their link related to land tenure. with agricultural production.

10

M. Løvschal et al.

Lynne A.  Kvapil’s paper (Chap. 8) focuses on terraced fields and an associated settlement in south-eastern Corinthia in Late Bronze Age Greece. The study combines archaeological surveys and excavations with mapping, soil analyses and OSL-dating. Similar to Romana Harfouche and Pierre Poupet’s paper, Kvapil sees large-scale land allotment and terraced fields as closely related to an intensification in agrarian production and cultivation practices. However, her paper explores and underlines the role of fields in enabling diverse cultivation techniques as well as diverse growing environments. Unlike the remaining authors, she also stresses issues of identity, the organisation and rhythms of labour and the creation of gendered agricultural spaces.

1.6

Final Remarks

We are currently facing a cascading crisis that crosses way beyond Europe, where radical, large-scale environmental change is accelerating due to humanity’s devastating impacts on nature, including climate change, flooding, rural depopulation, deforestation, infrastructural expansion and moorland transformation. Ancient fieldscapes will be lost amongst many of these transformations, which could be measured in years, maybe decades. With their loss, we lose parts of the biodiversity, cultural diversity and history embedded in them. In this context, it  remains critical that we reflect on what can we learn from long-term processes of landscape and environmental manipulation and transformation. This book builds on decades of archaeological research on ancient field systems. It particularly draws upon the last 20 years, when an increase in development-led and research excavations and mapping projects have dramatically expanded our knowledge in many regions of Europe. The new data provides opportunities for understanding the legacy of large-scale landscape changes superseding individual sites and local areas. Sufficient data is now available for characterization on a continental scale, taking account of chronologies and paleoenvironments. Additionally, the increased focus on digitization and open access has enabled supra-regional archaeological comparison in radical new ways than previously possible. In turn, the amount of information poses the risk that we ‘drown in data’ and that consequently any deeper understanding remains absent. The individual case studies in this book each highlight a series of issues that we believe could be raised to a further level as common goals to help us focus the research and investigation of prehistoric field systems in Europe. It is our hope that this book will provide an important step in enhancing our knowledge within this research field by highlighting the need of understanding and comparing processes of land allotment and use across existing geographical

boundaries and chronological periods, and by providing a more informed basis for facing the future challenges of human-driven large-scale landscape change. Acknowledgements  The idea for this book took shape following a 2015 EAA session with the title: “Cropland shaping: the spread, function and social implications of field systems across Europe during pre- and protohistory”. That session was initiated by Alessandro Vanzetti (Department of Ancient World Studies, University of Rome La Sapienza) who organised the session together with Stijn Arnoldussen (Groningen Institute for Archaeology, Groningen University), Mette Løvschal (Moesgaard Museum & Department of Archaeology and Heritage Studies, Aarhus University) and David McOmish (English Heritage). We would like to thank all organisers, participants and audience members of that lively session, where the seed for this book was sown, in particular Alessandro Vanzetti for his effort and enthusiasm. Since then, a broad collective of authors is to be thanked for their patience and understanding in the long process of peer-review and pre-press revisions. Finally, the EAA Themes editorial board (Kristian Kristiansen, Eszter Bánffy, Peter Attema  and Claes Uhnér) and Springer (Neelofar Yasmeen and Christi Lue) deserve our severe gratitude for coaching us towards the publication. Løvschal has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 853356) and Independent Research Fund Denmark (DFF–4180-00245).

References Allen, T. (2014). Managing the historic rural landscape: A sense of place. In J. Grenville (Ed.), Managing the historic rural landscape (pp. 163–172). Oxon/New York: Routledge. Antrop, M. (2006). Sustainable landscapes: Contradiction, fiction or utopia? Landscape and Urban Planning, 75, 187–197. Arnoldussen, S. (2018). The fields that outlived the Celts. The use-­ histories of Dutch later prehistoric field systems (Celtic fields or raakakkers). Proceedings of the Prehistoric Society, 84, 303–327. Arnoldussen, S., & van der Linden, M. (2017). Palaeo-ecological and archaeological analysis of two Dutch Celtic fields (Zeijen-Noordse veld and Wekerom-Lunteren): Solving the puzzle of local Celtic field bank formation. Vegetation History and Archaeobotany, 26(6), 551–570. Bech, J.  H., & Mikkelsen, M.  E. (2018). The Bronze Age settlement at Aas, eastern thy, Denmark. In J.-H.  Bech, B.  V. Eriksen, & K. Kristiansen (Eds.), Bronze Age settlement and land-use in Thy, Northwest Denmark, Vol. 1 (Jysk Arkæologisk Selskabs Skrifter 102) (pp. 477–503). Aarhus: Jysk Arkæologisk Selskab. Behre, K.-E. (2008). Landschaftsgeschichte Norddeutschlands: Umwelt und Siedlung von der Steinzeit bis zur Gegenwart. Neumünster: Wachholtz Verlag. Bonnie, R. (2010). Did Roman Cadastres exist in the Roman northwest? In A. Moore, G. Taylor, E. Harris, P. Girdwood, & L. Shipley (Eds.), TRAC 2009. Proceedings of the 19th theoretical Roman archaeology conference, Ann Arbor (USA) and Southampton (UK) (pp. 91–104). Oxford: Oxbow Books. Bradford, J. (1957). Ancient landscapes, studies in field archaeology. London: G. Bell & Sons. Brongers, J. A. (1973). 1833, Reuvens in Drenthe. Een bijdrage tot de geschiedenis van de archeologie/A contribution to the history of archeology. Bussum: Fibula-Van Dishoeck. Caravello, G. U., & Michieletto, P. (1999). Cultural landscape: Trace yesterday, presence today, perspective tomorrow for “Roman cen-

1  Europe’s Early Land Allotment: Questions of Time, Scale and Stewardship turiation” in rural Venetian territory. Human Ecology Review, 6(2), 45–50. Clavel-Lévêque, M., & Orejas, A. (Eds.). (2002). Atlas historique des cadastres d’Europe (Vol. 2). Luxembourg: Office des Publications Officielles des Communautés Européennes. Clemmensen, N. C. (2010). DHM - den digitale højdemodel - reliefkortet. In V.  Nielsen (Ed.), Oldtidsagre i Danmark. Sjælland, Møn og Lolland-Falster (Jysk Arkæologisk Selskabs skrifter 71) (pp. 25–28). Aarhus: Aarhus Universitetsforlag. Curwen, E., & Curwen, E. C. (1923). Sussex Lynchets and their associated field-ways. Sussex Archaeological Collections, 64, 1–65. Deleuze, G., & Guattari, F. (1980). A thousand plateaus: Capitalism and schizophrenia. London: Continuum. Dev, S. (2018). Application of micromorphology to study manuring practices: A case study from Bronze Age in Cornwall, UK. Global Journal of Archaeology & Anthropology, 6(3), 54–69. Enevold, R., Rasmussen, P., Løvschal, M., Olsen, J., & Odgaard, B. (2019). Circumstantial evidence of non-pollen palynomorph palaeoecology: A 5,500 year NPP record from forest hollow sediments compared to pollen and macrofossil inferred palaeoenvironments. Vegetation History and Archaeobotany, 28, 105–121. Fleming, A. (1987). Coaxial field systems: Some questions of time and space. Antiquity, 61, 188–202. Fleming, A. (2007). The Dartmoor reaves: Investigating prehistoric land divisions. Oxford: Windgather Press. Fowler, P.  J. (1983). The farming of prehistoric Britain. Cambridge: Cambridge University Press. Fowler, P. J. (2000). Landscape, plotted and pieced. Landscape history and local archaeology in Fyfield and Overton, Wiltshire. London: The Society of Antiquaries of London. Franceschelli, C. (2015). Riflessioni sulla centuriazione romana: paradigmi interpretativi, valenza paesaggistica, significato storico, Agri Centuriati. An International Journal of Landscape Archaeology, 12, 175–211. Galvin, K.  A., Reid, R.  S., Behnke, J., & Hobbs, R.  H. (2008). Fragmentation in semi-arid and arid landscapes. Berlin: Springer. Greer, A. (2012). Commons and enclosure in the colonization of North America. American Historical Review, 117, 365–386. Groenman-van Waateringe, W., & van Geel, B. (2017). Raised bed agriculture in Northwest Europe triggered by climatic change around 850 BC: A hypothesis. Environmental Archaeology, 22(2), 166–170. Gruffudd, P. (1995). Heritage as national identity: Histories and prospects of the national pasts. In D. T. Herbert (Ed.), Heritage, tourism and society (pp. 49–67). London: Mansell. Håkansson, N. T., & Widgren, M. (2014). Landesque capital: The historical ecology of enduring landscape modifications. Walnut Creek: Left Coast Press. Haraway, D. J. (2008). When species meet. Minneapolis: University of Minnesota Press. Hatt, G. (1949). Oldtidsagre (Arkæologisk-Kunsthistoriske Skrifter 2.1). København: Det Kongelige Danske Videnskabernes Selskab. Hesse, R. (2010). LiDAR-derived local relief models (LRM) – A new tool for archaeological prospection. Archaeological Prospection, 17, 67–72. Holst, M.  K., & Rasmussen, M. (2013). Herder communities: Longhouses, cattle and landscape organisation in the Nordic Early and Middle Bronze Age. In S.  Bergerbrant & S.  Sabatini (Eds.), Counterpoint: Essays in archaeology and heritage studies in honour of professor Kristian Kristiansen (pp.  99–110). Oxford: Archaeopress. Huisman, D.  J., Jongmans, A.  G., & Raemaekers, D.  C. M. (2009). Investigating Neolithic land use in Swifterbant (NL) using micromorphological techniques. Catena, 78, 185–197. Humme, A., Lindenbergh, R., & Sueur, C. (2006). Revealing Celtic fields from LiDAR data using kriging based filtering. In H. G. Maas & D.  Schneider (Eds.), Proceedings of the ISPRS Commission V

11

Symposium, Dresden, 25–27 September 2006 (pp.  1–5). Dresden. Conference paper: http://www.isprs.org/proceedings/XXXVI/ part5/. Johnston, R. (2005). Pattern without a plan: Rethinking the Bronze Age coaxial field systems on Dartmoor, South-West England. Oxford Journal of Archaeology, 24(1), 1–21. Klamm, M. (1993). Aufbau und Entstehung eisenzeitlicher Ackerfluren (“Celtic Fields”). I Stand der Forschung (Göttinger Bodenkundliche Berichte 102). Göttingen: Brunk Meyer. Kooistra, M. J., & Maas, G. J. (2008). The widespread occurrence of Celtic field systems in the central part of the Netherlands. Journal of Archaeological Science, 35, 2318–2328. Kostyrko, M., Kobiałka, D., Kajda, K., & Mlekuz, D. (2016). An archaeological flight further than post-processualism  – Seeking a non-anthropocentric perspective. AARGnews, 52, 71–79. Latour, B. (2004). Politics of nature: How to bring the sciences into democracy (trans: Porter, C.). Cambridge: MA: Harvard University Press. Lauer, F., Prost, K., Gerlach, R., Pätzold, S., Wolf, M., Urmersbach, S., Lehndorff, E., Eckmeier, E., & Amelung, W. (2014). Organic fertilization and sufficient nutrient status in prehistoric agriculture? Indications from multi-proxy analyses of archaeological topsoil relicts. PLoS One, 9(9), e106244. Løvschal, M. (2014). Emerging boundaries: Social embedment of landscape and settlement divisions in northwestern Europe during the first millennium BC. Current Anthropology, 55(6), 725–750. Løvschal, M. (2020). The logics of enclosure: Deep-time trajectories in the spread of land tenure boundaries in late prehistoric northern Europe. Journal of the Royal Anthropological Institute, 26(2), 365–388. Løvschal, M., Bøcher, P. K., Pilgaard, J., Amoke, I., Odingo, A., Thuo, A., & Svenning, J.-C. (2017). Fencing bodes a rapid collapse of the unique greater Mara ecosystem. Scientific Reports, 7, 41450. Available online: https://www.nature.com/articles/srep41450.pdf. Accessed 20 Feb 2020. Løvschal, M. & Gravesen, M.L. 2021 (in print). De-/fencing Grasslands: Ongoing Boundary and Unmaking in postcolonial kenya. LAND 10. Müller, G. (2013). Europe’s field boundaries. Hedged banks, hedgerows, field walls (stone walls, dry stone walls), dead brushwood hedges, bent hedges, woven hedges, wattle fences and traditional wooden fences. Stuttgart: Neuer Kunstverlag. Netz, R. (2004). Barbed wire: An ecology of modernity. Middletown: Wesleyan University Press. Nielsen, P. O. (1993). Settlement. In S. Hvass & B. Storgaard (Eds.), Digging into the past. 25 Years of archaeology in Denmark (pp. 92–95). Aarhus: Aarhus Universitetsforlag. Nielsen, N. H., & Dalsgaard, K. (2017). Dynamics of Celtic fields — A geoarchaeological investigation of Øster Lem Hede, Western Jutland, Denmark. Geoarchaeology, 32, 414–434. Nielsen, N. H., & Kristiansen, S. M. (2014). Identifying ancient manuring: Traditional phosphate vs. multi-element analysis of archaeological soil. Journal of Archaeological Science, 42, 390–398. Nielsen, N.  H., Kristiansen, S.  M., Ljungberg, T., Enevold, R., & Løvschal, M. (2019). Low and variable: Manuring intensity in Danish Celtic fields. Journal of Archaeological Science: Reports, 27, 101955. Ogden, L., Heynen, N., Oslender, U., West, P., Kassam, K.-A., & Robbins, P. (2013). Global assemblages, resilience, and Earth Stewardship in the Anthropocene. Frontiers in Ecology and the Environment, 11(7), 341–347. Oosthuizen, S. (2013). Beyond hierarchy: The archaeology of collective governance. World Archaeology, 45(5), 714–729. Pinto-Correia, T. (2000). Landscape identity, a key for integration. In B. Pedroli (Ed.), Landscape – Our home. Lebensraum Landschaft. Essays on the culture of the European landscape as a task (pp. 145– 150). Zeist/Stuttgart: Indigo/Freies Geistesleben.

12 Povinelli, E.  A. (2016). Geontologies: a requiem to late liberalism. Durham: Duke University Press. Rackham, O., & Moody, J.  A. (1992). Terraces. In B.  Wells (Ed.), Agriculture in ancient Greece (pp. 123–130). Stockholm: Astroms Forlag. Said, M.  Y., Ogutu, J.  O., Kifugo, S.  C., Makui, O., Reid, R.  S., & Leeuwa, J.  D. (2016). Effects of extreme land fragmentation on wildlife and livestock population abundance and distribution. Journal for Nature Conservation, 34, 151–164. Spek, T., Groenman-van Waateringe, W., Kooistra, M., & Bakker,  L. (2003). Formation and land-use history of Celtic fields in North-­ West Europe  – An interdisciplinary case study at Zeijen, The Netherlands. European Journal of Archaeology, 6(2), 141–173. Stabile, M. C. C., Guimarães, A. L., Silva, D. S., Ribero, V., Macedo, M. N., Coe, M. T., Pinto, E., Moutinho, P., & Alencar, A. (2020). Solving Brazil’s land use puzzle: Increasing production and slowing Amazon deforestation. Land Use Policy, 91, 104362. Steffen, W., Broadgate, W., & Deutsch, L. (2015). The trajectory of the Anthropocene: The great acceleration. The Anthropocene Review, 2(1), 81–98. Taylor, J. L. (2006). Negotiating the grassland: The policy of pasture enclosures and contested resource use in Inner Mongolia. Human Organization, 65, 374–386.

M. Løvschal et al. Teheripour, F., Hertel, T.  W., & Ramankutty, F. (2019). Market-­ mediated responses confound policies to limit deforestation from oil palm expansion in Malaysia and Indonesia. PNAS, 116(38), 19193–19199. Tsing, A.  L. (2015). The mushroom at the end of the world: On the possibility of life in capitalist ruins. Princeton: Princeton University Press. Upex, S.  G. (2002). Landscape continuity and the fossilization of Roman fields. Archaeological Journal, 159(1), 77–108. van Giffen, A. E. (1918). Begin van een onderzoek van’ de zogenaamde voormalige Romeinsche legerplaats en aangelegen grafheuvelveld te Zeijen. Nieuwe Drentse Volksalmanak, 36, 135–175. Vijay, V., Pimm, S.  L., Jenkins, C.  N., & Smith, S.  J. (2016). The impacts of oil palm on recent deforestation and biodiversity loss. PLoS One, 11(7), e0159668. Vinter, M. (2011). Rekonstruktion af det parcellerede dyrkningslandskab i Himmerland. Marker, græsning og bebyggelse belyst i tre mikroregioner. In N. A. Møller, S. S. Qvistgaard, & S. F. Jensen (Eds.), Nyt fra Vestfronten. Nord og vestjyske bebyggelser fra ældre jernalder (pp. 129–154). Copenhagen: The Saxo-Institute. Wickstead, H. (2007). Land division and identity in later prehistoric Dartmoor, South-West Britain: Translocating Tenure. PhD thesis. London: University College London.

2

Enclosing and Dividing Land: The Neolithic and Bronze Age Field Systems of Shetland Claire Christie

Abstract

The transition to agriculture is one of the fundamental transformations in human society, but the evolution, organization, and character of early farming landscapes of the Neolithic and Bronze Age in many regions is poorly understood. The rarely paralleled preservation of extensive prehistoric houses, field systems, and burial monuments in the West Mainland of Shetland affords unique opportunities for understanding prehistoric societies. Despite their impressive preservation, the remains on Shetland have received comparatively little archaeological attention. This paper presents the results of a programme of mapping, using high-resolution aerial photographs, that explores the extent and spatial distribution of upstanding prehistoric remains within the landscape of the West Mainland. The mapping has revealed extensive field systems and foci of activity providing insights into the organisation of prehistoric settlement and land-use. The assimilation of these results with existing data from recent excavations has allowed for a more detailed understanding of the development of the early farming landscapes of Shetland. Keywords

Shetland · West Mainland · Settlement · Land-use · Neolithic · Bronze Age · Aerial photography

Shetland is the northernmost archipelago in the British Isles, located 80  km north-east of Orkney and 320  km west of Norway (Fig.  2.1). It has a rich prehistoric archaeology in terms of extent and preservation and which can be character-

ised not by individual sites, but by preserved prehistoric landscapes. The finest examples are located on the West Mainland of Shetland, and comprise extensive field and settlement systems with field walls, lynchets, houses, and clearance and burial cairns littering the hillsides. The remains, preserved by the spread of peat and the lack of subsequent intensive plough based agriculture, are widely considered to date from the Neolithic – Bronze Age. Local antiquarian interest in the prehistoric remains of Shetland flourished from the mid-nineteenth century but outside of Shetland they remained largely unknown. The publication of the Royal Commission on the Ancient and Historical Monuments of Scotland Inventory of Sites in the Northern Isles (RCAHMS 1946) and the subsequent excavations by Charles Calder, in the 1950s and 60s, significantly enhanced the wider appreciation of the archaeology of Shetland (Murray and Christiansen 2013: 92). Calder’s excavations, along with the research conducted by Alasdair Whittle at Scord of Brouster in the 1980s, provide the foundations for our understanding of prehistoric Shetland (Whittle et  al. 1986). Turner (2012) has identified two distinct types of settlement enclosure: Homestead Enclosures and Multiple Field Systems. Turner’s work forms part of a wider corpus of research, both archaeological and palaeoenvironmental, that has sought to explore the nature of early farming in Shetland. This chapter presents both an overview of previous work and a preview of current mapping work, conducted from 2015 to 2017, which uses high resolution aerial photographs (orthophotographs) and targeted field survey to map the prehistoric settlements in an area of the West Mainland covering 160  km2. The mapping has revealed extensive remains of prehistoric sites, their relationships, connections to place and interactions with the wider landscape, and highlighted some of the potential problems of existing classifications and assumptions.

C. Christie (*) Headland Archaeology, Edinburgh, Scotland e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 S. Arnoldussen et al. (eds.), Europe’s Early Fieldscapes, Themes in Contemporary Archaeology, https://doi.org/10.1007/978-3-030-71652-3_2

13

14

C. Christie

Fig. 2.1  Map showing the location of The Shetland Islands 80km north-east of Orkney. This paper discusses the archaeological remains located on the West Mainland of Shetland with the survey area depicted. (© Crown copyright and database rights 2020 Ordnance Survey (EDINA, 100025252))

2.1

 he Archaeology of Prehistoric T Shetland: Houses, Fields and Function

In 1928, the Royal Commission on the Ancient and Historical Monuments of Scotland (RCAHMS) began their ‘Inventory of Sites in the Northern Isles’ (RCAHMS 1946). The work in Shetland was conducted by Charles Calder who returned to excavate a number of settlement sites on the West Mainland (Calder 1950, 1955–6, 1962). Few settlement and house sites had been identified prior to the work of Calder, with Jarlshof and Wiltrow (Curle 1934, 1936) being notable exceptions, and his excavations provided the first comprehensive assessment of the key characteristics and date of the houses. Calder’s excavations at Staneydale, Ness of Gruting and Gruting School allowed him to characterize the houses as being broadly sub-oval in

Fig. 2.2  The excavated houses at Gruting School and Stanydale displaying the oval form characteristic of many prehistoric houses in Shetland. The simplified plan of Ness of Gruting highlights the complexities of construction and internal layout. (Calder, 1955–6: 348 & 365, reproduced with kind permission of the Society of Antiquaries of Scotland)

shape with a variety of internal features and artefact assemblages suggestive of a Neolithic date (Fig.  2.2) (Calder 1955–6; Henshall 1955–6: 383). Calder recognised that the houses formed part of broader enclosure systems with associated irregular field systems ranging from single enclosures to conglomerations of multiple small fields (Calder 1955–6: 358). The result of his survey and excavation allowed Calder to present a lasting image of Neolithic life in Shetland with farmers living in individual crofts or townships tending livestock, cultivating crops and harvesting grain (Calder 1955–6: 358). The archaeological investigation of prehistoric Shetland advanced considerably in the 1970s with Alasdair Whittle’s field survey and excavations at Scord of Brouster, West Mainland (Whittle et al. 1986). The site comprises of three stone-built houses surrounded by eight irregular fields defined by stone walls and banks with associated features including lynchets, clearance cairns and a kerb cairn (Fig. 2.3) (Whittle et al. 1986: 4; Turner 2012: 84). The scale

2  Enclosing and Dividing Land: The Neolithic and Bronze Age Field Systems of Shetland

15

Fig. 2.3  Plan of the site at Scord of Brouster displaying the three excavated houses, kerb-cairn, clearance cairns, and the pattern of the central field. (Whittle et al. 1986: 3, reproduced with kind permission of Alasdair Whittle)

of the work allowed for the range of features on the site and the wider landscape to be analysed and explored. The application of radiocarbon dating at Scord of Brouster resulted in 14 dated samples being obtained from archaeological contexts (Fig. 2.4) (Whittle et al. 1986: 37). Although these dates have received criticism (Sheridan 2012), they indicate a Neolithic date for the majority of the features at Scord of Brouster with evidence for occupation into the Bronze Age in at least one of the houses. The excavation results, radiocarbon dates and pollen analysis were used to argue for a model of continuous occupation highlighting the development of the site over time with the field system being outlined from an early stage (Whittle et  al. 1986: 133–140). The scale of Whittle’s research and the wider landscape explorations allowed for the identification of further walls which spread out from the central system indeterminately into the landscape (Whittle et  al. 1986: 142). The analysis of pollen samples taken from under these walls indicates they may also date from the mid-  to  later third millennium BC and were constructed during early phases of clearance in the Scord of Brouster landscape (Whittle et al. 1986: 142). The results of the extensive programme of research conducted at Scord of Brouster led Whittle to conclude that the fields formed an infield and outfield system with an “arable nucleus, extensive grazing and dispersed settlement” (Whittle et al. 1986: 144). Whittle emphasised the

limited evidence for animal husbandry, arguing for an arable focus based upon the stone tool assemblage, soil analysis, and the cultivation of barley throughout the life of the settlement (ibid: 144; 146, Milles 1986: 120). The palynological assessments of cores taken from a nearby loch basin provide further evidence indicating woodland clearance beginning around 2700 BC and pasture becoming quickly established with early indications of cultivation (Keith-Lucas 1986: 116). The deteriorating environmental conditions towards the end of the occupation of the settlement are indicated by a notable reduction in cultivation, an expansion of pasture associated with deteriorating soil quality, and the growth of peat around the loch basin c.1500 BC (Whittle et al. 1986: 146). The model proposed by Whittle of mixed farming has been developed with subsequent work focusing on the varying degrees to which cereal cultivation and grazing contributed to farming strategies. In addition to the work at Scord of Brouster, Shetland has received considerable palaeoenvironmental attention (see Johansen 1975; Bennett et  al. 1992; Hulme and Shirriffs 1994; Edwards et  al. 2005) with the work of Edwards and Whittington (1998) on the West Mainland further contextualising several prehistoric landscapes. Sub-peat and sub-wall soil monoliths taken from Troni Shun, Pinhoulland and Ness of Gruting display varying evidence for grazing and trampling with the strong representation of heather, grasses, ribwort plantain, greater

16

C. Christie

Fig. 2.4  Radiocarbon dates from archaeological contexts at Scord of Brouster calibrated in OxCal v4.4. (Whittle et al. 1986: 37 recalibrated by the author)

plantain, daisy/thistle group (Asteroideae/Cardueae) and dandelion (Edwards and Whittington 1998: 12). There is some evidence for cereal cultivation in a limited number of samples (ibid: 14). The palynological analysis of cores taken from the Loch of Brunatwatt and the Mire of Troni Shun show a landscape with substantial woodland cover prior to a marked reduction in woodland after 3870 cal BC and indications for the intensification of arable activity with a mixed farming regime between 2460 and 1810  cal BC (ibid: 14). The results of the palaeoenvironment analysis led Edwards and Whittington to conclude that the “prehistoric landscape

of the West Mainland was devoted to pastoralism and that the area devoted to arable cropping was minor” (ibid: 16). This builds on the idea that the prehistoric houses and field systems were focused on arable agriculture as proposed by Whittle (1986), continuing the understanding of the field systems as functioning within a model of a mixed farming regime. A recognition and exploration of how sites ­developed and site based environmental work has significantly advanced our understanding of the function of the fields. In terms of field-system morphology, Charles Calder first recognised that the field systems associated with prehistoric

2  Enclosing and Dividing Land: The Neolithic and Bronze Age Field Systems of Shetland

houses varied from a single field to conglomerations of multiple fields (Calder 1955–6: 358). A comparison of the shape of the field systems allowed sites to be firmly classified into two distinct types: Homestead Enclosures and Multiple Field Systems (Turner 2012). Homestead Enclosures were defined as sites which present a consistent form with a single house contained within a single enclosure, whereas Multiple Field Systems represent a more variable form, but broadly consist of multiple houses surrounded by numerous conjoined irregular fields (Turner 2011, 2012).1 The simplicity of the Homestead Enclosures has led to suggestions that these are earlier in date, probably Neolithic, with Turner taking this further to propose that in some instances on the West Mainland Homestead Enclosures directly developed into Multiple Field Systems (Turner 1998, 2011, 2012 Turner et al. 2004). The evidence from the only excavated example of a Homestead Enclosure, Catpund, South Mainland, suggested that this site was Bronze Age in date (Ballin Smith 2005). The site consists of a single house located at the southern end of an irregular enclosure defined by a drystone dyke (ibid: 3). The site could not be radiocarbon dated but based upon comparative analysis it is believed to be of a slightly later date than House 1 Scord of Brouster, (2600– 1800 BC) (Whittle et al. 1986: 75; Ballin Smith 2005: 43). The later date combined with the pattern of occupation, reuse, and abandonment witnessed at Catpund suggests a variety of developmental models, including the coexistence of multiple types of field system, and insights into the longevity of Homestead Enclosures. Turner expanded upon her classification of the site types to explore the function of the fields for both Homestead Enclosures and Multiple Field Systems. On the basis of calculations of yield supplemented by the evidence from soil micromorphology, she concluded that Homestead Enclosures did not enclose an area sufficient to produce all the food required for the occupants, and as a corollary, people either grew crops and kept livestock outwith the enclosures or supplemented their diet with wild resources (Turner 2011, 2013: 30). Multiple Field Systems, using the same methods, were concluded to have the potential to support a small group of adults eating a cereal-based diet (Turner 2012, 2013: 31). Overall, Turner supports the assertions made in previous work that prehistoric farmers were practicing mixed agriculture but crucially places a greater importance on the use of wild resources, particularly marine resources, to supplement diet (Turner 2013: 33). This brief outline of previous work in Shetland has shown how we might begin to classify the prehistoric settlements of Shetland, explore their function, and propose models for their development. Nonetheless, discussions are These classifications will also be adopted within this paper as a means of differentiating between forms.

1 

17

constrained by the limited ability to date sites, resulting in the majority of such sites being described as broadly Neolithic/Bronze Age in date. The difficulties in dating the prehistoric remains in Shetland are well established due to the use of peat as fuel, the general dearth of material suitable for radiocarbon dating, and the lack of clear artefact typologies representing major constraining factors (Sheridan 2013). The problem of dating is intensified by longevity of construction methods, similarities in overall exterior form, and the wide, but non-­diagnostic, variety of internal layouts seen in prehistoric house forms (see Cracknell and Smith 1983; Downes and Lamb 2000; Ballin Smith 2005). Moreover, the difficulties in dating sites has been set against the relatively abundant (though not unproblematic) radiocarbon dates recovered from Scord of Brouster, resulting in this single site becoming the baseline to which both excavated and unexcavated sites are compared. Recent radiocarbon dating of barley grains recovered during Calder’s excavations at Ness of Gruting, a site Calder interpreted to be Neolithic, adds further complexity to the dating of sites. Four radiocarbon dates indicated that the structure was constructed c. 2200–2000 BC placing it in the Bronze Age2 (Sheridan 2012: 18; Barker et  al. 1971; Barcham 1978). This raises the possibility that other settlement sites across Shetland interpreted as Neolithic may be Bronze Age or even later in date, and raises further questions about how individual sites developed, the longevity of their use, the chronological relationship between features (primarily the houses and field systems), and the chronological relationship between sites which on the surface appear similar. Broader survey work has aided in exploring the complexities of the landscapes. In 2010, a programme of aerial mapping and field survey was conducted across a c. 21 km2 area to the north of the village of Wall, West Mainland by the RCAHMS. The landscape was mapped using high resolution orthophotographs complemented by field survey as part of the 2010 Shetland Survey. The features indicate a complex multi-period landscape with concentrations of prehistoric activity. The continued mapping of the West Mainland outlined here, building upon the work of the 2010 Shetland Survey, further highlights the complexities of Shetland’s prehistoric field systems.

The recent radiocarbon dates from calcined animal bone fragments, recovered from the main peat layer in the interior, have been interpreted as being affected by the uptake of old carbon due to being burnt in peat with the date from the barley grains favored for the construction of the building (Sheridan et  al. 2014: 215) (Radiocarbon Dates: 3857±33, OxA-X-2575-37, 4043±28, OxA-X-2579-42, 3964±3, OxA-X-2579-41).

2 

18

2.2

C. Christie

 apping the Neolithic and Bronze M Age Field Systems of Shetland

From 2015 to 2017, an area of c. 160 km2 has been mapped by the author using 0.25 m orthorectified vertical aerial photographs (supplied by Historic Environment Scotland). The strategy of the detailed mapping programme was to record all visible features regardless of apparent periods. The landscapes of Shetland are palimpsests of activity, with upstanding remains from prehistory to the modern period making it often difficult to untangle the complex remains. The initial results presented here focus on sites which can be identified as Homestead Enclosures, Multiple Field Systems, and a newly identified form of field system. The Homestead Enclosures and Multiple Field Systems mapped on the West Mainland display many of the traits outlined by Turner (2012). Homestead Enclosures represent the most consistent form composed of a sub-circular drystone enclosure encircling a single oval house. The simplicity and consistency observed in the Homestead Enclosures contrasts with the complexity and diversity of the Multiple Field Systems. Multiple Field Systems comprise conglomerations of multiple, conjoined, irregularly-shaped fields, containing one or more prehistoric houses. There is significant variability between sites in the shape, size and number of fields, and the number of houses. A number of Multiple Field Systems on the West Mainland including Scord of Brouster, Gallow Hill (HU 25682 51214), Pinhoulland (HU 25965 49909), Ness of Gruting (HU 28137 48332) and Stanydale (HU 28535 50240) have been previously mapped in detail (Mahler 2011; Turner 2012). This study, through mapping large areas and exploring the totality of the landscapes, has highlighted the presence of distinctive scales of field system.

2.3

Homestead Enclosures

The Homestead Enclosures mapped within the area of the West Mainland are consistent in form, with the sites at Bridge of Walls (HU 27070 51190) and Whirlie Burn (HU 22710 49940) proving typical examples. Bridge of Walls consists of a circular enclosure with a maximum extent of 39.3 m and the remains of an oval house situated within the enclosure to the northwest. The site at Whirlie Burn is less well-preserved, visible as a low bank with occasional protruding stones with a maximum extent of 49.2 m, the truncated remains of a possible house located to the western side (Fig. 2.5). The house has been truncated by the construction of a later planticrub (a small enclosure used for growing kale) directly on top of the remains: a common occurrence as they provide accessible building material and soil accumulation (Tait 2011).

The Homestead Enclosures mapped on the West Mainland show a preference for sloping ground towards the base of the hillside. For example, the site at Bridge of Walls is located on the south-east facing lower slopes of the Ward of Browland at 27 m aOD. This can also be witnessed at other sites with all being located on the lower slopes at elevations between 10 and 33 m aOD. The importance of elevation is difficult to ascertain given the uncertainties surrounding the details of Holocene sea level change in Shetland. There is evidence that sea level was significantly lower in the past with sea level curve models supplying a general overview (see Hoppe 1965; Lambeck 1993; Peltier et  al. 2002; Bondevik et  al. 2005). The location of many of the Homestead Enclosures places them in close proximity to areas of historic activity with sites being overlain or surrounded by later structures. The truncation of the prehistoric houses by later features has implications for our identification of Homestead Enclosures as the presence of a house within the enclosure is a defining characteristic. The truncated remains of possible houses can be identified at Bardister (HU 23145 50116), Stapness (HU 23141 48725), Burn of Forgrigarth (HU 25833 56365). Stapness has been severally impacted by later historic land-use with the construction of a planticrub and a spread of rubble in the interior obscuring possible prehistoric features. At Burn of Forigarth there is a large enclosure, with a maximum extent of 67.7 m, that has several internal features including clearance cairns and a possible house. The location of the enclosure at the base of a heather covered rocky hillside partly obscures these features, as well as further possible houses and field walls nearby. There are also numerous enclosures within the study area of similar size, shape and construction to the Homestead Enclosures without visible associated prehistoric buildings. A small number of these appear on the historic maps and tend to be visible as more substantial remains on the aerial photographs, perhaps indicating a more recent date for construction or reuse. There are numerous more ephemeral examples without any distinctive features but similar in form to both the historic and prehistoric examples.

2.4

Field Systems

Multiple Field Systems show a much greater variety of form. The mapping reveals two different scales with both extensive and more discrete examples apparent. Scord of Brouster provides the archetypal example with its distinctive field system composed of multiple conjoined fields. The walls which extend from the central system, identified during the 1970s (Whittle et al. 1986) and further explored by later surveys, are a distinctive feature of its overall form. The nearby site at Gallow Hill provides the most comparable example while those at Pinhoulland, Ness of Gruting, and Stanydale display

2  Enclosing and Dividing Land: The Neolithic and Bronze Age Field Systems of Shetland

19

Fig. 2.5  The location of the Homestead Enclosures of Whirlie Burn and Bardister on lower ground towards the base of the slope. The Multiple Field Systems of Clodister and Sclater’s Cuml are also depicted occupying the nearby hillsides (Left). A simplified plan of Whirlie Burn depicting the outline of the prehistoric house truncated by

a later planticrub surrounded by a sub-oval enclosure (Upper Right). Location of Homestead Enclosures discussed in the text (Bottom Right). (© Crown copyright and database rights 2020 Ordnance Survey (EDINA, 100025252))

some similar characteristics but a greater diversity in the morphology of the fields. These sites display significant variability in the shape, size and layout of the field but some common traits in overall composition and location are observed. Their extensive nature predisposes them to encompassing a greater range of gradients and aspects. The site at Kellister (HU 24425 56219) provides a particularly extreme example located on a hillside with a steep gradient of c. 3/20, from 5 to 37 m aOD with a less dramatic example provided by Leeans (HU 22203 49026) with a gradient of c.1/10. Kellister is composed of five, irregular, conjoined central fields with a house located towards the centre (Fig.  2.6). The site is located on a predominantly south-west-facing hillside that slopes down towards the Loch of Kellister. The fields spread along the hillside with further walls extending from the central systems to the north and east. Walls which appear to extend away from the core of irregular fields for many hundreds of metres are a key

feature of all the extensive field systems. In areas with a higher concentration of sites, such as between Scord of Brouster and Gallow Hill, the walls appear to form a linear pattern across the landscape. The walls follow the contours of the hillside spreading along and over the hillside becoming fragmentary and overlain by later features making the associations unclear. Those which extend from the field systems have been argued to directly relate to the primary field systems (Whittle et al. 1986; Turner 2012: 34). If this is the case, they expand the scale of the systems and bring them into direct contact with surrounding archaeological features. At Kellister, there are three burial cairns near the field system with one cairn lying within a field. The close proximity between cairns and fields is also a key feature of other Multiple Field Systems with burial cairns in close proximity to Scord of Brouster, Trolligarts, to the south of Gallow Hill and overlooking Staneydale temple site. Since the chronological relationship between these features cannot be ascertained through

20

Fig. 2.6  Plan of the Multiple Field System at Kellister, overlain on the 5m DTM, depicting the prehistoric house surrounded by irregular fields with further walls extending across the landscape. The field system is located on sloping ground with burial cairns in close proximity. (© Crown copyright and database rights 2020 Ordnance Survey (EDINA, 100025252))

Fig. 2.7  Oblique photograph looking west over the site of Troni Shun with the prehistoric features outlined. (Image by C. Christie)

C. Christie

the mapping, this must remain a key question for future work. The Multiple Field Systems have been a focus of discussions on Neolithic and Bronze Age settlement in Shetland. The broad overall definition, and an acceptance of a high level of variation as a characteristic, allows for a multitude of sites to be defined as Multiple Field Systems. However, the survey work reveals field systems of varying form which cannot be easily classified within the existing framework, specifically those composed of a smaller number of fields which, in many cases, appear to expand from a primary field. The site at Troni Shun (HU 25018 50724) provides a clear example of this, with a primary enclosure and two adjoining fields (Fig. 2.7). Troni Shun is located within a valley with the Loch of Brunatwatt to the north and the mire of Troni Shun to the south on a natural rise with commanding views south towards the modern village of Walls. The overall form and location of Troni Shun at the base of the slope is a pattern which can be observed at other sites, most clearly Burn of Houll (HU 22620 48940), but also South Houllan (HU 29390 55430). The extent and location of the fields differs from that of the more extensive Multiple Field Systems and there is no evidence for walls emanating from these smaller examples. The location of the sites at the base of the hillsides combined with the lack of extending walls also limits the interaction between these sites and surrounding archaeological features. The mapping has revealed that on the West Mainland of Shetland there is a high level of variation within the definitions of the prehistoric field systems identified by Turner

2  Enclosing and Dividing Land: The Neolithic and Bronze Age Field Systems of Shetland

(2012). The typology of Homestead Enclosures and Multiple Field Systems allows for sites to be identified and discussed within a broader interpretational framework. However, this variation on the Multiple Field System form highlights the difficulty in developing strict classifications of landscape features which are unlikely to represent single phase development. The identification of possible subgroups, and new forms, adds detail to this variation and allows us to begin to comprehend its significance.

2.5

 nclosing and Dividing Land: E The Neolithic and Bronze Age Field Systems of Shetland

The detailed mapping of the West Mainland of Shetland reveals the extent and complexity of the prehistoric field systems and confirms that Homestead Enclosures can, indeed, be identified as a distinctive field type. While Multiple Field Systems can also be identified in their broadest sense, the survey reveals striking diversity in form and scale. The comparative assessment of these field systems, and the identification of a greater variety of settlement forms, enriches the current typo-chronological models. The interpretation of temporal patterns observed in the mapping is limited by difficulties in determining a high-resolution chronology for Shetland with all sites being defined as Neolithic/Bronze Age. However, this should not detract from the potential of Shetland field systems to provide unique opportunities for understanding early communities. A striking feature of the Homestead Enclosures, particularly when compared to the other field systems within the study, is the uniformity of their form. If accepted as a coherent earlier site type these characteristics inform the interpretation that they represent an initial stage of defining and investing in a place before developing into Multiple Field Systems (Turner 2012). In contemplating possible developmental models, the excavation results from Catpund suggests an alternative interpretation with evidence for the coexistence of multiple types of field system (Ballin Smith 2005). Therefore, Homestead Enclosures can also be interpreted as a contemporary settlement form to the larger field systems, perhaps with a distinct role or function. The enclosures have been predominately explored as ‘fields’ within the prevailing models of a mixed farming regime. Indeed, there is artefactual evidence for cultivation (stone ard points) within the enclosures supported by the results of micromorphological analysis of soil samples taken from within the enclosure at Houlland, Nesting (Ballin Smith 2005: 43; Turner 2012). This has been discussed solely in terms of cereal cultivation yet a multitude of edible plants

21

could have been grown, tended, and benefited from the shelter provided by enclosure (Edwards 1998; Jones 2000; Johnston 2005). Enclosure as a means to define an area devoted to a single crop or livestock is arguably a relatively modern concept. Livestock movements can be managed without the need for enclosure by shepherds or as evidenced by the Bronze Age tethering posts at the Sumburgh Runway House (Downes and Lamb 2000). Furthermore, the enclosures did not contain an area sufficient to produce the crops or support the livestock required by the occupants (Turner 2011). As such, Homestead Enclosures can perhaps best be described not in terms of a ‘field’, for the production of a single crop or solely the grazing of animals, but as a means of delineating the space and containment of activities relating to the house. The evidence from Catpund attests to a variety of activities which could have been taking place including leather processing and the manufacture of wooden or bone artefacts as indicated by the quartz tool assemblage (Ballin Smith 2005: 44). The societal implications of defining an area so directly related to the house may have also been seen in terms of controlling access, defining space and establishing permanence or a “routine of returns” if a more seasonal use of such sites or wider landscape use is envisaged (Whittle 1996: 176). Thus, Homestead Enclosures may have represented fixed and defined spaces with both functional and social roles within a landscape which was used more widely to graze animals, grow crops and gather resources. Here, the constraints of our limited understanding of date, development and relationship come into play as the act of enclosing land and home within an earlier system of land use may be very different from that of a contemporary or later settlement system. Multiple Field Systems contrast with the bounded units of the Homestead Enclosures with conglomerations of multiple irregular fields and boundaries that extend across the surrounding ground. The Multiple Field Systems, within the prevailing models of Neolithic and Bronze Age land use in Shetland, have been discussed in terms of the intensification of agriculture and the predominance of either cereal ­cultivation or grazing. The development of the fields and the extent to which they represent a coherent system is a key concern. Whittle argues that at Scord of Brouster the irregular layout of the fields is the result of the methods of cultivation not a misleading reflection of the continuity and duration of occupation (1986: 143). It is instead proposed that much of the system may have been outlined from an early stage, functioning as an ‘arable nucleus’ within a landscape used more widely to graze animals (Whittle et  al. 1986: 143). Turner’s application of shape analysis provides a possible sequence of primary, secondary and tertiary fields within the Multiple Field Systems, if it is accepted that they

22

developed from Homestead Enclosures (Turner 2012: 490). In her assessment, primary fields can be identified as convex fields resembling Homestead Enclosures with the larger fields being secondary (ibid: 490). Whittle’s arguments are echoed in Turner’s analysis of the potential yields of the fields as the entire area enclosed by the field system is included in calculations implying it functioned as a whole. The presence of boundaries which extend from the nucleus of the settlement were recognised at Scord of Brouster, in Turner’s assessment and mapped for the first time in extenso by the RCAHMS (Whittle et al. 1986; Turner 2011). Their relationship to the ‘arable nucleus’ or inner field system has been argued by both Whittle and Turner to represent an ‘outfield’ system perhaps related to the pressures on summer grazing and the delineation of pasture (Whittle et al. 1986; Turner 2011). A key result of the mapping was the recognition of the frequency of walls which extend for considerable distances, some over 1  km, either terminating at seemingly random points or continuing between sites. The presence of walls which appear to emanate directly from the field systems can be witnessed in areas with a high concentration of settlements but also at sites such as Kellister. The walls at Kellister extend to further divide the hillside, a feature also seen at Scord of Brouster where they continue in discontinuous fragments to the south to the site of Gallow Hill. When mapped in extenso across the wider landscape the Fig. 2.8  Linear pattern of the field boundaries which extend across the landscape. (© Crown copyright and database rights 2020 Ordnance Survey (EDINA, 100025252))

C. Christie

walls appear to form a linear pattern of boundaries and divisions (Fig. 2.8). If interpreted as a representation of, or part of, a continuous system, then the extension of the field systems show the possibility of interconnectivity between sites and a much more expansive and connected farmed Neolithic/Bronze Age landscape. This, as previously proposed, could form a coherent ‘infield and outfield system’ or an extension of the sites with boundaries to further divide the landscape. The juxtaposition between the irregular fields and the liner boundaries may indicate the negotiation of tenure between groups across what may have been a busy, managed landscape. Tenure within this would have been complex, involving issues of access, allocation of resources, history and the expression of a community’s and an individual’s connection to place (Johnston 2001: 185). However, the seemingly coherent pattern and clear relationship displayed in the survey data should be treated with caution. The limited excavation and dating of features allows for a number of possible relationships to be reasonably considered. Firstly, that the field walls and boundaries taken as a whole do indeed form a coherent ‘infield and outfield’ representative of the same system of land use. Equally, the boundaries may still represent a coherent pattern but of an earlier or later system of land use into which the Multiple Field Systems were incorporated. As later boundaries, the walls could be considered to abut the field systems as opposed to emanating from them.

2  Enclosing and Dividing Land: The Neolithic and Bronze Age Field Systems of Shetland

23

Alternatively, any semblance of coherency may be misleading with the boundaries resulting from unrelated phases of use, remodeling, and replacement. A cautionary tale of the implications can be found at An Sithean, Islay where field survey indicated the presence of two types of bank and two ‘fields’ of possible early date. Excavation and environmental sampling of sections across the field walls revealed that the two ‘fields’ were, in fact, formed of disparate elements dating from the Late Bronze Age to the post-medieval period (Barber and Brown 1984: 186) (Fig. 2.8). Recognition of the complexities of the prehistoric field systems brings into sharp focus the central issue in discussions of Neolithic and Bronze Age Shetland  – chronology. The few excavated examples, and the well-established difficulties in radiocarbon dating, have resulted in sites being broadly defined as Neolithic/Bronze Age with very limited chronological resolution. This constrains the interpretation of the temporal patterns observed in the mapping as it limits our ability to understand the development of individual sites and ascertain the chronological relationship between field systems and other possibly contemporary archaeological features. An overall model of field system development based on the progression from simple to complex forms has been proposed. The identification of small-scale Multiple Field System forms during this program of mapping adds further complexity to this developmental sequence. The sites at Troni Shun, Loch of Brunatwatt and South Houllan are composed of a smaller number of conjoined fields and are located towards the base of hillsides. These smaller-scale field systems could be viewed as the link between Homestead Enclosures and Multiple Field Systems representing a stage in this unilineal progression. However, given the indications of possible contemporaneity the multiplicity of field systems may represent differing ways of occupying land that are not chronologically distinct. The possibility of the coexistence of simple and complex forms of field system invites further questions surrounding subsistence strategies in the Neolithic and Bronze Age of Shetland. The mapping allows for the full spectrum of variations to be identified and interpreted with caution given the lack of supplementary information. The continued investigation of these smaller-scale field systems will be a key element of future research as they raise questions about our understanding of the development of field systems in Shetland.

Homestead Enclosures and Multiple Field Systems display a variety of characteristics and appear to indicate changing and developing settlement and land use strategies. This during the Neolithic and Bronze Age can be seen as a patchwork of activity structured by boundaries and tenurial systems. The difficulties in dating sites in Shetland has resulted in Scord of Brouster becoming the baseline to which all other sites, both excavated and unexcavated, are compared. This in many respects is unavoidable, but it does result in giving undue emphasis to morphological similarities and the creation of overarching models of settlement and land use rather than looking at the differences between sites and landscapes. The high-level mapping and survey of these sites in many respects compounds this issue as it presents us with a diachronic perspective. The landscapes appear to be full of features, cairns and field systems with apparent connections and similarities. Yet, the chronological relationships between these features is far from certain and while they have been broadly dated to the Neolithic and Bronze Ages, there are significant technological, societal and environmental changes which have yet to be teased apart. The nature of these changes, their impact and the chronological relationship between site and the extent to which sites coexisted can only be explored and established through further excavation and the application of alternative dating methods to build a more detailed chronology. The remains in Shetland appear to conform to an image of Neolithic and Bronze Age life with fields for crops, boundaries for livestock, orderly settlements with homes protected from the wild; precursors to the substantial settlements of later prehistory (Gibson 2003: 136; Malone 2001: 48). However, while the stone construction does give a greater sense of permanence, perhaps misleadingly, the mapping reveals the striking levels of variation within sites which have been be broadly categorised as ‘similar’. The problems in defining the relationship between features and their chronological development should not detract from the astonishing potential of Shetland to explore entire landscapes and ask detailed questions about life in the Neolithic and Bronze Age. Taken collectively the prehistoric features in Shetland including, houses, field systems, clearance and burial cairns, burnt mounds, cists and standing stones present an image of a vibrant and busy landscape of rarely paralleled preservation.

2.6

References

Conclusions

Mapping of the West Mainland using high resolution aerial photographs and targeted field survey allows these systems to be placed within a broader landscape context. The results of this macro scale analysis reveal the extent to which the sites present a picture of activity across the landscape. The

Ballin Smith, B. (2005). Catpund: A Prehistoric House in Shetland (Scottish archaeological internet report 7). https://archaeologydataservice.ac.uk/archiveDS/archiveDownload?t=arch-­310-­1/dissemination/pdf/sair7.pdf. Accessed 11 Sept 2015. Barber, J., & Brown, M. (1984). An Sithean, Islay. Proceedings of the Society of Antiquaries of Scotland, 114, 161–188.

24 Barcham, R.  C. (1978). A lost radiocarbon date for Shetland. Proceedings of the Society of Antiquities Scotland, 110, 502–506. Barker, H., Burleigh, R., & Meeks, N. (1971). British Museum natural radiocarbon measurements VII. Radiocarbon, 13(2), 157–188. Bennett, K.  D., Boreham, S., Sharp, M.  J., & Switsur, V.  R. (1992). Holocene history of environment, vegetation and human settlement on Catta Ness, Lunnasting. Shetland Journal of Ecology, 80(2), 241–273. Bondevik, S., Mangerud, J., Dawson, S., Dawson, A., & Lohne, Ø. (2005). Evidence for three North Sea tsunamis at the Shetland Islands between 8000 and 1500 year ago. Quaternary Science Reviews, 24, 1757–1775. Calder, C. S. T. (1950). Report on the excavation of a Neolithic temple at Stanydale in the Parish of Sandsting, Shetland. Proceedings of the Society of Antiquaries of Scotland, 84, 185–205. Calder, C.  S. T. (1955–6). Report on the discovery of numerous Stone Age house-sites in Shetland. Proceedings of the Society of Antiquaries of Scotland, 89, 340–397. Calder, C.  S. T. (1962). Cairns, Neolithic Houses and Burnt Mounds in Shetland. Proceedings of the Society of Antiquaries of Scotland, 96, 37–86. Cracknell, S., & Smith, B. (1983). Archaeological investigations at Mavis Grind, Shetland. Glasgow Archaeological Journal, 10, 13–39. Curle, A.  O. (1934). An account of further excavation at Jarlshof, Sumburgh, Shetland, in 1932 and 1933. Proceedings of the Society of Antiquaries of Scotland, 68, 224–319. Curle, A.  O. (1936). Account of excavations of an iron smelting and of an associated dwelling and tumuli at Wiltrow in the Parish of Dunrossness, Shetland. Proceedings of the Society of Antiquaries of Scotland, 70, 153–169. Downes, J., & Raymond, L. (2000). Prehistoric Houses at Sumburgh in Shetland: Excavation at Sumburgh Airport 1967–74. Oxford: Oxbow Books. Edwards, K. J. (1998). Detection of human impact on the natural environment: Palynological views. In J. Bayley (Ed.), Science in archaeology: An agenda for the future (pp.  69–88). London: English Heritage. Edwards, K. J., & Whittington, G. (1998). Landscape and environment in prehistoric West Mainland, Shetland. Landscape History, 20, 5–17. Edwards, K. J., Whittington, G., Robinson, M., & Richter, D. (2005). Palaeoenvironments, the archaeological record and cereal pollen detection at Clickimin, Shetland, Scotland. Journal of Archaeological Science, 32(12), 1741–1756. Gibson, A. (2003). What do we mean by Neolithic settlement? Some approaches, 10 years on. In I.  Armit, E.  Murphy, E.  Nelis, & D.  Simpson (Eds.), Neolithic settlement in Ireland and Western Britain (pp. 136–145). Oxford: Oxbow Books. Henshall, A. S. (1955–6). APPENDIX II: Pottery and stone implements from Ness of Gruting. Proceedings of the Society of Antiquaries of Scotland, 89, 381–397. Hoppe, G. (1965). Submarine peat in the Shetland Islands. Geografiska Annaler, 47A, 195–203. Hulme, P.  D., & Shirriffs, J. (1994). The Late-glacial and Holocene vegetation of the Lang Lochs Mire area, Gulberwick, Shetland: A pollen and macrofossil investigation. New Phytologist, 128(4), 793–806. Johansen, J. (1975). Pollen diagrams from the Shetland and Faroe Islands. New Phytologist, 75, 369–387. Johnston, R. (2001). Land and society: The Bronze Age cairnfields and field systems of Britain. Unpublished PhD thesis, Newcastle University. Johnston, R. (2005). A social archaeology of garden plots in the Bronze Age of northern and western Britain. World Archaeology, 37(2), 211–223.

C. Christie Jones, G. (2000). Evaluating the importance of cultivation and collecting in Neolithic Britain. In A. S. Fairbairn (Ed.), Plants in Neolithic Britain and beyond (pp. 78–84). Oxford: Oxbow Books. Keith-Lucas, M. (1986). Neolithic impact on vegetation and subsequent vegetational development at Scord of Brouster. In A. Whittle, M. Keith-Lucas, A. Milles, B. Noddle, S. Rees, & J. C. C. Romans (Eds.), Scord of Brouster: An early agricultural settlement in Shetland (pp. 92–118). Oxford: Oxford University Committee for Archaeology, Monograph 9. Lambeck, K. (1993). Glacial rebound of the British Isles – I. Preliminary model results. Geophysical Journal International, 115, 941–959. Mahler, D. L. (2011). Shetland-the border of farming 4000-3000 BC, some features of the Neolithic of Shetland. In D.  L. Mahler & C. Andersen (Eds.), Farming of the edge. Cultural landscapes of the north: Short papers from then network meeting in Lerwick, Shetland September 7th–10th (Vol. 2010, pp. 6–18). Copenhagen: National Museum of Denmark. Malone, C. (2001). Neolithic Britain and Ireland. Gloucestershire: Tempus. Milles, A. (1986). Charred remains of barley and other plants from Scord of Brouster. In A.  Whittle, M.  Keith-Lucas, A.  Milles, B. Noddle, S. Rees, & J. C. C. Romans (Eds.), Scord of Brouster: An early agricultural settlement in Shetland (pp. 119–122). Oxford: Oxford University Committee for Archaeology Monograph 9. Murray, J., & Christiansen, C. (2013). Those who came before: Shetland’s antiquarians and the Neolithic. In D. L. Mahler (Ed.), The border of farming. Shetland and Scandinavia, Neolithic and Bronze Age farming: Papers from the symposium in Copenhagen September 19th to the 21st 2012 (pp. 92–102). Copenhagen: National Museum of Denmark. Peltier, W. R., Shennan, I., Drummond, R., & Horton, B. (2002). On the postglacial isostatic adjustment of the British Isles and the viscoelastic structure of the Earth. Geophysical Journal International, 148, 443–475. RCAHMS. (1946). Royal Commission on the Ancient and Historical Monuments of Scotland, twelfth report with an inventory of the Ancient Monuments of Orkney and Shetland. Vol. III, Inventory of Shetland. Edinburgh: H. M. Stationery Office. Sheridan, A. (2012). Neolithic Shetland: A view from the “mainland”. In D. L. Mahler (Ed.), The border of farming and the cultural markers: Short papers from then network meeting in Lerwick, Shetland September 5th–9th 2011 (pp. 6–36). Copenhagen: National Museum of Denmark. Sheridan, A. (2013). Plus ca change…developments in Shetland, 2500-­ 1800 BC. In D. L. Mahler (Ed.), The border of farming. Shetland and Scandinavia, Neolithic and Bronze Age farming: Papers from the symposium in Copenhagen September 19th to the 21st 2012 (pp. 47–72). Copenhagen: National Museum of Denmark. Sheridan, A., Garrow, D., Mason, O., Schulting, R., Snoeck, C., Bradley, R., Waddington, C., Evershed, R., Whelton, H., Cook, G., & Naysmith, P. (2014). Radiocarbon dates associated with the Scottish History and Archaeology Department, National Museums Scotland, 2013/14. Discovery and Excavation in Scotland, 15, 213–217. Tait, I. (2011). Post-medieval reuse of Neolithic sites in Shetland. In D. L. Mahler & C. Andersen (Eds.), Farming of the edge. Cultural landscapes of the north: Short papers from then network meeting in Lerwick, Shetland September 7th–10th (Vol. 2010, pp.  70–79). Copenhagen: National Museum of Denmark, Copenhagen. Turner, V. E. (1998). The shaping of Shetland. Lerwick: The Shetland Times. Turner, V. E. (2011). From homestead enclosure to farm? Field development in Shetland in the Neolithic period. In D.  L. Mahler & C. Andersen (Eds.), Farming of the edge: Cultural landscapes of the north: Short papers from then network meeting in Lerwick, Shetland

2  Enclosing and Dividing Land: The Neolithic and Bronze Age Field Systems of Shetland September 7th–10th 2010 (pp.  19–31). Copenhagen: National Museum of Denmark. Turner, V.  E. (2012). Location, form and function in Shetland’s Prehistoric field systems. Unpublished PhD thesis. University of Stirling. Turner, V. E. (2013). Living on the land? In D. L. Mahler (Ed.), The border of farming. Shetland and Scandinavia, Neolithic and Bronze Age farming: Papers from the symposium in Copenhagen September 19th to the 21st 2012 (pp. 24–34). Copenhagen: National Museum of Denmark. Turner, V.  E., Chrystall, F., Simpson, I.  A., & Guttmann, E. (2004). Form and function in Shetland prehistoric field systems. In R.  A.

25

Housley & G.  Coles (Eds.), Atlantic connections and adaptions. Economies, environments and subsistence bordering the North Atlantic (Symposia of the Association for Environmental Archaeology No. 21, pp. 120–127). Oxford: Oxbow Books. Whittle, A. (1996). Europe in the Neolithic: The creation of new worlds. Cambridge: Cambridge University Press. Whittle, A., Keith-Lucas, M., Millies, A., Noddle, B., Rees, S., & Romans, J. C. C. (1986). Scord of Brouster: An early agricultural settlement on Shetland. Oxford: Oxford University Committee for Archaeology Monograph 9.

3

Formation, Use and Chronology of Celtic Fields: New Perspectives from the Groningen Celtic Field Research Programme S. Arnoldussen

Abstract

Targeted excavations of Dutch Celtic fields (Dutch: raatakkers) have yielded much-needed data on the ways in which Celtic fields developed and were used over time. This type of later prehistoric field system is common to gently undulating upland landscapes consisting of Saalian and Weichselian deposits, which were frequently affected by podzolisation. As direct dating of the use-period of Celtic fields is difficult, the Groningen Celtic field research programme employs combined archaeological, AMS- and OSL-dating of bank sediments. Using such dates, use-histories starting in the Middle Bronze Age and lasting well into the Roman era could be shown. Detailed palaeo-botanical analyses of bank sediments suggest that soil and plants from wetland parts were taken to settlements to be used as byre-bedding, after which they were enriched with household refuse (sherds, ash, charcoal) and carted-off to the fields to be used as manure. The locations of such settlements remain difficult to establish, as the chronological relations between settlement features and locations of banks are often unclear. Moreover, despite the fact that botanical macro-remains and pollen inform us on the types of crops cultivated, details of (changes in) the agricultural cycles still escape us. Keywords

Netherlands · Celtic fields · Raatakkers · Dating · Palaeo-botanical · Manuring · Excavations

S. Arnoldussen (*) Groningen Institute of Archaeology, Groningen University, Groningen, The Netherlands e-mail: [email protected]

3.1

 ater Prehistoric Agricultural L Landscapes in the Low Countries

Whereas the excavations of later prehistoric barrows (e.g. Theunissen 1999: 41–46; Bourgeois 2013: 3), hoards (e.g. Verlaeckt 1996; Fontijn 2003) and settlements have long traditions in the Low Countries (Theunissen 1999, 116; Bourgeois et al. 2003), the agricultural landscapes that supported the communities reflected by such site types are rarely targeted for research, thus failing to address their interrelations and embeddedness within the wider (agri)cultural landscape fabric. Despite the fact that later prehistoric field systems are frequently still preserved as tangible elements in the present-­day landscape, only few investigations into their age, agricultural regimes and particular morphology have taken place. Notwithstanding the early realisation the later prehistoric communities of the Low Countries were primarily subsistence farmers (e.g. Bursch 1951: 115; van Giffen 1947: 524; De Laet 1958: 131; Butler 1969: 26–28), the gardens, fields and grazing grounds vital to later prehistoric sustainability are seldom the focus of archaeological research. Or, freely after Louwe Kooijmans (2000: 324): “However important monuments, ritual places and cemeteries might appear, we should realize that 90 per cent of life revolved in and around [the fields, gardens and meadows of] the settlements” (my addition). This paper presents data from a series of recently investigated Dutch later prehistoric field systems known as raatakkers or Celtic fields, in order to unravel as yet poorly known characteristics such as their dating, palaeoecology and palaeo-economy. In this paper, the area and period under study reflect pragmatic choices in availability and coherence of relevant data. In chronological terms, later prehistory is equated here to the Bronze Age and Iron Age periods, or c. 2000 BC – AD 0, or Reinecke Br. A1 to La Tène D2 periods (Arnoldussen and Jansen 2010: 380 fig. 1), but some evidence from the preceding Neolithic periods and later Roman era will be considered as well. In geographical terms, the Low Countries are taken

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 S. Arnoldussen et al. (eds.), Europe’s Early Fieldscapes, Themes in Contemporary Archaeology, https://doi.org/10.1007/978-3-030-71652-3_3

27

28

S. Arnoldussen

here to represent the zone of 0–50 m altitude spanning from northwesternmost France to Belgium, The Netherlands and parts of German Nordrhein-Westfalen and Niedersachsen up to the Elbe estuary (Fig.  3.1). Whilst it is evident that this lowland zone continues across the Elbe into Schleswig-­ Holstein and southern Jutland – and that prehistoric field systems occur there too (Arnold 2002, 2011; this volume) – the rich data-set of investigated Celtic fields located in the above defined zone allow for a coherent and representative discussion of later prehistoric field systems in the Low Countries, even if based here mostly on Dutch data. The lowland region thus defined encompasses a variety of zones with variable geological histories and subsequent environments: it comprises major river systems active throughout the Holocene (Scheldt, Meuse, Rhine) and equally dynamic coastal zones (Arnoldussen and Fokkens 2008: 18–25). Inland of these coastal zones, inactive creek systems and peatlands presented somewhat less dynamic environments to later prehistoric farming communities (loc.

cit.). In the southeastern and northernmost areas, vast coversand landscapes were present, in the latter region interspersed with glacial deposits of Saalian age (ibid.).

3.2

 relude: A Long-term Perspective P on Fields and Subsistence – And Its (in)Visibility

The introduction and piecemeal acceptance of agriculture in the period of 5300–2000 cal BC (van Gijn and Kooijmans 2005: 208; Amkreutz 2013: 412–13) did not herald an immediate increase in the numbers of agricultural plots (fields, gardens, pastures) known through archaeological research. For example, for the better part of this trajectory, the evidence for crop-cultivation locations is indirect (Amkreutz 2013: 329–30). The oldest reliably identified agricultural field plot from the Low Countries is Swifterbant S4, where a hoe-cultivated plot dated to 4300–4000 cal BC

Flögeln

Zeijen Hijken Westeinde

Vaassen Hilversum - Hoornboegse heide Wekerom Wekerom

Someren Kolisbos

Herkenbosch

0

100 kilometres

Fig. 3.1  Schematic distribution of evidence for field systems (dark brown: Neolithic and Bronze Age fields and field systems; lighter brown: Celtic field systems; after Brongers 1976; Klamm 1993; with

additions). Archaeologically investigated Celtic fields are indicated with star-shaped symbols

3  Formation, Use and Chronology of Celtic Fields: New Perspectives from the Groningen Celtic Field Research Programme

was uncovered on a creek-levee of the eponymous culture group (Cappers and Raemaekers 2008; Huisman and Raemaekers 2014). It is not until the Funnel Beaker Culture period (c. 3500–3000 cal BC) that field plots come into view again  – this time as sites with ard-marks such as at Groningen – Winschoterdiep (Kortekaas 1987; Lanting and van der Plicht 2000: 67; Arnoldussen and van Os 2015: 17). On settlement sites datable to the Single Grave Culture period (c. 3000–2400 cal BC), such ard marks are more frequently encountered (Fokkens 2008: 20; Drenth et al. 2008: 158; 171) yet structures indicating the sizes of plots remain absent. The larger extents of excavations of later Bronze Age settlements – combined with a (desire for) better boundary visibility during that period (c. 1500–1000 cal BC) – renders agricultural “plots” visible for the first time. At Hoogkarspel – Watertoren, an excavation strategy of following ditch trajectories resulted in the identification of various “plots”, of which one showed ard marks (Bakker et al. 1977: 194 fig. 7). Such ditch-enclosed ard-mark plots range between 0.02 and 5 hectares in area and are frequently identified on Bronze Age settlements with sufficient preservation (Arnoldussen 2008: 259 note 279), and an average of 1.8 hectares arable per household (range 1–3 ha; Van Amerongen 2016: 168) has been suggested. In the Dutch river area wattlework fences and ard marks have been uncovered during several settlement excavations, but determining the exact sizes and locations of agricultural plots remains difficult regardless (Arnoldussen 2008: 243–53; 257 note 277; 421–24, cf. Van Amerongen 2016: 167). This pattern holds at a wider scale too: with over 60 Bronze Age settlements excavated (Arnoldussen and Fokkens 2008), and with the Bronze Age economy of true mixed-farming (interdependent livestock rearing and crop-cultivation; Louwe Kooijmans 1993: 104) being well-studied (Clason 1999; Brinkkemper and van Wijngaarden-Bakker 2005; Van Amerongen 2016), the scarcity of information on fields proper may testify to their original fluid or ephemeral definition in the past, but is also affected by heritage management decisions in which settlement cores rather than peripheries (where fields may be expected) are favoured. According to traditional views, it is only with the start of the Iron Age (c. 800 – 12 BC) that field plots, as part of the embanked field systems known as Kammerflüre in German, Oltidsagre in Danish and raatakkers in Dutch, come into clearer view. Their suspected Iron Age date was however based on few, and mostly older, excavations. In the period between 1918 and 1949, Dutch archaeologist Albert Egges van Giffen excavated several such “heathen encampments’, only gradually realising their true nature as later prehistoric field systems (e.g. van Giffen 1918, 1936, 1940a, 1949). In the following decades, few targeted investigations of Celtic fields in the Low Countries took place, but in the 1970s

29

important volumes on the excavations at Flögeln (Zimmermann 1976) and Vaassen (Brongers 1976) were published. Again, nearly three decades lapsed before the embanked field systems were once more targeted (e.g. Spek et al. 2003, 2009). In the period 2010–2017, several Dutch Celtic fields have been subjected to research excavations as part of a research program by Groningen University’s Institute of Archaeology and the present author, which allows a recent and more precise characterisation than before of these field systems.

3.3

Celtic Fields and Where to Find Them

For the present paper, raatakkers or Celtic fields are defined as later prehistoric embanked field systems, in which low ( 10 cm), it could be determined that the banks hardly contained any stones in the largest (> 10  cm) size-grade. This proves that the Celtic field banks were not constructed from stones tossed to the sides of fields by farmers during ploughing (cf. Nielsen and Dalsgaard 2017: 430). Stones of smaller sizes that were present show that the banks also were not constructed solely by (or of) driftsand. An intake of non-­local soil (possibly from wetter parts of the landscape) was suggested by micromorphological analysis: hydromorphous iron formations (native to wet landscape conditions) and clods of non-local soils could be observed at Zeijen and Wekerom (Arnoldussen and van der Linden 2017: 562, cf. Nielsen and Dalsgaard 2017: 423). The palynological data also contained indicators for wetland landscapes: pollen of sedges, bullrush and burr-reed as well as freshwater algae such as Debarya, Spirogyra and Zygnemataceae were found in the banks (Arnoldussen and van der Linden 2017: 560; 562, cf. Nielsen and Dalsgaard 2017, 431; Nielsen et  al. 2019). The presence of dung is indicated through spores of coprophilous fungi such as Sporormiella, Cercophora, Podospora and Sordaria (ibid.), albeit that it cannot be determined whether this represents dung carted-off to the fields or droppings from animals penned within (stubble) fields. Combined, the composition of the Dutch Celtic field banks represents a mixture of sherds, charcoal, burned plant seeds, non-local soil and dung. The ‘recipe’ of this mixture can be reconstructed as follows: (a) sods from wetter parts of the landscape were brought to the settlement to be used as byre bedding, (b) excrement-enriched sods were mixed with household rubbish, which was (c) carted-off to the fields to be used as fertilizer. Each of these steps leaves distinct proxy signals: the first step is evidenced by the clods and non-local soil formation in thin sections, and via wetland pollen/algae (cf. Fokkens 1998: 120; Behre 2008: 155), whereas the second step is documented by the coprophilous dung spores, sherds, charcoal from the hearth and burned plant seeds (Arnoldussen and van der Linden 2017: 563). Essentially, this is an agricultural system similar to that of the plaggen-

34

wirtschaft known from the Medieval period (cf. Kroll 1975; Heidinga 1988), yet here dated to the final millennium BC. Explaining the composition of the banks is not the same as explaining why and how the banks evolved. Arnoldussen and Van der Linden (2017: 564; 566) have argued that for the Low Countries, the (1) omnipresence and uniform outlook of Celtic field banks, (2) their composition (supra) and (3) their centuries deep historical trajectories (supra), are best explained as resulting from soil adhering to uprooted field weeds which were tossed to the sides of fields. Presumably, such uprooted field weeds were initially tossed against wattlework fences that delineated individual plots, yet over time the organic accumulations to either side of the fence took over the demarcating role previously played by the fences (Liversage et al. 1987: 21; Klamm 1993: 40; 42; Spek et al. 2003: 167; Arnoldussen and de Vries 2014: 100–101 fig. 12). In such a scenario, local tilled soils were enriched with a mix of household refuse and manure with byre bedding (sods) that was intended to fertilize fields (Bradley 1978: 272), but through the repeated cycles of weeding and harvesting, tiny amounts of this mixture transported in plant root clusters would ever so slowly accumulate at the fields’ edges. The suggested (very slow!) accumulation speed for Celtic field banks in the Low Countries (i.e. c. 30 years/cm; supra), is in line with uprooting of fields weeds as a small but constant (and ubiquitous) actor in bank aggradation (cf. Curwen and Curwen 1923: 31–32; Harsema 1980: 20; Jankuhn 1958: 181; Klamm 1993: 44). Every field plot had to be cleared of weeds and harvested at various points in the agricultural cycle, which may explain the uniformity in bank dimensions and heights within a Celtic field (even if allowing for more use-trajectories of individual fields or clusters of field plots; cf. Klamm 1993: 44; Zimmermann 1995: 293; Gerritsen 2003: 175; 177; Nielsen et al. 2019).

S. Arnoldussen

100 litre bank volume), which – in tandem with the absence of wild species  – does not seem to reflect in-situ burned fields or vegetation (Arnoldussen and Smit 2017: 57). Furthermore, the burned state of crops that do not require charring as part of their preparatory stages (e.g. dechaffing, roasting) such as millet and flax, suggests that this represents kitchen waste rather than crops grown nearby (Arnoldussen and van der Linden 2017: 562). The concurrence of such burnt cultivars with charcoal and sherds (supra) again supports the interpretation of a dung and household debris mixture being used to fertilize fields. This means, however, that – quite counter-intuitively – analysis of charred cultivars from Celtic fields banks can only indirectly (by proxy) inform us about Celtic field agriculture: the cultivars recovered from the banks do reflect the consumption patterns of the later prehistoric communities working the field plots, but do not inform us whether or where such crops were grown in the adjacent field plots. For samples from Celtic field banks (and agricultural layers from the field plots within) pollen is a far better indicator for the types of crops grown locally and agricultural strategies applied (e.g. fallow, manuring, clearance) compared to macro-botanical remains. This even holds if one factors in risks posed by mixing (ploughing, digging, bioturbation) and contamination (downwash of pollen and non-local influx) of pollen records (cf. Nielsen and Dalsgaard 2017: 425). That said, the pollen locally present is likely to comprise a mixture of (1) locally cultivated plants and fields weed communities, (2) pollen of nearby grassland/fallow plots, (3) pollen transported as part of the manuring mix (wetland sods, herbivore fodder) and (4) younger intrusive pollen. For all pollen identified, a critical review of their possible origin(s) is required. At Zeijen, pollen of cultivated species (mainly cerealia of Hordeum/triticum type and flax) could amount to 11% of the pollen sum and species of fallow arable land (Plantago lan3.6 Celtic Field Economy ceolata and Pteridium aquilinum) were common (Arnoldussen and van der Linden 2017: 559–560). At Reconstruction efforts for specifics of the agricultural sys- Wekerom, pollen of Hordeum/triticum type and Triticum tems at play in Celtic fields, must rely on integrated palaeo-­ type was found with arable weeds (e.g. Artemisia, botanical and palaeo-ecological methodologies comprising Convolvulus arvensis-type, Papaver rhoeas-type, studies of macro-botanical remains as well as pollen analy- Scleranthus and Spergula arvensis; Arnoldussen and van der ses. Fine mesh sieving of bank sediments at Zeijen Linden 2017: 561–62). At both sites, grassland pollen may (Arnoldussen 2012: 44–46), Wekerom (Arnoldussen and amount to c. 40–50% of the pollen sums (Arnoldussen and Scheele 2014: 60–61), Someren (Arnoldussen, Schepers van der Linden 2017: 559; 561), hinting at the importance and  Maurer 2016) and Westeinde (Arnoldussen and Smit that grassland plots for grazing (in their own right, or as part 2017) has resulted in the recovery of charred seeds and plant of fallow cycles) may have had in Celtic field agriculture (cf. remains of various cultivars (Fig.  3.4; comprising wheat Arnoldussen 2018: 310; Bradley 1978: 272). At Someren, (emmer and bread wheat), barley, millet, flax and spurrey). It grassland percentages were fluctuating but could reach simiwould be erroneous, however, to assume that these remains larly high values (Arnoldussen, Schepers and Maurer 2016: reflect the local crop-composition of the fields adjacent to the 29–30), suggesting that conversions of plots for crop-­ banks from which the samples were obtained. First, their fre- cultivations into grassland plots (and vice versa) were comquency of occurrence is very low (c. 1–5 charred seeds per mon (ibid.).

++

pollen

1-6.3%

0-2.6%

1+ rachis i.

pollen

0.5-6%

pollen

2.5-7.8%

0-0.5%

is

45-65%

2.1-10.9%

er os

m iu

oc ae Ph

Ty p

ha

an

ar gn

er ac

ea

e

0-3.3%

C yp

ta an Pl +

Sp

go

ae

in

G ra m

er gu

Sp

6%

40-60%

Waterbolk, 1949 + 0-3%

+

Arnoldussen, 2012: 44-51

+

0-1%

1.5-15.3% 117-209% 0-0.5%

1

0-4.2%

53-64%

8.5-22.1% 2.3-9.4%

0.3%

2.7-5.2%

0-0.5%

rachis i.

Westeinde

macro

2+ rachis i.

Someren

pollen

0.8-2.7%

0.4-0.8%

Valthe - HW

pollen

+

+

Meinweg

pollen

+

+

Casparie, 1976: 106-107 Arnoldussen & Scheele, 2014: 60-77

1

macro

Spek, et al. 2003 Arnoldussen, 2012: 44-51

0-0.8%

1?

+

35

ns

m si

la

us

6% +

+ +

ar ve

ita

tis

ac ile Li

nu

m

cu

m

ce

ni

le

Pa

ca Se

m

re a

iv st um

H or de

ae m

cu Tr iti

le

um

m cu co di Wekerom

cu

1-10%

pollen

macro

Vaassen

m

lia Zeijen - NV

Tr iti

C er ea pollen

eu

m

um

3  Formation, Use and Chronology of Celtic Fields: New Perspectives from the Groningen Celtic Field Research Programme

Arnoldussen & Scheele, 2014: 60-77

2

1

1

1

Arnoldussen & Smit, 2017 18-59%

0.8-2.2%

Arnoldussen, Schepers & Maurer, 2016

0.8-2.% +

+

Fens & Arnoldussen, 2015: 66 Arnoldussen, et al. 2014: 17

Fig. 3.4  Plant macroremains and pollen of cultivars and selected wetland species from Dutch Celtic fields (Cerealia = cereals (indet.), Triticum dicoccum = Emmer wheat, Hordeum = Barley, Secale cereale = Rye, Panicum miliaceum  =  Millet, Linum usitatissimum  =  Flax/

Linseed, Spergula arvensis  =  Spurrey, Gramineae  =  Grasses, Plantago = Common plantain, Cyperaceae = Sedges, Sparganium = Bur-­ reed, Typha = Bulrush, Phaeoceros = Hornworts). After: Arnoldussen and Smit 2017, 58 fig. 3.

The above observations pay little respect to the still ample voids in our knowledge of how Celtic fields functioned from farmers’ perspectives. Did they make use of crop-cycles and what (variation in) frequency and duration of fallow may be assumed (cf. Jankuhn 1958: 203; Fowler 1983: 112; Klamm 1993: 50; 80; Odgaard 1985: 127; Nielsen and Dalsgaard 2017: 415)? What percentage of fields within a Celtic fields complex may have laid waste as part of fallow cycles and for how long? Behre (2008: 115) assumes that only up to 10% of the Flögeln field plots were in simultaneous use for crop-­ cultivation (again allowing for ample grazing and an extensive use-regime (cf. Becker 1971: 97–98; Groenman-van Waateringe 1980: 364–66; Liversage et  al. 1987: 80–81; Klamm 1993: 81; Løvschal and Holst 2014: 8; Smith 1996: 214; Spek et  al. 2003: 165–66; Odgaard 1985: 127; Zimmermann 1976: 88–89). Solid data on fallow duration (e.g. via percentages of biennials) and nutrient conditions (e.g. depletion risks and manuring requirements (cf. Gebhardt 1976: 100; Liversage et al. 1987: 59; Odgaard 1985: 127) are however still lacking (Arnoldussen 2018: 308; 322).

dating the field systems (supra), means that establishing detailed biographies for individual Celtic fields is often difficult. Therefore, it is here only possible for the lowlands data to address such issues in general ‘phases’: the before and during of Celtic field usage. The fact that the pedology of primary soils uncovered underneath Celtic field banks is generally disturbed and homogenized (at Zeijen even hoe-marks were detected; Arnoldussen 2012: 29) suggests that these landscapes may have been cleared and worked well before the advent of the Celtic field economy (op.cit: 28; 34). At Wekerom and Someren, no intact primary pedology could be documented (underneath the banks primary soils formations have been homogenized down into the BC-horizon; Arnoldussen and Scheele 2014: 32–52; Arnoldussen, Schepers and  Maurer 2016: 24 fig. 2). At Westeinde, homogenisation down to the BC-horizon of the primary soil was also the norm, albeit that here in one test-pit on a bank location a primary podzol had been preserved underneath the bank. Whereas the soil homogenisation described above could hint at tillage from pre-Celtic field periods (cf. Brongers 1976: 48–50), the nature and dating of pre-Celtic field activities often remains unclear. At Wekerom, some Bell Beaker sherds were sieved from the banks (Arnoldussen and Scheele 2014: 80, cf. van Klaveren 1986: 48) and at Vaassen pits and a four-post outbuilding with Early Bronze Age barbed-wire stamp decorated sherds were found during the excavation (Brongers 1976: 45; 50; 104), indicating that in these locations domestic activities (or even settlements) from the Late Neolithic to Early Bronze Age periods are to be expected. For Zeijen  – Noordse veld, usage of the later Celtic field

3.7

 Cultural-Landscape Approach A to Celtic Fields: The Before and During

A diachronic approach, towards the prior and contemporaneous usage(s) of the cultural landscapes in which Celtic fields in the Low Countries developed, can inform us about what prehistoric communities themselves considered ‘appropriate’ locations and life-histories for Celtic fields. Unfortunately, the modest research intensity and difficulty in

36

complex is documented from the Middle Neolithic onwards: the lowermost homogenized layer with hoe-marks in trench 1 was OSL-dated to 3390–2790 BC (Wallinga and Versendaal 2013b: 6; Arnoldussen 2012: 29) and within 100  m of  the Celtic field banks the Funnel Beaker Period passage grave D5 is situated (Arnoldussen 2012: 21). Possibly, a linear arrangement of Bronze Age barrows (and/or the road accompanying the barrows) delimited the Celtic field complex of Zeijen to the west – it may even have steered the fields’ main axis of the orientation (Waterbolk 1977: 6(168)– 8(180);14(186), cf. Jager 1985: 186; Nielsen et  al. 2017: 389). Combined, the observations at Zeijen, Wekerom and Vaassen indicate that  – like elsewhere in Northwestern Europe (cf. Nielsen and Dalsgaard 2017: 14) – Celtic fields could be established in landscapes that were used previously by prehistoric communities for domestic, funerary, and possibly also agricultural purposes: evidently, no tabula rasa situation was preferred. For non-agricultural activities taking place amidst the Celtic fields during their usage, the interplay of houses and Celtic fields must be discussed. Whilst ample sites have yielded house-plans and granary-type outbuildings during excavations, the spatial and chronological relationships between houses and plots remains difficult (e.g. at Zuidveld, Zeijen, Wekerom; Vaassen, Peelo, Hijken, Westeinde: van Giffen 1940b; Waterbolk 1977: 8(180), 19(188)–19(191); fig. 3.4; 8; Arnoldussen and Scheele 2014: 15 fig. 8; Brongers 1976: 52 fig. 6; Kooi and de Langen 1987; Arnoldussen and De Vries 2014, 2017). The spatial relationship between house structures and bank/field locations can be characterized as awkward and ambiguous (Fig. 3.5; bottom): in spite of artist’s reconstructions that show us otherwise, houses appear frequently to be situated partly overlapping the banks proper – as if their placement was done irrespective of the banks (albeit that the orientation of the houses is generally similar to those of nearby banks). Unfortunately, stratigraphic relations between houses and banks have not been documented for any of these excavations, which means we can only speak of ‘overlap’ between houses and banks, instead of identifying which preceded which. Celtic field banks in those cases were mostly preserved only as slight elevations (e.g. van Klaveren 1986: 11) or greyish discolourations of the excavation level (e.g. Arnoldussen and de Vries 2014: 100–01). Determining whether postholes of houses were cut through this greyish (bioturbation) layer, or alternatively, whether the greyish tint developed after the construction of houses was impossible from the legacy excavation data. Whereas intuitively a sequence of houses placed atop of banks seems plausible (Arnoldussen and De Vries 2017: 80), reverse or alternating sequences are known from later prehistoric fields elsewhere (e.g. ploughed-out houses: Liversage et  al. 1987: 61; 79).

S. Arnoldussen

Some credibility to the interpretation of houses being placed onto or into banks is offered by the dating of the house structures: most such houses are dated to the Middle- and Late Iron Age (Brongers 1976: 51; Arnoldussen and Scheele 2014: 19; Arnoldussen and de Vries 2014: 99; Waterbolk 2009: 64). Yet, Early Iron Age dates for specific houses amidst the Celtic fields at Peelo and Hijken are secure (Kooi and de Langen 1987: 58(158); Arnoldussen and de Vries 2014: 92–95). Moreover, whilst their spatial placement in relation to banks seems awkward, the orientation of the houses appears to respect and reflect that of banks nearby (Arnoldussen and De Vries 2017: 87). The much earlier (Middle Bronze Age) starting dates advocated here for bank aggradation imply that habitation pertaining to the initial use-phases of such Celtic fields has not yet been uncovered: this fact may be explained by low research intensity but could equally well reflect meaningful shifts in residential locations (more distant from the fields in the later Bronze Age and amidst the fields in the Iron Age periods). The scarcity of material culture from the earlier (Bronze Age) use-­ phases during excavations of the Celtic field banks (cf. Taayke 1996: 173; 186; 188) moreover suggests that adding settlement debris to the manuring mix may be an innovation occurring only after the Late Bronze Age (Arnoldussen 2018: 322). Evidently, the life-histories of Celtic fields seem to present detailed but idiosyncratic narratives, in which interplays of domestic structures or activities and agricultural developments have become fossilized. Unfortunately our narratives of use-histories of individual Celtic fields or (supra-)regional groups of Celtic fields, still require ample additional research in order to provide a level of detail suitable to describe later prehistoric variabilities in agricultural practices in the Low Countries.

3.8

 ields of Opportunity: What Is Left F to Learn?

The present evaluation of later prehistoric field systems in the Low Countries has addressed their economy, dating and (cultural) landscape context, but myriad other equally relevant aspects have gone unconsidered. Starting from the largest spatial scale, archaeologists are yet to describe and explain inter-regional differences in forms (stone banks, lynchets, ditches etc.; Klamm 1993: 9–30; English 2013; Johnston 2013; Nielsen and Dalsgaard 2017: 414–15) and similarities in morphological syntax (Løvschal 2014: 731; Nielsen et al. 2017) of later prehistoric field systems at the West-European scale. One level down, the evident morphological similarities in outlook for Celtic fields in parts of the Low Countries with a different geogenesis merits more research (Arnoldussen 2018: 309, cf. English 2013:

3  Formation, Use and Chronology of Celtic Fields: New Perspectives from the Groningen Celtic Field Research Programme

Wekerom - Lunteren

37

457750

559500

Peelo - Kleuvenveld

100 m

174000

559250

0

A

0

B 235250

100 m

235500

547000

Hijken - Hijkerveld

0

C 228500

Fig. 3.5  Top: Cut-outs of artist reconstructions of Celtic fields, showcasing the suspected interrelations of houses and fields (left: © Gemeente Ede, right: © Drents Museum). Bottom: Overview of excavated Celtic fields with Iron Age habitation (A: Wekerom – Lunteren: after van Klaveren 1986; Arnoldussen and Scheele 2014: 15 fig. 8), B: Peelo – Kleuvenveld (after: Kooi and de Langen 1987; Kooi 1997), C: Hijken – Hijkerveld (after: Harsema 1974, 1991: 23 fig. 2; Arnoldussen

100 m

229000

and de Vries 2014: 101 fig. 12), all to the same scale). White areas and outlines represent the excavated areas, the locations of reconstructed and observed Celtic field banks are depicted in halftone brown. Iron Age houses and outbuildings are depicted in red and the black polylines represent fence lines (note that for Peelo and Hijken barrows also present have been omitted from the plans)

38

15): what can explain interregional traditionality in morphological patterns? Does it reflect supra-regionally shared ideas on proper tenure – and if so, why are differences visible at supra-regional level (cf. Nielsen et al. 2017: 402)? At the spatial scale of individual Celtic field complexes, two topics are particularly understudied: the first issue concerns models for Celtic field-complex development and the second topic concerns settlement-field interrelations. Gerritsen (2003: 167–78) argued that information on whether (and at what scales) Celtic fields were ‘planned’ and at which point in time banks accumulated (from the onset, gradually, or at the end-phase, cf. Spek 2004: 147–49) was insufficient to choose between tentative models proposed (Gerritsen 2003: 174–77). We presently have no wellsubstantiated ideas on  through what processes or developmental stages Celtic fields accumulated: is there evidence for initial block- or strip planning which later coagulated or expanded into aggregate Celtic field-complexes (cf. Brongers 1976: 57; Nielsen et al. 2017: 396) or should other developmental mechanisms be explored (cf. Anker 2017: 13; Schrijver 2011; Nielsen et  al. 2017: 389–92)? Also, scholars of later prehistoric field systems should strive to explore the social systems behind resultant bank morphologies: what groups of people were involved in which stages of the process? Here, I have argued that uprooting was a generative process for bank construction that could have small-scale decentralized social units (e.g. households) as their main proponents and actors (Arnoldussen and van der Linden 2017: 566), regardless of any coordination in terms of use-rights and use-ways at higher social levels (cf. Johnston 2005; Yates 2007: 135–36 for nuanced views). In simpler terms: whereas the overall layout (orientation and extent of Celtic fields boundaries) may have been initially planned by leadership or community councils above the household level (cf. Johnston 2005, 3; 17; Nielsen et  al. 2017: 406), the formalization of such ­boundaries into banks that ever so slowly grew in height, was the results of individual households that routinely performed the same tasks (e.g. clearing field weeds) across the full extent of the aggregate Celtic field complex. However, to distill (changes in) long-term tenure regimes from Celtic field developmental trajectories, a detailed chronology is required. Presently, our chronological resolution and numbers of dated locations per aggregate Celtic field-compex are still too low to identify directions and modes of expansion, nucleation, or contraction within Celtic fields spanning up to hundreds of hectares (cf. Nielsen et al. 2017: 386; 393; 404–06). The development of more fine-grained chronologies or phasing of Celtic fields is unfortunately an prerequisite to arrive at social inferences on the dynamics of tenerual regimes applied – information that is in part encoded in bank age and overall morphology (cf. Nielsen et al. 2017: 385–86).

S. Arnoldussen

As to the second topic, I feel that the interrelations between domestic activities and house-sites and Celtic field agriculture also deserve additional scientific attention. Why do we fail to identify house-sites for the (Bronze Age) start of the Celtic field systems but are Iron Age houses commonly found? Were later house-sites placed in parts of the aggregate Celtic field-complex that at those times laid waste? What can explain the awkward placement of such houses in relation to the locations of banks – particularly as shared orientation of banks and houses suggests crude contemporaneity? Lastly, whilst I have tried in the above to discuss particulars of the agronomy of the later prehistoric Celtic fields, many such details still escape us. Basic information such as alternations or cycles applied to crops grown, the duration and frequency of fallow or grassland plots or the specifics of planting, weeding and manuring (cf. Nielsen et al. 2019) all require more detailed study. A fuller and more detailed understanding of the role that Celtic field agriculture played for communities in the Low Countries between 1200  BC and AD 200 hinges on continued study of these remarkable sites using interdisciplinary approaches: pedologists, palaeobotanists, geochemists, agronomists and scholars of (pre)historic tenure are as much needed as archaeologists for this.

References Amkreutz, L. W. S. W. (2013). Persistent traditions: A long-term perspective on communities in the process of Neolithisation in the Lower Rhine Area (5500–2500 cal BC). Leiden (PhD Thesis): Sidestone press. Anker, E.  F. A. (2017). Celtic fields and where to find them. Initial location parameters and spatial development patterns identified. Groningen (unpublished MA thesis). Arnold, V. (2002). Spuren urgeschichtlicher Ackersysteme (“Celtic Fields”) in Dithmarschen, Schleswig-Holstein. In U.  Masemann (Ed.), Forschungen zur Archäologie und Geschichte in Norddeutschland. Festschrift für Wolf-Dieter Tempel zum 65. Geburtstag (pp. 443–452). Wümme: Rotenburg. Arnold, V. (2011). Celtic Fields und andere urgeschichtliche Ackersysteme in historisch alten Waldstandorten Schleswig-Holsteins aus Laserscan-Daten. Archäologisches Korrespondenzblatt, 41, 439–455. Arnoldussen, S. (2008). A living landscape. Bronze Age settlement sites in the Dutch river area (c. 2000-800 BC). Leiden (PhD thesis): Sidestone press. Arnoldussen, S. (2012). Het Celtic field te Zeijen  – Noordse veld: kleinschalige opgravingen van wallen en velden van een laat-­ prehistorisch akkersysteem (Grondsporen 16). Groningen: Groninger Instituut voor Archeologie. Arnoldussen, S. (2018). The fields that outlived the Celts. The use-­ histories of Dutch later prehistoric field systems (Celtic fields or raakakkers). Proceedings of the Prehistoric Society, 83, 303–327. Arnoldussen, S., & Bouman, A. (2011). Van de wal en het veld: nieuw archeologisch onderzoek aan Celtic fields. Paleo-aktueel, 22, 45–54.

3  Formation, Use and Chronology of Celtic Fields: New Perspectives from the Groningen Celtic Field Research Programme Arnoldussen, S., & de Vries, K. M. (2014). Of farms and fields: The Bronze Age and Iron Age settlement and Celtic field at Hijken  – Hijkerveld. Palaeohistoria, 55/56(2013/2014), 85–104. Arnoldussen, S., & Fokkens, H. (2008). Bronze Age settlement sites in the Low Countries: An overview. In S. Arnoldussen & H. Fokkens (Eds.), Bronze Age settlement sites in the Low Countries (pp. 17–40). Oxford: Oxbow books. Arnoldussen, S., & Jansen, R. (2010). Iron Age habitation patterns on the Southern and Northern Dutch Pleistocene coversand soils: The process of settlement nucleation. In M.  Meyer (Ed.), Haus  – Gehöft  – Weiler  – Dorf. Siedlungen der vorrömischen Eisenzeit im nördlichen Mitteleuropa. Internationale Tagung an der Freien Universität Berlin vom 20.-23. März 2009 (Berliner Archäologische Forschungen 8) (pp. 379–397). Rahden: Verlag Marie Leidorf. Arnoldussen, S., & Scheele, E.  E. (2014). De Celtic Fields van Wekerom: kleinschalige opgravingen van wallen en velden van een laat-prehistorisch raatakkersysteem (Grondsporen 18). Groningen: Groningen Institute of Archaeology (GIA). Arnoldussen, S., & Smit, L. (2017). Zat er nog wat in? Archeobotanisch onderzoek aan een Drentse raatakker. In G.  Aalbersberg, M.  Schepers, & S.  E. Boersma (Eds.), Paleo-Palfenier: Met Rita van Egypte tot Ezinge (pp. 53–63). Groningen: Barkhuis. Arnoldussen, S., & van der Linden, M. (2017). Palaeo-ecological and archaeological analysis of two Dutch Celtic fields (Zeijen-Noordse Veld and Wekerom-Lunteren): Solving the puzzle of local Celtic field bank formation. Vegetation History and Archaeobotany, 26(6), 551–570. Arnoldussen, S., & van Os, B. J. H. (2015). The potential of lacquer-­ peel soil profiles for palaeo-geochemical analysis using XRF analysis. Catena, 128, 16–30. Arnoldussen, S. & Vries, K. M. de 2017. A plan in place? Celtic field habitation at Westeinde (prov. Drenthe, The Netherlands), Lunula Archaeologia protohistorica, 25: 79–89. Arnoldussen, S., Scheele, E. E., & de Kort, J. W. (2014). Briefrapport verkennend veldonderzoek Celtic field Herkenbosch – ‘De Meinweg’ (Grondsporen 19). Groningen: Groningen Institute of Archaeology (GIA). Arnoldussen, S., Schepers, M., & Maurer, A. (2016). Celtic fields in Brabant: Wat stuifmeel en zaden kunnen vertellen. Paleo-aktueel, 27, 23–31. Bakels, C. C. (1997). The beginnings of manuring in Western Europe. Antiquity, 71, 442–445. Bakker, J.  A., Brandt, R.  W., van Geel, B., Jansma, M.  J., Kuijper, W.  J., van Mensch, P.  J. A., Pals, J.  P., & IJzereef, G.  F. (1977). Hoogkarspel-Watertoren: Towards a reconstruction of ecology and archaeology of an agrarian settlement of 1000 BC. In B. L. van Beek, R.  W. Brandt, & W.  Groenman-van Waateringe (Eds.), Ex Horreo (Cingula 4) (pp. 187–225). Amsterdam: Universiteit van Amsterdam. Becker, C.  J. (1971). Früheisenzeitliche Dörfer bei Grøntoft, Westjütland, 3, Vorbericht: Die Ausgrabungen 1967-68. Acta Archaeologica, 42, 79–110. Behre, K.-E. (2000). Frühe Ackersysteme, Düngemethoden und die Entstehung der Nordwestdeutschen Heiden. Archäologisches Korrespondenzblatt, 30(1), 135–151. Behre, K.-E. (2008). Landschaftsgeschichte Norddeutschlands: Umwelt und Siedlung von der Steinzeit bis zur Gegenwart. Neumünster: Wachholtz Verlag. Blume, H.-P., Hoffmann, D., & Reichstein, J. (1987). Ein bronzezeitlicher Auftragsboden in Rantum. In G.  Kossack, O.  Harck, & J.  Reichstein (Eds.), Archsum auf Sylt, 2: Landwirtschaft und Umwelt in vor- und frühgeschichtlicher Zeit (Römisch-Germanische Forschungen 44) (pp. 33–45). Mainz am Rhein: La Baume. Bourgeois, Q. P. J. (2013). Monuments on the horizon. The formation of the barrow landscape throughout the 3rd and 2nd millenium BC. Leiden (PhD thesis): Sidestone press.

39

Bourgeois, I., Cheretté, B., & Bourgeois, J. (2003). Bronze Age and Iron Age settlements in Belgium. An overview. In J.  Bourgeois, I. Bourgeois, & B. Cheretté (Eds.), Bronze Age and Iron Age communities in North-Western Europe (pp. 175–190). Brussel: Koninklijke Vlaamse Academie van België voor wetenschappen en kunsten. Bradley, R. (1978). Prehistoric field systems in Britain and North-West Europe  – A review of some recent work. World Archaeology, 9, 265–280. Brinkkemper, O., & van Wijngaarden-Bakker, L. H. (2005). All-round farming. Food production in the Bronze Age and the Iron Age. In L. P. Louwe Kooijmans, P. W. van den Broeke, H. Fokkens, & A. L. van Gijn (Eds.), The prehistory of the Netherlands (II) (pp.  491– 512). Amsterdam: Bert Bakker. Brongers, J. A. (1976). Air photography and Celtic field research in the Netherlands (Nederlandse Oudheden 6). Groningen (PhD. thesis). Bursch, F.  C. (1951). De westerse mens ontdekt zijn wereld. Leiden: Sijthoff. Butler, J.  J. (1969). Nederland in de Bronstijd (Fibulareeks 31). Bussum: Fibula-Van Dishoeck. Cappers, R. T. J., & Raemaekers, D. C. M. (2008). Cereal cultivation at Swifterbant? Current Anthropology, 49(3), 385–402. Casparie, W. (1976). Palynological investigation of the Celtic field near Vaassen, The Netherlands. In J. A. Brongers (Ed.), Air photography and celtic field research in The Netherlands (Nederlandse Oudheden 6) (pp. 105–113). Amersfoort: Rijksdienst voor het Oudheidkundig Bodemonderzoek. Clason, A.  T. (1999). What’s new in the Bronze Age ? The livestock and Gamebag of the Bronze Age farmers in the Western and Central Netherlands. In H.  Sarfatij, W.  J. H.  Verwers, & P.  J. Woltering (Eds.), Discussion with the Past. Archaeological studies presented to W.A. van Es (pp. 35–40). Zwolle: SPA/ROB. Clemmensen, N.  C. (2010). DHM  – den digitale højemodel  – reliefkortet. In V.  Nielsen (Ed.), Oldtidsagre i Danmark. Sjælland, Møn og Lolland-Falster (Jysk Arkæologisk Selskabs skrifter 71) (pp. 25–28). Aarhus: Aarhus Universitetsforlag. Curwen, E., & Curwen, E. C. (1923). Sussex Lynchets and their associated field-ways. Sussex Archaeological Collections, 64, 1–65. de Hingh, A. E. (2000). Food production and food procurement in the Bronze Age and Early Iron Age (2000–500 BC) (Archaeological Studies Leiden University 7). Leiden (PhD thesis). De Laet, S. J. (1958). The Low Countries (Ancient peoples and places 5). New York: Frederick A. Praeger, Inc. Drenth, E., Brinkkemper, O., & Lauwerier, R. C. G. M. (2008). Single Grave Culture Settlements in the Netherlands: the state of affairs anno 2006. In W. Dörfler & J. Müller (Eds.), Umwelt– Wirtschaft – Siedlungen im dritten vorchristlichen Jahrtausend Mitteleuropas und Südskandinaviens (Offa-Bücher N.F. 84) (pp.  149–181). Neumünster: Wachholtz. English, J. (2013). Pattern and Progress: Field systems of the second and early first millennia BC in Southern Britain (BAR British series 587). Oxford: Archaeopress. EU. (1985). Soil map of the European Communities at 1:1 000 000. The Commission of the European Communities, directorate general for agriculture, coordination of agricultural research. Luxembourg: The Office for Official Publications of the European Communities. Fens, R. L., & Arnoldussen, S. (Eds.). (2015). Een opgraving nabij de hunebedden D36 en D37 te Valthe (Grondsporen 23). Groningen: Groningen Institute of Archaeology (GIA). Fokkens, H. (1998). Drowned Landscape. The Occupation of the Western Part of the Frisian-Drenthian Plateau, 4400 BC – AD 500. Assen (PhD thesis): Van Gorcum. Fokkens, H. (2008). The temporality of culture changes. In H. Fokkens, B.  J. Coles, A.  L. van Gijn, J.  P. Kleijne, H.  H. Ponjee, & C.  G. Slappendel (Eds.), Between foraging and farming. An extended broad spectrum of papers presented to Leendert Louwe Kooijmans

40 (Analecta Praehistorica Leidensia 40) (pp. 15–24). Leiden: Faculty of Archaeology. Fontijn, D.  R. (2003). Sacrificial Landscapes. Cultural biographies of persons, objects and ‘natural’ places in the Bronze Age of the Southern Netherlands, c. 2300-600BC (Analecta Praehistorica Leidensia 33/34). Leiden (PhD Thesis): Leiden University. Ford, S., Bowden, M., Mees, G., & Gaffney, V. (1988). The date of the ‘Celtic’ field systems on the Berkshire Downs. Britannia, 19, 401–404. Fowler, P.  J. (1983). The farming of prehistoric Britain. Cambridge: Cambridge University Press. Gebhardt, H. (1976). Bodenkundliche Untersuchung der eisenzeitlichen Ackerfluren von Flögeln-Haselhörn, Kr. Wesermünde. Probleme der Küstenforschung im südlichen Nordseegebiet, 11, 91–100. Gebhardt, H. (1982). Phosphatkartierung und Bodenkundliche Geländeuntersuchungen zur Eingrenzung historischer Siedlungsund Wirtschaftsflächen der Geestinsel Flögeln, Kreis Cuxhaven. Probleme der Küstenforschung im südlichen Nordseegebiet, 14, 1–10. Gerritsen, F. (2003). Local identities. Landscape and community in the late prehistoric Meuse-Demer-Scheldt region (Amsterdam Archaeological Studies 9). Amsterdam: Amsterdam University Press. Groenman-van Waateringe, W. (1979). Nogle aspekter af Jernalderens agerbrug i Holland og NV Tyskland. In H.  Thrane (Ed.), Fra Jernalder til Middelalder. Beretning fra et symposium 17.-19. maj 1979 afholdt af Odense universitet (Skrifter fra Historisk institut 27) (pp. 75–86). Odense: Universitas Othiniensis. Groenman-van Waateringe, W. (1980). Weeds. In M. Ryan (Ed.), The origins of metallurgy in Atlantic Europe: Proceedings of the fifth Atlantic colloquium Dublin, 30th march to 4th April 1978 (pp. 363– 368). Dublin: Stationary office. Groenman-van Waateringe, W., & van Geel, B. (2017). Raised bed agriculture in Northwest Europe triggered by climatic change around 850 BC: A hypothesis. Environmental Archaeology, 22(2), 166–170. Guttmann, E. B., Simpson, I. A., & Davidson, D. A. (2005). Manuring practices in antiquity: A review of the evidence. In M. Brickley & D. Smith (Eds.), Fertile ground: Papers in honour of Susan Limbrey (pp. 68–76). Oxford: Oxbow Books. Harsema, O.  H. (1974). Archeologisch onderzoek op het Hijkerveld, gemeente Beilen. Voorlopig bericht over de opgravingen 1969 en 1970. Nieuwe Drentse Volksalmanak, 91, 161–168. Harsema, O. H. (1980). Drents boerenleven van de bronstijd tot de middeleeuwen (Museumfonds Publicatie 6). Assen: Drents Museum. Harsema, O.  H. (1991). De bronstijd-bewoning op het Hijkerveld bij Hijken. In H.  Fokkens & N.  Roymans (Eds.), Nederzettingen uit de bronstijd en de vroege ijzertijd in de Lage Landen (Nederlandse Archeologische Rapporten 13) (pp. 21–29). Amersfoort: Rijksdienst voor het Oudheidkundig Bodemonderzoek. Hatt, G. (1931). Prehistoric fields in Jylland. Acta Archaeologica, 2, 117–158. Hatt, G. (1949). Oldtidsagre (Arkæologisk-Kunsthistoriske Skrifter 2.1). København: Det Kongelige Danske Videnskabernes Selskab. Heidinga, H. A. (1988). Climate and plaggen soils. In W. Groenman-­ van Waateringe & M. Robinson (Eds.), Man-made soils (BAR international series 410) (pp. 21–33). Oxford: Archaeopress. Hesse, R. (2010). LiDAR-derived local relief models (LRM) – A new tool for archaeological prospection. Archaeological Prospection, 17, 67–72. Huisman, D.  J., & Raemaekers, D.  C. M. (2014). Systematic cultivation of the Swifterbant wetlands (The Netherlands ): Evidence from Neolithic tillage marks (c. 4300–4000 cal. BC). Journal of Archaeological Science, 49, 572–584. Jager, S. W. (1985). A prehistoric route and ancient cart-tracks in the gemeente of Anloo (Province of Drenthe). Palaeohistoria, 27, 185–245.

S. Arnoldussen Jankuhn, H. (1958). Ackerfluren der Eisenzeit und ihre Bedeutung für die frühe Wirtschaftsgeschichte. Berichte der Römisch-­ Germanischen Kommission, 37/38(1956/1957), 148–214. Jensen, J. (1982). The prehistory of Denmark. London: Methuen & Co.. Johnston, R. (2005). Pattern without a plan: Rethinking the Bronze Age coaxial field systems on Dartmoor, South-West England. Oxford Journal of Archaeology, 24(1), 1–21. Johnston, R. (2013). Bronze Age fields and land division. In H. Fokkens & A. F. Harding (Eds.), Oxford handbook of the European Bronze Age (pp. 311–327). Oxford: Oxford University Press. Klamm, M. (1993). Aufbau und Entstehung eisenzeitlicher Ackerfluren (“Celtic Fields”). I Stand der Forschung (Göttinger Bodenkundliche Berichte 102). Göttingen: Brunk Meyer. Kooi, P.  B. (1997). Het project Peelo. het onderzoek van het Kleuvenveld (1983, 1984), het burchtterrein (1980) en het Nijland (1980), met enige kanttekeningen bij de resultaten van het project. Palaeohistoria, 37/38(1995/1996), 417–479. Kooi, P.  B., & de Langen, G.  J. (1987). Bewoning in de vroege ijzertijd op het Kleuvenveld te Peelo, gem. Assen. Nieuwe Drentse Volksalmanak, 104, 151–165. Kooistra, M. J., & Maas, G. J. (2008). The widespread occurrence of Celtic field systems in the central part of the Netherlands. Journal of Archaeological Science, 35, 2318–2328. Kortekaas, G.  L. G.  A. (1987). Een laat-neolithisch akkercomplex in de Oosterpoortwijk te Groningen. Groninger Oudheden, 19, 108–124. Kristensen, H.  K. (1933). Digevoldinger. Jyske Samlinger, 5(1), 234–245. Kroll, H. J. (1975). Ur- und Frühgeschichtlicher Ackerbau in Archsum auf Sylt. Eine botanische Großrestanalyse. Kiel (PhD Thesis): Christian-Albrechts Universität. Lanting, J. N., & van der Plicht, J. (2000). (1999/2000). De 14C chronologie van de Nederlandse Pre- en Protohistorie III: Neolithicum. Palaeohistoria, 41/42, 1–110. Liversage, D., Munro, M. A. R., Courty, M.-A., & Nørnberg, P. (1987). Studies of a buried Iron Age field. Acta Archaeologica, 56, 55–84. Louwe Kooijmans, L. P. (1993). The Mesolithic/Neolithic transformation in the lower Rhine Basin. In P. Bogucki (Ed.), Case studies in European prehistory (pp. 95–145). Boca Raton: CRC Press. Louwe Kooijmans, L. P. (2000). Living in the Neolithic: Habitation and landscape. In A. Ritchie (Ed.), Neolithic Orkney in its European context (McDonald institute monographs) (pp. 323–331). Cambridge: McDonald Institute, Cambridge. Løvschal, M. (2014). Emerging boundaries: Social embedment of landscape and settlement divisions in Northwestern Europe during the first millennium BC. Current Anthropology, 55(6), 725–750. Løvschal, M., & Holst, M. K. (2014). Repeating boundaries: Repertoires of landscape regulations in southern Scandinavia in the Late Bronze Age and pre-Roman Iron Age. Danish Journal of Archaeology, 3(2), 95–118. Meylemans, E., Creemers, G., De Bie, M., & Paesen, J. (2015). Revealing extensive protohistoric field systems through high resolution LIDAR data in the northern part of Belgium. Archäologisches Korrespondenzblatt, 45, 197–213. Miller, N. F., & Gleason, K. L. (1994). Fertilizer in the identification and analysis of cultivated soil. In N. F. Miller & K. L. Gleason (Eds.), The archaeology of garden and field (pp.  25–43). Philadephia: University of Philadephia Press. Müller, S. (1911). Vendsyssel-Studier. Aarbøger for nordisk oldkyndighed og historie, 1911, 255–259. Nielsen, N. H. (2007). RIM 9140 – Øster Lem Hede. Sønder Lem sogn, Hind herred, Ringkøbing Amt, Sted-nr: 180412 Sb-nr: 163. Aarhus (unpublished interim report): Aarhus Universitet. Nielsen, N.  H., & Dalsgaard, K. (2017). Dynamics of Celtic fields  – A geoarchaeological investigation of Øster Lem Hede,Western Jutland, Denmark. Geoarchaeology, 32, 414–434.

3  Formation, Use and Chronology of Celtic Fields: New Perspectives from the Groningen Celtic Field Research Programme Nielsen, N.  H., Kähler Holst, M., Gadd, A.  C., & Kähler Holst, K. (2017). The layout and internal development of Celtic fields: Structural and relative chronological analyses of three Danish field systems. European Journal of Archaeology, 21(3), 385–410. Nielsen, N.  H., Kristiansen, S.  M., Ljungberg, T., Enevold, R., & Løvschal, M. (2019). Low and variable: Manuring intensity in Danish Celtic fields. Journal of Archaeological Science: Reports, 27, 12pp. Odgaard, B.  V. (1985). A pollen analytical investigation of a Bronze Age and pre-Roman Iron Age soil profile from Grøntoft, Western Jutland. Journal of Danish Archaeology, 4(1), 121–128. Reichmann, C. (1982). Ein Bronzezeitliches Gehöft bei Telgte, kr. Wahrendorf. Archäologisches Korrespondenzblatt, 12, 437–449. Roessingh, W. (2014). Huisplattegronden uit de midden-bronstijd van Bovenkarspel-Het Valkje, West-Friesland. In E.  M. Theunissen & S. Arnoldussen (Eds.), Metaaltijden 1. Bijdragen in de studie van de metaaltijden (pp. 67–80). Leiden: Sidestone Press. Scheele, E.  E., & Arnoldussen, S. (2012). De wallen van Wekerom (Gl.): een midden-Nederlands Celtic field onderzocht. Paleo-­ aktueel, 23, 23–32. Schrijver, P. (2011). Niet enkel ruiten en raten. Een analyse van de culturele informatie die besloten ligt in de ruimtelijke morfologie van Nederlandse Celtic fields. Groningen (unpublished Ba thesis). Smith, G. (1996). Archaeology and environment of a Bronze Age cairn and prehistoric and Romano-British field system at Chysauster, Gulval, near Penzance, Cornwall. Proceedings of the Prehistoric Society, 62, 167–219. Sørensen, P. H. (1984). Luftarkeologie: Hvilke oplysninger giver soil marks? In H. Thrane (Ed.), Danks Landbrug i oldtid og middelalder (Skrifter fra Historisk institut 32) (pp. 30–40). Odense. Spek, T. (2004). Het Drentse esdorpenlandschap: een historisch-­ geografische studie. Amersfoort: Rijskdienst voor het Oudheidkundig Bodemonderzoek. Spek, T., Groenman-van Waateringe, W., Kooistra, M., & Bakker, L. (2003). Formation and land-use history of Celtic fields in North-­ West Europe  – An interdisciplinary case study at Zeijen, The Netherlands. European Journal of Archaeology, 6(2), 141–173. Spek, T., Snoek, M., van der Sanden, W.  A. B., Kosian, M., van der Heijden, F., Theunissen, L., Nijenhuis, M., Vroon, H., & Greving, K. (Eds.). (2009). Archeologische waardering van Celtic fields in Drenthe (Rapportage Archeologische Monumentenzorg 141). Amersfoort: Rijksdienst voor het Cultureel Erfgoed. Taayke, E. (1996). Die einheimische Keramik der nördlichen Niederlande, 600 v.Chr. bis 300 n.Chr. Groningen (PhD thesis): Casparie Heerhugowaard bv. Theunissen, E. M. (1999). Midden-bronstijdsamenlevingen in het zuiden van de Lage Landen. Een evaluatie van het begrip ‘Hilversum-­ cultuur’. Leiden (PhD thesis): Leiden University. Toms, H. (1911). The problem of ancient cultivations. Antiquary, 47, 411–417. Van Amerongen, Y. F. (2016). Wild West Frisia: The role of domestic and wild resource exploitation in Bronze Age subsistence. Leiden (PhD thesis): Sidestone press. van Giffen, A. E. (1918). Begin van een onderzoek van ‘de zogenaamde voormalige Romeinsche legerplaats en aangelegen grafheuvelveld te Zeijen. Nieuwe Drentse Volksalmanak, 36, 135–175. van Giffen, A.  E. (1936). Een versterking naar Romeinsch patroon op het Noordscheveld bij Zeijen, Gem. Norg. Nieuwe Drentsche Volksalmanak, 54, 123–129. van Giffen, A.  E. (1940a). De zgn. Romeinsche of heidensche legerplaats op het Noordsche Veld bij Zeijen, gem. Vries. Nieuwe Drentsche Volksalmanak, 58, 200–202. van Giffen, A. E. (1940b). De zgn. heidensche legerplaats te Zuidveld bij Sellingen, gem. Vlachtwedde. Verslag Museum van Oudheden Groningen, 1939, 86–93.

41

van Giffen, A.  E. (1944). Opgravingen in Drenthe tot 1941. In J.  Poortman (Ed.), Drente, een handboek voor het kennen van het Drentsche leven in voorbije eeuwen (pp.  393–568). Boom: Meppel. van Giffen, A.  E. (1947). Oudheidkundige perspectieven, in het bijzonder ten aanzien van de Vaderlandsche Prae- en Protohistorie. In H. E. van Gelder, P. Glazema, G. A. Bontekoe, H. Halbertsma, & W. Glasbergen (Eds.), Een kwart eeuw oudheidkundig bodemonderzoek in Nederland. Gedenkboek A.E. van Giffen, Directeur van het Biologisch-Archaeologisch Instituut der Rijksuniversiteit te Groningen 1922  – 17 juni  – 1947 (pp.  497–544). Meppel: Boom & Zoon. van Giffen, A. E. (1949). Het Noordse Veld bij Zeijen, Gemeente Vries. Opgravingen in 1944. Nieuwe Drentse Volksalmanak, 67, 93–148. van Giffen, A. E. (1950). De nederzettingsoverblijfselen in het Bolleveen en de versterking, de zgn. “legerplaats” aan het Witteveen op het Noordse veld, beide bij Zeijen, gem. Zeijen. Nieuwe Drentsche Volksalmanak, 68, 89–122. van Gijn, A.  L., & Kooijmans, L.  P. L. (2005). Early- and middle Neolithic: Introduction. In L. P. Louwe Kooijmans, P. W. van den Broeke, H. Fokkens, & A. L. van Gijn (Eds.), The prehistory of the Netherlands (pp. 203–218). Amsterdam: Bert Bakker. Van Ginkel, E. (1987). Opgraven of bewaren: 25 jaar archeologische monumentenzorg in Nederland 1961-1986. Leiden/Amersfoort: Zorn/R.O.B. van Klaveren, H.  W. (1986). Celtic Field en nederzettingssporen bij de Vijfsprong, gemeente Ede, Leiden University (Unpublished MA thesis). Vanmontfort, B., Langohr, R., Marinova, E., Nicosia, C., & Van Impe, L. (Eds.). (2015). Een archeologische evaluatie en waardering van Celtic Fields in het Kolisbos (Neerpelt, provincie Limburg) (EPA-­ Rapport 50). Leuven: Eenheid Prehistorische Archeologie. Verlaeckt, K. (1996). Between river and barrow: A reappraisal of Bronze Age metalwork found in the Province of East-Flanders (Belgium) (British archaeological reports, international series 632). Oxford: Tempus Reparatum. Wallinga, J., & Versendaal, A. J. (2013a). NCL-7213 quartz luminescence dating report Wekerom – Lunteren. Wageningen: Netherlands Centre for Luminescence Dating. Wallinga, J., & Versendaal, A.  J. (2013b). NCL-7212 Quartz Luminescence Dating Report Zeijen – Noordse veld. Wageningen: Netherlands Centre for Luminescence Dating. Waterbolk, H.  T. (1949). Palynologisch onderzoek van grafheuvels en oud akkerland op het Noordse Veld bij Zeijen. Nieuwe Drentse Volksalmanak, 67, 126–147. Waterbolk, H.  T. (1977). Opgravingen rond het Witteveen op het Noordse Veld bij Zeijen, gem. Vries (1949-1953). Nieuwe Drentse Volksalmanak, 94, 5(177)–31(203). Waterbolk, H.  T. (2009). Getimmerd verleden. Sporen van voor- en vroeghistorische houtbouw op de zand- en kleigronden tussen Eems en IJssel. Groningen: Barkhuis. Wortelboer, R. (2014). Van Bronstijd tot moderne tijd. Een reis door de tijd aan de hand van lokale hoogteverschillen in de Hoorneboegse Heide bij Hilversum. Archeologica Naerdincklant, 2014(2), 13–18. Yates, D. T. (2007). Land, power and prestige Bronze Age field systems in southern England. Oxford: Oxbow. Zimmermann, W.  H. (1976). Die eisenzeitlichen Ackerfluren  – Typ “Celtic Field” – von Flögeln-Haselhörn, Kr. Wesermünde. Probleme der Küstenforschung im südlichen Nordseegebiet, 11, 79–90. Zimmermann, W. H. (1995). Ackerbau in ur- und Frühgeschichtlicher Zeit auf der Geest und in der Marsch. In H.-E.  Dannenberg & H.-J.  Schulze (Eds.), Geschichte des Landes zwischen Elbe und Weser (Schriftenreihe des Landschaftsverbandes der ehemaligen Herzogtümer Bremen und Verden 7) (pp. 289–315). Stade.

4

A Large-Area Prehistoric Cultural Landscape in the Sachsenwald Forest Near Hamburg Volker Arnold

Abstract

The Sachsenwald Forest 25 km east of Hamburg covers an area of more than 62 km2, and is situated on an Old Drift plateau which is only intersected by a few valleys. Apart from some peripheral forestation, it has been woodland at least since the thirteenth century, probably since the Migration Period. Around 750 burial mounds are registered, partly dating from the late Bronze Age and early Iron Age. LiDAR data resulted in the detection of large Celtic field areas. With omission of heavily disturbed or altered forest areas, 50 km2 of woodland remains, which contain Celtic fields encompassing an area of over 15 km2. Most of these correspond with interspersed or surrounding burial mounds, and represent a large unfragmented and sophisticated cultural landscape. The plots vary greatly in form and size and seem to disappear towards the edges of the system so that it can be assumed that the originally cultivated area was even larger. A first attempt has been made to analyse the structures, their chronological sequence and their position in the context of newly discovered Celtic fields in Northern Germany. Keywords

LiDAR · Sachsenwald · Celtic fields · Chronology

4.1

Introduction

The Sachsenwald Forest is situated some 25  km east of Hamburg and, excluding the separate Rülau Forest, comprises a vast forested area of c.  62  km2 (Fig.  4.1). Except from a few zones, mainly along its south-western boundary, which were reforested in the course of the nineteenth cenV. Arnold (*) Berliner Strasse 61, Heide/Holstein, Deutschland

tury, it has probably been a forest since the Migration Period, and certainly since the thirteenth century. Up to recent times, archaeological surveys have revealed some Stone Age long barrows and about 750 burial mounds (Gütschow 2014). A review of the LiDAR data obtained around 2006 led to the discovery of large Celtic fields areas. On the other hand it has become clear that some of the previously registered burial mounds need to be revised (Erlenkeuser 2009, Arnold 2011: 447–49, figure 11–15; 2012a: 44–45, figure 22–24; 2012b: 4; Gütschow 2014).1 Particularly many of the long barrows without chambers registered at that time turned out to be Celtic field embankments. The Sachsenwald Forest covers a sandy and sandy-loamy Old Drift plane, from 16 m above sea level along the Bille River up to 74 m in its south-east to the east of Dassendorf village. Celtic fields areas are at altitudes from 23 to 60 m. Most of the Sachsenwald Forest appears to be flat and sparsely structured. Only the valleys of the Bille River, the Schwarze Au creek and the lower parts of some secondary valleys are more prominent (Fig. 4.2). Some sub-areas of the Sachsenwald can be identified when searching for traces of prehistoric cultivation in the surface relief. In addition to recent disturbances (railways, motorways, roads and forest road constructions, building areas, large-scale soil disturbance due to forestry activities), mainly the creation of medieval ridge and furrow-type cropland has  destroyed existing structures or at least covered them. If all such disturbed areas are subtracted from the total forest plain as shown in Fig. 4.2, about 50 km2 of forest area remains. Around 15 km2 of this is covered with Celtic fields, which amounts to a little more than 31%. The largest coherent subarea of Celtic fields has a surface area of a little more than 5 km2. This very favourable data set allows some conclusions about the history, context and former use of the structures, even though nearly no archaeological fieldwork has yet been carried out (and will only be possible in small-­ Surprisingly Gütschow (2014) does not mention Celtic fields.

1 

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 S. Arnoldussen et al. (eds.), Europe’s Early Fieldscapes, Themes in Contemporary Archaeology, https://doi.org/10.1007/978-3-030-71652-3_4

43

44

areas in a forest). Except from the northern zone, the Celtic fields correlate with burial mounds which occur scattered among them or which lie in their vicinity, often in groups. The Celtic fields are by no means homogenous, but rather are highly variable in form and size. They range from large, highly irregular plots to smaller, more or less square or rectangular plots. More pronounced main embankments can often be seen. In many cases, the plots become less pronounced towards the outer edges of the system and gradually disappear, suggesting that the primary cultivated area was much larger than is recognizable from the data.

Fig. 4.1  Situation of the Sachsenwald Forest and two comparative locations in Schleswig-Holstein, Germany. The border between Old Drift (left) and Young Drift (right) morainic landscape is marked Fig. 4.2  Sachsenwald, Kr. Hgtm. Lauenburg. Rectangles with numbers mark the detailed areas discussed later. LiDAR surface data © LVermGeo SH

V. Arnold

4.2

Sachsenwald Celtic Fields

In the following, seven exemplary sub-areas are discussed which are representative for the different characteristics of the Sachsenwald Celtic fields. In all cases the original laser raster of 1 × 1 m was reduced to 2 × 2 m to minimise data noise. The display combines contour-shading with a coloured mean deviation. Sections are exaggerated by a factor of 10 and averaged from 11 equidistant parallel sections with an overall width of 10 m. At the Brahmhorst site (Fig. 4.2, no. 1 and Fig. 4.3) there are well-preserved Celtic fields in its centre. Their boundaries are more or less regularly aligned. The structures gradually disappear towards the west and the south. Some prominent burial mounds appear at Celtic field embankments. Other minor elevations, mostly to the south-west of the road, are caused by forestry. The section does not clarify all of the cut boundary embankments. A group of small burial mounds of early Iron Age type is situated to the south-­ east of the section at the opposite side of the valley, where a Jastorf cremation urn was found (Kersten 1951: 404, LA 248–266). In the eastern part of the Buschhege site (Fig. 4.2, no. 2 and Fig. 4.4) there are Celtic fields intermingled with numerous burial mounds which are mostly located within the plots without any distinct arrangement. The section cuts a prehistoric road which is accompanied by embankments which are formed in the same manner as the boundary embankments of the plots (see also Erlenkeuser 2009: 18). A counterpart of it is the longer, but narrower, road through the well-known Øster Lem Hede Celtic fields in western Jutland (Hatt 1949:

4  A Large-Area Prehistoric Cultural Landscape in the Sachsenwald Forest Near Hamburg

45

Fig. 4.3 Sachsenwald-­ Brahmhorst (see Fig. 4.2, no. 1 for location) with section N-S. Celtic fields area, cut by a railway (top right) and a motorway (bottom left). The terrain section is exaggerated by a factor of 10 and averaged from 11 parallel sections with an overall width of 10 m, as with all following sections. w: embankments. Data © LVermGeo SH

Fig. 4.4 Sachsenwald-­ Buschhege East (see Fig. 4.2, no. 2 for location) with section W-E. The crescent shaped embankment situated to the east of the prehistoric road and north of the section is of recent age, as is the railway at the northern edge. Data © LVermGeo SH

92–108, pl. X, Nielsen and Dalsgaard 2017). Charcoal from the eastern road wall showed a C14 age of 91 BC–75 AD. Traces of the prehistoric road do not continue

either to the north or to the south, and also not outside the image. More strongly pronounced embankments incorporate larger plots which often have highly irregular forms. Their

46

further divisions remain largely unclear. Occasionally small faint parallel strips can be recognized, reminiscent of ridge and furrow fields. Around the Kupferberg site there are Celtic fields with reasonably regularly aligned plots (Fig.  4.2, no. 3 and Fig.  4.5). Some burial mounds lie predominantly at the boundaries of plots. The Celtic field boundary embankments are more prominent in two sub-areas and gradually disappear towards the edges, so that they are only adumbrated in many parts of the illustration. Two neighbouring burial mounds in the southern part caused the construction of a plot boundary embankment transverse to the surrounding system. The mounds are therefore now connected by an embankment. The section crosses seven long rectangular plots which are slightly more pronounced than the remainder. The troughlike form of these plots is recognizable, but is not very pronounced. Well-preserved Celtic fields are situated in the western part of the Langenbruch site (Fig. 4.2, no. 4 and Fig. 4.6). Partly they consist of irregular large plots, and partly of smaller and more regular plots. Some main embankments can be seen, predominantly aligned from north to south. Parts of the Celtic field system, including in the south-­eastern part of the section, are superimposed by ridge and furrow fields. In places, small dunes have developed in the ridge and furrow areas. In many of the larger plots faint structures may Fig. 4.5 Sachsenwald-­ Kupferberg (see Fig. 4.2, no. 3 for location) with section N-S. Data © LVermGeo SH

V. Arnold

be recognised, similar to ridge and furrow fields. A distinct ridge and furrow field near the eastern edge to the south of the large burial mound is enclosed by an obviously more recent embankment, which cannot be distinguished from the embankments which were common in the eighteenth and nineteenth centuries. It remains ambiguous whether the group of points north-northwest of the burial mounds are grave mounds of an early Iron Age type (‘Buckelgräber’). The section shows no trough-like form of the plot which is cut. In the northern part of the Schadenbek site there are inhomogeneous Celtic fields, but almost no burial mounds (Fig. 4.2, no. 5 and Fig. 4.7). To the north of the motorway, large structures or main embankments are recognizable, which are generally divided by much less pronounced embankments. To the south of the motorway, a remarkably regular Celtic field system can be seen in the middle and eastern parts of the illustration. This system seems to be constructed systematically, using more small rectangular or quadratic plots, yet gradually disappears towards the edges. Two sections do not clarify all of the cut embankments, because they are not very pronounced. A 14C-measurement of charcoal resulted in a problematic young date of 892–1015 AD. Some more pronounced main embankments in the Hasenbekshorst site (Fig.  4.2, no. 6 and Fig.  4.8) enclose areas which seem to be secondarily divided into long rectan-

4  A Large-Area Prehistoric Cultural Landscape in the Sachsenwald Forest Near Hamburg Fig. 4.6 Sachsenwald-­ Langenbruch West (see Fig. 4.2, no. 4 for location) with section NW-SE. Data © LVermGeo SH

Fig. 4.7 Sachsenwald-­ Schadenbek North (see Fig. 4.2, no. 5 for location) with two sections. Celtic fields, cut by a motorway. The structures around the scalebar are the result of recent forest cultivation. Small rings below 20 m diameter are remains of recent charcoal kilns. Data © LVermGeo SH

47

48

V. Arnold

Fig. 4.8 Sachsenwald-­ Hasenbekshorst (see Fig. 4.2, no. 6 for location) with two sections N-S. Celtic fields, cut by a railway. Numbers: enclosed sub-­areas in ha. Z: Zone of soil removal in order to raise the neighbouring bank. Data © LVermGeo SH

gular plots. Their lateral embankments are barely recognizable. The central cut shows no trough-like form of the plot which is cut. Instead, a small zone of material removal along the embankment seems to be visible. This is more distinct in the second north-eastern section. A zone to the east of the Schwarze Au valley is cultivated with ridge and furrow fields of medieval type, which in one place also extends to the west of the Au. Numerous presumably small burial mounds (‘Buckelgräber’) are situated in the north-western corner, while only few larger burial mounds are distributed over the remaining area.

4.3

Celtic Field Land Forms

Some main embankment enclosures are prominent at Hasenbekshorst (Fig. 4.8). These enclose areas of 6.7 ha and 13.5 ha, the larger one to the west being divided into a northern area of 7 ha and a southern area of approximately 6.5 ha by another main embankment. Another main embankment enclosure covers 9.7 ha, which however covers poorly visible ridge and furrow fields. These areas approximately correspond to a medieval ‘Oxgang’ (‘Hufe’; cf. Hägermann and Hedwig 1991), a variable surface measure of cropland sufficient for the self-supply of a farming family. Similar areas can be found in some isolated and less fragmented Celtic fields in the terminal moraine debris to the north of Schleswig (see below). Celtic fields with noticeably long regular rectangular plots can be found at the Stangenteichshorst site (Fig.  4.2, no. 8 and Fig.  4.9). The section shows very slightly pro-

nounced boundary embankments and accordingly the plots show faint trough-like forms. The structures disappear towards the edges and are interspersed with several larger burial mounds. In the south-western corner of the illustration more irregular Celtic fields occur, which continue to the west towards the Buschhege site (Fig. 4.4). With regard to the long and narrow rectangular plots, cross-ploughing as is commonly supposed for the cultivation of shorter rectangular or square Celtic field plots seems implausible. Ploughing in a transverse direction would cause too frequent turning of the team harnessed for ploughing. It should be examined if a turning mouldboard plough (‘Streichbrettpflug’) was used here without producing distinct ridge and furrow fields. Mouldboard ploughs were in use since the first century BC in German marshlands, and an increasing number of mouldboard ploughmarks were detected dating to the later Roman Iron Age (Larsen 2015). The embankments which mark the boundaries of the Celtic field plots are generally less pronounced in the Sachsenwald region than in many other Celtic fields, for instance in Angeln and Schwansen (Young Drift  morainic landscape, see Fig.  4.10), or Dithmarschen (Old Drift, see Fig.  4.11; Arnold 2011: 450 and figure 16, Arnold 2012b: 10) or outside of Schleswig-Holstein (for example in the north-eastern Netherlands; Arnoldussen, Chap. 3, this volume). Also, trough-like sections which are mostly recognizable in the latter examples (Figs. 4.10 and 4.11) are not – or are only slightly – developed in most Sachsenwald Celtic fields. The author believes that the size of the embankments and the extent of trough-like formations would increase with time

4  A Large-Area Prehistoric Cultural Landscape in the Sachsenwald Forest Near Hamburg

49

Fig. 4.9 Sachsenwald-­ Stangenteichshorst (see Fig. 4.2, no. 8 for location) with section W-E. Celtic Fields with long rectangular plots cut by the railway line. Data © LVermGeo SH

Fig. 4.10  Ornum Forest near Kosel, Kr. Rendsburg-­ Eckernförde, SchleswigHolstein. Fragmentary Celtic field system with distinct trough-like plots. A 14C-date from an embankment resulted in a date of 247–393 AD, which closely corresponded to a nearby excavated settlement. Data © LVermGeo SH

and intensity of use, without proof from data obtained nearby. In the author’s opinion, the (minor) erosion within the plots and the increasing filling towards the boundaries of the plots is mainly caused by the transfer of uprooted weeds and adhered soil in the course of cross-ploughing. While turning the harnessed ploughing team, weeds and soil were dropped or were removed at this point. So the existing

embankments were enlarged or embankments gradually developed, as could be demonstrated in the loess or alluvial clay areas where embankments known as ‘Ackerberge’ (field embankments; cf. Hartke 1954) developed until the twentieth century. Also a recent miniature ploughing experiment gave indications to a gradual raising of embankments caused just by crosswise ploughing (Arnold 2018: 26 fig. 8). On the

50

V. Arnold

embankments like Langenbruch-West or Hasenbekshorst cultivation of only the banks seems to be implausible. In Schleswig-Holstein there are only four minor areas where Celtic fields are preserved nearly unfragmented and are not extensively fragmented by more recent forest or cropland boundaries.2 In spite of more fragmented preservation five isolated crop plots can be postulated in the terminal moraine debris north of Schleswig in the Idstedt-Karrenberg Forest and in the eastern Falkenberg forest (Loewe 1998: 156: Idstedt LA 31 and 203: Lürschau LA 46). These cover areas of between 4 and 10 ha, which approximately corresponds to the values found in Sachsenwald-Hasenbekshorst.

4.4

Fig. 4.11  Odderade, Kr. Dithmarschen, Schleswig-Holstein. Fragmentary Celtic field  system with distinct trough-like plots in the so-called Primula Wood. A 14C date from an embankment section yielded a date of 666–770 AD, another, also from an embankment section yielded a date of 359–106 BC. Data © LVermGeo SH

other hand, especially the Sachsenwald Celtic fields provide many indications for primary raised embankments. In ­addition, some of the earth excavated elsewhere and deposited here for soil improvement may have affected the relief. Assuming this, it could take many centuries of cropland use to develop such pronounced structures. Most of the Sachsenwald Celtic fields with their considerably less pronounced structures, seem to have been in use less intensively or for a shorter period of time. New substantial investigations regarding the emergence of these embankments have recently been published (e.g. Arnoldussen and Vries 2014; Arnoldussen and van der Linden 2017; Nielsen and Dalsgaard 2017). Most recently an investigation of a Celtic field embankment in the Riesewohld (Odderade, Dithmarschen, see Figure 1) resulted in a life-span of several hundred years from two datings of carbonized hazelnut shells, reaching from late Bronze Age at the bottom to late pre-Roman Iron Age at the top (Arnold 2018: 27 fig. 5). Bank cultivation is discussed by Gebhardt (1976), Zimmermann (1976), Behre (2000) and more recently by Groenman-van Waateringe and van Geel (2017). At least in the case of irregularly shaped

Implications

Investigations of the age and duration of use of the Sachsenwald Celtic fields are lacking and urgently recommended because it cannot be assumed that all of the features are of the same age. The two previously mentioned 14C dates are by no means sufficient. Also the remaining 32 dates from Schleswig-Holstein Celtic fields (Arnold n.d.) mainly cover the period from the first millennium BC to the first millennium AD. These only provide an initial indication and in individual cases they are problematic from a methodological point of view. Finds from older burial mound excavations date from the late Bronze Age as well as from the early Iron Age (Gütschow 2014: 40). Future research is necessary to prove how long cropland usage took place in the centuries AD, and whether the long rectangular plots are the most recent prehistoric structures. For this reason, the Sachsenwald Celtic fields would provide an outstanding data basis. The Sachsenwald offers a well preserved, large scale and differentiated cultural landscape which is largely unfragmented. This can be seen and experienced only on closer examination. Another large and well preserved cultural landscape with Celtic fields in an ancient forest can only be found in Germany in the Jasmund forest on the island of Rügen.3 However, this only comprises about one third of the Sachsenwald area. The Jasmund Celtic fields are much more coherent than those in the Sachsenwald, although the landscape relief is much more pronounced. They cover at least 12 km2, which is nearly 60% of the forest. Another large Celtic field accumulation was actually found at the Isar cobIn the Außelbek Forest near Ülsby in Anglia (Loewe 1998: 390–392, Ülsby LA 24; Klamm 1995; Menke 1995; Arnold 2011: 440–41 and figure 1-3; 2012a: 34–36 and figure 2-4; 2012b: 7), in the Schierenwald forest in western Holstein, in the Hahnheide Forest north-east of Sachsenwald (Arnold 2012b: 8) and in reforested heathland near Hartenholm in the middle of Holstein (Erlenkeuser 2009: 19; Arnold 2011: 10). 3  Information kindly provided by Fred Ruchhöft, Naturmuseum Goldberg; Becker (2017). 2 

4  A Large-Area Prehistoric Cultural Landscape in the Sachsenwald Forest Near Hamburg

ble terraces immediately south of Munich covering about 9  km2 and cut by a Roman road from about 50  BC (Arnold 2020). These cultural landscapes provide impressive evidence of a second agricultural revolution, which compares to the ‘Neolithic Revolution’ in terms of its importance: the change from permanent pasture to the indoor housing of animals (at least in the winter), which is reflected in changed house layouts, combined with a change from nomadic agriculture to more sedentary agriculture. The latter seems only achievable by continuous enhancement of soil fertility by applying suitable soil material on the one hand and by spreading stable manure mixed with house litter on the other. Nevertheless this second agricultural revolution seems to be enhanced by climatic deterioration in the first half of the last millennium BC. It appears to be confined to parts of northern and western Europe as well as to the western, northern and eastern parts of central Europe. It certainly did not proceed simultaneously and in the same manner everywhere. Later agricultural changes such as the overall introduction of mould-board ploughs or the rise of the so-called Plaggenesch cultivation with ‘perennial’ rye growing are of minor importance on the whole, although they caused further alterations to the landscape. Only the nineteenth century agricultural innovations have had a comparable or greater importance in cultural history. Acknowledgements  The author wishes to thank Gerard Parkinson for his kind assistance with the translation.

References Arnold, V. (n.d.) 14C-dated celtic fields in the Dithmarschen moraine land/Eight ‘celtic fields’ of Schleswig-Holstein: LIDAR data seen in 3D by redgreen glasses / Germany’s largest celtic fields area at the Sachsenwald and the Guttau and Schierenwald celtic fields. Conference poster. https://www.academia.edu/35183815/14C-­ dated_celtic_fields_in_the_Dithmarschen_moraine_land_Eight_ celtic_fields_of_Schleswig-­Holstein_LIDAR_data_seen_in_3D_ by_redgreen_glasses_Germanys_largest_celtic_fields_area_at_ the_Sachsenwald_and_the_Guttau_and_Schierenwald_celtic_fields Arnold, V. (2011). Celtic Fields und andere urgeschichtliche Ackersysteme in historisch alten Waldstandorten Schleswig-Holsteins aus Laserscan-Daten. Archäologisches Korrespondenzblatt, 41, 439–455. Arnold, V. (2012a). Laserscandaten als Prospektionshilfe zur punktuellen Untersuchung von urgeschichtlichen Ackersystemen insbesondere des Jungmoränengebietes in Schleswig-Holstein. In A. Stobbe & U.  Tegtmeier (Eds.), Verzweigungen, eine Würdigung für A.  J. Kalis und J. Meurers-Balke (Frankfurter Archäologische Schriften 18) (pp. 33–47). Bonn: Habelt. Arnold, V. (2012b). Schleswig-Holstein neu entdecken. Spuren der Erdgeschichte und Archäologie in Laserscandaten. Alle Poster

51

der Ausstellung des Museums für Archäologie und Ökologie Dithmarschen, Albersdorf. Heide: Volker Arnold. Arnold, V. (2018). Spuren urgeschichtlicher Siedlungsaktivitäten und Ackerwirtschaft in Dithmarscher Bauernwäldern nach Laserscandaten. Dithmarschen, 3, 21–30. Arnold, V. (2020). Älter als die Römer? Bisher übersehene Spuren einstiger Beackerung unter Bayerns Wäldern. In J. Hamberger (Ed.), Festschrift zur Ruhestandsversetzung von LWF-Präsident Olaf Schmidt (Forstliche Forschungsberichte München 218) (pp. 8–18). Freising: Zentrum Wald-Forst-Holz Weihenstephan. Arnoldussen, S., & van der Linden, M. (2017). Palaeo-ecological and archaeological analysis of two Dutch Celtic fields (Zeijen-Noordse Veld and Wekerom-Lunteren): Solving the puzzle of local Celtic field bank formation. Vegetation History and Archaeobotany, 26(6), 551–570. Arnoldussen, S. & Vries, K.M. de (2014). Of farms and fields: The Bronze Age and Iron Age settlement and Celtic field at HijkenHijkerveld. Palaeohistoria, 55/56: 85–104. Becker, M. (2017). GIS gestützte Identifizierung von Celtic Fields und weiteren Kulturlandschaftsrelikten unter Wald im Nationalpark Jasmund mit hilfe von Laserscanning-Aufnahmen und weiteren Geodaten. Greifswald: Unpublished bachelor thesis. Behre, K.-E. (2000). Frühe Ackersysteme, Düngemethoden und die Entstehung der nordwestdeutschen Heiden. Archäologisches Korrespondenzblatt, 30, 135–151. Erlenkeuser, H. (2009). Landschaft neu entdecken, Airborne-­ Laserscandaten in der archäologischen Denkmalpflege. Archäologische Nachrichten aus Schleswig-Holstein, 15, 17–19. Gebhardt, H. (1976). Bodenkundliche Untersuchung der eisenzeitlichen Ackerfluren von Flögeln-Haselhörn, Kr. Wesermünde. Probleme der Küstenforschung im südlichen Nordseegebiet, 11, 91–100. Groenman-van Waateringe, W., & van Geel, B. (2017). Raised bed agriculture in Northwest Europe triggered by climate change around 850 BC: A hypothesis. Environmental Archaeology, 22(2), 166–170. Gütschow, E. (2014). Gut beschützt: Die Grabhügel des Sachsenwaldes. Archäologische Nachrichten aus Schleswig-Holstein, 20, 36–40. Hatt, G. (1949). Oldtidsagre (Arkæologisk-Kunsthistoriske Skrifter 2.1). København: Det Kongelige Danske Videnskabernes Selskab. Hägermann, D., & Hedwig, A. (1991). Hüfe. In N.  Angermann & R.-H.  Bautier (Eds.), Lexikon des Mittelalters. Bd.5 (p.  154). München: Deutscher Taschenbuch Verl. Hartke, W. (1954). A propos des crêtes de labour (“Ackerberge”) comme indice de l’ancienneté des limites agraires. Bulletin de l’Association de Géographes Français, 245/246, 196–198. Klamm, M. (1995). Archäologische und bodenkundliche Untersuchungen der eisenzeitlichen Ackerflur im Gehege Ausselbek bei Ülsby, Kreis Schleswig-Flensburg. Offa, 52, 29–43. Kersten, K. (1951). Vorgeschichte des Kreises Herzogtum Lauenburg (Die vor und frühgeschichtlichen Denkmäler und Funde in Schleswig-Holstein II). Neumünster: Wachholz. Larsen, L.  A. (2015). Muldfjælsplovens tidlige historie fra yngre romersk jernalder til middelalder. Kuml, Årbog for Jysk Arkæologisk Selskab, 64, 165–200. Loewe, G. (1998). Kreis Schleswig. Archäologische Denkmäler Schleswig-Holsteins VIII. Neumünster: Wachholz. Menke, B. (1995). Vegetations- und Bodenentwicklung im Bereich der celtic fields im Gehege Ausselbek bei Ülsby, Kreis Schleswig-­ Flensburg. Offa, 52, 7–28. Nielsen, N.  H., & Dalsgaard, K. (2017). Dynamics of Celtic fields  – A geoarchaeological investigation of Øster Lem Hede, Western Jutland, Denmark. Geoarchaeology, 56, 1–21. Zimmermann, W.  H. (1976). Die eisenzeitlichen Ackerfluren  – Typ »Celtic Field« – von Flögeln-Haselhörn, Kr. Wesermünde, Probleme der Küstenforschung im südlichen Nordseegebiet, 11: 79–100.

5

A Sheep’s Eye View: Land Division, Livestock and People in Later Prehistoric Somerset, UK Clare Randall

Abstract

Keywords

Fields and field systems in later prehistoric British archaeology have generally been discussed in relation to territory or land tenure. They are also frequently assumed to relate purely to arable agriculture. Alongside this, we also tend not to situate livestock animals within landscapes. Increasingly, morphological features of fields can be identified as having use in animal handling. Consequently field system morphology, and changes to layouts over time, enable their re-­examination in relation to pastoral and arable husbandry (and the interplay between them), and consideration as to why differing approaches may have been adopted within the same landscape at different times. This provides models which, focussing on pastoral husbandry, are potentially applicable to a range of places and periods. The second and first millennia BC bounded landscapes surrounding the hillfort at Cadbury Castle, Somerset, UK, reveal an intimate relationship between the occupiers of the hillfort, sites in its environs, livestock, and the landscape. A series of different forms of land division and organisation from the earlier Bronze Age onwards can be compared with both faunal and plant macro-­fossil data from within that landscape. Different forms of layout appear to reflect different types of strategy and approach in later prehistoric farming. During the second and first millennia BC changes can be observed between different forms of highly extensive pastoral farming and closely integrated and intensive systems. The explanation would seem to be more social than practical in origin, but discerning this is reliant on large scale field survey, and integration of multiple strands of information.

Fields · Livestock · Prehistoric · Landscape · Husbandry

C. Randall (*) Department of Archaeology and Anthropology, Bournemouth University, Poole, Dorset, UK e-mail: [email protected]

5.1

Introduction

In southern Britain, in common with neighbouring areas of northern Europe, the landscape has been structured and demarcated in a variety of ways from the later prehistoric period onwards. British fields have tended to be discussed as evidence of the relationship between social control and subsistence, and in relation to issues of tenure and territory (e.g. Barker and Gamble 1985; Breen 2008; Chadwick 2008a, b; Evans 2008; Fowler 1978; Johnston 2000, 2001; Kitchen 2001; Yates 2007), although consideration of Bronze Age divisions on Dartmoor in particular has included the interplay between tenure and land use (Fleming 1978a, b, 1983, 1985, 1988, 1998; Wickstead 2008). As Løvschal (2014: 728) notes, perceived and socially accepted boundaries are different from physical, dividing structures, which adds nuance to the issue. In addition, whilst we should not deny physical demarcation of spaces as relating to concepts of ownership and land holding, and see them purely in terms of functionality, we should reconsider their role in the lived experience of those who constructed and used them. Fields have often tended to be assumed to be the arena of production of arable crops, with the management of livestock being a lesser issue, although new interpretations are beginning to be adopted (e.g. English 2013; Evans 2009; Masefield 2015; Pryor 1996). The necessity of pulling together more diverse lines of enquiry has been proposed in order to enhance understanding of people’s everyday lives (Chadwick 2008a, b: 205). This paper seeks to consider how land division might be structured around what went on within spaces, by relating the evidence from faunal assemblages to the evidence of the fields. This paper discusses the needs and preferences of the

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 S. Arnoldussen et al. (eds.), Europe’s Early Fieldscapes, Themes in Contemporary Archaeology, https://doi.org/10.1007/978-3-030-71652-3_5

53

54

C. Randall

animals, the parameters offered by grassland and its management, and identification of features within field systems which relate to the practicalities of livestock handling and management. A selection of various likely physical manifestations of different approaches to husbandry methods provides a series of possible models. These are applied to one example in south-­west Britain, where contemporary faunal assemblages and landscape could be compared. The environs of Cadbury Castle, Somerset, UK, is one location where a series of field systems have been examined in some detail (Tabor 2008), and there is both a large faunal dataset (Randall 2010a) and archaeobotanical data (De Carle 2014) available for comparison. A series of bounded systems commenced some time after c. 1700  BC with subsequent landscape reorganisation throughout the second and first millennia BC and into the first millennium AD. The morphology of these various arrangements has commonality with examples of similar dates across the south-­west of Britain and beyond, but the South Cadbury environs offer an opportunity to examine a sequence spanning more than a millennium. The topography, elevations, and soils are variable within a relatively small area, and offer a range of agricultural possibilities. Faunal remains from Cadbury Castle, a hill which attracted activity from the early Neolithic including the creation of a hillfort in the first millennium BC, and sites examined as part of the South Cadbury Environs Project, date from the second half of the second millennium BC to the first millennium AD.

5.2

Background

Aside from considerations of ‘design’ (such as co-­axial or irregular examples), fields possess an inherent utility. Consequently, whilst acknowledging the different likely combinations of topographic and environmental influences, we should rehabilitate the practicality of fields. We may then more readily identify the aspects of organisation and use which inform more nuanced understandings of their social role. Likewise, we need to challenge our assumptions about livestock animals. Most people in the developed world have little experience of cattle, sheep, goats and pigs, which means that most archaeologists encounter them generally as datasets which may say more about how they were consumed and deposited than how they lived. Zooarchaeology focuses on attempting to identify the mechanisms and strategies utilised to achieve ‘products’ – primary products (meat) and secondary products (e.g. milk, hides, horn). However, animals had their own lives, their own relationships with landscape and people. We should therefore accept livestock as social actors alongside people (Chadwick 2008a, b: 205; Giles 2012). By situating animals within the landscapes they inhabited, and have been altered for them, we may gain some insight into the broader motivations of the people who shaped those

places. Whilst boundaries may have a range of socio-­­ conceptual meanings to the people who constructed them, it is their physicality with which livestock animals interact. Also in regulating the movement of livestock, contingent changes occur to the movement and perceptions of space by people. Consequently, where boundaries were organised with the needs of livestock husbandry in mind, changes in organisation over time are of interest where the inherent physical parameters remained largely unaltered. As Brück (2000) indicates, alterations in subsistence practice were not necessarily the cause of change but contingent upon it. A reflexive relationship exists between choices of the organisation of the pastoral landscape and other social factors. We should also consider matters of intensity and scale. We should be careful of our use of the term ‘intensification’, as it is often used in place of what might more properly be called ‘expansion’ (Morrison 1996: 587). In simple terms ‘intensive’ approaches seek to produce as great, or greater, production from a smaller area (Netting 1974: 39; Tiffen et al. 1994: 29), whilst ‘extensive’ livestock production relies on fewer animals per hectare. The definition needs to take into account the changes to labour input, depending on the type of activity undertaken (cf. Leach 1999; Wickstead 2008: 74). Extensive cultivation and animal husbandry are complementary as both involve larger areas of land, infrequent cropping or less intense management (Bogaard 2005: 179), whilst intensive systems are likely to be highly controlled, complex and integrated with arable cropping. The drivers for these two differing approaches (and a range of variations in between) are likely to be highly specific to time and place, but relate to the inherent carrying capacity of land, land availability, availability and distribution of labour, and a range of socially derived preferences and factors. However, understanding the scale of an individual system, and how it relates to other neighbouring systems, depends on being able to propose an entire ‘unit’, with the problems of identification which that entails (cf. Evans 2009: 61). It should also be pointed out that, contrary to some suggestions relating to the shape and regularity of systems being associated with chronology or seasonality of use (Fowler 1983:128–9; Harvey 1980:44), even ‘planned’ systems may involve a long period of development, alteration and accretion (cf. Chadwick 2008b; Barnatt 2008: 50), and the explanation for form is as likely to be rooted in their use as in any chronological assumptions or aesthetic principles.

5.3

Livestock in the Landscape – Understanding the Needs of the Animals

Despite the perceptions engendered in the modern developed world which treats livestock as a commodity product, livestock are not, and probably never have been, walking larders;

5  A Sheep’s Eye View: Land Division, Livestock and People in Later Prehistoric Somerset, UK

neither are they an assemblage of potential handbags, knife handles or clothing. The practice of zooarchaeology encourages us to view the data in terms of products and consumption, but we also need to consider livestock as living entities, occupying and moving through space. In order to understand their full role in ancient societies, we should think of them as not purely a means to an end, a product, but a living creature with a range of needs and preferences. People in the past, in common with modern farmers, would have had a detailed knowledge of those needs, and will have responded to them. Therefore, the choices surrounding husbandry are likely to have reflected far more nuanced considerations than our somewhat crude approaches to identifying the aims of livestock production. Also, by understanding the practical necessities of husbandry we gain a greater appreciation of the expertise required to create management systems. This provokes greater consideration of the range of choices and decisions open to people. If we make the assumption that daily care for animals, and inhabiting landscapes alongside them, was an almost universal experience for people in the past, this potentially changes the way in which we understand how expertise in animal husbandry may have been viewed, and how the associated tasks may have had a profound role in structuring society. On the surface of it, livestock have fairly straightforward needs, primarily food, water and, dependent on location and climate, a degree of shelter. However, different species are better suited to particular environments. Whilst past domesticated animals probably differed in some respects to modern livestock, and data on analogous rare breeds, such as Soay sheep (Ryder 1981: 184) are limited, the fundamental aspects of biology and behaviour are likely to have been similar (Randall 2010a: 29–30). Tolerances for climate, reproductive cycles, herd structures, animal behaviour, availability of fodder, and other aspects of management, would immediately impose an annual seasonal and daily rhythm on communities. Access to water is crucial but differs by species. Cattle consume more water than pigs, goats or sheep, which obtain most of what they require from vegetation (Goodwin 1973; Goodwin 1979; Salmon 1981; Reynolds 1987). This affects the numbers and location in which they can be kept, or how often they need to be moved. Livestock need a constant supply of fodder year round, with particular needs at certain times of year (e.g. tupping, lambing and calving (Fraser 1947: 40; Goodwin 1979: 84–85; Henderson 1944: 82–83; Smith Thomas 2005: 125)). Foddering and grazing are also different for each species. Cattle and sheep are primarily grazers but will tolerate a range of other feed, whilst goats thrive on a range of browse. Pigs can graze, but as omnivores will also turn over the soil in search of tubers, invertebrates and small vertebrates, will forage for nuts and seeds, and can be fed on waste (Allen 1910; Goodwin 1973; Manolson et al. 1988; Masseti 2007; Papachristou et  al. 2005; Ross 1989;

55

Seymour 2003; Smith Thomas 2005). This in turn affects the location of suitable grazing. In addition, different herbivores have different methods of eating. Cattle eat longer grass than sheep, wrapping their tongues around it to rip it up; sheep nibble, so prefer shorter, refined species (van Wijngaarden-­­ Bakker 1998: 176; Blake 1990: 18), which manifests as grazing preferences (Putfarken et al. 2008). Goats can assist in reclamation of scrub, grazing more marginal land whilst leaving the grass (Bryce and Wagenaar 1985: 94; Halliday and Halliday 1988, Papachristou et al. 2005: 146; Morand-­­ Fehr 2005: 27). Topography, exposure, soil types and rainfall provide broad parameters which influence plant communities, choice of livestock species, and matching these to the available grazing. This seems to be reflected in regional patterns in British prehistoric faunal assemblages (e.g. Hambleton 1999; Serjeantson 2011). In addition, a wide range of livestock diseases are related to grazing and other husbanding practice; parasites and some infections are directly related to re-­infection from pasture due to the life cycle of the organisms as well as damp ground conditions (Fraser 1988: 213– 4; Salmon 1981: 146–7; Manolson et al. 1988: 187–9). Some conditions can be managed with simple treatments, for example foot cleaning and fleece clipping, which are assisted by pens and stalls. On the other hand, close housing of stock can have implications for respiratory illness and infectious diseases (e.g. Bryce and Wagenaar 1985; Defra 2002; Fitzherbert 1534; Olmos et  al. 2009; Schütz et  al. 2009; Street 1942; Wall et al. 1993). These considerations immediately bring about decisions regarding the stocking density for any given area, suitability or possibility of winter grazing, foddering with crop by-­ products or on stubbles, or the development of other strategies, including which landscapes were exploited or not. Choices are not necessarily environmentally determined, but are framed by the parameters which the physical landscape provides. The range of possible choices might include high degrees of mobility and utilisation of highly extensive approaches (smaller numbers of animals grazing over a wide area), utilisation of crop residues or fodder storage, or selective seasonal culling to conserve fodder. All have possible repercussions for other economic activities (arable cropping, exploitation of wild resources), organisation of labour and mobility (mobile and transhumant lifestyles), settlement, tenure, territory and society.

5.4

 anaging Grazing – Understanding M the Contribution of the Land

Management of grassland is an area of considerable expertise for modern livestock farmers and crucial for animal health and yields. However, we have not tended to consider it as an area of technical expertise for ancient farmers. We

56

cannot assume that prehistoric farmers did not understand the nutritional and health related differences between different forages and pastures. Pasture is not just ‘grass’ but a complex mixture of grasses and herbaceous species, each with its own range of ecological preferences. Some assist fertility by fixing nitrogen or concentrating minerals and nutrients, whilst others are unpalatable and invasive (Bryce and Wagenaar 1985: 12). Acid soils and wet land favours grass species which are nutritionally poor (Houghton Brown and Powell-­Smith 1984: 167; Seymour 2003: 166). It is possible to manage some of these problems by relatively simple technologies such as burning dead grass, repeated mowing, ploughing up, cutting or pulling weeds before they seed, encouraging drainage, or marling with chalk if available (Archer 1988: 25). However, strict management of grazing itself can manage invasive weeds, and prevents development of coarse and unpalatable clumps (Bryce and Wagenaar 1985: 13–15). Spring grazing encourages a wider range of palatable species to produce more foliage when grazed (Gibson et  al. 1987: 42). It can be used to target specific weed types (de Bruijn and Bork 2006; Harrison and Bardgett 2008: 208). Grassland varies in its components not only by climatic region and within specific topographical/geological landscapes, but highly locally, over a matter of meters. Grazing herbivores interact with the sward depending on palatability and accessibility, selectively grazing out their preferred plants, as well being influenced by elevation, exposure and distance from water (Frame 1992: 176–7; Fraser 1947: 103; Ganscopp and Bohnert 2009: 110–111; Searle et  al. 2007, Searle and Shipley 2008: 128–9). On the other hand, over-­­ grazing can lead to reduction of grazable perennial plants, poaching, soil erosion and run-­off (Jones and Dowling 2005). This is particularly exacerbated if pigs are grazed (Allen 1910: 118; Goodwin 1973; Hodgkinson et  al. 2009: 222, 225). Due to the differing grazing preferences of the species, and consequential varying degrees of localised nitrogen input, different combinations of livestock can cause highly variable changes to the sward (Esmail 1991; Harrison and Bardgett 2008: 205; Prins and Fritz 2008: 192), which may require management. Grazing practice also has an important role in managing parasite problems, as some parasite lifecycles occur partly on the ground (Salmon 1981: 146–7). Animals lose weight or do not thrive with a parasite load, and the loss of condition is quite visible, enabling an association between a particular area of ground and poor stock condition. Management can include regular rotation of grazing, allowing animals onto fresh, rested pasture (Fraser 1947: 157–8; Salmon 1981: 49, 146–7; Walker et  al. 1989), mixed grazing of cattle and sheep, or alternate grazing between species, as different parasites attack different livestock. Reducing the stocking density reduces the likelihood of re-­infection (Alderson 1988:

C. Randall

136). Liver fluke in particular are persistent, but prefer wet locations, so restriction to seasonal use might be sufficient to avoid re-­infestation, whilst saltmarsh is beneficial in deterring both liver fluke and foot rot (Salmon 1981: 146–7; Stallibrass 1996: 59). All of these issues were open to observation by ancient farmers, and the development of expertise to respond to it, including arranging the spaces used for husbandry to facilitate a style of management.

5.5

Fields as Infrastructure

A selection of ethnographic and modern developed world approaches to livestock husbandry were used in considering the use of physical changes to the landscape for animal management.1 Structures are generally associated with close handling (assisting with calving/lambing, managing injury or disease, slaughter), grazing management and movement through spaces. In crude terms, most animals in highly extensive and mobile pastoralist and transhumant systems require a near constant human presence at close quarters (Chang and Tourtellotte 1993; Kelly 2000; Lewthwaite 1981). Using physical infrastructure (fences, hedges, ditches, walls) can prevent animals from straying, provide water, and manage grazing with reduced human presence; labour is invested in the structures themselves, and therefore activity can be more episodic (Randall 2010a; cf. Wickstead 2008: 74). However, management of grazing in a more restricted space (assuming a concern for future resource availability), would have required different types of intervention, either in directing livestock to particular areas, and providing an infrastructure to manage their grazing behaviour. Fields by themselves do not indicate intensification (Wickstead 2008) but reflect changes in the deployment of human effort. Optimisation of labour was an underpinning feature of eighteenth century developments in farming in Britain which led to radical reorganisations of fields into the nineteenth century (Taylor 1975: 119, 140), but need not be the case in all examples. Fields obviously had a role in arable cropping. Indeed, maximisation of grazing quality, fodder and arable crops for human consumption leads to greater demands for manure, so intensive cultivation is complementary with intensive livestock management (Bogaard 2005: 179). However, specific elements indicate animal handling, but have little benefit in arable cropping. If land boundaries, fields, gates and pens are seen as equipment (integrated husbandry practices to provide

1  E.g. Albarella et al. 2007; Amorosi et al. 1998; Chang 1993; Chang and Tourtellotte 1993; Doyle 1870; Fitzherbert 1534; Frame 1992; Fraser 1947; Goodwin 1979; Haas et  al. 1998; Halstead 1996; Henderson 1944; Kelly 2000; Lake 1989; Lewthwaite 1981; Rasmussen 1993; Street 1942; Tani 2002.

5  A Sheep’s Eye View: Land Division, Livestock and People in Later Prehistoric Somerset, UK

a system), we can regard them as a form of technological choice (cf. Pfaffenberger 1988, 1992) or as Løvschal (2014: 727) describes it, spatial technology. Fields can be agricultural ‘fixed equipment’. Components of field systems can be understood as practical devices directly related to parts of the process of livestock production, tailored to the localised requirements of livestock, provision of water, management of grazing and fodder, handling and shelter. So, how to identify such ‘infrastructure’? Firstly, it needs to adequately and reliably control the movement of animals, primarily to contain them. Consequently stock-­proof boundaries are a necessity (Alderson 1988: 117; Bryce and Wagenaar 1985: 95; Hart 1994: 38; Smith Thomas 2005: 14). These might take a number of forms but essentially need to be of a certain height/width and robustness, to control access and egress. Whilst the features we excavate may not fulfil those criteria we should be wary of dismissing them. The original height of stone walls, and the very existence of hedges and fences can be very difficult to discern, and lack of this has been used to interpret boundaries in conceptual terms (e.g. Johnston 2005: 219). The most frequently excavated later prehistoric boundaries in southern Britain are ditched, although earthwork banks are also frequent. Some ditches would in some places be crucial for drainage, for both arable and grassland as well as provide an effective barrier to livestock, especially when flooded. Where this was not the case, ditch digging would also provide material for a well drained bank needed for the establishment of hedge plants (Pryor 2006: 84; Maclean 2000: 47). Narrowly separated and concentric ditches may well, as suggested for Fengate, East Anglia, represent embanked hedge lines (Evans 2009: 245). Traditional modern laid hedges are stock proof when maintained with simple equipment. Suitable hedge species readily establish from cuttings (Seymour 2003: 138), and these are frequently identified in preserved wood assemblages. Convincing evidence for the presence of hedging associated with later prehistoric fields and enclosures is now being identified from a variety of south-­western British sites, including Hillfarrance, Somerset (Jones 2006). Woodland management would provide wood for suitable post and rail fencing and wattle hurdles. There is evidence for fences in various contexts across the south-­west of Britain throughout the Bronze Age and Iron Age, including fence lines and wood from possible post-­and-­rail fences at Wotter Common, Shaugh Moor (Orme and Morgan 1982; Smith et al. 1981: 269). Lynchets on hillsides are a by-­product of ploughing and a useful indicator of arable cultivation, (Fowler and Evans 1967: 298) but fences or hedges would render them stock-­proof, and assist and explain the locations of soil accumulation. ‘Infrastructure’ also includes stock handling equipment. These components can be regarded as diagnostic to a system designed around the needs of livestock. It has been argued

57

that rectangular spaces relate to cross-­ploughing with an ard (Harvey 1980: 42–3), but cultivation marks have also been identified in non-­rectilinear spaces (Fowler 1981: 19). Whilst gates positioned in the corners of fields make no practical difference to arable cropping, in rectilinear fields the sides of the land parcel create a funnelling effect useful in stock handling (Pryor 2006: 101). The provision of angled field sides, facilitate handlers to work effectively with the visual parameters and flight response of the animals to move them with ease (Grandin 1980; Grandin and Deesing 2008; Randall 2010a: 108). Gates in field corners are therefore a compelling indicator of pastoral use, albeit with the caveat that curvilinear systems can also be utilised effectively, as evidenced by modern work on low stress livestock handling (Grandin and Deesing 2008; Weller 1982: 67). Races, sorting gates, and close handling pens (Fig. 5.1) have a recognisable morphology (Randall 2010a: 105–112) and can be related to the physical size and numbers of the primary livestock species (Grandin and Deesing 2008; Pryor 1996, 2006). Where buildings occur in association, their use for animal housing should be considered (Randall 2010a: 109–111). The arrangement of enclosed spaces can also directly relate to use. Grazing management presupposes that a series of parcels will be available, and animals easily transferred from one space to the next. In modern British livestock farming this is referred to as paddock grazing, with ‘creep’ grazing a variant which allows lambs first access by use of variable sized gates/barriers (Goodwin 1979: 97–8, 173–4, 191–98). In recent years the method has been reintroduced in large fields as ‘strip’ grazing which controls sequential access to a fodder crop across a field by use of a moveable barrier. In modern systems this is generally an electric fence, but is achievable using wood hurdles. Fields also provide the possibility of ‘aftermath’ grazing of stubbles after cropping (Forbes 1998: 30) and reintroduction of highly valuable nutrients (Bakels 1997; Barker 1985: 51; Guttmann 2005). Sheep in particular are a very efficient converter of plant matter to dung (Lewthwaite 1981: 61). Whilst livestock can utilise spaces used for arable cropping without stock proof boundaries, via tethering, this is only suitable for management of small herds. Cattle are often

D

A B

E C

F

Fig. 5.1  Commonly seen stock handling features. A  =  corner gate B = funnel entrance C = track D pen E = race F = sorting/drafting gate

58

tethered individually on cultivated ground in modern Greece, and goats in Britain and northern Europe are generally tethered due to their capacity to eat through fences and hedges (Bryce and Wagenaar 1985; Goodwin 1973). In both cases these animals are kept in small numbers. The variable application of medieval southern English strip farming provided a number of integrated functional attributes, including regulation of grazing and cropping (Campbell 1981: 113–4; Campbell and Godoy 1992: 100), with spatial arrangements complemented by social practices. The relationship of livestock husbandry and arable cropping can be potentially complex, given the possibilities of deliberate production of cut fodder (e.g. hay) for seasonally housed animals, or ‘zero grazing’ systems (which involves keeping animals separate from the land which produces the fodder, and in the past might apply particularly to pigs), which reduce waste (Salmon 1981: 51). Neither is the distinction between crops for food, fodder or fuel straightforward, with the potential for arable crops to be food for people, fuel, or animal feed dependent on crop, plant part, quality, available quantity, and time of year (e.g. Jones 1998: 96). With regard to nutrients, we should also consider the relationship of the ‘inside’ and the ‘outside’ of a system. Historically in Britain, and into the modern period, moors, commons, woodland grazing, wetlands and flood meadows have been important (Poulsen 1997: 119). In the modern period, the integration of these elements is best exemplified by upland hill farms which utilise a combination of ‘inbye’ land (smaller areas of better quality lower lying land capable of an arable crop), ‘outfields’ (providing managed grassland), and the ‘common’ grazing where there is generally free rein at low stocking densities over a wide but defined area. Allowing animals access to wider areas not only provides additional grazing at times when fields might be being used for arable crops, but also potentially enables a transfer of nutrients from ‘outside’ to ‘inside’ especially where livestock are penned overnight on arable stubbles and grazed ‘outside’ during the day. This introduces not only the possibility of a number of daily, weekly and seasonal movements between the enclosed and unenclosed space but has implications for the diurnal or seasonal deployment of labour. Where movement of animals around and beyond a system of land parcels is frequent, the appropriate infrastructure is likely to be designed for it, not only via tracks and droves allowing access to the land parcels themselves, but also in the form of funnel-­shaped entrances into the field system to facilitate the seasonal, periodic or even daily, movement of stock (Randall 2010a: 108). These types of arrangements have been recognised in various places, and were discussed with respect to the Dartmoor Reaves (Wickstead 2008). In Yorkshire the later Bronze Age/Early Iron Age linear earthworks defined areas of high pasture, with an interest in regulating movement between it and land at lower elevations, which contin-

C. Randall

ued as a seasonal practice into the recent past (Giles 2012: 56–57). Related to the daily or seasonal rhythm of tasks of livestock care, foddering and movement, is the relationship of settlement to fields. Housing for humans gives us an indication of the demands of labour by the system employed.

5.6

Model Systems

The considerations outlined above can be distilled down to a range of model arrangements. These were prompted by a variety of modern examples from both traditional and more industrialised systems and from historical, largely European approaches. A highly simplified series of hypothetical systems for livestock management arising from these observations is shown in Fig. 5.2. Isolated structures in an unbounded landscape (Fig. 5.2a) provide stock-­handling facilities. The extensive cattle herding practiced by the semi-­ nomadic Masai, relies upon enclosures with integrated pens, paddocks and houses (e.g. Shahack-­Gross et  al. 2004). Transhumant cattle and sheep herders in Sardinia and the Greek Pindos mountains utilised small enclosures for milking and handling livestock, which incorporate huts and pens (Mientjes 2004; Chang 1993; Chang and Tourtellotte 1993), whilst small stock pens for sheep ‘stells’ were frequently used in the Sottish Highlands (Weller 1982), and small stone structures within enclosures used for lambing across the southern Mediterranean and Near East (Tani 2002). The general absence of physical barriers, but provision of potential places for close handling of livestock suggests an open terrain and an extensive, mobile system of livestock management. Large parcels (Fig. 5.2b) might encompass a range of terrain and ecological niches. This may include some stock-­ handling features and  – whilst facilitating more controlled grazing and potential reduction in day to day human input – still represent an extensive system, most likely focussed on grassland management. Large terrain-­oblivious areas which

A

C

B

D

Fig. 5.2  Hypothetical systems

5  A Sheep’s Eye View: Land Division, Livestock and People in Later Prehistoric Somerset, UK

might include livestock handling features cater to the range of needs outlined above, but within the individual parcel. The nineteenth century Enclosure Movement in southern Britain generally involved the creation of field boundaries, in some cases rearranging the landscape, and sometimes where none had previously existed or creating larger plots. This was prompted by changes in land holding, but produced more coherent units of land which provided arable production and/ or pastoral husbandry, dependent on location and conditions (Grigg 1982), and reduced labour input whilst increasing rotational grazing and greater foddering potential in one parcel. Clustered isolated blocks (Fig. 5.2c) of smaller, similarly sized land parcels more suitable to arable production, with stock handling features, suggest a more tightly controlled livestock management system, which might utilise folding of animals on arable stubbles or rotational grazing. The southern British nineteenth century field systems often have internally coherent arrangements of land parcels and stock handling features (e.g. Doyle 1870), and enable an integrated operation as a coherent unit, addressing a range of more specialised production goals. The temporary seasonal ­deployment of infrastructure can be seen in the use of temporary lambing pens and penning on arable stubbles used in lowland central southern England in the nineteenth and twentieth centuries (Street 1942, 42). Hierarchical spaces with races, runs, sorting gates and pens are common in modern industrialised systems (e.g. Grandin and Deesing 2008). Extensive modern hill farmers in Britain utilise a combination of ‘inbye’ land, ‘outfields’ and ‘common’ grazing which is characterised by use of the enclosed land for arable in the summer and overwintering stock (Fraser 1947, 84). A similar approach was utilised in the twentieth century by Mennonite farms in Manitoba, Canada (de Garis de Lisle 1982). Associated settlement, adjacent or within the system and greater evidence for arable crops suggest a more integrated arable-­pastoral system, with the scale of the operation indicated by the size of the overall unit. Tracks and funnel entrances providing for movement of stock into a wider unbounded landscape indicate use of unenclosed grazing, and dependent on terrain, potential for seasonal changes in land use. This extensive grazing would require stock to be accompanied, or ownership and control negotiated in other ways, and organised around the timing of arable production. Larger contiguous blocks of bounded land (Fig.  5.2d) may result from the aggregation of smaller units or exhibit coherence or planning. This on a larger scale may include a range of parcel sizes suitable for a variety of purposes, suited to the terrain or soils, stock-­handling features, and buildings. Where the complexity demands tracks and droves interior to the system to facilitate movement of people and animals within and between its elements, closely managed rotational

59

grazing and close integration with arable cropping is suggested, but on a greater overall scale. These generalised models do not reflect the ways in which field systems would respond to local topography, pre-­­existing anthropogenic landmarks and other places of significance. We never see the entire picture, and certainly not in sharp chronological focus. No system that lasted any duration is likely to have been static, but adapted, developed, extended, partly abandoned and reworked.

5.7

 pplying the Models – Cadbury A Castle and its Environs, Somerset, UK

In southern Britain, later prehistoric field systems can be argued to be relatively well understood, although often their full extent, range of boundary forms (ditches, fences hedges or walls), or chronology has not been fully defined. A limited number offer the opportunity to distinguish a sequence of field systems of different forms within the same landscape throughout the Bronze Age and Iron Age. This is desirable as it addresses the issue of the morphology of a system being dictated by the form of the land itself (its geology, aspect, elevation and topography) or by a ‘design’ preference related to chronology. In addition, it is rare to have this information in combination with data on the contemporary animal populations which likely inhabited these field systems. Some lack, or have limited faunal data such as the mid second millennium BC fields at Bestwall, Dorset (Ladle and Woodward 2009), Mucking, Essex (Done 2016: 201; Rajkovača 2016: 433) or Yarnton, Oxfordshire (Mulville and Robinson 2016; Mulville et  al. 2011; Pelling 2011). Others have a shorter chronology, such as the predominantly Bronze Age landscapes of Fengate, East Anglia (Evans 2009). In other cases, investigation has occurred in great detail, but over more limited areas such as Mucking (Evans et  al. 2016) or Yarnton (Hey et al. 2011, 2016). Cadbury Castle, Somerset, is one location where a large area has been systematically examined, and there is both a large faunal dataset and archaeobotanical data is available for comparison for most periods (Fig. 5.3). The site is a multivallate Iron Age hillfort, with underlying Neolithic and later Bronze Age activity, and subsequent Roman, post-­ Roman and medieval reuse. The hillfort was excavated between 1966–70 and 1973 by Leslie Alcock (1967, 1968a, b, 1970, 1971, 1972, 1980, 1995), and analysis of the first millennium BC material was carried out in the 1990s (Barrett et  al. 2000). Re-­examination of the Neolithic material was recently completed (Tabor and Randall 2018). The South Cadbury Environs Project (SCEP) was founded in 1992 to carry out a study of the surrounding prehistoric landscape, and examined a number of sample areas within an 8 km by

60

C. Randall

Fig. 5.3  Cadbury Castle location, and sites mentioned in the text Somerset

South Cadbury N

4

South 3 Cadbury

2 1

7 8

6

5 1. Weston Bampfylde 2. Milsoms Corner 3. Homeground 4. Woolston 5. Hicknoll Slait 6. Crissells Green 7. The Moor 8. Parsonage Farm 9. Sigwells 10. Sheep Slait

9

10

8 km square around the hillfort. Using geophysical survey, systematic plough zone sampling, test pits and area excavation, SCEP identified a variety of previously unknown occupation, field systems, boundaries, and settlements dating from the Early Neolithic to the medieval period (Tabor 2002, 2004a; b; Tabor 2008; Tabor and Randall 2018). Sites were both closely adjacent to and contemporary with the various phases of activity on the hill. A series of field systems, interspersed with phases of unbounded landscape, commenced from around 1700  BC. Cadbury Castle and several of the SCEP sites produced faunal assemblages (c. 100,000 fragments from the hill, c. 20,000 from SCEP sites; Randall 2010a). These assemblages related to the Early Neolithic (Tabor and Randall 2018), the Middle and Late Bronze Age and through the Iron Age, although no earlier Bronze Age material was available. Cadbury Castle occupies a steep-­sided isolated limestone hill, situated to the north of a limestone escarpment. It is surrounded by steep valleys on Yeovil Sands with a lower lying rolling landscape of heavy clay soils to the west and north, where it more distantly overlooks the wetlands of the Somerset Levels. Locally there are a number of springs feeding streams which encircle the hill and characterise the surrounding valleys, whilst 2  km to the north, the River Cam runs east-­west and forms a natural boundary. The topography, elevations, and soils are therefore highly variable within

0

1km

a relatively small area, and offer a range of agricultural possibilities. There is a notable similarity between the faunal assemblages from sites in the environs and the hillfort, suggesting that these can be regarded as closely representative of the local domestic animal populations and their consumption and disposal (Fig. 5.4). The changing organisation of the landscape over nearly two millennia can be considered against the proposed models (Fig. 5.5). The summit of Cadbury Castle was a focus of activity during the Early Neolithic with the digging of pits and creation of possible post-­built structures. The SCEP site of Milsom’s Corner on the west flank of the hill produced evidence of near-­contemporary use, around the thirty-­sixth century BC (Tabor and Randall 2018). The faunal assemblage from the hill was characterised by a large proportion of pig remains, which fits with utilisation of a largely wooded landscape (Randall 2018), although general information on the local environment is limited. The earliest indication of a structure used to facilitate livestock husbandry in the Cadbury landscape is a small sub-­rectangular ditched enclosure with an annexed pen and possible sorting gate at Card’s Piece, Woolston (Tabor 2008: 53; Fig. 5.6). This structure probably dates to before c. 1700 BC and likely existed in an area of more open landscape, although environmental information is scant. No faunal remains were available but contemporary assemblages in the region tend to reflect cattle with few

5  A Sheep’s Eye View: Land Division, Livestock and People in Later Prehistoric Somerset, UK Fig. 5.4  Relative abundance of the main livestock species, Cadbury Castle

61

70 60

% NISP

50 40 30 20 10 0

sheep, although the samples are small and occur infrequently (Randall 2016). The scale of the structure could be consistent with cattle or sheep, and may fit with Model ‘A’ (Figs. 5.2 and 5.6), of an extensive pastoral approach with a considerable level of mobility. The first major attempt at systematic division of the landscape occurred in the later part of the first half of the second millennium BC.  The dating of these systems is inexact, based on relationships with later, better dated features (Tabor Tabor 2004a, b; Tabor 2008: 49; 52, 61), but is consistent with an inception around 1700  BC, in keeping with other examples in south-­west England. Contemporary flint and pottery was generally widely distributed in the landscape (Tabor 2008: 48–9). Linears on the high ground and lighter soils at Sigwells, Charlton Horethorne, consisted of long, straight, parallel ditches spaced c. 100 m apart, across a limestone and sandstone plateau, and were traceable for at least 1  km, following an apparently terrain oblivious course. Similar boundaries also occur in low-­lying locations with heavy soils, such as Crissells Green in the base of the South Cadbury Valley (Tabor Tabor 2004a, b), Sparkford and Weston Bampfylde, on heavy clay (Tabor 2008: 51), and at Parsonage Farm, Sutton Montis, in a low lying, wet area (Randall 2011). Together these suggest a co-­axial system aligned broadly north-­west to south-­east which covered an area of at least 3 km around the south of the Cadbury hill, covering varied topography and soils, extending to the west and north of the hill. In all cases water was easily available within the bounded areas, reducing the need for daily stock movement. None of the excavated examples of ditches have proved particularly deep or broad, and may not have been stock proof on their own. The ditches may have provided a source of bank material for hedging as suitable scrub species were present in many locations throughout later prehistory (De Carle 2014: 142). The Sigwells system (Fig. 5.6), which is the best understood section, included a possible race. Tabor (2008: 49) has suggested it could be of use in marshal-

Cattle Sheep/goat Pig

ling sheep, but the scale might imply cattle (Randall 2010a: 143). The layout could function equally well for the extensive running of cattle or sheep, but it would enable regulation of extensive grazing. No contemporary settlement has been identified which is in itself potentially telling, and contributes to a lack of faunal material. The wider southern British picture is one of cattle with an increasing number of sheep (Serjeantson 2011: 96). There is little evidence for any focus on arable production. A small single Early Bronze Age sample from Woolston provided a few indeterminate cereal grains, weed seeds and a nut shell (De Carle 2014: 146). The form of this system appears to be one of large scale, open areas, broadly suitable for either cattle or sheep, but with limited evidence for arable cropping or intensity of production and is consistent with Model ‘B’ (Figs. 5.2 and 5.5). These uncomplicated and coaxial systems which covered areas spanning kilometres were superseded by an entirely different approach. An arrangement of settlement and field systems at Milsom’s Corner (Fig. 5.6) dating from the fourteenth–thirteenth centuries BC indicates a more small scale and nucleated approach. Just to the north of the earlier linears at Parsonage Farm, the arrangement was situated on well drained and fertile soil. It comprised a cluster of small fields and paddocks with buildings within it and a spinal trackway which opened via a funnel entranceway onto the lower lying wetter ground, with easy access to water. Utilisation of lower lying areas may have been limited and imply seasonal use. The small animal bone assemblage gives a relatively even abundance of cattle and sheep with some pig. This is a typical approach for the Middle Bronze Age of southern England, in which there was no specific emphasis on one species or product (Randall 2010a: 144–45). The archaeobotanical evidence also indicates a range of crops: barley and emmer wheat as well as flax (De Carle 2014: 76, 147). The form of fields, on better soil, and facilitation of grazing on the poorer land, appears to indicate a small scale but integrated pastoral and arable system, as suggested in

62

C. Randall

Earliest Bronze Age

Limited control of animals by physical barriers Limited investment in infrastructure Need for human presence with livestock Limited evidence of arable crops No evidence of settlement Regional indication of cattle economy with some sheep

Pen Open grazing

Earlier Bronze Age

F2

F1

Control of animals over broad area with physical barriers Considerable investment in infrastructure Reduced need for human presence with livestock Limited evidence of arable crops No evidence of settlement Regional indication of cattle economy with some sheep

F3 F4

F5

FX

Field and number

Middle Bronze Age

Control of animal access to defined fields and routeways to open grazing Investment in infrastructure Variable need for human presence with livestock on a diurnal or seasonal basis Clear evidence for arable crops Settlement within the system Faunal evidence of a non-specialised mixed cattle and sheep economy

F1 F2

House within a paddock, garden or yard

F3

Open grazing FX

Field and number

Later Bronze Age

Control of animals over a broad area with physical barriers Investment in infrastructure Reduced need for human presence with livestock Clear evidence of arable crops Settlement adjacent to system Faunal evidence of cattle, increasing sheep and focus on pig consumption

F2

F1 F3 F4

F5

FX

Field and number

Early Iron Age

Limited control of animals by physical barriers No investment in infrastructure Need for human presence with livestock Evidence of arable crops Settlement nucleated within hillfort Faunal evidence of cattle, increased sheep and focus on pig consumption

Open grazing

Middle to Late Iron Age F7 F1

F8 F6

F3 F2

Control of animal access to defined fields and routeways to open grazing, with multple units, over a wide area Considerable investment in infrastructure Variable need for human presence with livestock on a diurnal or seasonal basis Clear evidence for arable crops, possible use of cattle for traction Settlement within the system Faunal evidence of a non-specialised mixed cattle and sheep economy

F5

House within a paddock, garden or yard Open grazing

F4

FX Field and number

Fig. 5.5  Changing models of organisation in the South Cadbury environs

5  A Sheep’s Eye View: Land Division, Livestock and People in Later Prehistoric Somerset, UK

63

South Cadbury Valley

Sigwells

0

200m

Sheep Slait

Card’s Piece, Woolston

Milsom’s Corner

0

100m

Sigwells

0

10m

0

500m

N

0

200m

Features

Fig. 5.6  Bronze Age Sites in the South Cadbury environs: Sigwells, Milsom’s Corner and Card’s Piece

Model ‘C’ (Figs. 5.2 and 5.5). A slightly later enclosure at eastern Sigwells however, appears to have existed in a largely unenclosed landscape, although it was aligned on and reused one of the earlier boundaries (Tabor 2008: 61). Here the faunal assemblage had sheep as the more abundant species by some margin (Randall 2010a: 145), but the particular uses of the enclosure for metalworking (Tabor 2008) and obviously structured nature of deposition on the site may have skewed the record. Sheep were utilised in at least one specific consumption event. The enclosure itself does not fit with any of the agriculturally focussed models proposed, which supports its identification as having a particular and different purpose. This picture fits with a model of focussed areas of production within a less intensively utilised and unbounded landscape. The earlier first millennium saw these systems go out of use whilst new construction of boundaries appears to have been rare. The distribution of Late Bronze Age pottery was limited to an area directly around Cadbury Castle, suggesting a focus of settlement at the time that the hillfort commenced its development (Tabor 2008: 77–78; Barrett et al. 2000). At Sheep Slait, to the south, during the Late Bronze Age-­Early Iron Age, boundaries were associated with a ringwork enclosure, of a type better known in the Thames Valley

N

Features

0

200m

Fig. 5.7  Iron Age Sites in the South Cadbury environs: Sheep Slait, Cadbury Valley and Sigwells

(Tabor 2008: 93–94; cf. Yates 2007: 128). The boundaries enclosed large areas of land, rather than small aggregated or nucleated parcels (Fig. 5.7). Although understanding of the full system is hampered by the effects of modern ploughing, this system in many respects fits Model ‘B’ (Figs. 5.2 and 5.5). However, the association with settlement and the evidence of arable cropping suggests more aspects of Model ‘C’. There was a shift in the arable economy in the later Bronze Age with an increase in the importance of spelt wheat, with spring sowing indicated (De Carle 2014: 76, 131–32). This might also allow more grazing time over winter (Jones 1981: 104). The cropping regime suggested by archaeobotanical material from Sheep Slait is an intense one (De Carle 2014: 158), which maximises land use, and which appears at odds with the structure of the landscape. The location of arable production may be debatable, and the morphology of the model potentially misleading. The con-

64

temporary land use around Cadbury Castle, appears to be completely unenclosed, and more in keeping with Model ‘A’, so the interaction of components over a wider area was variable. In the Cadbury Castle animal bone assemblage, cattle and sheep/goat were of similar abundance with pig a minority species in Late Bronze Age contexts and into the earlier Iron Age. A maintenance approach was applied to sheep culling enabling exploitation for both primary and secondary products, at least for sheep/goat (Randall 2010a: 150–152) and ensuring continuity in the flock/herd. However, in Early Iron Age contexts pig was particularly well represented, and this is echoed by the Sheep Slait assemblages. This probably relates to specific consumption practices, but may also fit with availability of a less regulated landscape in some areas, where pig keeping would not be disruptive. The Cadbury Castle Early Iron Age faunal assemblage shows an increase in sheep/goat abundance at the proportional expense of cattle (Fig.  5.4; Randall 2010a: 159). However, the marked change came in the mid first millennium. The volume of faunal data was much expanded both in the Cadbury Castle and SCEP sites, reflecting both the zenith of the hillfort development but also activity at a greater number of settlement sites in the environs. The relative abundance of sheep/goat increased further, with cattle reduced, and pig a minor player. This chimes entirely with the characterisation of the period as the ‘sheep age’ in southern Britain (Albarella 2007). The number of goats represented was very low, which may relate to their disruptive behaviour. There is some evidence for the utilisation of cattle for dairy production and traction (presumably related to arable production), but a notable factor is the culling strategy employed for sheep/goat. This reflected a clear herd management strategy of culling out surplus young stock to manage foddering requirements and probable deliberate selection of breeding animals for flock management reasons (Randall 2018). Once the approach was established it remained stable until the end of the first millennium BC (Randall 2010a: 166, 184–187). The landscape showed a floruit of field systems which occurred throughout the South Cadbury Valley, on the Sigwells plateau (Fig.  5.7) to the south, as well as on the north flanks of Cadbury hill, and to the west. All areas and soils were used, including for the first time, areas of heavy clay (Tabor 2008), although there was a preference for more level ground. Each of these field systems was positioned with access to water either within it or nearby. These Iron Age systems were extended arrangements of multiple enclosures of varying sizes and incorporating numerous stock handling features including races, droveways, as well as funnel entrances from what appears to have been open grazing. Houses occur within the layouts, along with smaller features which appear to be pens (Randall 2010a: 175–179), and this would fit with Model ‘D’ (Figs. 5.2 and 5.5). The number of paths and droveways indicates that

C. Randall

moving animals appears to have been an everyday concern, either for water or grazing management. A large droveway, 40 m wide, north of Cadbury Castle, which crossed a substantial tributary of the River Cam, probably had Middle Iron Age origins, and continued in use into the Romano-­British period (Randall 2014). The droveway may have been used to bring stock in from outside the system, presumably for consumption within the hillfort. This raises intriguing questions about the relationship of the hillfort with its wider hinterland and the possible seasonal grazing of the wetlands to the north-­west. The likelihood that arable cropping was closely related in systems of this type is clearly supported in this case, as the archaeobotanical data show an increase in absolute numbers of specimens, but also changes with the introduction of oat, rye, celtic bean and pea, and more emphasis on spelt at Sigwells and emmer elsewhere. The usefulness of legumes in nitrogen fixing suggests that this may have become important in the Cadbury environs in the Middle Iron Age (De Carle 2014: 129), which supports a more intensive use for the field systems. There is a lack of flax in the Iron Age samples which fits the picture elsewhere (De Carle 2014: 150), possibly relating to the availability of wool from the animal economy (Lambrick and Robinson 2009), and fits with the focus on sheep seen in the faunal assemblages. The weed assemblages appear to also indicate a shift, either to autumn sown crops or a less intensive regime, although many of them are possibly associated with areas previously in cultivation or field margins and different soil types (De Carle 2014: 124; 131). Given the other evidence however, an interpretation of a change to autumn sowing might be preferred. Autumn sown crops also tend to be more productive (Ellis and Russell 1984). Storage capacity in the form of pits increased dramatically within the hillfort but also in several locations in the surrounding landscape, particularly at Sigwells and to the east of the hillfort at Hicknoll Slait (Tabor 2008). The extensive areas of pits imply a dramatic change in arable production, but could also have been used for storage of animal fodder. Increasingly through the last few centuries BC, these pits became the receptacles for a range of complex structured deposits often including the articulated or accumulations of disarticulated remains of livestock animals, which suggest changes to the ideological uses of animals. In the latest Iron Age that there was an observable increase in the frequency and complexity of highly structured deposition involving livestock and other domestic animals, human remains and other artefacts and materials (Randall 2010b; Jones and Randall 2010). There was a clear change in the meaning of domesticated animals which went beyond their economic contribution, but may also have reflected it. The Iron Age systems were altered, added to, with parts abandoned over time, but provide a general picture of a highly complex system which lasted several hundred years

5  A Sheep’s Eye View: Land Division, Livestock and People in Later Prehistoric Somerset, UK

before parts began to go out of use in the decades before the Roman invasion. The stabilities in the relative abundance of sheep/goat and the culling strategy throughout the late first millennium BC are striking (Randall 2010a: 187). However, there were nuances, such as changes in the rates of particular pathologies in sheep/goat that might have related to alterations in husbandry and a propensity to over-­grazing (Randall 2018). As fields went out of use, sheep appear to have been more prone to degenerative joint problems potentially related to walking further or standing on steep or hard ground. They were however less prone to oral pathologies caused by over-­ grazing. A drift towards a less intense approach might be conjectured. Some field systems appear to have weathered the Roman invasion, but many fell out of use before the turn of the first millennium AD.  Despite several settlements in the area, including Castle Farm, Sigwells, and Parsonage Farm, there is little evidence for new land division. The exception is part of the Cadbury Valley system and at Woolston. Preliminary examination of the faunal remains, however, seems to indicate a broadly similar relative abundance of species as those seen in the later Iron Age, although the culling strategies are not yet clear. The aims of production appear to have remained the same but it took place in a much more open landscape. The post-­Roman period is even more elusive. Despite the clear refurbishment of the hillfort in the fifth–sixth centuries (Alcock 1995), burials dating to the seventh century at Hicknoll Slait (Davey 2005), and likely activity of a similar date at Castle Farm (Davey 2005: 50; Randall 2018), no boundaries of this date have been identified (Davey 2005). A small assemblage of material from post-­Roman contexts on Cadbury may not be representative, but seems to show very similar abundance of livestock as the later Iron Age assemblages (Randall 2018), although with the addition of slightly more wild species such as deer.

5.8

Time and Place – Discussion

Each phase in the sequence of field system layouts has parallels in the south-­west of Britain, so the South Cadbury fields are far from unique. However, they are distinguished by the possibilities of tracing changes throughout the second and first millennia BC, and beyond, within one defined area. The focus of activity around the Cadbury hill in the later prehistoric period was most likely a combination of the variety of opportunities afforded by the topography, soils, and water availability, combined, as time went on, with the associations of previous use. We need to consider then why the different fields (and periodic lack of them) were appropriate at different times. Some of this may be related to changing population density, both of people and animals, but in this case the

65

evidence is elusive. The large areas of land encompassed by boundaries from about 1700 BC – if understood as extensive grazing areas with as yet unidentified foci of arable production – may not have supported a greater number of people than the open country approach of earlier periods. The increased evidence for fields was not matched by extensive evidence of buildings or settlement. Therefore the imperative may have been around concepts of territoriality but also with the effect of regulating the degree and timing of labour involvement. Buildings first appear around the fourteenth century BC but in limited locations, and associated with fields. The more localised fields with an integrated approach seen at Milsom’s Corner, were more complex but small scale, focussed on one location and utilising both good arable soils and nearby probably seasonal grazing opportunities. It also seems that there were not numerous foci of this kind in the district. The scale of landscape coverage may have changed, but the level of population involved perhaps did not. It is interesting then that the increase in evidence of people and settlement in the earlier Iron Age, nucleated on Cadbury Castle and its immediate flanks, with the beginnings of a focus on sheep production, produced little substantive land division. The effect climatic changes in the south-­west of Britain during the earlier first millennium BC (Christie 1986: 105; Quinnell 1988: 10) is debateable, but the evidence from Cadbury Castle implies a greater population in the South Cadbury landscape in this period compared with before. It was apparently possible to cater for a larger local population within a fully extensive farming system, but this would have had implications for the organisation of daily tasks such as minding animals and the relationships between those carrying them out and other members of the population, as well as the as yet unidentified location of arable production. The variation in contemporary approach within a few kilometres is indicative of social complexities associated with land holding and the emergence of the hillfort. The nature of the ringwork at Sheep Slait indicates that Cadbury Castle was not the only significant locale in this period. The increasing intensity and integration of production visible in the faunal data from the middle of the first millennium BC is contemporary with the appearance of large numbers of boundary features. There appears to be a relationship between complexity of field systems and a shift in intensity of occupation and production, but with the addition of a more defined ‘product’, sheep. The versatility of sheep in providing wool, enhancing manuring for cropping and potential for more storable products, combined with evidence of storage, indicates a step change from previous approaches to exploiting the South Cadbury landscape. Presumably, communities attempted to produce more from

66

the same space and with a greater definition of the aims of production. The way in which labour was organised and negotiated, and land controlled or owned must have changed dramatically in comparison to the previous period. Also, the range and depth of expertise needed to maintain these transformed husbandry practices over the long term would have been considerable, with greater knowledge of specific culling strategies, grazing, arable crops and soils. These changes also went hand in hand with the development of the hillfort, so the reorganisation and creation of highly integrated and probably highly productive local fields indicates a co-­­ dependence of the hillfort and its immediate environs, with hints of connections to a broader hinterland. This chimes with the envisaging of hillforts emerging within a landscape of fields and settlement and possibly as the foci of an integrated productive system (Barrett 1999: 254), but also indicates a different social organisation facilitating the deployment of labour and identifying areas of responsibility for land, crops and animals. The re-­orientation of the landscape at the end of the first millennium similarly relates to the use of the hillfort space and articulation of new structures into the first millennium. Each of the approaches seen in the South Cadbury landscapes can be seen to be ‘effective’ ways of exploiting that landscape, dependent on the aims of those arranging them. The flexibility of mixed farming, both in terms of livestock species, and in combining arable and pastoral agriculture means that it is achievable within a wide range of scales and intensity. The Cadbury Bronze Age landscapes indicate at least two different approaches and the earlier Iron Age a third approach. Each of these would have provided much the same outputs but required differing social negotiations of rights and responsibilities. In contrast the later Iron Age landscape is oriented around highly focussed production, both in terms of scale and type, resulting in much greater complexity and regulation. The landscape reflects the greater complexity of the aims of husbandry. The social roots of this step change must relate to the consumption needs of the burgeoning hillfort in this particular case. The series of Cadbury fields demonstrates there will always be a variety of ways of responding to the combination of possibilities and constraints supplied by the particular locale, livestock animals, and available crops. At different times within this one landscape different ways of making a living were implemented, all of them to one degree or another successful given their longevity. There was no single ‘right’ way of carrying out livestock husbandry integrating it with arable cropping and neither was there a chronological ‘progression’, an evolution from simple to complex. At different times extensive and intensive approaches were ‘right’, and that was something socially determined rather than dictated by the place itself.

C. Randall

5.9

Conclusion

The physical properties of certain elements of field systems can not only help us to appreciate the framework of the economy, but also the daily, weekly and seasonal activities associated with it and consequent ordered movement of people, animals within a space. Re-­articulation of the components of the space within the landscape, responded to the natural form and opportunities of that landscape, as well as referencing natural and anthropogenic landmarks. Changes in the ways that field systems were configured in the Cadbury landscape might be thought of as transformations in practical husbandry and social husbandry. Where field systems are complex and integrate various types of highly regulated production, there is also an implication that there would be a necessity for more complex negotiation of rights and responsibilities, access to the products, control of the resource, deployment of labour, and the ways in which status might accrue through demonstrated knowledge and expertise. The attitude of people to their animals is likely to have been more complex where considerations as to the livestock’s welfare was integral to the way in which the landscape was organised and structured and where most people on a daily, seasonal and annual basis inhabited that space with and because of those animals. It might offer some explanations as to why animals increasingly featured in practices in the South Cadbury landscape which resulted in highly complex structured deposition in the later Iron Age. The physical proximity, relationships and meaning of animals placed at the heart of a society will have become part of its ideology; some of the depositional practices observed may be the result of status and meaning being mediated through animals as the core of the community. Comparison of the model landscapes with that around Cadbury Castle has demonstrated that they can be a useful tool in appreciating that their form is not necessarily dictated by landform or chronology, but assists in framing questions with respect to intensity of activity and integration with other lines of enquiry. However, examples from within the Cadbury Castle environs do indicate that form alone is not diagnostic of an agricultural approach, as the archaeobotanical data are in this example at times at odds with the model suggested by the landscape layout. This is however an extremely useful prompt to examine more closely the potential location of activities. Nevertheless, better understanding of how a particular layout or system could be used may assist in situations where faunal assemblages are absent due to preservational conditions. If we can understand land use and division in a more nuanced fashion, we gain a whole new perspective on the social organisation of livestock and arable production and

5  A Sheep’s Eye View: Land Division, Livestock and People in Later Prehistoric Somerset, UK

understand fields as whole entities, including the minutiae of their inception, use, development and abandonment; we can think about what went on inside fields, not just what they looked like from the outside, and adjust our research strategies accordingly. Another implication is that we underestimate the technical knowledge and expertise needed to carry out livestock husbandry, an expertise which arguably we in the modern developed world have a tendency to undervalue. We should perhaps look for the role of past farming expertise, or even specialists, in the different approaches to dealing with livestock production apparent within landscapes and decision making evident from faunal assemblages. Acknowledgements  Thanks are due to Prof Mark Maltby, Dr. Ellen Hambleton, Dr. Richard Tabor, and Dr. Mike Allen for their assistance with this project, which was originally funded by a Bournemouth University Studentship award. Great appreciation goes to Tara Fairclough for tidying up my figures to far greater effect that I could ever achieve; my attempts at design constantly amuse. Dr. Cheryl Green and Richard McConnell cast helpful eyes over the text, and I am indebted to the reviewers for their detailed feedback. Any lingering nonsense is entirely my own. My lasting appreciation goes to all of the volunteers of the South Cadbury Environs Project and subsequently the South Somerset Archaeological Research Group, including dear friends who have departed to fresher pastures, and without whom the data would never have been available.

References Albarella, U. (2007). The end of the Sheep Age: people and animals in the late Iron Age. In C. Haselgrove & T. Moore (Eds.), The Late Iron Age in Britain and beyond (pp.  393–406). Oxford: Oxbow Books. Albarella, U., Manconi, F., Vigne, J.-D., & Rowley-Conwy, P. (2007). Ethnoarchaeology of pig husbandry in Sardinia and Corsica. In U. Alberella, K. Dobney, A. Ervynck, & P. Rowley-Conwy (Eds.), Pigs and humans: 10,000 years of interaction (pp.  285–307). Oxford: University Press. Alcock, L. (1967). A reconnaissance excavation at South Cadbury Castle, Somerset, 1966. Antiquaries Journal, 47, 47–51. Alcock, L. (1968a). Cadbury Castle 1967. Antiquity, 42, 47–51. Alcock, L. (1968b). Excavations at South Cadbury Castle, 1967: A summary report. Antiquaries Journal, 48, 6–17. Alcock, L. (1970). South Cadbury excavations, 1969. Antiquity, 44, 46–49. Alcock, L. (1971). Excavations at South Cadbury Castle, 1970: A summary report. Antiquaries Journal, 5, 1–7. Alcock, L. (1972). By South Cadbury is that Camelot…’: the excavation of Cadbury Castle 1966–1970. London: Thames and Hudson. Alcock, L. (1980). The Cadbury Castle sequence in the first millennium BC. Bulletin of the Board of Celtic Studies, 28, 656–718. Alcock, L. (1995). Cadbury Castle, Somerset: The early medieval archaeology. Cardiff: University of Wales Press. Alderson, L. (1988). Sheep and goats. In K. Thear & A. Frazer (Eds.), The complete book of raising livestock and poultry (pp. 104–139). London: Pan Books. Allen, T. (1910). Profitable pig breeding and feeding. London: Upcott Gill. Amorosi, T., Buckland, P. C., Edwards, K. J., Mainland, I., McGovern, T. H., Sadler, J. P., & Skidmore, P. (1998). They did not live by grass

67

alone: the politics and palaeoecology of animal fodder in the North Atlantic region. Environmental Archaeology, 1, 41–54. Archer, J. (1988). Crop nutrition and fertiliser use. Ipswich: Farming Press Ltd.. Bakels, C. C. (1997). The beginnings of manuring in western Europe. Antiquity, 71, 442–445. Barker, G. (1985). Prehistoric farming in Europe. Cambridge: Cambridge University Press. Barker, G., & Gamble, C. (1985). Beyond domestication: A strategy for investigating the process and consequence of social complexity. In G. Barker & C. Gamble (Eds.), Beyond domestication in prehistoric Europe: Investigations in subsistence archaeology and social complexity (pp. 1–31). London: Academic. Barnatt, J. (2008). From clearance plots to ‘sustained’ farming: Peak District fields in prehistory. In A.  M. Chadwick (Ed.), Recent approaches to land allotment (British archaeological reports (international series) 1875) (pp. 41–68). Oxford: British Archaeological Reports. Barrett, J. C. (1999). The mythical landscapes of the British Iron Age. In W. Ashmore & A. B. Knapp (Eds.), Archaeologies of landscape: Contemporary perspectives (pp.  253–265). Oxford: Blackwell Publishers Ltd. Barrett, J.  C., Freeman, P.  W. M., & Woodward, A. (2000). Cadbury Castle, Somerset: The later prehistoric and early historic archaeology. London: English Heritage. Blake, F. (1990). Organic farming and growing (2nd ed.). Marlborough: The Crowood Press. Bogaard, A. (2005). ‘Garden agriculture’ and the nature if early farming in Europe and the Near East. World Archaeology, 37(2), 177–196. Breen, E. (2008). Encounters with place in prehistory: Writing a case study for Shipman Head Down, Isles of Scilly. In A. M. Chadwick (Ed.), Recent approaches to the archaeology of land allotment (British archaeological reports (international series) 1875) (pp. 97–109). Oxford: British Archaeological Reports. Brück, J. (2000). Settlement, landscape and social identity: the Early-­ Middle Bronze Age transition in Wessex, Sussex and the Thames Valley. Oxford Journal of Archaeology, 19(3), 273–300. de Bruijn, S. L., & Bork, E. W. (2006). Biological control of Canada thistle in temperate pastures using high density rotational cattle grazing. Biological Control, 36, 305–315. Bryce, D., & Wagenaar, A. (1985). Grassland smallholding. Lampeter: Llanerch Enterprises. Campbell, B. (1981). The regional uniqueness of English field systems? Some evidence from eastern Norfolk. The Agricultural History Review, 29(1), 16–28. Campbell, B., & Godoy, R.  A. (1992). Commonfield agriculture: The Andes and medieval England compared. In D.  W. Bromley (Ed.), Making the commons work: Theory, practice and policy (pp. 99–127). San Francisco: Institute for contemporary studies. Chadwick, A.  M. (2008a). Fields for discourse? Toward more self-­ critical, theoretical and interpretive approaches to the archaeology of field systems and land allotment. In A. M. Chadwick (Ed.), Recent approaches to the archaeology of land allotment (British archaeological reports (international series) 1875) (pp.  205–237). Oxford: Archaeopress. Chadwick, A.  M. (2008b). Introduction. In A.  M. Chadwick (Ed.), Recent approaches to the archaeology of land allotment (British archaeological reports (international series) 1875) (pp.  1–24). Oxford: Archaeopress. Chang, C. (1993). Pastoral transhumance in the southern Balkans as social ideology: Ethnoarchaeological research in Northern Greece. American Anthropologist, 95(3), 687–703. Chang, C., & Tourtellotte, P. A. (1993). Ethnoarchaeological survey of pastoral transhumant sites in the Grevena region. Greece Journal of Field Archaeology, 20(3), 249–264.

68 Christie, P.  M. (1986). Cornwall in the Bronze Age. Cornish Archaeology, 25, 81–110. Davey, J. (2005). The Roman to Medieval transition in the region of South Cadbury Castle, Somerset. In British archaeological reports (British series) 399. Oxford: Archaeopress. De Carle, D. (2014). Changing Arable Subsistence in Bronze and Iron Age South West Britain. University of Sheffield Unpublished PhD thesis. De Garis de Lisle, D. (1982). Effects of distance on cropping patterns internal to the farm. Annals of the Association of American Geographers, 72(1), 88–98. Defra. (2002). Code of recommendations for the welfare of livestock: Sheep. London: Defra Publications. Done, G. (2016). Animal bone. In C.  Evans, G.  Appleby, & S.  Lucy (Eds.), Lives in Land: Mucking excavations by Margaret and Tom Jones 1965–1978: Prehistory, Context and Summary (CAU landscape archives: Historiography and fieldwork (6/mucking 6)) (pp. 201–202). Oxford: Oxbow Books. Doyle, M. (1870). Cottage farming: Or how to cultivate from two to twenty acres, with chapters on the management of cows, pigs and poultry. London: Groombridge and Sons. Ellis, R., & Russell, G. (1984). Plant development and grain yield in spring and winter barley. The Journal of Agricultural Science, 102, 85–95. Evans, H. (2008). After the axe: Ways into the upland landscapes of Cumbria. In A.  M. Chadwick (Ed.), Recent approaches to the archaeology of land allotment (British archaeological reports (international series)) (Vol. 1875, pp.  122–142). Oxford: Archaeopress. English, J. (2013). Pattern and progress: Field systems of the second and early first millennia BC in Southern Britain (British archaeological reports (British series) 587). Oxford: British Archaeological Reports. Esmail, S.  H. M. (1991). Multi-species grazing by cattle and sheep. Rangelands, 13(1), 35–37. Evans, C. (2009). Fengate Revisited: Further Fen-edge excavations, Bronze Age fieldsystems and settlement and the Wyman Abbott/ Leeds archives (CAU landscape archives: Historiography and fieldwork (1)). Cambridge: Cambridge Archaeological Unit. Evans, C., Appleby, G., & Lucy, S. (2016). Lives in land. Mucking excavations by Margaret and Tom Jones 1965–1978: Prehistory, context and summary (CAU landscape archives: Historiography and fieldwork (6/mucking 6)). Oxford: Oxbow Books. Fitzherbert, A. (1534). The boke of husbandry. London: Thomas Berthelet. Fleming, A. (1978a). The prehistoric landscape of Dartmoor Part 1: South Dartmoor. Proceedings of the Prehistoric Society, 44, 97–123. Fleming, A. (1978b). The Dartmoor reaves. In H.  C. Bowen & P.  J. Fowler (Eds.), Early land allotment (British archaeological reports (British series) 48) (pp.  17–21). Oxford: British Archaeological Reports. Fleming, A. (1983). The prehistoric landscape of Dartmoor Part 2: North and East Dartmoor. Proceedings of the Prehistoric Society, 49, 195–241. Fleming, A. (1985). Land tenure, productivity and fields systems. In G. Barker & C. Gamble (Eds.), Beyond domestication in prehistoric Europe: Investigations in subsistence archaeology and social complexity (pp. 129–146). London: Academic. Fleming, A. (1988). The Dartmoor Reaves: Investigating prehistoric land divisions. London: Batsford. Fleming, A. (1998). Prehistoric landscape and the quest for territorial pattern. In P. Everson & T. Williamson (Eds.), The archaeology of landscape (pp. 42–66). Manchester: Manchester University Press. Forbes, H. (1998). European agriculture viewed bottom-side upwards: Fodder-and forage-provision in a traditional Greek community. Environmental Archaeology, 1, 19–34.

C. Randall Fowler, P. J. (1978). Pre-medieval fields in the Bristol region. In H. C. Bowen & P. J. Fowler (Eds.), Early land allotment (British archaeological reports (British series) 48) (pp.  29–47). Oxford: British Archaeological Reports. Fowler, P. J. (1981). Wildscape to landscape: ‘Enclosure’ in prehistoric Britain. In R. Mercer (Ed.), Farming practice in British prehistory (pp. 9–54). Edinburgh: Edinburgh University Press. Fowler, P.  J. (1983). The farming of prehistoric Britain. Cambridge: Cambridge University Press. Fowler, P. J., & Evans, J. G. (1967). Plough marks, lynchets and early fields. Antiquity, 41, 289–301. Frame, J. (1992). Improved grassland management. Ipswich: Farming Press Books. Fraser, A. (1947). Sheep production. Edinburgh: Thomas Nelson & Sons. Fraser, A. (1988). Horses. In K. Thear & A. Frazer (Eds.), The complete book of raising livestock and poultry (pp. 196–217). London: Pan Books. Ganscopp, D. C., & Bohnert, D. W. (2009). Landscape nutritional patterns and cattle distribution in rangeland pastures. Applied Animal Behavioural Science, 116, 110–119. Gibson, C. W. D., Dawkins, H. C., Brown, V. K., & Jepsen, M. (1987 Spring). Grazing by sheep: Effects of seasonal changes during early old field succession. Vegetatio, 70(1), 33–43. Giles, M. (2012). A forged glamour: Landscape, identity and material culture in the Iron Age. Oxford: Windgather Press. Goodwin, D.  H. (1973). Pig management and production. London: Hutchinson. Goodwin, D. H. (1979). Sheep management and production (2nd ed.). London: Hutchinson. Grandin, T. (1980). Observations of cattle behavior applied to the design of cattle-handling facilities. Applied Animal Ethology, 6(1), 19–31. Grandin, T., & Deesing, M. (2008). Humane Livestock Handling. North Adams: Storey Publishing. Grigg, D. (1982). The dynamics of agricultural change the historical experience. London: Hutchinson. Guttmann, E. B. A. (2005). Midden cultivation in prehistoric Britain: arable crops in gardens. World Archaeology, 37(2), 224–239. Haas, J. N., Karg, S., & Rasmussen, P. (1998). Beech leaves and twigs used as winter fodder: examples from historic and prehistoric times. Environmental Archaeology, 1, 88–86. Halliday, J., & Halliday, J. (1988). Goats. In K. Thear & A. Frazer (Eds.), The complete book of raising livestock and poultry (pp.  73–103). London: Pan Books. Halstead, P. (1996). Pastoralism or household herding? Problems of scale and specialisation in early Greek animal husbandry. World Archaeology, 28(1), 20–42. Hambleton, E. (1999). Animal husbandry regimes in Iron Age Britain: A comparative study of faunal assemblages from British Iron Age sites (British archaeological reports (British series) 282). Oxford: Archaeopress. Harrison, K.  A., & Bardgett, R.  D. (2008). Impacts of grazing and browsing by large herbivores on soils and soil biological properties. In I.  J. Gordon & H.  H. T.  Prins (Eds.), The ecology of browsing and grazing (Ecological Studies 195) (pp.  201–216). Heidelberg: Springer. Hart, E. (1994). Sheep: A guide to management. Marlborough: The Crowood Press. Harvey, N. (1980). The industrial archaeology of farming in England and Wales. London: BT Batsford. Henderson, G. (1944). The farming ladder. London: Faber and Faber. Hey, G., Bell, C., Dennis, C., & Robinson, M. (2016). Yarnton Neolithic and Bronze Age settlement and landscape (Thames Valley Landscapes Monograph 39). Oxford: Oxford Archaeology.

5  A Sheep’s Eye View: Land Division, Livestock and People in Later Prehistoric Somerset, UK Hey, G., Booth, P., & Timby, J. (2011). Yarnton Iron Age and Romano-­ British settlement and landscape (Thames Valley Landscapes monograph 35). Oxford: Oxford Archaeology. Hodgkinson, S. M., López, I. F., & Navarrete, S. (2009). Ingestion of energy, protein and amino acids from pasture by grazing European wild boar (Sus scrofa L.) in a semi-extensive production system. Livestock Science, 122, 222–226. Houghton Brown, J., & Powell-Smith, V. (1984). Horse and stable management. Oxford: BSP Professional Books. Johnston, R. (2000). Field systems and the Atlantic Bronze Age: Thoughts on a regional perspective. In J. C. Henderson (Ed.), The prehistory and early history of Atlantic Europe (British archaeological reports (international series) 861) (pp.  47–55). Oxford: Archaeopress. Johnston, R. (2001). ‘Breaking new ground’: Land tenure and fieldstone clearance during the Bronze Age. In J. Brück (Ed.), Bronze Age landscapes tradition and transformation (pp. 99–109). Oxford: Oxbow Books. Johnston, R. (2005). A social archaeology of garden plots in the Bronze Age of northern and western Britain. World Archaeology, 37(2), 211–223. Jones, G. (1981). The Carbonised plant remains. Somerset Levels Papers, 7, 33–35. Jones, G. (1998). Distinguishing food from fodder in the archaeobotanical record. Environmental Archaeology, 1, 95–98. Jones, J. (2006). Plant macrofossil remains. In T. Gent and S. Reed, A Bronze Age field and figure from Hillfarrance. Proceedings of the Somerset Archaeological and Natural History Society, 150, 20–24. Jones, R., & Dowling, P. (2005). Sustainability and economics of temperate perennial Australian grazing systems. Agriculture, Ecosystems and Environment, 106, 359–376. Jones, S., & Randall, C. E. (2010). Death, destruction and the end of an era: The end of the Iron Age at Cadbury Castle, Somerset. In M. Sterry & A. Tullett (Eds.), In search of the Iron Age: Proceedings of the Iron Age Research Student Seminar 2008 (pp.  165–184). Leicester: Leicester University Press. Kelly, F. (2000). Early Irish farming. Dublin: Dublin Institute for Advanced Studies. Kitchen, W. (2001). Tenure and territoriality in the British Bronze Age: A question of varying social and geographic scales? In J.  Brück (Ed.), Bronze Age landscapes: Tradition and transformation (pp. 110–120). Oxford: Oxbow Books. Ladle, L., & Woodward, A. (2009). Excavations at Bestwall Quarry, Wareham 1992–2005 volume 1: The prehistoric landscape (Dorset Natural History and Archaeological Society monograph 19). Dorchester: DNHAS. Lake, J. (1989). Historic farm buildings an introduction and guide. Blandford: Blandford Press. Lambrick, G. H., & Robinson, M. A. (2009). The Thames through time. The archaeology of the gravel terraces of the upper and middle Thames: The Thames Valley in later prehistory 1500 BC – 50 AD. Oxford: Oxford Archaeology Press. Leach, H.  M. (1999). Intensification in the Pacific: a critique of the archaeological criteria and their application. Current Anthropology, 40, 311–339. Lewthwaite, J. (1981). Plains tails from the hills: Transhumance in Mediterranean archaeology. In A.  Sheridan & G.  Bailey (Eds.), Economic archaeology (British archaeological reports (British series) 96) (pp. 57–66). Oxford: British Archaeological Reports. Løvschal, M. (2014). Emerging boundaries Social embedment of landscape and settlement divisions in Northwestern Europe during the First millennium BC. Current Anthropology, 35(6), 725–750. Maclean, M. (2000). Resource management: Hedges. Tonbridge: Farming Press. Manolson, F., Fraser, D. V. M., & Fraser, A. (1988). Cattle. In K. Thear & A. Fraser (Eds.), The complete book of raising livestock and poultry (pp. 159–195). London: Pan Books.

69

Masefield, R. (2015). Origins, development and abandonment of an Iron Age village: Further archaeological investigations for the Daventry International Rail Freight Terminal. Crick & Kilsby, Northamptonshire 1993–2013. (DIRFT Volume II). Masseti, M. (2007). The economic role of Sus in early human fishing communities. In U.  Alberella, K.  Dobney, A.  Ervynck, & P. Rowley-Conwy (Eds.), Pigs and humans: 10,000 years of interaction (pp. 156–170). Oxford: Oxford University Press. Mientjes, A.  C. (2004). Modern pastoral landscapes on the island of Sardinia, Italy. Recent pastoral practices in  local versus macro-­ economic and macro-political contexts. Archaeological Dialogues, 10(2), 161–190. Morand-Fehr, P. (2005). Recent developments in goat nutrition and application: a review. Small Ruminant Research, 60, 25–43. Morrison, K. D. (1996). Typological schemes and agricultural change: Beyond Boserup in precolonial South India. Current Anthropology, 37(4), 583–608. Mulville, J., Ayres, K., & Smith, P. (2011). Animal bone. In G.  Hey, P. Booth, & J. Timby (Eds.), Yarnton: Iron Age and Romano-British settlement and landscape (Thames Valley landscapes monograph 35) (pp. 487–522). Oxford: Oxford Archaeology. Mulville, J., & Robinson, M. (2016). Food production and consumption. In G. Hey, C. Bell, C. Dennis, & M. Robinson (Eds.), Yarnton Neolithic and Bronze Age settlement and landscape (pp. 135–154). Oxford: Oxford Archaeology. Netting, R.  M. (1974). Agrarian ecology. Annual Review of Anthropology, 3, 21–56. Olmos, G., Boyle, L., Hanlon, A., Patton, J., Murphy, J.  J., & Mee, J. F. (2009). Hoof disorders, locomotion ability and lying times of cubicle-housed compared to pasture-based dairy cows. Livestock Science, 125, 199–207. Orme, B.  J., & Morgan, R.  A. (1982). Prehistoric wood from Wotter Common, Shaugh Moor. In N.D.  Balaam, K.  Smith and G.  Wainwright, The Shaugh Moor Project: Fourth report  – environment, context and conclusion. Proceedings of the Prehistoric Society, 48, 220–227. Papachristou, T. G., Dziba, L. U., & Provenza, F. D. (2005). Foraging ecology of goats and sheep on wooded rangelands. Small Ruminant Research, 59, 141–156. Pelling, R. (2011). Charred plant remains from Cresswell Field. In G. Hey, P. Booth, & J. Timby (Eds.), Yarnton: Iron Age and Romano-­ British settlement and landscape (Thames Valley landscapes monograph 35) (pp. 523–533). Oxford Archaeology: Oxford. Pfaffenberger, B. (1988). Fetishised objects and humanised nature: Towards an anthropology of technology. Man, 23(2), 236–252. Pfaffenberger, B. (1992). Social anthropology of technology. Annual Review of Anthropology, 21, 491–516. Poulsen, B. (1997). Agricultural technology in medieval Denmark. In G.  Astill & J.  Langdon (Eds.), Medieval farming and technology (pp. 115–146). New York: Brill. Prins, H. T., & Fritz, H. (2008). Species diversity of browsing and grazing ungulates: Consequences for the structure and abundance of secondary production. In I. J. Gordon & H. H. T. Prins (Eds.), The ecology of browsing and grazing (Ecological studies 195) (pp. 179– 200). Heidelberg: Springer. Pryor, F. (1996). Sheep, stockyards and field systems: Bronze Age livestock populations in the Fenlands of eastern England. Antiquity, 70, 313–324. Pryor, F. (2006). Farmers in prehistoric Britain (2nd ed.). Stroud: Tempus. Putfarken, D., Dengler, J., Lehmann, S., & Härdtle, W. (2008). Site use of grazing cattle and sheep in a large-scale pasture landscape: A GPS/GIS assessment. Applied Animal Behaviour Science, 111, 54–67. Quinnell, H. (1988). The local character of the Devon Bronze Age and its interpretation in the 1980s. Proceedings of the Devon Archaeological Society, 46, 1–12.

70 Rajkovača, V. (2016). Faunal remains. In C.  Evans, G.  Appleby, & S.  Lucy (Eds.), Lives in land. Mucking excavations by Margaret and Tom Jones 1965–1978: Prehistory, context and summary (CAU landscape archives: Historiography and fieldwork 6) (pp. 432–435). Oxford: Oxbow Books. Randall, C.  E. (2010a). Livestock and landscape: the exploitation of animals in the south west of Britain in later prehistory. Unpublished PhD thesis, Bournemouth University. Randall, C. E. (2010b). More ritual rubbish? Exploring the taphonomic history, context formation processes and ‘specialness’ of deposits including human and animal bone in Iron Age pits. In M. Maltby & J.  Morris (Eds.), Social environmental archaeology: Integrated studies of ritual (British Archaeological reports (international series) 2077) (pp. 83–102). Oxford: Archaeopress. Randall, C.  E. (2011). Fieldwork undertaken by the South Somerset Archaeological Research Group in 2010. In C. J. Webster, Somerset Archaeology. Proceedings of the Somerset Archaeological and Natural History Society, 155, 229–233. Randall, C.  E. (2014). Fieldwork undertaken by the South Somerset Archaeological Research Group in 2013. In C.  J. Webster, ed. Somerset Archaeology 2014. Proceedings of the Somerset Archaeological and Natural History Society, 158, 146. Randall, C.  E. (2016). Faunal Remains. In E.  Lupprian, ed. Pits and pottery – some evidence for Beaker activity at Bryanston School. Proceedings of the Dorset Natural History and Archaeological Society, 137, 210–211. Randall, C. E. (2018). The faunal remains. In R. Tabor, & C.E. Randall, Early Neolithic pits at Cadbury Castle and an adjoining temporary occupation site at Milsom’s Corner, South Cadbury. Proceedings of the Somerset Archaeological and Natural History Society, 161, 33–40. Rasmussen, P. (1993). Analysis of goat/sheep faeces from Egolzwil 3, Switzerland: Evidence for branch and twig foddering of livestock in the Neolithic. Journal of Archaeological Science, 20, 479–502. Reynolds, P. J. (1987). Ancient farming. Aylesbury: Shire Archaeology. Ross, P. (1989). Goats: A guide to management. Marlborough: The Crowood Press. Ryder, M.  J. (1981). Livestock products: Skins and fleeces. In R. Mercer (Ed.), Farming practice in British prehistory (pp. 182– 209). Edinburgh: Edinburgh University Press. Salmon, J. (1981). The goatkeepers guide (2nd ed.). Newton Abbot: David and Charles. Schütz, K.  E., Rogers, A.  R., Cox, N.  R., & Tucker, C.  B. (2009). Dairy cows prefer shade that offers greater protection against solar radiation in summer: Shade use, behaviour, and body temperature. Applied Animal Behaviour Science, 116, 28–34. Searle, K. R., & Shipley, L. A. (2008). The comparative feeding behaviour of large browsing and grazing herbivores. The Ecology of Browsing and Grazing, 117–148. Searle, K. R., Hobbs, N. T., & Gordon, I. J. (2007). It’s the “foodscape”, not the landscape: using foraging behavior to make functional assessments of landscape condition. Israel Journal of Ecology & Evolution, 297–316. Serjeantson, D. (2011). Review of animal remains from the Neolithic and Early Bronze Age in Southern Britain (4000–1500 BC) (English Heritage research department report 29). London: English Heritage.

C. Randall Seymour, J. (2003). The new complete book of self sufficiency the classic guide for realists and dreamers. London: Dorling Kindersley. Shahack-Gross, R., Marshall, F., Ryan, K., & Weiner, S. (2004). Reconstruction of spatial organization in abandoned Maasai settlements: Implications for site structure in the pastoral Neolithic of East Africa. Journal of Archaeological Science, 31, 1395–1411. Smith, K., Coppen, J., Wainwright, G.  J., & Beckett, S. (1981). The Shaugh Moor Project: the third report  – settlement and environmental investigations. Proceedings of the Prehistoric Society, 47, 205–273. Smith Thomas, H. (2005). Getting started with beef and dairy cattle. North Adams: Story Publishing. Stallibrass, S. (1996). Animal bones. In R. P. J. Jackson & T. W. Potter (Eds.), Excavations at Stonea, Cambridgeshire (pp.  587–612). London: British Museum Press. Street, A.  G. (1942). Round the year on the farm. London: Oxford University Press. Tabor, R. (2002). South Cadbury environs project interim fieldwork report 1998–2001. Bristol: Centre for the Historic Environment University of Bristol. Tabor, R. (2004a). Cadbury Castle: Prehistoric pottery distribution in the surrounding landscape. Proceedings of the Somerset Archaeological and Natural History Society, 147, 29–39. Tabor, R. (2004b). South Cadbury environs project interim fieldwork report 2002–2003. Bristol: Centre for the Historic Environment University of Bristol. Tabor, R. (2008). Cadbury Castle the hillfort and landscapes. Stroud History Press. Tabor, R., & Randall, C.  E. (2018). Early Neolithic pits at Cadbury Castle and an adjoining temporary occupation site at Milsom’s Corner, South Cadbury. Proceedings of the Somerset Archaeological and Natural History Society, 161, 1–48. Tani, Y. (2002). Early techniques as a forerunner of milking practices. In J. Mulville & A. K. Outram (Eds.), The zooarchaeology of fats, oils, milk and dairying (pp. 114–120). Oxford: Oxbow Books. Taylor, C. (1975). Fields in the English landscape. London: Dent. Tiffen, M., Mortimore, M. J., & Gichuki, F. (1994). More people, less erosion: Environmental recovery in Kenya. Chichester: Wiley. Walker, J.  W., Heischmidt, R.  K., & Dowhower, S.  L. (1989). Some effects of a rotational grazing treatment on cattle preference for plant communities. Journal of Range Management, 42(2), 143–148. Wall, R., French, N. P., & Morgan, K. L. (1993). Sheep blowfly population control: Development of a simulation model and analysis of management strategies. Journal of Applied Ecology, 30(4), 743–751. Weller, J. (1982). History of the farmstead. London: Faber and Faber. Wickstead, H. (2008). Theorising tenure: Land division and identity in later prehistoric Dartmoor, South-West Britain. In British archaeological reports (British series) (Vol. 465). Oxford: Archaeopress. Wijngaarden-Bakker, L. (1998). Animal husbandry in a wetland: The case of Assendelft. In C. M. Mills & G. Coles (Eds.), Life on the edge: Human settlement and marginality (Oxbow monograph 100) (pp. 173–178). Oxford: Oxbow Books. Yates, D. (2007). Land power and prestige: Bronze Age field systems in southern England. Oxford: Oxbow Books.

6

Terraced Crop Fields in the Eastern Pyrenean Mountains (France): The View from Pedoarchaeology Romana Harfouche and Pierre Poupet

Abstract

Long-­term dynamics of terraced fields are undervalued as archaeology traditionally expressed little interest and faced difficulties in dating these features. However, sediments and soils underpin the agricultural production process and are repositories of landscape dynamics, particularly in mountainous regions where the quality of information is enhanced by the steep slopes, with the presence of thick pedosedimentary stratifications due to colluvial deposits. Crop fields and societies which shape them can be examined through agrarian archaeology, which incorporates the study of soils and palaeosols. The Eastern Pyrenees case-­study questions the visibility of farming practices at high altitude in landscapes that were considered as pastures for sheep and goat grazing in ancient times, as they are today. A pedoarchaeological survey and an excavation project were undertaken between 1500 and 2000 m altitude. There, the mapping of terraced areas and subsequent identification, characterization and dating of palaeosols brought to light a period of cultivation on the sunny slope of the granitic Carlit Mountain as early as the Late Neolithic and the Bronze Age (mainly from 2890–2620  cal. BC to 2028–1874 cal. BC). Keywords

Pedoarchaeology · Mountains · Mediterranean · Late Neolithic · Bronze Age · Crop processing · Radiocarbon dating

R. Harfouche (*) CNRS et Université Panthéon-Sorbonne, UMR 7041 ArScAn (Archéologie et Sciences de l’Antiquité), Robiac-Rochessadoule, France

6.1

 ultivated Fields, Soils and Palaeosols C in Terraced Mountains

Hilly landscapes, often close to the sea, are a major component of the Mediterranean region. The necessity for farmers to build the slopes with terraces for agriculture gave them an economical and environmental importance so that they became a characteristic image of the Mediterranean, especially when terracing is associated with vineyards (Alcaraz 1999: 476–99; Bonardi 2010). These human landscapes were studied for a long time by geographers, agronomists, historians and anthropologists (e.g. Ambroise et  al. 1993; Asins Velis 2009; Zufferey-­Périsset 2012). The long term spatiotemporal dynamics of terraced landscapes are paradoxically little studied.

6.1.1 The Dating of Terraces Agrarian archaeology showed little interest in terraces. Studies focused on vineyards (identifiable through the pits of vine stock), that date primarily from the Roman period while only few examples are from Late Iron Age (e.g. in Southern France: Boissinot 1995; Daveau 2007). Archaeological excavations also focused on the ditch systems – often interpreted in a restrictive way – as an antique cadaster without examining their agronomical and environmental signification (e.g. Chouquer 1996). A major reason for this lack of archaeological studies is the difficulty to date a field built with a terrace wall or with a grassed embankment. The hypothetical link between a dated settlement and a nearby field system is far from sufficient to date agricultural terraces. A stratigraphic relationship between the terrace wall and the wall of a dated construction (e.g. a protohistoric house/farm) may be established, but often in rural areas field systems are not directly connected to an excavated site. Other possibilities of investigation are offered to the archaeologist, depending on the

P. Poupet Soil Science, CNRS, Robiac-Rochessadoule, France © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 S. Arnoldussen et al. (eds.), Europe’s Early Fieldscapes, Themes in Contemporary Archaeology, https://doi.org/10.1007/978-3-030-71652-3_6

71

72

state of conservation of the terrace and on its evolutionary history (Harfouche 2007: 37–43, Harfouche 2017: 13–16). One way to proceed is to conduct archaeological excavations behind the terrace walls still visible in the landscape. It uncovers the pedosedimentary deposits that were preserved from erosion due to the presence of the wall. Often, the archaeologist finds under the current field, a more ancient terrace covered by colluvial and/or anthropogenic deposits (e.g. in the present case-­study on the Carlit Mountain). The primary chronological information is provided by the stratigraphic study of these superimposed phases of terracing. The interpretation of the soil and palaeosol related to the terraces‘walls is conducted on the field. It is then completed by a sampling of charcoals from soil horizons for radiocarbon dating. In appropriate environmental conditions and when charcoal is lacking, Optical Stimulated Luminescence dating on sediment layers or on palaeosol horizons (cf. Arnoldussen, this volume) may also provide a terminus ante quem for the buried terrace, but this method is still expensive compared to radiocarbon dating. When the terrace field still visible in the landscape is abandoned but not covered by sedimentary deposits, dating arguments may be provided from the interpretation of the soil profile in close relation to the construction techniques of the wall. As the cultivator builds the wall of their terrace, the sediments turned upside down are already pedogenetized. This precise moment corresponds to what the soil scientist calls ‘une remise à zéro de l’horloge de l’évolution pédologique de ces matériaux’ (Poupet 2009: 153): the point zero of the pedogenesis (soil formation) of the sediments behind the terrace wall. The archaeologist has to determine this moment on the time scale as it corresponds to the date of the construction of the wall. On the Cycladic island of Delos, the excavation of a terrace revealed a heap of stones at the foot of the inner face of the wall to ensure drainage (Poupet 2015). The presence of these stones means the start of a new pedogenesis at the moment of the construction of the terrace wall, as the builders had to move the sediments down to the foot of the wall to be able to put down these drainage stones. The characteristics of the soil profile, that evolved protected behind the wall, allow the pedologist to give an estimate of its age and, in so doing, an estimate of the age of the terrace field.

6.1.2 B  ronze Age Fields in the Mediterranean Mountains The archaeological knowledge of Bronze Age croplands in the Mediterranean mountains is very limited compared to the rich bibliography on pre-­and protohistoric fields in Northern Europe, while data in Central Europe is equally scarce (although valuable work had been done on early farming

R. Harfouche and P. Poupet

inferred from paleobotanical and micromorphological data; Bogaard 2004). Particularly in the Channel-­North Sea region famous British and Danish studies were done since the 1980s (Bradley 1978; Lerche and Steensberg 1980; Fleming 1988; Vinter, this volume; Johnston, May McOmish, this volume), extended by recent archaeological and LiDAR data (Løvschal and Holst 2014; Arnoldussen and Vries 2017; Arnold, this volume, Arnoldussen, this volume). In this light, the spectacular agrarian structures dated to the Late Neolithic (c. 6.2 ka cal BP) and Early Bronze Age that were uncovered in southern Italy under the deposits caused by the activity of the volcanic complexes of Somma-­Vesuvius and Campi Flegrei remain exceptional (Saccoccio et  al. 2013; Saccocio, this volume). In the Alpine mountains, the often cited protohistoric rock engravings of Val Fontanalba in Mont Bego and the Bedolina rockfaces in Val Camonica represent probably plots and field systems connected by paths (Blain 1998; Magail and Giaume 2005). Recent archaeological work has been done on some aspects of the exploitation of the Alpine mountain resources, but soils and possible existing croplands were not investigated (Walsh et al. 2007; Mocci et al. 2008; Walsh and Mocci 2011). Similar studies, linking archaeological excavations on pastoral huts and paleobotanical investigations, were conducted in the Massif Central mountains targeting the relations of societies to livestock and to breeding practices (Miras et  al. 2003, 2004; Servera Vives et  al. 2014). The remarkable multidisciplinary studies that were conducted in Spain, integrating archaeological excavations of terraced fields with pollen analyses, concerned the history of agricultural landscape, though not earlier than Roman times (Orejas and Ruiz del Arbol 2015), while medievalists attributed the shaping of the Spanish Mediterranean mountains to Arabo-­ Andalousian populations (Kirchner 2015). It is uncontested that the slopes of the Balearic islands and of the Iberian Peninsula were cultivated at this period. Nevertheless, the presence of several Bronze Age archaeological sites poses the problem of the existence of more ancient crop fields. The hypothetical relation between Bronze Age settlements and terraces should be investigated, yet unfortunately the fields were never excavated.

6.1.3 Bronze Age Agricultural Terraces This lack of archaeological data on pre-­and protohistoric terraces is shared by many countries where excavations of cropland landscapes are lacking. On the hilly slopes of the city of Nîmes in Southern France, terraces with stone walls and a field with ploughing marks were excavated but they date from the Iron Age (Poupet and Harfouche 2000, 2007; Harfouche 2007: 122–26). The association of field systems to Bronze Age societies in Cycladic Greece is often based on the density of settlement through history (Harfouche 2007:

6  Terraced Crop Fields in the Eastern Pyrenean Mountains (France): The View from Pedoarchaeology

150), but seldom do they rely on the archaeology of croplands. For the Cycladic island of Delos that was first inhabited from Early Cycladic II period (c. 2500 bc) up to the Late Bronze Age (c. 1500 bc) a study linking soil science to archaeology gives relevant evidence suggesting the existence of a phase of construction of terraced croplands in the Southern part of the island predating the Classical Period (Poupet 2015). Bronze Age agricultural terraces were moreover discovered on the islands of Pseira (Crete), of Kythera (between Peloponnese and Crete) and—albeit on a more modest scale—on the nearby small island of Antikythera (Betancourt et al. 2004; Krahtopoulou and Frederick 2008; Bevan et al. 2013). In the two latter islands, the stratigraphic and the pedological observations of the sections behind the walls were systematically completed with analysis on samples from the soil horizons for granulometry/particle size, magnetic susceptibility, and calcium carbonate content. Dating relied on radiocarbon analysis of charcoals, cross-­ checked with pottery fragments when available in the pedosedimentary deposits. OSL dates were also applied on the island of Antikythera. The multidisciplinary study that was led on this small island illustrates very clearly an important methodological aspect: soil analysis in the laboratory do not control the results of the macroscopic observations made in the field, but they can be performed to confirm them. The most relevant analyses on the palaeosol horizons (Ab/B/R) linked to the Bronze Age terrace involve magnetic susceptibility and loss-­on-­ignition (Bevan et al. 2013: 268 fig. 12). On one hand, the result of the magnetic susceptibility shows an enhancement in the lower part of the palaeosol suggesting the presence of a B-­ horizon of accumulation, that was already detectable on the field by the observation of the structure and of the colour of the horizon (the ferromagnetic elements are responsible of the reddish colour visible on the soil profile). On the other hand, the results of the loss-­on-­ ignition analyses show a predominant quantity of organic matter in the upper part of the palaeosol  – suggesting the presence of a buried A horizon – that was easily recognizable from the macroscopic observation of its dark colour and of its structure and texture. The OSL dating of the B-­horizon of the palaeosol indicates that the sand grains from this deposit were last exposed to sunlight between 1891 bc and 875 ad (ibid.). For this part of Europe, iconographic evidence is available as well. The wall-­painting Miniature Freize of Akrotiri (West House, Room 5, south wall, Thera, Crete), is a rare example of a pictorial representation of a terraced Mediterranean landscape dating back to the Late Bronze Age (Harfouche 2007: 152; Fig. 6.1). It had not been completed at the time of the volcanic eruption in the seventeenth century bc (Doumas 1999). The fresco depicts a series of hills suggesting a spatial organization of different parts of the mountainous territory behind the coastal towns. The closest

73

slopes are pastures with herds of cattle, flocks of sheep and goats accompanied by herdsmen and shepherds who drive the ovine and caprine animals into a circular pen. In the section of the frieze called Flotilla, wild animals stand on the farthest reliefs illustrated by a scene of deer being chased by a lion. The cultivated fields are situated immediately on the outskirts of the town. A terraced system formed by five visible levels of fields is built up against the slope in a position dominating the shoreline, toward which the boats are navigating. The perfectly aligned plantations are visible on the plots, but it is impossible to identify the plant species. Each field is supported by a wall and the whole system is surrounded by an enclosure-­wall (as currently in Mediterranean landscapes, possibly constructed from stones moved in field clearance). The nearest archaeological field systems to the mountainous Western Mediterranean region were uncovered in the Alps. There, soils with marks of cross-­ploughing are dated back to the Neolithic, but it is not always easy to assess whether these marks are the product of prehistoric, or later, agriculture (Poupet and Harfouche 2007). Either way, in the Swiss Alps, terraced fields seem to have been built since the second millennium bc (Cantons of Neuchâtel, Valais, Central Grisons and Engadine; Perret 1950; Zoller 1998: 161–64), but unfortunately the study led by A. Abderhalden-­Raba in Lower Engadine (Ramosch, Canton of Grisons) using geophysical and soil analysis data to demonstrate a Bronze Age terraced system is still unpublished. The above brief and non-­exhaustive overview shows that the Mediterranean mountains were largely neglected by archaeologists of croplands and fields, despite the fact that paleobotanical data exists for contemporary settlements. This lack of interest in these regions is also due to a certain bias in the data, as developer-­led archaeology (that brings the highest quantity of materials in Western Europe) is largely confined to lower altitudinal ranges (e.g. Poupet 1998; Poupet and Harfouche 2000, 2007). In historical and archaeological literature, ancient croplands are traditionally reconstructed in the plains and in the valley bottoms where the best fertile soils are assumed. The presumed usage of Mediterranean mountains is often limited to forest exploitation and pastoral activities, as in the case of the Pyrenees (Carozza et  al. 2005). This narrow perception that reduces mountainous landscapes to a stock of timber, game and grass, prevents the consideration/detection of other kinds of relationships between nature and past societies. Nevertheless, other approaches focusing on non-­restrictive patterns were developed at the same time and have changed the perspectives on prehistoric agriculture in the Alps (Martin 2010; Jacomet et al. 2016). In both mountains, the Pyrenees and the Alps, the teams employed a multidisciplinary approach, with a wider range of disciplines for the latter. What made the difference is that the team working in the Alps from the start of

74

R. Harfouche and P. Poupet

Fig. 6.1  The Late Bronze Age wall-­painting Miniature Freize of Akrotiri (West House, Room 5, south wall, Thera, Crete) with detail of a terraced landscape. (© Photo R. Harfouche)

the study dropped the paradigm that states mountainous landscapes are exclusively devoted to pasture because their soils are unsuitable for cultivation. There, the data was considered and discussed from an unprejudiced point of view.

6.1.4 Archaeology and Soil Science The aim of this contribution is to show that agricultural lands and the communities which shape them can also be examined from the perspective of cropfield archaeology; an approach that relies on the integrated study of archaeological remains as well as studies of soils and palaeosols. Such a multidisciplinary approach, linking archaeology with soil science, brings to light the long-­term developments of agricultural production systems in tandem with the long-­term environmental changes influenced by human activities. Fundamentally, sediments and soils underpin agricultural production processes and moreover can provide repositories of shifts and changes in landscape usage and the management of natural resources. Care is not only devoted to the study of pedological profiles but also to soils and palaeosols linked to archaeological stratification in order to understand

human impacts (extension and abandonment of cultivated land, forest clearing, breeding, etc.) on past environments and to highlight the mechanisms involved in landscape-­and soil degradation due to biophysical factors (mainly climatic events/variations). This archaeoagronomical approach builds from an innovative method that was already applied to several Mediterranean regions, in France (Languedoc, Massif Central, Pyrenees and Corsica), in the Greek Cyclades (island of Delos), in Jordan and in Mount Lebanon (Poupet 1998, 2015; Harfouche 2007; Poupet and Harfouche 2000; Harfouche and Poupet 2013; Harfouche et  al. 2015). The pedoarchaeological approach advocated here allows researchers to date terraces and to reconstruct the evolution of the agroecosystems by emphasizing and studying the interactions between socioeconomic factors and natural variables, but also through the integration of agronomic variables (soil protection by retaining walls, vegetation type, manuring, drainage and irrigation). The fieldwork is based on several archaeological excavations behind the walls of terraces and on the interpretation of the stratigraphy, soil and paleosol profiles related to these terraces. The pedological work relies above all on the macroscopic observation and on

6  Terraced Crop Fields in the Eastern Pyrenean Mountains (France): The View from Pedoarchaeology Carlit

France

Cerdagne

l ro Ca

Andorra

Campcardos Enveitg

Llivia Llo

e gr Se

the interpretation of soil features on the field (type of horizons, nature of transitions between them, textural and structural characteristics of each horizon, colors using the Munsell Chart, pedofeatures). Comparisons are made between different soil profiles in various geomorphological contexts: natural/anthropogenic soils, soils of ancient/current terraces, soils developed under a coniferous forest/grassland, et cetera. It is completed by radiocarbon dates on charcoal mainly sampled from the upper horizons of palaeosols. Depending on the problem to resolve and on the environmental conditions, additional sampling can be done for chemical analysis of horizons (e.g. Active Carbonate-­ CaCO3, Organic Matter-­OM, pH, Nitrogen-­N, Phosphorus-­P, Potassium-­K), micromorphology, sedimentology or palaeobotany (e.g. anthracology, phytholits, palynology, carpology). Nevertheless, it is not necessary to implement a systematic broad spectrum of expensive analyses to demonstrate the scientific quality of the study. If the project objectives are precise, the use of wide ranges of analyses is not always imperative. It will not by default bring more credibility to archaeological conclusions obtained by an expert ­macroscopic observation which answers  – in most of the cases – the archaeological question asked. Besides, let us not forget that the stratigraphic interpretation strongly influences the choice of sampling and of the types of analyses that will make sense to perform in a laboratory. On our Pyrenean project, mindful of the (limited) assigned financial means to the archaeological research project, we made the choice to conduct analyses in the laboratory only if it was necessary. The heart of the research on ancient terracing is formed by the pedoarchaeological interpretation on the field. Soil science is a natural science – a science of observation – as it consists primarily of a macroscopic interpretation of soil profiles. This already gives information on various agronomic and environmental aspects like hypotheses on manuring, irrigation, the general type of vegetation cover, or the pedological consequences of the construction techniques (Poupet 2009; Harfouche 2015). Essentially, it is the set of converging macroscopic observations that allows us to argue in favour of a particular hypothesis. It is also important to stress that the timeframes of archaeoagronomical studies are much more precise than those of geomorphological analyses generally undertaken within geoarchaeological studies. Consequently, the temporality of pedoarchaeology is much closer to that of human societies. The relevance of our investigation is however not confined to the past: through the study of relict landscapes, input for sustainable management of soil and water resources in regions that are threatened or already affected by desertification and land degradation may come to the fore. Although the approach followed here is diachronic, in what follows we will focus on the Later Prehistoric and Protohistoric periods. Particularly the Bronze Age period

75

Puigcerdà

gre Se

Puigmal

0

Spain Serra del Cadi

Néolithique-Bronze

Molina

1300

1600

2000

5 km Bronze 2600 m

Fig. 6.2  Location map of the Carlit Mountain in Cerdanya (France-­­ Spain) and of the Neolithic to Bronze Age settlements. (© R. Harfouche and P. Poupet 2008)

appears to be crucial in understanding the genesis of extensively terraced landscapes, including those at high altitude.

6.2

A mountain Only for Shepherds?

Cerdanya is an historical territory divided between France and Spain (Fig. 6.2). It corresponds to the upper valley of the Segre river at 1200 m altitude. It forms an elevated plain surrounded by villages located on the foothills of 3000 m high mountains. Production activities are now distributed according to altitude: the upper plain, up to the village zone, is cultivated with cereals (Triticum aestivum, T. spelta, Secale cereale, Hordeum vulgare) and fodder crops (e.g. Secale cereale with Vicia sativa or V. villosa: Ruas and Rendu 2005: 152). The slopes above the villages are devoted to cattle grazing and are organized according to the vegetation stages and to the different pastoral activities. From 1200 to 1400 m altitude, plant formations of the supramediterranean stage are dominated by Quercus pubescens flora. This is followed by the Pinus sylvestris mountainous stage up to 1800  m, which corresponds to the hamlets zone. There, houses are surrounded by woodlands, pastures and presently abandoned terraced fields that are used for cattle rearing since the middle of the twentieth century (Rendu 2003; Conesa 2012). This is the intermediate grazing zone with pens and stone enclosures. At a higher level, the subalpine stage, between 1800 and 2200  m altitude, is a characteristic zone where broom (Cytisus sp.) and juniper (Juniperus communis, Juniperus nana) vegetation dominates, with bearberry (Arctostaphylos uva-­ursi) and Rhododendron sp., and also small woods of local pines (Pinus uncinata). The highest areas, above 2200 m altitude, are covered by the extending lawns (Festuca gautieri, Deschampsia flexuosa, Trifolium alpinum) of the asylvatic alpine stage. Our study area on the

76

Carlit granitic mountains, is situated between 1500  m and slightly over 2000 m (coniferous forest phytoclimatic level) altitude, in a forest that is today reduced in size to interspersed groves. The Cerdanya mountains were previously studied by a team of historians with an anthropologist, an archaeologist, a biologist and geographers, through fieldwork and using historical written and iconographical records (e.g. Davasse et al. 1997; Vannière et al. 2001; Rendu 2003; Conesa 2012). For fifteen years, these researchers wrote the long history of the mountain being limited in use to solely forestry and livestock rearing activities (mostly sheep and goat) since the Neolithic period (Rendu 2003). They see it, in ancient times as it is today, made of large areas devoted to herd breeding, a summer pasture for sheep and goats, oxen or horse transhumance, depending on the historical period. From the Neolithic to the Middle Ages, archaeological data underline the presence of sheeps/goats in the mountain, while from the fourteenth to the nineteenth centuries written sources describe the herd as comprising a majority of sheep (and goats) with 10% of oxen and 5% of equidae; oxen constantly increasing since the mid-­twentieth century become dominant today with some horses (Rendu 2003; Conesa 2012: 344). The current landscape, with its open vegetation of broom and juniper, is the result of forest degradation in Modern times instead, according to historical ecology (Davasse et al. 1997). It is not only due to wood exploitation, but mainly to forest fires for grazing as cattle farmers regularly burn the vegetation on the slope to improve the quality of the grasslands. As pastoral activities decline, a slow forest regeneration can be noticed, particularly in riparian forest. Several stone constructions, interpreted as pastoral sites (enclosures, stone corridors for milking ewes and huts) were discovered during archaeological surveys and were partly excavated in the subalpine and alpine levels (i.e. over 1800 m; Rendu 2003). As stone buildings were very ruined (in most cases only the first stone course is still visible), they were identified as pastoral huts through comparison with the still standing modern constructions that have a similar small area inside the first stone course. At the same time, palaeobotanical data obtained from pollen and charcoal, cross-­checked with the density of pastoral sites, led to a history of the vegetation cover which was linked exclusively to breeding activities and climate variations (Davasse et al. 1997; Galop 1998; Vannière et al. 2001; Bal et al. 2008). The forest is opened-­up as soon as 4000 bc, but mainly since 3300 bc and throughout the Late Neolithic and Bronze Age periods (Galop 1998: 65–69; Vannière et al. 2001). Although pollen of cereals are present at 2110  m height, palaeobotanists postulate that crops were actually grown below, in the valley fields, but that pollen was brought up into the mountains via the fleece of sheep (Vannière et al. 2001: 32, 37). The archaeological and paleobotanical investigation of agricultural plots has mainly

R. Harfouche and P. Poupet

been confined to areas situated at lower altitudes (Harfouche 2010, 2015), as it is usually assumed that ancient societies exploited these regions with priority for their food economy (Harfouche 2007). A large majority of the archaeological sites surveyed are identified as Neolithic and Medieval/Modern occupation sites and only two sites are dated to the Bronze Age period (for the total of 20 sites excavated; Rendu 2003). Moreover, one of these concerns a big square house (situated on the Pla de l’Orri, at 2100  m altitude) that is dissimilar to a shepherd’s hut, but looks rather like a Bronze Age farm (infra). During the excavation of a hut at almost 2000  m altitude, medieval carbonized remains of cereals, mostly rye, were uncovered, and interpreted as sheaves (due to the presence of by-­products of cereal processing like chaff remaining from sieving before storage) from nearby cultivation (Ruas and Rendu 2005: 154–55), which was otherwise disregarded as being marginal and temporary. How are we to ­reconcile a strongly pastoral interpretation of the mountain’s crop production with this evidence of local plant cultivation? Is it reasonable to imagine herders of earlier periods climbing the mountains whilst carrying their sheaves harvested previously in the valley? In 2003, the pedoarchaeological research developed in the area (Harfouche 2005, 2006) started to reveal rapidly growing sets of data for the Bronze Age, hinting at a more permanent and intensive human agricultural occupation. These results could be positively reconnected to the evidence coming from pollen and charcoal (Vannière et  al. 2001; Ruas 2003; Ruas and Rendu 2005; Bal 2006), in suggesting that cultivation activities had left their traces even at high altitude, over 1600  m asl. Therefore, after 2007 the model of occupation was progressively changed, and the existence of chronological phases with permanent settlement supported by a combination of agricultural and pastoral activities became accepted, also in accordance with other sources (Harfouche 2007; Martin 2010). This view has come to be shared also by the researchers formerly supporting a model of occupation of the mountain area either relying only on a solely pastoral economy, or making use of some form of shifting agriculture (Bal et  al. 2008; Rendu et  al. 2009, 2015). Thus, the paradigm that the mountains did not support multi-­year agriculture at the mountainous superior levels or subalpine levels, has now been left.

6.3

Under the Grassland, the Crop Fields

The search for ancient cropfields by our team started in 2003 with surveys, paying close attention to soils, landforms, and non-­anthropogenic alignments of granitic blocks among the rocky outcrops or linear ridges in the pastures (Figs. 6.3 and 6.4). The latter features are the remains of ancient terraces

6  Terraced Crop Fields in the Eastern Pyrenean Mountains (France): The View from Pedoarchaeology

77

Fig. 6.3  Alignment of granitic blocks belonging to the wall of an abandoned agricultural terrace on the Carlit Mountain (Orri d’en Corbill, 1950 m asl). (© Photo R. Harfouche 2003)

Fig. 6.4  Detail of the exterior facing of a terrace wall (terrace 1 on Fig. 6.5, valley of Brangoli; 1700 m asl). (© Photo R. Harfouche 2009)

walls enclosing old abandoned fields  – which were erased from the cultural memory of the last few generations of the local population, for whom the mountains have always been a land(scape) of pastures. We looked at the Cerdanya mountains using a novel approach, entwining the archaeology of these enigmatic alignments of granitic rocks with soil history as part of a project directed by R. Harfouche for the French Ministry of Culture and Communication. We recorded several sets of terraces built on the sunny side of the watershed amounting to almost 40% of slope. These large terraced areas of more than 100 ha in size are found between 1700 m and 2000 m altitude. The courses of stone of the retaining wall (terrace face) were built by the farmers on top of glacial deposits including granite rocks, and also between natural occuring boulders (Fig. 6.5). The terraces are situated on the slope where the soils are originally the thickest due to the presence of the

boulders: up to 80 cm thickness for the soil of a terrace compared to a maximum of 30 cm for a natural Rankosol (according to the French soil classification in Baize and Girard 2009: 276) on the same slope. For the first time in the region, our pedoarchaeological approach took into account soils and palaeosols as pedosedimentary archives to reconstruct the long environmental history. The selection of the locations investigated proved essential. We targeted those spots where the pedosedimentary deposits were thickest in order to maximize their information value. Trenches were excavated using a mechanical spider digger that is well-­adapted to rocky steep slopes and that allows documentation of deep sections into the pedosedimentary cover. Yet, every archaeological trench that was opened behind a (ruined) wall revealed the presence of a more ancient terraced field with its associated palaeosol, buried under the current terrace (Figs. 6.6 and 6.7). Thirty-­six

78

R. Harfouche and P. Poupet

oli ng B ra

Fig. 6.5  Map of a (presently abandoned) terraced area with the location of the excavations (valley of Brangoli; 1700 m asl). The walls of the terraces are built in connection with natural boulders. (© R. Harfouche 2009)

n ou (m

23

pen

n

t

ai

22

to r

21 19

20

r ent)

18

track

Cerd 35 7

12

16

13

17

14

11 1

Cerd 32

15 5

2

6

Cerd 34

Cerd 33

8

3 4

10

9

7

pen

pen 0

Fig. 6.6  Excavation behind the terrace wall revealing the presence of a more ancient agricultural terrace wall with its palaeosol buried under the later field, currently abandoned (Cerd.3.2; 1950 m asl). (© Photo R. Harfouche 2003)

50 m

pedoarchaeological trenches were dug, letting us highlight the parcelling of the landscape, thanks to the recurrent features identified (Harfouche and Poupet 2013). Close to the pastoral settlements of various chronological periods, several abandoned walls had been built to retain soils (Harfouche 2005, 2006, 2015). The terraced fields on the steep slopes were later replaced by a vegetation cover of Cytisus and Juniperus with a forest regeneration of Pinus uncinata. Two major points are relevant to the stratigraphy/soil profile behind the terrace wall: the quality of palaeosols and the existence of a long history of anthropogenic versus pedosedimentary phenomena. The mountainous landscapes were shaped by the interconnection of successive morphodynamic processes (such as accelerated or slow and steady erosion versus sedimentary or pedosedimentary colluvial deposits), with layers of different pedogenetic phases overlapping with one another (polycyclic or polygenetic soils), and ultimately supporting human constructions related to agricultural use. Excavated solums are often composed of a palaeosol with a very dark organo-­mineral horizon (Ahb/Sal/C/M; Ahb/M; Ahb/Sal/C/R; Ab/S/R) connected to a lower wall of terrace; the palaeosol is covered by pedosedimentary deposits with one or two pedogenesis separated by one (or more) erosion phase. The most recent soil related to the upper terrace is generally composed of a succession of Oh/Ah/(E)/S-­horizons or O/Eh/E/BP on top of the Ahb-­horizon. The soils that

6  Terraced Crop Fields in the Eastern Pyrenean Mountains (France): The View from Pedoarchaeology Fig. 6.7  Section of an excavation trench behind an abandoned terrace, showing the former field and its palaeosol (Cerd.24; 1750 m asl). (© R. Harfouche and P. Poupet 2007)

79

SW

NE

terrace 2

Ab

Ah A Sal

terrace 1

20 cm

Sal

Ab

C

unexcavated

developed in the context of the ancient croplands appeared to be quite different from the non-­anthropogenic soils, in the fact that they did not have the defining pedological characteristics of the soils that naturally develop on the granitic rock at this altitude, such as Alocrisols (A/Sal/R) and Rankosols (A/R; Baize and Girard 2009). Not only do they have a very dark coloured Ab-­horizon (generally a Ahb), but also the structure of the Ahb-­ horizon has often more developed aggregates than the non-­anthropogenic A or Ah-­horizon in natural solums. The terraces soils were also more varied with the presence of Anthroposols (possibly of a plaggen type; infra), Organosols (OH/Aho/C) and Podzosols (A/E/BP/C) at different stages of development, forming a significant pedological mosaic in terms of agronomic practices: e.g. Organsols with their thick organo-­mineral horizon (Aho) due to an enrichment in organic matter, or some Podzosols that developed under irrigated terraces in Roman times and evolved as irrigation is responsible of soil leaching (Ahb/E/ Gr), while the non-­irrigated soils in the immediate environment were Organosols and Alocrisols (Harfouche 2015). Plough-­marks could not be detected on the top of palaeosols due to the use of the mechanical spider digger to open the trenches down to the rock. Observations could only be done on the pedosedimentary profile of the deep sections behind the terrace wall. Manual excavation would have taken too much time regarding the funding allowed to the project, and we chose to increase the number of test-­pits to get a wide range of stratigraphical data in varied environmental and topographical conditions instead.

6.4

The Spread of Terraced Field Systems

Soils can act as archives of landscape changes, and through integration of pedological characteristics of palaeosols and soils with archaeological data (stratigraphy) and radiochronology (14C AMS dating) it is possible to date the cropfields. Dated charcoal incorporated in organo-­mineral horizons of palaeosols linked with ancient terraces highlight major periods of agricultural expansion (at the expense of the forest) from the Neolithic to the Early Middle Ages (Figs. 6.8 and 6.9; Harfouche 2015). The most prominent period of terrac-

Fig. 6.8  Charcoal lens on the top of a palaeosol behind the wall of an ancient terrace (Cerd.12; 1800 m asl). (© Photo P. Poupet 2004)

ing at the carlit Mountain seems to be from the Late Neolithic/Early Bronze Age, around 2890–2620  cal  BC (4185±35 BP, Vera-­3255) up to 2028–1874 cal BC (3575±35 BP, Poz-­ 9153) (Harfouche 2005, 2010; Harfouche and Poupet 2013). Another period of lesser prominence is the Late Bronze Age. A large square building of about 40  m2 (Rendu et al. 2009) was excavated on the Pla de l’Orri, not far from the crop fields at 2100  m altitude. Two phases of occupation dating back to the Bronze Age were inferred from the radiocarbon dates (1876–1625  cal  BC and 1426– 1126  cal  BC; Rendu et  al. 2012). The plan of the earlier phase of settlement is unknown due to the architectural elements from latter occupation (Rendu et  al. 2012), yet this construction appears to be different from the 10m2 to 20m2 structures traditionally identified as pastoral huts (e.g. Rendu 2003: 348–50; 550), and it looks rather like a Bronze Age farm. The main aspects hinting at a permanent farm-­like occupation is the size of the construction (which is much larger than the summer pasture shepherd’s huts, as mentioned earlier), the bi-­absidial plan in the fifteenth to twelfth

80

R. Harfouche and P. Poupet cal. y. BC/AD 7000

6000

5000

4000

30 samples dated for 17 pedoarchaeological excavations

8000

3000

2000

1000

-1+1

1000

2000

Late Neolithic Early Bronze Age Middle Bronze Age II Late Bronze Age

Late Iron Age Roman

8000

7000

6000

5000

4000

3000

2000

1000

-1+1

Early Middle Ages

1000

2000

Fig. 6.9  Distribution of 14C AMS dates on charcoal concentrations sampled from the upper horizons of the lower terraces palaeosols, behind the terracing walls. (© R. Harfouche and P. Poupet 2011)

centuries bc with a large wooden posthole (Rendu 2004 report unpublished) that has no equivalent on the other excavated sites of the Carlit Mountain, and the presence on this site of Bronze Age pottery sherds (Rendu et  al. 2012; Bousquet et al. 2012: 155) that are very scarce in all the pastoral huts (Rendu 2003). Unlike agricultural remains from historical periods (e.g. radiocarbon dates do not give evidence of Roman terraces over 1700 m asl; Harfouche 2015), Bronze Age terracing is present nearly everywhere where topographical conditions allow it, from 1690  m to 2000  m altitude (Fig.  6.10). Obviously, the location of terraced fields did not depend on elevation. It is also clear that Bronze Age communities chose the soils they wanted to cultivate as they neglected other parts of the slope. They preferred a sunny exposure and the locations with naturally occurring boulders that are linked to the originally thickest soils. This selective usage is not a minor issue in the history of agriculture, as it shows that the construction of terraced fields at these altitudes and on these soils was not necessitated by a demographic growth that generated a need for more arable. On the contrary, Bronze Age terracing appears to be a deliberate and strategic practice that was based on thorough knowledge of the agricultural potential of mountainous soils. The evidence for prehistoric arable at high altitudes, among which the wide grasslands, throws a new light on

plant cultivation patterns (Harfouche and Poupet 2013). Local Bronze Age diets did evidently not consist solely of dairy, meat and wild plants gathered in the mountains. Barley (Hordeum vulgare, Hordeum vulgare var. nudum) and wheat (Triticum aestivum/durum, Triticum dicoccum) were consumed in a Bronze Age cave (Cova d’Anes, Prullans, 1200 m asl) that was inhabited from 1500 to 1200 bc, on the sunny slope of the Campcardos Mountain (Fig.  6.2), West of the Carlit Mountain (Alonso i Martínez 1995). Poaceae like barley (Hordeum vulgare) and naked wheat (Triticum ­aestivum/ durum/turgidum) comprised a large part of the diet of another Bronze Age site in Cerdanya; the settlement of Llo at 1630 m asl on the Puigmal Mountain (Fig.  6.2), together with Fabaceae (species of the pea family; Pisum sativum) and were most likely cultivated in the surrounding environment (Erroux 1983; Ruas et  al. 2009). Rye (Secale cereale) is a common crop in Roman times and in the Middle Ages (supra; on the Carlit Mountain cultivated around 1900 m asl) but it could have also been cultivated in Southern France in the Neolithic and the Bronze Age (as suggested by palaeobotanical studies; Marinval 2009), although the rye seeds that were uncovered on the nineteenth century bc site of Llo were interpreted as wild plants (Ruas et al. 2009). Pollen of cereals are recorded on the Carlit Mountain at 2110 m height, in the peat-­bog of Pla de l’Orri, as soon as 3500 to 2300 cal. BC (Galop 1998: 67). That is consistent with an increase of fires

6  Terraced Crop Fields in the Eastern Pyrenean Mountains (France): The View from Pedoarchaeology

81

Fig. 6.10  Extension of the proved Bronze Age terraced areas in their agropastoral context on a mosaic of orthophotographs from the IGN (National Geographic Institute). (© R. Harfouche and P. Poupet)

recorded from the microcharcoal analysis in the peat-­bog (Vannière et al. 2001: 38). As we know since 2003 – after these palaeobotanical analysis were published – that a set of terraces existed in the Late Neolithic/Bronze Age, we suggest that these fires and the evidence for cereals were related to farming activities in the mountain like wood clearing for cultivation. Furthermore, microcharcoal sampled from the soils and palaeosols (notably charcoal-­ rich horizons and charcoal lenses on top of Ab horizons) of the terraces we excavated give also evidence for deforestation by fire at this period (Bal 2006; Bal et al. 2008). On one hand, autumnal sowing, vegetative cycle and phyllotherm are consistent with the climate around 2000 m altitude. On the other hand, the warmer and dryer Bronze Age climate may have also played a role in the shaping of the mountain for crop production and in yield stability especially during the Early Bronze Age as the glaciers retreat in Europe from 4200 to 3800 BP (Mayewski et  al. 2004: 250). The 3500–2500 BP period (Mayewski et al. 2004) was another warm and dry climatic stage in Late Bronze Age, ending in North-­West Europe around 2700 BP/850 cal BC (van van Geel and Magny 2002). In the Northern Pyrenees, major snowmelt periods were also recorded around 4210 and 2640 BP (Simonneau et al. 2013). Recent archaeological studies support this hypothesis about the likely influence of climate trends on the shaping of

agricultural landscapes at high altitude. Terraced fields were excavated east of our study area on a sunny slope of the Carlit Mountain at 1650  m asl (Vilalta; Rendu et  al. 2015). The main phase of terracing dates back to the Middle Ages and was linked to the development of a medieval village but the stratigraphy and the radiocarbon dates suggest an earlier phase of cultivation from the end of the Middle Bronze Age to Late Bronze Age (c. 1350  cal  BC; Rendu et al. 2015: 474–75).

6.5

 he Significance of Mountainous Crop T Fields in Terms of Social and Territorial Organization

A systematic exploration of soil archives has revised and refined our views of some well-­established patterns of agriculture in mountainous regions. Archaeological and palaeobotanical analyses of agricultural practices in these zones had hitherto mainly targeted the low-­lying areas, as it is traditionally assumed that ancient societies exploited primarily these regions in their food economy. We argue, however, that a multidisciplinary approach, linking archaeology to soil science, and a long-­term perspective, may foreground the reality and importance of arable situated on higher-­ altitude

82

(>1700 m) slopes and its subsequent impact on the organization of territories and communities in such landscapes.

6.5.1 S  oil Improvement at High Altitude and Agricultural Know-How Evidently, Bronze Age cultivators knew how to protect their crops from soil erosion by building retaining walls that countered colluviation. They moreover could actively improve the fertility of such locales. The mineralization of the organic matter in the upper horizons of soils developed on mountainous acid rock is poor and very slow, due to the altitude (Legros 2007: 68–69; Jamagne 2011: 389–92). In such environmental settings, soils belonging to the groups of unsaturated Organosols and of humic Alocrisols develop. At Orri d’en Corbill, c. 1950 m height, a Late Bronze Age drystone hut (3107±45 BP, 1491–1261 cal BC, Ly-­8222; Rendu 2003: 260–64) with two small enclosures adjoining the Bronze Age terraced fields were excavated (Harfouche 2005, 2010). The pedoarchaeological trench we opened in the 30  m2 enclosure next to the hut revealed a soil of 1.20  m thick (2805±35 BP, 1050–830  cal  BC, Vera-­2952; 2825±35 BP, 1130–840 cal BC, Vera-­3251) with a very thick Aho-­horizon of more than 50 cm, full of organic matter (Harfouche 2015; Harfouche and Poupet 2013). This Late Bronze Age pedological profile proved quite different from the characteristics of the common non-­anthropogenic soils that usually developed on the granitic areas (e.g. Alocrisol, Rankosol, ocric Podzosol; Jamagne 2011: 388–95). Deliberate manuring could be inferred from the thickness of the organic horizon, but the project funding did not allow us to proceed to a geochemical detection. This soil may have been improved with animal derived material over many agricultural cycles. It could also be the result of repeated application of organo-­­ mineral materials taken from the upper horizons of the soils in the nearby woods, then this soil would be a Plaggen soil type (Legros 2007: 85; Langohr 2001). Archaeological ­excavations could not determine whether this small enclosure was a cultivated plot where crops were protected from livestock by a stone wall next to the hut or if it was only linked to the fencing-­in of animals. The accumulation of organic materials could have been the result of manuring that helped improve the (sustainability of seasonal) yields, particularly on the smaller cultivated surfaces. In another set of terraces we excavated, c. 1730  m height, the same type of soil developed on terraced fields that were cultivated from the second half of the Middle Ages (Harfouche 2015). Early Bronze Age manuring on terraced fields at high altitude was evidenced by the co-­occurrence of both geochemical and archaeological data in Mount Lebanon, where terraced fields were excavated nearby a third millennium settlement, c. 1200  m. altitude (Harfouche et  al. 2015). The presence of

R. Harfouche and P. Poupet

phosphorus concentrations in the palaeosols of the terraces correlates with the occurrence of degraded ceramic sherds as a component of urban domestic waste that was transported from the settlement and spread in the surrounding fields to fertilize them (cf. Arnoldussen, this volume). The same co-­ occurrence of manuring signal was evidenced on the Greek island of Pseira, where Bronze Age manuring (with systematic deposition of human or porcine derived fecal matter) relied on the detection of diagnostic organic chemical compounds in the soil of the terrace and on the presence of ceramic waste (Bull et al. 2001). This agricultural practice is another major observation brought to fore by the pedoarchaeological approach. It means that the conceptual misinterpretation of mountain agronomy being held back by soils so thin and poor that they only host grasslands, can be disproved. Moreover, the influx of organic matter into terraced soils provides information on the relationships between pastoral activities and cultivation of the mountainous slopes: livestock was grazed or penned in the croplands/stubble fields. This could take many forms, ranging from a rotation of crops and pasture on terraces, a long-­­ lasting grazing on the same terraced plot after the harvest (until the depletion of the fodder resource) or by a periodical practice of controlled stalling in a small pen – like the dry stone walled pen at Orri d’en Corbill – with a certain amount of litter to provide the maximum quantity of manure. Another way of integrating pastoral and cultivation activities could have been a controlled leading of herds as it is today, by following transhumance pathways up to the final pens that are situated at a higher altitude above the terraced fields. Then the sown fields on the way to summer pastures needed to be protected by temporary and removable enclosures of branches when the herd is grazing close to it, as it can be seen in many countries where pre-­ industrial agricultural practices are still in use (Harfouche 2005). In the Eastern Pyrenees fields extended up to 2000 m height, and over this limit only pastures seem to have been in use.

6.5.2 S  ettlement and Land-Use Patterns from Bottom-Up and Top-Down In the Cerdanya area, agriculture relied on terraced fields that were ‘built to last’ and the mountainsides definitely formed crop production areas even at high altitude. The supposed shifting cultivation inferred from archaeological data or ancient texts in Southern France is affected by a false debate due to methodological inadequacies and semantic inaccuracies (Poupet and Harfouche 2000; Vital 2008; Harfouche 2015). It was postulated that in Late Neolithic and Bronze Age Cerdanya settlements were divided into villages, hamlets and isolated farms as people lived in a few ‘important’ (without any precision on the type) sites located in the plain,

6  Terraced Crop Fields in the Eastern Pyrenean Mountains (France): The View from Pedoarchaeology

and that they moved periodically for slash-­and-­burn agriculture, with seasonal settlements in mountainous caves when they managed marginal resources (Rendu 2003: 422). In the plain, the slash-­ and-­ burn agriculture is supposed to be responsible for a horizontal ‘mobility’ of settlements related to the exhaustion of soil fertility. A ‘vertical’ mobility related to transhumance from the plain to the faraway mountainous grasslands, with semi-­permanent settlements and fields in the mountain, would have been in use in the seventeenth to twelfth centuries bc (Rendu et al. 2015). This proposed interpretation of alleged double (i.e. vertical and horizontal) mobility of settlement suffers from several weaknesses: there are no convincing arguments for a horizontal mobility (within the plain) as there are no Neolithic or Bronze Age archaeological site extensively excavated in the plain (that can be identified to a village rather than a hamlet or a farm). The vertical mobility (from the bottom zone to the upper mountain) that could be linked to a pastoral economy and to slash-­and-­burn practices was never recorded in archaeological and paleoenvironmental studies (Harfouche 2010). The ‘large semi-­ permanent settlements’ in the seventeenth to twelfth centuries bc in the mountain (Rendu et al. 2015: 476) refer to only one example: the farm-­like site at Pla de l’Orri, c. 2100 m asl, for which there is no scientific evidence of a non-­permanent occupation (Rendu et  al. 2012). The same team of archaeologists and historians recognize ‘permanent fields in a cleared area, cultivated intensively’ close to the Bronze Age site of Llo, c. 1630  m asl (Rendu et  al. 2015: 476, referring to carpology results (Ruas et al. 2009)). As a consequence of this they stated recently that ‘the evidence shows that in the second millennium cal. B.C., there was a specific management of various forms of agro-­ pastoral exploitation at different altitudes suggesting the existence of territorial organization’ (Rendu et  al. 2015: 478). Yet the arguments of the carpology study in favour of permanent fields on the site of Llo rely on the presence of by-­products of cereal processing (chaff from cereal sieving before storage) and on the correlation of ecological requirements of identified weeds with the surrounding environmental ­conditions. Identical arguments were used to infer a non-­­ permanent/shifting agriculture on the terraces of Orri d’en Corbill (Ruas and Rendu 2005; supra). Conversely, since the early seventies many studies have argued in favour of permanent cropfields, cultivated according to well-­known practices since the Early Neolithic in Central Europe (e.g. Bogaard et al. 2013; Styring et al. 2016) and during the Late Neolithic in the Alpine foothills and the Jura mountains (Jacomet et al. 2016). In such cases, a labour-­intensive management of plots and the use of manuring over long periods could alternate with short fallow periods during which the fields were grazed. The mapped sizes of protohistoric terraces on the Carlit massif suggest that these fields were built for a small com-

83

munity economy, as the terraced areas on the sunny slopes slightly exceed 100  ha in size. It correlates with farm-­like type of settlement excavated at Pla de l’Orri. Surveys in the mountains never provided archaeological evidence for a Late Neolithic/Bronze Age village or hamlet between 1700 m and 2000 m altitude where the terraces were found. Below that height, the Medieval and modern hamlets (Rendu 2003; Conesa 2012), standing around 1500 to1600 m asl, and their surroundings were never excavated. The limited extent of Late Neolithic/Bronze Age terracing is also an argument in favour of crop fields rather than cultivation or facilitation for animal fodder production, as the non-­terraced grasslands of the mountain were vast enough to provide sufficient winter hay. In this context, terraces appear as a counterpoint, a resource for significant production of subsistence cultivation of human food (starting with cereals) at a small economy scale. This minimal local subsistence could have been coupled with intensive herding in specific places like at Orri d’en Corbill where small enclosures were built next to a Late Bronze Age hut. Our recent investigations in the Carlit mountains shine a new light on the agricultural usage and relationships between the mountain’s upper ranges and the lower lands, between complementary pastoral and farming activities, supplemented by hunting and gathering. In spite of the great impact of geomorphological processes recorded in the deposits during the Holocene period, land-­use and the occupation of specific areas depended primarily on the capacity of early societies to adapt to these constantly evolving environments. These societies intervened through the control of erosion and the construction of terracing walls, as well as through active manuring. The territorial organization of slopes and high-­ altitude areas (previously deemed unsuitable for permanent settlement and intensive agriculture) shows the ability of Late Neolithic and Early Bronze Age small groups to adapt their internal social and economical structure to the living conditions in this particular milieu. Understanding long term pedosedimentary evolution and its impacts on archaeological features is instrumental in reconstructing ancient human settlement patterns and in understanding the society-­ environment co-­ evolution here, and could disprove old-­fashioned and reductionist models for early farming and mountain territories. Contrary to traditional views, there is no archaeological evidence that terracing higher-­ altitude zones resulted from a demographic growth at this time. By studying mountainous slopes in their own right (e.g. from the mountain itself) rather than from the lowlands looking up, our investigations have shown that this milieu was one of a wide range of productive and economically viable landscapes for Bronze Age societies. Therefore, it seems essential to take into account the pedoarchaeological data in any multidisciplinary study of soil-­vegetation-­­ terracing dynamics. If not, soil maps will only yield

84

information on the recent pedological cover with a lack of information concerning long term human-­ environment changes. Acknowledgements  The authors would like to thank the editors for the opportunity given to share their insights into the shaping of ancient agricultural landscapes, and to the reviewers for their valuable comments that helped them greatly in improving their contribution. This work was supported by the French Ministry of Culture and Communication under its research program 31 ‘Anthropisation et aménagement des milieux durant l’Holocène’; the French National Center of Scientific Research (CNRS) under the laboratory TRACES (UMR 5608, Toulouse).

References Alcaraz, F. (1999). Les terrasses méditerranéennes entre terroirs et paysages (nord-ouest du bassin méditerranéen). Thèse de Doctorat en Géographie à l’Université de Toulouse II. Alonso i Martínez, N. (1995). Estudi de Llavors i fruits dels jaciments arqueològics de la cova d’Anes (Prullans, la Cerdanya) i de la cova de les Portes (Lladurs, el Solsonès) i el seu context pirinenc. In Cultures i medi, de la prehistòria a l’edat mitjana, Xe Col.loqui internacional d’arqueologia de Puigcerdà, homenatge al Professor J. Guilaine (pp. 97–104). Puigcerdà: Institut d’Estudis Ceretans. Ambroise, R., Frapa, P., & Giorgis, S. (1993). Paysages de terrasses. Edisud: Aix-en-Provence. Arnoldussen, S., & de Vries, M. (2017). A plan in place? Celtic field habitation at Westeinde (prov. Drenthe, the Netherlands). LUNULA. Archaeologia protohistorica, 25, 79–89. Asins Velis, S. (2009). El paisaje agrario aterrazado: Diálogo entre el medio y el hombre en Petrer (Alicante). Valencia: Publicacions de la Universitat de València. Baize, D., Girard, M.-C. (coord). (2009). Référentiel pédologique 2008. Paris: éditions Quæ. Bal, M.-C. (2006). Constructions et dynamiques des espaces et des terrasses agro-pastoraux en zone intermédiaire des Pyrénées du Néolithique à nos jours (Cerdagne, Pays Basque et Pays de Sault). Approche archéoenvironnementale par la pédoanthracologie [online]. Thèse de Doctorat en géographie, Université de Toulouse. Available at https://tel.archives-­ouvertes.fr/tel-­00132105. Accessed 1 Dec 2017. Bal, M.-C., Harfouche, R., Poupet, P., Campmajo, P., & Rendu, C. (2008). Archaeo-environmental studies of cultivation terraces in the Enveig Mountain (Cerdagne) in eastern Pyrenees (France). Use of pedo-anthracology. In G. Fiorentino & D. Magri (Eds.), Charcoals from the past: Cultural and palaeoenvironmental implications, IIIrd international meeting of anthracology (pp.  11–23). Oxford: Archaeopress. Betancourt, P.  P., Davaras, C., & Hope Simpson, R. (Eds.). (2004). Pseira IX: The Pseira Island survey. Part 2: The intensive surface survey. Philadelphia: INSTAP Academic Press. Bevan, A., Conolly, J., Colledge, S., Frederick, C., Palmer, C., Siddall, R., & Stellatou, A. (2013). The long-term ecology of agricultural terraces and enclosed fields from Antikythera. Greece. Human Ecology, 41, 255–272. Blain, A. (1998). Les parcellaires incisés au val Fontanalbe (Mont Bégo). In A. D’Anna & D. Binder (Eds.), Production et identité culturelle (pp. 463–466). Sophia Antipolis: éditions APDCA. Bogaard, A. (2004). The nature of early farming in central and South-­ east Europe. Documenta Praehistorica, 31, 49–58. Bogaard, A., Fraser, R., Heaton, T.  H. E., Wallace, M., Vaiglova, P., Charles, M., Jones, G., Evershed, R. P., Styring, A. K., Andersen,

R. Harfouche and P. Poupet N. H., Arbogast, R.-M., Bartosiewicz, L., Gardeisen, A., Kanstrup, M., Maier, U., Marinova, E., Ninov, L., Schäfer, M., & Stephan, E. (2013). Crop manuring and intensive land management by Europe’s first farmers. Proceedings of the National Academy of Sciences of the United States of America, 110(31), 12589–12594. Boissinot, P. (1995). L’empreinte des paysages hellénistiques dans les formations holocènes de Saint-Jean-du-Désert (Marseille). Méditerranée, 3-4, 33–40. Bonardi, L. (2010). Les paysages viticoles en terrasses: Espaces de convergence. In J. Perard & M. Perrot (Eds.), Rencontres du Clos-­ Vougeot 2009. Paysages et patrimoines viticoles (pp.  129–140). Dijon: UNESCO and Centre Georges Chevrier. Bousquet, D., Campmajo, P., Crabol, D., & Rendu, C. (2012). Évolution de la poterie du Bronze moyen au Bronze final. Éléments pour une chronologie de l’anthropisation d’un espace montagnard (Cerdagne, Pyrénées-Orientales). Documents d’Archéologie Méridionale, 35, 151–174. Bradley, R. (1978). Prehistoric field systems in Britain and north-west Europe. A review of some recent work. World Archaeology, 9(3), 265–280. Bull, I. D., Betancourt, P. P., & Evershed, R. P. (2001). An organic geochemical investigation of the practice of manuring at a Minoan site on Pseira Island, Crete. Geoarchaeology: An International Journal, 16(2), 223–242. Carozza, L., Galop, D., Marembert, F., & Monna, F. (2005). Quel statut pour les espaces de montagne durant l’âge du Bronze ? Regards croisés sur les approches société-environnement dans les Pyrénées occidentales. Documents d’Archéologie Méridionale, 28, 7–23. Chouquer, G. (Ed.). (1996). Les formes du paysage. Tome 2  Archéologie des parcellaires. Paris: Errance. Conesa, M. (2012). D’herbe, de terre et de sang. La Cerdagne du XIVe au XIXe siècle. Perpignan: Presses Universitaires. Davasse, B., Galop, D., & Rendu, C. (1997). Paysages du Néolithique à nos jours dans les Pyrénées de l’Est d’après l’écologie historique et l’archéologie pastorale. In J. Burnouf, J.-P. Bravard, & G. Chouquer (Eds.), La dynamique des paysages protohistoriques, antiques, médiévaux et modernes (pp.  577–599). Sophia Antipolis: éditions APDCA. Daveau, I. e. (2007). Port Ariane (Lattes, Hérault): Construction deltaïque et utilisation d’une zone humide lors des six derniers millénaires (p. 20). Lattara: Lattes. Doumas, C. (1999). The Wall-paintings of Thera. Athens: The Thera Foundation Petros M. Nomikos. Erroux, J. (1983). Graminées et légumineuses sur le site de Llo. In P.  Campmajo (Ed.), Le site protohistorique de Llo (Pyrénées-­ Orientales) (pp.  128–129). Centre d’Études Préhistoriques Catalanes: Perpignan. Fleming, A. (1988). The Dartmoor reaves: Investigating prehistoric land divisions. London: Batsford. Galop, D. (1998). La forêt, l’homme et le troupeau dans les Pyrénées. 6000 ans d’histoire de l’environnement entre Garonne et Méditerranée. Toulouse: GEODE, Laboratoire d’écologie terresre, FRAMESPA. van Geel, B., & Magny, M. (2002). Mise en évidence d’un forçage solaire du climat à partir de données paléoécologiques et archéologiques: la transition Subboréal-Subatlantique. In H.  Richard & A.  Vignot (Eds.), Equilibres et ruptures dans les écosystèmes durant les 20 derniers millénaires en Europe de l’Ouest (pp. 107–122). Besançon: Presses Universitaires de Franche-Comté. Harfouche, R. (2005). Retenir et cultiver le sol sur la longue durée: Les terrasses de culture et la place du bétail dans la montagne méditerranéenne. Anthropozoologica, 40(1), 45–80. Harfouche, R. (2006). Soil care and water management on Mediterranean slopes: An archaeopedological approach. Arqueología Espacial, 26, 311–339. Harfouche, R. (2007). Histoire des paysages méditerranéens terrassés: aménagements et agriculture. Oxford: Archaeopress.

6  Terraced Crop Fields in the Eastern Pyrenean Mountains (France): The View from Pedoarchaeology Harfouche, R. (2010). Agriculture en terrasses à haute altitude au cours de l’Âge du Bronze dans les Pyrénées orientales (Massif du Carlit). In A. Beeching, E. Thirault, & J. Vital (Eds.), Économie et société à la fin de la Préhistoire (Vol. 34, pp. 125–145). Lyon: Documents d’Archéologie en Rhône-Alpes et en Auvergne. Harfouche, R. (2015). Une agriculture à haute altitude dans la montagne méditerranéenne. Regards pédoarchéologiques sur le Massif du Carlit (Pyrénées-Orientales, France). In R. Harfouche & P. Poupet (Eds.), Du Mont Liban aux Sierras d’Espagne: Sols, eau et sociétés en montagne. Autour du projet franco-libanais CEDRE “Nahr Ibrahim” (pp. 251–269). Oxford: Archaeopress. Harfouche, R. (2017). Archéologie et histoire des maçonneries à pierres sèches. In L. Cagin (Ed.), Pierre sèche, théorie et pratique d’un système traditionnel de construction (pp. 8–31). Paris: Eyrolles. Harfouche, R., Poupet, P. (2013). Approche pédoarchéologique des espaces de production agricole à l’Âge du Bronze dans les montagnes méditerranéennes (exemples des Pyrénées-Orientales et de la Haute-Corse, France). Préhistoires Méditerranéennes [online], 4. Available at: http://pm.revues.org/748. Accessed 23 Mar 2017. Harfouche, R., Poupet, P., Darwich, T., Verdin, P., Shaban, A., Khater, C., Goslar, T., & Faour, G. (2015). Nouveaux regards multidisciplinaires sur la montagne de l’hinterland de l’antique Byblos (Jbail). In R.  Harfouche & P.  Poupet (Eds.), Du Mont Liban aux Sierras d’Espagne: Sols, eau et sociétés en montagne. Autour du projet franco-libanais CEDRE “Nahr Ibrahim” (pp.  73–110). Oxford: Archaeopress. Jacomet, S., Ebersbach, R., Akeret, Ö., Antolín, F., Baum, T., Bogaard, A., Brombacher, C., Bleicher, N.  K., Heitz-Weniger, A., Hüster-­ Plogmann, H., Gross, E., Kühn, M., Rentzel, P. H., Steiner, B. L., Wick, L., & Schibler, J. M. (2016). On-site data cast doubts on the hypothesis of shifting cultivation in the late Neolithic (c. 4300–2400 cal. BC): Landscape management as an alternative paradigm. The Holocene, 26(2), 1858–1874. Jamagne, M. (2011). Grands paysages pédologiques de France. Paris: éditions Quæ. Kirchner, H. (2015). Gestion de l’eau et espaces irrigués en al-Andalus. In R.  Harfouche & P.  Poupet (Eds.), Du Mont Liban aux Sierras d’Espagne: Sols, eau et sociétés en montagne. Autour du projet franco-libanais CEDRE “Nahr Ibrahim” (pp.  207–219). Oxford: Archaeopress. Krahtopoulou, A., & Frederick, C. (2008). The stratigraphic implications of long-term terrace agriculture in dynamic landscapes: Polycyclic terracing from Kythera Island, Greece. Geoarchaeology: An International Journal, 23(4), 550–585. Langohr, R. (2001). L’anthropisation du paysage pédologique agricole de la Belgique depuis le Néolithique ancien. Apports de l’archéopédologie. Etude et Gestion des Sols, 8(2), 103–118. Legros, J.-P. (2007). Les grands sols du monde. Lausanne: Presses polytechniques et universitaires romandes. Lerche, G., & Steensberg, A. (1980). Agricultural tools and field shapes. Copenhagen: National Museum of Denmark. Løvschal, M., & Holst, M.  K. (2014). Repeating boundaries  – Repertoires of landscape regulations in southern Scandinavia in the late Bronze Age and pre-Roman Iron Age. Danish Journal of Archaeology, 3(2), 95–118. Magail, J., & Giaume, J.-M. (Eds.). (2005). Le site du Mont Bego de la Protohistoire à nos jours. Nice: Serre éditeur. Marinval, P. (2009). Le seigle, une plante ambiguë durant le Néolithique et la Protohistoire: Adventice ou céréale ? In De Méditerranée et d’ailleurs… Mélanges offerts à Jean Guilaine (pp.  445–462). Toulouse: Archives d’Écologie Préhistorique. Martin, L. (2010). Agriculture et  alimentation en milieu montagnard durant le Néolithique: nouvelles données carpologiques dans les Alpes françaises du Nord. Thèse de Doctorat, Université de Paris I et Universität Basel.

85

Mayewski, P. A., Rohling, E. E., Curt Stager, J., Karlén, W., Maasch, K. A., David Meeker, L., Meyerson, E. A., Gasse, F., van Kreveld, S., Holmgren, K., Lee-Thorp, J., Rosqvist, G., Rack, F., Staubwasser, M., Schneider, R. R., & Steig, E. J. (2004). Holocene climate variability. Quaternary Research, 62, 243–255. Miras, Y., Guenet, P., Surmely, F., Michelin, Y., Tible, R., Walter-­ Simonnet, A.-V., & Richard, H. (2003). Histoire de l’environnement et des dynamiques agro-pastorales dans le Massif du Cantal: étude pluridisciplinaire de la tourbière de Roussy (Saint-Projet-de-Salers, France). Quaternaire, 14(4), 265–278. Miras, Y., Laggoun-Defarge, F., Guenet, P., & Richard, H. (2004). Multi-disciplinary approach to changes in agro-patoral activities since the Subboreal in the surroundings of the “narse d’Espinasse” (Puy de Dôme, French Massif Central). Vegetation History and Archaeobotany, 13, 91–103. Mocci, F., Walsh, K., Richer, S., Court-Picon, M., Talon, B., Tzortzis, S., Palet-Martinez, J.-M., & Bressy, C. (2008). Archéologie et paléoenvironnement dans les Alpes méridionales françaises: Hauts massifs de l’argentiérois, du Champsaur et de l’Ubaye, Hautes-­ Alpes et Alpes-de-Haute-Provence, Néolithique final-début de l’Antiquité. In M. Desmet, M. Magny, & F. Mocci (Eds.), Du climat à l’homme dynamique holocène de l’environnement (pp. 253–272). Grenoble: EDYTEM. Orejas, A., & Ruiz del Arbol, M. (2015). L’eau et la terre. Exploitation et gestion des ressources dans les zones minières au nord-ouest de l’Hispanie. In R.  Harfouche & P.  Poupet (Eds.), Du Mont Liban aux Sierras d’Espagne: Sols, eau et sociétés en montagne. Autour du projet franco-libanais CEDRE “Nahr Ibrahim” (pp. 235–249). Oxford: Archaeopress. Perret, S. (1950). Terrasses de culture et tertres allongés protohistoriques dans les forêts du pied du Jura neuchâtelois. Archives suisses d’anthropologie générale, 15(1), 42–71. Poupet, P. (1998). Protohistoric and ancient agriculture in the Mediterranean region: Dominion and improvment of soil. In 16th World Congress of Soil Science, IUSS Symposium 45 ‘Attitudes to soil care and land use through human history’. Montpellier: edited on CD Rom. Poupet, P. (2009). La Pédologie. In J.-P. Bravard, C. Cammas, P. Nehlig, P.  Poupet, P.-G.  Salvador, & J.  Wattez (Eds.), La Géologie. Les sciences de la Terre appliquées à l’archéologie (2nd revised ed., pp. 127–179). Paris: Errance. Poupet, P. (2015). Pédogenèse et taphonomie d’un espace agricole de la Grèce antique: Les sols et l’eau sur l’île de Délos. In R. Harfouche & P. Poupet (Eds.), Du Mont Liban aux Sierras d’Espagne: Sols, eau et sociétés en montagne. Autour du projet franco-libanais CEDRE “Nahr Ibrahim” (pp. 147–185). Oxford: Archaeopress. Poupet, P., & Harfouche, R. (2000). À la recherche des champs protohistoriques en Languedoc méditerranéen (France): Gestion des sols et moyens de production. In R.  Buxo & E.  Pons (Eds.), Els productes alimentaris d’origen végétal a l’edat del Ferro de l’Europa Occidental: de la produccio al consum (pp.  221–236). Museu d’Arqueologia de Catalunya: Girona. Poupet, P., & Harfouche, R. (2007). Établir et travailler le champ pour le paysan de la Méditerranée pré-industrielle: Contributions de l’archéologie et de la science du sol à l’histoire des pratiques. In R. Bourrigaud & F. Sigaut (Eds.), Nous labourons. Techniques de travail de la terre, hier et aujourd’hui, ici et là-bas (pp.  61–71). Nantes: éditions du Centre d’Histoire du Travail. Rendu, C. (2003). La montagne d’Enveig, Une estive pyrénéenne dans la longue durée. Canet: éditions Trabucaire. Rendu, C., Bille, E., Conesa, M., Calastrenc, C., Campmajo, P., & Crabol, D. (2009). Margins and centers in the shaping of the Pyrenean slopes. Medieval dynamics within the long-term perspective. In Medieval rural settlement in marginal landsapes, Ruralia VII (pp. 235–251). Turnhout: Brepols.

86 Rendu, Ch., Campmajo, P. Crabol, D. (2012). Etagement, saisonnalité et exploitation des ressources agro-pastorales en montagne à l’âge du bronze. Une possible “ferme d’altitude” à Enveig (Pyrénées-­ Orientales). Bulletin de l’Association pour la Recherche sur l’Âge du Bronze, (2012), 58–61. Available at: https://hal.archives-­­ ouvertes.fr/halshs-­00945435. Accessed 2 Dec 2017. Rendu, C., Passarrius, O., Calastrenc, C., Julia, R., Llubes, M., Illes, P., Campmajo, P., Jodry, C., Crabol, D., Bille, E., Conesa, M., Bousquet, D., & Lallemand, V. (2015). Reconstructing past terrace fields in the Pyrenees: Insights into land management and settlement from the Bronze Age to the Early Modern era at Vilalta (1650 masl, Cerdagne, France). Journal of Field Archaeology, 40(4), 461–480. Ruas, M.-P. (2003). Des céréales et des fruits dans le niveau incendié de la cabane 81. In C.  Rendu (Ed.), La montagne d’Enveig, une estive pyrénéenne dans la longue durée (pp. 393–413). Canet: éditions Trabucaire. Ruas, M.-P., & Rendu, C. (2005). Glanes et cultures en haute montagne d’après les restes de graines et de fruits carbonisés de deux sites médiévaux de Cerdagne et du Capcir (Pyrénées-Orientales). In A. Catafau (Ed.), Les ressources naturelles des Pyrénées du Moyen Âge à l’Époque Moderne. Exploitation, gestion, appropriation (pp. 147–184). Perpignan: Presses Universitaires. Ruas, M.-P., Bouby, L., & Campmajo, P. (2009). Agriculture en montagne cerdane au Bronze final: Les données carpologiques de Llo-Lo Lladre (Pyrénées-Orientales). In De Méditerranée et d’ailleurs… Mélanges offerts à Jean Guilaine (pp. 638–660). Toulouse: Archives d’Écologie Préhistorique. Saccoccio, F., Marzocchella, A., & Vanzetti, A. (2013). The field system of Gricignano d’Aversa (southern Italy) and the agrarian impact in the Piana Campana, ca. 3900 cal BP. Quaternary International, 303, 82–92. Servera Vives, G., Miras, Y., Riera, S., Julià, R., Allée, P., Orengo, H., Paradis-Grenouillet, S., & Palet, J. P. (2014). Tracing the land use history and vegetation dynamics in the Mont Lozère (Massif Central, France) during the last 2000 years: The interdisciplinary

R. Harfouche and P. Poupet study case of Countrasts peat bog. Quaternary International, 353, 123–139. Simonneau, A., Chapron, E., Courp, T., Tachikawa, K., Le Roux, G., Baron, S., Galop, D., Garcia, M., Di Giovanni, C., Motellica-Heino, M., Mazier, F., Foucher, A., Houet, T., Desmet, M., & Bard, E. (2013). Recent climatic and anthropogenic imprints on lacustrine systems in the Pyrenean Mountains inferred from minerogenic and organic clastic supply (Vicdessos valley, Pyrenees, France). The Holocene, 23(12), 1764–1777. Styring, A., Maier, U., Stephan, E., Schlichtherle, H., & Bogaard, A. (2016). Cultivation of choice: New insights into farming practices at Neolithic lakeshore sites. Antiquity, 90, 95–110. Vannière, B., Galop, D., Rendu, C., & Davasse, B. (2001). Feu et pratiques agro-pastorales dans les Pyrénées-Orientales: le cas de la Montagne d’Enveitg (Cerdagne, Pyrénées-Orientales, France). Sud-­ Ouest Européen, 11, 29–42. Vital, J. (2008). Architectures, sociétés, espaces Durant l’Âge du bronze. Quelques exemples dans le bassin rhodanien. In J. Guilaine (Ed.), Villes, villages, campagnes de l’Âge du bronze (pp. 179–201). Paris: Errance. Walsh, K., Mocci, F., & Palet-Martinez, J.-M. (2007). Nine thousand years of human landscape dynamics in a high altitude zone in the southern French Alps (Parc National des Ecrins, Hautes-Alpes). Preistoria Alpina, 42, 9–22. Walsh, K., & Mocci, F. (2011). Mobility in the mountains: Late third and second millennia alpine societies’ engagements with the high-­ altitude zones in the Southern French Alps. European Journal of Archaeology, 14(1–2), 88–115. Zoller, H. (1998). Les Alpes et le versant sud des Alpes. In S. Hochuli, U.  Niffeler, & V.  Rychner (Eds.), La Suisse du Paléolithique à l’aube du Moyen Age: Âge du Bronze (pp. 162–163). Société suisse de préhistoire et d’archéologie: Bâle. Zufferey-Périsset, A.-D. (Ed.). (2012). Murs de pierres. Murs de vignes. Sierre-Salquenen and Gollion: edited by Musée valaisan de la Vigne et du Vin and by Infolio editions.

7

Fields and Farming-Systems in Bronze Age Scotland Stratford Halliday

Abstract

Traces of prehistoric agriculture are found widely in the Scottish landscape, but, unlike in southern England, there is little evidence of enclosed fields. Thus, the interpretation of enclosed English field-systems as evidence for the intensification of agricultural production in the Middle Bronze Age carries an implication of non-­intensive production further north. This paper analyses the contemporary domestic and agricultural components of the Scottish landscape to argue that they indicate in extenso systems of farming in which stock and crops were largely kept separate, thus rendering permanently enclosed fields unnecessary. Rather than being entirely sedentary, settlements and their cultivated fields moved at intervals in a dynamic system where a location was sequentially occupied, unoccupied and re-occupied on numerous occasions, and the surrounding ground was cultivated or reverted to pasture as appropriate. Similar sequences of occupation and reoccupation may be observed in southern England, and the differences in the farming-systems in operation may not be as great as has been suggested. Keywords

Scotland · Field systems · Burnt mounds · Boundaries

7.1

Introduction

Some 50 years ago, Stuart Piggott wrote a classic essay on the pre-Roman native economies in northern Britain. If nothing else, it was notable for the near total absence of any evidence for agriculture north of a line drawn roughly between

S. Halliday (*) Rath Manach, Peeblesshire, UK

The Wash and the Severn Estuary. There were no field-­ systems to speak of, no grain storage-pits, and no significant assemblages of animal bones from excavated settlements (Piggott 1958: 8 Map 1; 10 Map 2). And yet, in a masterpiece of argument, the negative northern values of every strand of southern evidence were entwined in the enduring image of the footloose Celtic cowboy, roaming whither his herds should take him. The classical trope of the meat-eating savage as the counterpoint of bread-eating citizen was apparently played out on the ground, and the southern farmers lived out their settled existence in farmsteads surrounded by systems of small square fields, practising an economy ‘associated with intensive corn-growing and distinctive methods for the preparation and storage of the threshed grain’ (Piggott 1958: 12). The cowboys dogged archaeological thinking throughout the 1960s and 70s, despite the recording of extensive traces of agriculture in the uplands by the Ordnance Survey (OS) Archaeology Division. Most of these were speculatively assigned by Richard Feachem to the third and second millennia BC, tailing off into the first millennium in the face of a deteriorating climate, and thus creating a convenient gap through which the cowboys could continue riding as ‘the highlanders came to rely on stock-raising at the expense of crop cultivation’ (Feachem 1973: 347–49). The threads upon which any of these interpretations hung were slender indeed, and could not be dispelled until new evidence and radiocarbon dates became available in the late 1970s (Fowler 1981: 169). A series of excavations at Tormore and Machrie on Arran (Barber 1997), Cul a’Bhaile on Jura, Argyll (Stevenson 1984), and Green Knowe, Peeblesshire (Feachem 1961; Jobey 1980), thrust the settlement record firmly back into the Bronze Age, with an apparent abandonment in the final centuries of the second millennium BC. Ground-breaking work on the moors of Arran also revealed a complex and largely hidden agricultural history (Barber 1997). Further work at Achany Glen near Lairg, Sutherland, revealed that c­ ultivation played a significant part in every episode of settlement in this

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 S. Arnoldussen et al. (eds.), Europe’s Early Fieldscapes, Themes in Contemporary Archaeology, https://doi.org/10.1007/978-3-030-71652-3_7

87

88

highland glen (McCullagh and Tipping 1998). Furthermore, palynological studies consistently reveal evidence of cereals being grown widely during every period from the Neolithic onwards. In an extreme example recently recorded at Cocklawhead in the Cheviot Hills, cereals may have been grown between 2100 and 1250 cal BC at an altitude of 500 m on the ridge now forming the boundary between England and Scotland, though it has apparently left no visible archaeological traces (Tipping 2010: 116). Arable farmers were alive and well in northern Britain. Work elsewhere, particularly in The Fens (reviewed in Evans 2009) and Dartmoor (Fleming 2008), had also proved a revelation, both in the extent of some of the enclosed landscapes and in securing their Bronze Age origins. And despite the unfortunate Celtic tag of the downland field-systems, these clearly went back long before Piggott imagined, just as Herbert Toms had argued in the 1920s (Bradley 1989; see also Evans 2009: 265–66). The most wide-ranging review of these fieldsystems and their context has been presented by David Yates (2007). Taking ideas developed in Scandinavia (Kristiansen 1987), he suggested that southern England controlled the north-south connections of a prestige goods economy in an international core-periphery network. The system of exchange in the economy was manifested in large quantities of deposited bronze metalwork, and underpinned by the intensification of agricultural production represented by the field-systems (Yates 2007: 1–4). The distribution of the latter (Yates 2007: 111, Fig. 12.2), in a distant echo of Piggott’s maps, re-establishes that same notional boundary roughly from The Wash to the Severn estuary. Set against the pattern of commercial evaluations (Yates 2007: 109, Fig.  12.1), the distribution itself appears undeniable, but there are some key questions to be asked. Are the field-systems in themselves a valid measure of the intensification of production, thus creating a signature difference north and south of this line, or are they merely the result of variations and choices exercised more locally in response to other factors? And in any case how do we evaluate the intensity of production of any field-­system independently of the fact of its existence? This runs much deeper than superficial comparisons between fields on Dartmoor, the Marlborough Downs and the Fen-edge, because if there are no independent measures of production in these systems, how can comparisons be made with practice in northern Britain? (For a critique, see Wickstead 2007: 111–36.) Bearing these questions in mind, my purpose here is to ask a question of Scottish archaeology (Fig. 7.1) about the evidence that has now accrued and how it might best be interpreted. At its core it will be argued that the character of the settlements is as much a guide to the nature of the agricultural systems in operation as any of the traces of fields, and that the absence of enclosed fields simply indicates that

S. Halliday

Fig. 7.1  Map of Scotland showing the principal locations mentioned in the text

stock and growing crops were kept separate in the landscape. Furthermore, close examination of all the evidence suggests that long-term continuity of permanent settlement was not invested in single locations, but that settlements, fields and pastures were moving round the landscape in short- and long-term cycles. In essence, if Bronze Age farming practice is to be understood, it needs to be studied at the landscape scale and all the fragments, including the spaces between them, need to be integrated into a single dynamic system. The focus will be Bronze Age, but this perspective has implications both before and after, and it will necessarily trespass across adjacent millennia. The text is broken into a series of sections detailing the character of the domestic and agricultural components that make up the Bronze Age landscape, beginning with its extent, and thereafter discussing the character of the settlement evidence, burnt mounds, and the evidence for fields and enclosures. The final sections integrate these strands into a system that was designed to exploit the totality of the landscape.

7  Fields and Farming-Systems in Bronze Age Scotland

7.2

 he Extent of the Bronze Age T Landscape

89

were also occupied in both the first and the second millennia BC.  This is demonstrably so at Kintore, and implied more ambiguously elsewhere. An undated post-ring at the Late Until comparatively recently Bronze Age settlements were Iron Age settlement at Seafield West, Inverness, for example, largely confined to the uplands, manifested in groups of hut-­ enclosed a pit dated 1260–920 cal. BC, within which a mould circles and stone clearance heaps surviving beyond the upper fragment for a bronze sword was recovered (Cressey and limits of medieval and later cultivation (Fig.  7.2). More Anderson 2011). recent commercially-driven excavation, however, has begun Kintore, Aberdeenshire, with its large sample of round-­ to redress this balance (see Phillips and Bradley 2004: 20 houses spread over some 35 ha, perhaps gives a clearer idea Illus. 1). The change can be tracked annually through reports of the proportion of these lowland buildings that are likely in Discovery and Excavation, Scotland (DES). In contrast to to be Bronze Age. The excavation report (Cook and Dunbar previous decades, following 1990 there are only 2  years 2008) takes into account a total of 36  in the immediate when fewer than five sites with unenclosed round-houses neighbourhood, 22 of which have returned radiocarbon have been excavated (1992 and 1994), and in many years the dates. Of these, one is possibly Late Neolithic, four are figure is closer to ten. And while some are single buildings, Middle Bronze Age, six Late Bronze Age, and 11 Iron Age; others are extensive groups: for example, at least 27 at Forest 14 are undated. In crude terms, there seems to be a relatively Road, Kintore, Aberdeenshire (DES 2001: 11; Cook and equal split between Bronze Age and Iron Age buildings. By Dunbar 2008); about 30 at Birnie, Moray (Hunter 2006: way of comparison, of the seven upland hut-circles fully 21–6; Hunter 2007: 12–15); nine at Pitlethie Road, Leuchars, excavated in Achany Glen, six were Middle Bronze Age and Fife (DES 2004: 63–4); seven at Midross, Loch Lomond one was Late Iron Age (McCullagh and Tipping 1998). Of (DES 2005:  36); up to five at Oldmeldrum, Aberdeenshire the 55 hut-­circles noted in the survey phase of this project, (DES 2005: 35; White and Richardson 2010); five at 39 were sampled, providing a series of termini post quem Meadowend, Clackmannan (DES 2006: 47); five at Gogar and termini ante quem dates for another 12. Of these, all bar Mains and four nearby at Newbridge, Midlothian (DES two fall in the second millennium BC. At first sight, at least, 2008: 72); and 14 at Auchterarder, Perth & Kinross (DES the lowland settlements are providing a wider chronological 2008: 139). All are lowland settlements, though often Bronze span of occupation than some of their upland counterparts, Age buildings can only be distinguished from Iron Age ones though there may be underlying regional patterns in play with radiocarbon dates post excavation. Many of the sites (infra). The Kintore excavations are the first time such a large area has been laid bare in the bottom of a river valley where the soils and climate militate against the formation of cropmarks. Such areas are largely mute in settlement distributions, but clearly similar densities of remains should be expected in equivalent locations elsewhere. Away from these areas, which are inevitably focused on modern centres of population, it is linear development projects along road and pipeline corridors that often reveal the wider distribution of Bronze Age settlements in the lowlands, reaching out to meet the general distribution of Early Bronze Age cairns and burials. The Southwest Scotland Gas Interconnector Pipeline through Galloway, for example, in another landscape that has a fairly patchy cropmark record, revealed two or three timber round-houses at Blairhall Burn, near Amisfield to the north-­east of Dumfries. Six radiocarbon dates place all of them in the second millennium BC (Strachan et  al. 1998: 60–67). In addition, two burnt mounds were excavated nearby, producing broadly similar radiocarbon dates. Numerous other burnt mounds were located along the corridor (Maynard 1993), extending a distribution otherwise known only from survey in the Fig. 7.2  Middleton Muir, Perth & Kinross; a typical example of hut-­ adjacent uplands (Cowley 2011: 45–47; Halliday 1990; circles, banks and small cairns surviving on the moors beyond the upper limit of medieval and later cultivation. (After RCAHMS 1990: 64, RCAHMS 1994: 10–11; 1997, 100–02). At the western end Fig. 147)

90

of the pipeline, at Ross Bay in the Stewartry, at least one more round-house with dates of 1530–1290  cal  BC and 1520–1250  cal  BC was uncovered, while charred grain from a small cluster of pits and post-holes some distance away returned dates of 1020–820  cal  BC and 1000– 790 cal BC (Ronan and Higgins 2005). In some areas of the lowlands burnt mounds are also becoming as ubiquitous as the round-houses, albeit in varying densities. The overall pattern is heavily skewed by fieldwork, in which various field projects form particular concentrations (Cowley 2011: 45–7, Fig.  7.3), but whereas new discoveries in Dumfries and Galloway or Sutherland run into hundreds (Halliday 1990; RCAHMS 1993: 11–12; RCAHMS 1994: 10–11; RCAHMS 1997: 100–102), survey elsewhere has recorded much smaller numbers: in the Kale Water, Roxburgh, only four; of a larger area in north-east Perthshire, 28 (RCAHMS 1990: 5); of Glen Clova, Angus, 22; and of the whole of Donside, Aberdeenshire, a mere eight. This variable density is also evident in the commercial work, which has thrown up only a single example in Angus (DES 2008: 28) and two in Aberdeenshire (Strachan and Dunwell 2003). None were found in the extensive excavations around Kintore and none in the A1 road corridor through East Lothian, though at Thorneybank, on the line of the Dalkeith bypass in Midlothian, a pit filled with fire-­cracked stones was dated 2140–1880  cal  BC (Rees 2002: 318–19). In contrast, evaluations and excavations near Girvan, in southern Ayrshire, and around Inverness have found numerous examples dating 2600–1200  cal  BC (Donnelly and MacGregor 2005; Banks et al. 2008; Cressey and Strachan 2003). The one element consistently missing from mainland Scotland is evidence of any extensive field enclosures. The typical hut-circle group in the uplands is usually juxtaposed to no more than a scatter of clearance features and the occasional lynchet, and the much-reproduced aerial photograph of a field-system at Drumturn Burn, Perth & Kinross, is literally unique (Fig. 7.3). Outside Shetland, any coherent enclosures are few and far between. This has not changed with the results of aerial reconnaissance, and nor from commercial work. In this respect, the general absence of field ditches from pipeline and road corridors is probably telling. The length of Middle Bronze Age ditch at Howmuir, East Lothian, where analysis of the silts implied manured cultivation or occupation nearby (Lelong and MacGregor 2007: 121), was the sole example along the AI corridor. Neither was there any evidence of a field-system around the settlements at Kintore (Cook and Dunbar 2008). Whatever remains of fields there may have been in the lowlands, they certainly were not routinely enclosed with ditches, and they have been eradicated by determined and aggressive ploughing over the last two centuries, if not before.

S. Halliday

Fig. 7.3  Drumturn Burn, Perth & Kinross; plan of the hut-circles and fields. (After RCAHMS 1990: 48 Fig. 124.10C)

7.3

The Character of Settlement

When the first radiocarbon dates were returned from the work on Arran (Barber 1997), they showed a long chronology of occupation spanning the greater part of the second millennium BC; one of the hut-circles on Arran, Tormore 10/1, appeared to have been occupied continuously for hundreds of years, with Beaker pottery in its earlier phases and a date of 1435–900 cal BC from burnt wattles marking its catastrophic end. Thereafter, a major desertion seemed to have taken place here and elsewhere. By 1983 a catastrophe of disease, starvation and strife had been visited upon the Bronze Age population (Burgess 1985), with a volcanic winter following on not far behind (Burgess 1989). A core-­periphery model underlay these simplistic environmental explanations. When the going was good, people were able to expand from the best lowland soils into the poorer uplands, remaining there until such times as occupation proved untenable and they retreated whence they came. The settlements themselves were understood to be ‘farms’ and were occupied for centuries; and ‘soil exhaustion’ was a convenient fallback to explain their desertion.

7  Fields and Farming-Systems in Bronze Age Scotland

By the late 1990s, however, it had become clear that these conventional models of Bronze Age settlements as enduring farmsteads no longer fitted the data coming forward from the publication of the landscape projects on Arran (Barber 1997) and at Achany Glen, Lairg (McCullagh and Tipping 1998). My own misgivings (Halliday 1999, 2007) were also shared by John Barber and Anne Crone, who following similar lines of argument independently challenged the assumptions that settlements were occupied continuously for long periods of time (Barber and Crone 2001). Jo Brück was also asking a similar range of questions of Middle Bronze Age settlements in southern England (Brück 1999). The sources of the misgivings were twofold, drawing firstly upon settlements preserved in wetlands, and secondly on a relatively small number of excavated hut-circles in dryland environments. The evidence from the wetlands includes dendrochronological evidence, which allows the fine resolution of the chronology of construction, maintenance and replacement (Barber and Crone 2001). Whereas dryland stratigraphy typically gives glimpses of the past at the points of change, the wetland equivalents allow the elapse of time to be measured between every stage, be it the duration of occupation or a period of desertion. Typically, occupations in the wetlands are relatively short and intermittent. The second strand of the argument stemmed from hut-circle Tormore 10/1 (Barber 1997). Closer analysis had demonstrated that the stratified sequence of as many as eight separate buildings was interrupted by erosion deposits washed down off the walls, while in a major break the eroded tail of the bank had also been cultivated (Barber 1997: 6–25). Far from continuous, this was a sequence in which the walls had probably been left open to the skies between every reconstruction, at the very least implying periods of desertion of the structure itself, if not of the surrounding land. This sort of interrupted sequence is not unique. It is found in successive buildings set both concentrically and eccentrically to each other and is demonstrated unequivocally where the stages in a sequence are stratigraphically separated by a cultivated soil (Halliday 2007). This is the case with two concentrically superimposed timber round-houses buried beneath colluvium at Suisgill, Sutherland (Barclay 1985), and two phases of a succession of stone walled hut-circles at Cnoc Stanger, Caithness (Mercer 1996). To summarise the key conclusions, concentric and eccentric successions of superimposed structures do not confer either continuity or long duration of occupation. Both are aspects of settlement that need to be challenged and measured wherever the stratigraphy and samples offer the opportunity (Fig. 7.4). Unfortunately, by far the majority of the lowland round-­ houses have been variously reduced to scatters of post-holes, hearths, post-rings, fragmentary ring-grooves and shallow sunken hollows where parts of the floors have been worn into relatively soft subsoils. The record is partial and the laudable

91

Fig. 7.4  Balnabroich, Perth & Kinross; excavation of hut-circles in Scotland suggests that occupations were typically brief and episodic. The juxtaposition of hut-circles with other monuments of widely ranging date, here comprising two ring-cairns (a), a stone circle and standing stone (b), several large burial cairns (c), two clusters of early medieval longhouses (d), and two post-medieval farmsteads (e), is probably further evidence of episodic occupation rather than continuous settlement. (After RCAHMS 1990: 34 Fig. 108A)

attempt to classify different types amongst those discovered at Kintore and describe their architectural evolution (Cook and Dunbar 2008: 89; 321–29) is deeply flawed by differential erosion. This same problem makes it difficult to assess the complexity of individual round-houses in order to get some idea of the history of their maintenance and repair. At a well-preserved hut-circle at Navidale, Sutherland, the duration of the building’s occupation was tested with four singleevent samples, including one from a post-­ abandonment deposit, but returned almost identical dates bracketing the period 1400–1200 cal BC (Dunbar 2007: 158–59). The only stratigraphic dataset with the potential to contradict this is from Cladh Hallan, on South Uist in the Western Isles. Here the sequence of buildings contains 14 separate floors spanning some 700  years from about 1100–400  BC (Parker Pearson et al. 2004: 70), and essentially represents the period following on from the major desertion of settlement seen at Achany Glen on the mainland. Intermittent sequences are also likely to be characteristic of the rather earlier settlement remains found on Shetland, where the typical moorland settlements contain oval buildings divided up into cells. These Shetland settlements are often surrounded by small systems of irregular fields, which, uniquely in Scotland, are linked into much more extensive landscape enclosures (Turner 2011; see Christie, this volume). The general trend of an imprecise chronology at Scord of Brouster, in the interior of West Mainland (Whittle et al. 1986; Sheridan 2012: 20), suggests these were abandoned

92

rather earlier in the second millennium BC than Achany Glen some 300 km to the south-west. Their floruit seems to be Early Bronze Age rather than Middle Bronze Age, and they have clearly not been exposed to the same process of churning and recycling that evidently takes place on the typical mainland hut-circle group later in the second millennium BC. This churning is demonstrated most clearly at Achany Glen, where Middle Bronze Age hut-circles were ploughed over in the Middle Bronze Age, and the only traces of Early Bronze Age structures were found deeply buried in sediment traps. In one case, what were interpreted as some of the internal features of a building were recovered from beneath a cultivated soil on which a cairn with a kerb had been constructed; a shallow pit containing a Food Vessel and beads from a shale bracelet or necklace was cut into the tilled soil (McCullagh and Tipping 1998: 84–91). The intensity of Middle Bronze Age occupation and cultivation is clearly the reason why Early Bronze Age settlements are largely invisible in the Highlands, and explains why they also continue to elude us elsewhere. Apart from the earlier horizon of abandonments on Shetland, the main examples come from the machair — coastal plain formed by wind-blown calcareous sand — of the Western Isles, where they are also associated with cultivated fields (Parker Pearson et  al. 2004: 45–52; Parker Pearson 2012: 23–25; Sharples 2012), and on Islay in the Inner Hebrides, at Ardnave (Ritchie and Welfare 1983) and Kilellan Farm (Ritchie 2005). Visibility of settlements and fields at any period, so it would seem, is tied up in two factors: deposition and erosion. Environments in which sediments are generally accreting, such as in the machair or the plaggen soils in parts of Orkney and Shetland, are preserving both Early Bronze Age and Neolithic structures, though they remain hidden until exposed by erosion or excavation. The hillsides of mainland straths and glens, however, are typically shedding sediments, partly through natural agencies, but also in combination with subsequent settlement and cultivation. Here it is the Middle Bronze Age settlements abandoned at the end of the second millennium BC that are most extensive and remain visible on the surface, while traces of earlier periods only survive in so far as they are represented by large, robust monuments, such as burial-cairns, or where they are sealed into sediment traps. The sequence demonstrated at Achany Glen is probably fairly typical of the peatlands of the north and west of Scotland, though on the drier hills of south-eastern and eastern Scotland the process of use and re-use kept churning through the first millennium BC.  In the south-east, it is clearly the enclosed Iron Age settlements that have obliterated their Bronze Age predecessors, while in north-eastern Perthshire and Angus many hut-circle groups seem to have been re-occupied in the Late Bronze Age and later. Nevertheless, in the places where the mainstay of the visible settlement remains are Middle Bronze Age, it seems likely

S. Halliday

that the desertions witnessed at the end of the second millennium BC saw occupation focused into the most favourable positions in their localities, in the Highlands in the very places that have been subsumed into the pattern of post-­ medieval townships and their surrounding fields. In these locations Middle Bronze Age settlements were probably obliterated in the fields of Late Bronze Age settlements, which in their turn fell victim to Iron Age cultivation before being subsumed into the fields of medieval and later townships. This is the multi-layered spectrum of differential destruction that has been operating across four millennia to preserve and emphasise the extent of Middle Bronze Age settlement. This same bit of landscape history can be detected at Kintore. Whereas the Achany Glen settlement was not re-­ occupied until the Late Iron Age, at Kintore there are a series of round-houses with dates falling throughout the first millennium BC. In effect, Kintore is everything that we should expect of an area into which settlement contracted at the end of the second millennium BC, right down to the ardmarks preserved in five of the round-houses, which testify to the process of attrition in action, albeit that the ardmarks themselves are undated. Nevertheless, two are Middle Bronze Age buildings (Alexander 2000; Cook and Dunbar 2008: 93–95: RH26), as might be expected, two Late Bronze Age (Cook and Dunbar 2008: 97–103: RH10 & 11), and the fifth is undated (Cook and Dunbar 2008: 105–7, RH12). It is not simply that the process of churning and degradation of the earlier buildings seen in the Middle Bronze Age at Achany Glen was being played out at Kintore, but that it was also visited there on Late Bronze Age round-houses. The general desertion at the end of the second millennium BC is also detectable in the radiocarbon chronology at Kintore, but here it does not seem to have entailed any change in the subsequent system of occupation and land-use.

7.4

 he Landscape Context of Burnt T Mounds

Burial cairns and clearance heaps aside, the most durable components of the Bronze Age landscape seem to be burnt mounds. While those in the Northern Isles often occur with buildings (see Moore and Wilson 1999: 231–35), across mainland Scotland the majority are not in themselves the remains of settlements, but nevertheless manifest patterns of activity that recur in both highland and lowland settings throughout the Bronze Age.1 Two were excavated in Achany Glen, and two others were sampled. The dates derived from these clearly overlap with those from the hut-circles in the The overall range of dates from burnt mounds, however, extends from the Neolithic to the early medieval period.

1 

7  Fields and Farming-Systems in Bronze Age Scotland

area (McCullagh and Tipping 1998: 72–77 Fig. 55), but two of them were also used in the Late Bronze Age, after the hut-­ circles had been deserted. In neither case was the burnt mound being established de novo, so much as a location that had already been used probably continued to serve the same purpose. Of the three dates from one, the earliest of 2030– 1695 cal. BC comes from the lower fill of a pit partly filled with fire-cracked stones, while the main body of the mound produced a date of 1585–1320 cal. BC, and its upper deposits one of 1050–815 cal. BC. The eastern part of the second mound produced dates of 1615–1165  cal. BC and 1420– 1135 cal. BC, but a pit for a trough had also been cut through the mound and a date of 1145–865 cal. BC again came from its uppermost deposits. The simple calculation of the size of these relatively small mounds, and the most optimistic figures for the durability of igneous rocks (see Moore and Wilson 1999: 227–28), would extend the life of a burnt mound of some 20 m3 (about 10 m by 5 m by 0.5 m) to about 1000 boiling events, but even at this level of use most burnt mounds would not sustain daily boilings by a population in permanent residence for any length of time. As seems often the case, the Achany Glen mounds appear to lie within a settled landscape, but by virtue of their preferred positions immediately adjacent to sources of water, they stand slightly separate from the nearest hut-circles. It is difficult to make this comparison in the lowland landscapes, but there is little evidence of settlement in the immediate vicinity of the burnt mounds at either Girvan, Ayrshire, or Beechwood Farm, Inverness (above; Donnelly and MacGregor 2005; Banks et al. 2008; Cressey and Strachan 2003). The range of dates from Girvan, however, gives the impression that while the same wetland environment was being exploited repeatedly for in excess of a thousand years, the individual troughs and mounds were fairly short-lived. At Achany Glen the excavators linked the burnt mounds to the exploitation of poorer land (McCullagh and Tipping 1998: 208), and associated them with nearby hut-circles. One of the most striking aspects of the palynological record in Achany Glen, however, is that the pollen core taken adjacent to the excavated settlement reveals relatively little evidence of settlement and cultivation during the Middle Bronze Age. Indeed, if the occupation no more than 150 m away was continuous over the eight centuries represented by the radiocarbon dates from the hut-circles, there is precious little to show for it in the data from this core. Relatively minor minerogenic inwashing picks up from about 2200  BC, though the earliest of the clearance heaps date from considerably earlier. From about 1800 BC birch and hazel woodlands were being cleared, and an increase in charcoal seems to represent domestic occupation. Grazing indicators, however, are prevalent throughout and no cereal-type pollen grains are recorded, and this general level of activity persists long after the desertion of the hut-circles until the end of the

93

first millennium BC (McCullagh and Tipping 1998: 190–91; 209). No great weight should be placed on the lack of palynological evidence for cereals, but the general emphasis on the grazing indicators here must surely reflect the major use of the landscape, both during the occupation of the hut-­ circles and after. And if this is the background to the use of the burnt mounds after the more general desertion of the hut-circles, then why not the case before? In essence, the activities with which they were associated in the Middle Bronze Age, and for that matter the Early Bronze Age, related to the pastures rather than the visible settlements and their fields; likewise in the lowlands around Inverness and Girvan, the latter including saltmarsh. In both uplands and lowlands burnt mounds are surely monuments that are associated with the exploitation of the wider landscape, rather than the apparently settled landscapes within which they are so often found (e.g. RCAHMS 1994: 17–18 Fig. 15; McIntyre 1998).

7.5

 he Character of Fields T and Enclosures

With only a few exceptions, the traces of fields in mainland Scotland are incoherent. The system of at least three large enclosures covering perhaps 10  ha beneath the Moss of Achnacree remains unique (Carter and Dalland 2005), while the extensive enclosed landscapes in Shetland (Christie, this volume) are not found elsewhere. At Achany Glen, having sectioned one of several lynchets (Dyke 1) in three places and extended one of these sections in plan, the excavators observed that: ‘Transects across these linear features tend to simplify the stratigraphic complexity and also tend to encourage the identification of concordance in the observed stratigraphic sequences’ (McCullagh and Tipping 1998: 80). This complexity is probably the product of piecemeal action and accretion, and these lynchets simply identify one recurring edge of an area where cultivation has taken place. Season to season, however, or episode to episode, individual plots may have been extended, contracted or shifted altogether, in much the same manner seen in some of the systems of Iron Age cultivation rigs  – known as cord rig  – in the Southern Uplands. At sites such as Hut Knowe North and Scowther Knowe (Fig. 7.5) in the northern Cheviots, plots of these narrow cultivation rigs have been variously cut by plough-scars marking the edges of smoothed plots, or indeed lie within larger patches of smoothed ground (Halliday 1993). Although untested by excavation, this survey evidence suggests that the cord rig plots were not permanent features, but that their shape and extent changed with every episode of cultivation. This sort of practice would explain why there are no other visible boundaries in places like Achany Glen. There was certainly ground lying fallow or abandoned since an earlier

94

S. Halliday

Fig. 7.5  Scowther Knowe, Scottish Borders; low scarps (a) cut across the bottom of several plots of cord rig on the slopes below this small settlement, almost certainly resulting from separate, unridged, episodes of cultivation cutting across them. The edges of other episodes of cultivation are marked with dashed lines (b) beyond the edges of the main plot of cord rig in the saddle linking the summit of the spur occupied by the settlement to the higher ground on the north

Fig. 7.6  The Ord, Balrownie, Aberdeenshire; a typical group of clearance heaps and short stony banks around four hut-circles. Now divided ENE and WSW by a modern fence, all the ground to the S had already been ploughed for forestry when the plan was surveyed by RCAHMS in 1984 and now lies below dense coniferous forest. (After Historic Environment Scotland KCD/132/2 PO)

phase of occupation, because the carbonised seeds recovered from turves in the walls of several of the hut-circles indicate that they were cut from ground that had been lightly manured (McCullagh and Tipping 1998: 168–69), but there is none of the discipline of fixed-edge field-systems in this landscape. The only fixed points are the dumps of cleared stones, and where linear heaps coalesced on slopes, so sediment traps were occasionally created to form a few irregular lynchets. The other edges of cultivated plots otherwise flexed from episode to episode of cultivation and were not conceived as elements of formal enclosures. As such the cultivated plots were as intermittent as the occupation seen in hut-circles such as Tormore 10/1. Despite soils up to 1 m deep in the lynchets, there is no need to invoke continuous use over centuries to explain their formation, rather than intermittent episodes of intensive cultivation at intervals spread over the Early and Middle Bronze Age. While this may be generally true, there may also be exceptions in the more extreme environment of the Northern

Isles. Around many bays punctuating the coast of Shetland, for example, there are areas of artificially deepened plaggen soils with long chronologies that go back at least as far as the Bronze Age, such as at Old Scatness (Dockrill 2007: 34–37). Comparable soils are even more extensive on Orkney. They tend to be seen in section, rather than as fields in plan, but these are field-systems nonetheless. How those in Shetland relate to the settlements in the peatlands of the interior (Christie, this volume) is unclear, though it may not be unduly deterministic in such a harsh environment to suggest that the more favourable locations were exploited rather more continuously than their inland counterparts. Shetland apart, the general lack of stockproof enclosures around mainland settlements suggests that the main grazings were typically at some distance from the settlements (Fig. 7.6). This is a pragmatic solution to the happy coexistence of growing crops and grazing beasts, though for over half the year the beneficial effects of animal dung would have been concentrated on the pastures. Manuring was cer-

7  Fields and Farming-Systems in Bronze Age Scotland

tainly well understood from at least the Bronze Age if not before (Shepherd and Tuckwell 1977; McCullagh and Tipping 1998: 169; 171; Guttmann et  al. 2004; Dockrill 2007: 34–7; 386–7), and apart from the physical addition of material, we should expect that plots were dunged by the beasts on the hoof. This is the historically documented practice of tathing, in which after harvest beasts were folded onto the next season’s arable (see RCAHMS 2008: 21–3; Dixon 2011, 2016; Dodgshon 2011a, 2011b). By day they grazed surrounding pasture, but by night they would be brought back into temporary tathe-folds, thus transferring nutrients from pasture to arable. This is a system that was surely not lost on prehistoric farmers and may well explain the multiperiod stake fences and ardmarks found at Machrie Moor, though the manuring regime there also seems to have included seaweed brought a distance of at least 2 km from the nearest shore (Haggarty 1991). A burnt hurdle fence associated with cultivated soils, dating from the late second millennium BC, has also been found at Rattray, Aberdeenshire (Murray et al. 1992).

95

Rather, to see elements of the system as pragmatic solutions to mixing beasts with unenclosed fields, with the added benefit of exploiting the wider potential of the landscape. It is exactly this principle that lies behind the seasonal grazing of saltmarsh pastures along the Severn Estuary in the Middle Bronze Age (Britton and Müldner 2013; Bell 2013: 319–22), and there is every reason to anticipate that this sort of structured mobility, as Francis Pryor has termed it (2006: 93–94), formed part of agricultural practice in the Scottish landscape. Like later shieling practice, this is also likely to have required the exercise of rights by communities, establishing not only their right to the tenure of the lands they were currently occupying, but also of the lands they were grazing. Rights of tenure are unlikely to be detectable in archaeological data, and yet they are perhaps implicated in the routine discovery of burials and other deposits inserted into hut-­ circles and clearance heaps. At Cladh Hallan, for example, a series of foundation burials has been recorded beneath the hut-circles, but there were also other small deposits of artefacts on the floors – a pair of bronze chisels in one, a broken bronze bracelet in another – that appeared to mark the clo7.6 A Landscape System in Action sure of the structures prior to subsequent rebuilding (Parker Pearson et al. 2004: 74–77). At Achany Glen three cremation All these various strands of evidence – the settlements, burnt deposits were found in truncated pits adjacent to one of the mounds, clearance heaps, and fields  – seem to suggest a hut-circles, while another cremation was contained within a much more dynamic system of land-use than any conven- pit in the lower of the two phases of a clearance heap tional model of sedentary farming, in which the attendant (McCullagh and Tipping 1998: 91–94). Both here and on settlements are usually invested with a sense of permanence Arran burials were recovered also from more formal and the occupation of the upland margins only takes place ­monuments that on the surface were indistinguishable from when the lowlands are full. In this dynamic system, all the other clearance heaps (McCullagh and Tipping 1998: 84–91; different activities that sustained the long-term settlement of Barber 1997: 38–42; 88–91). In another clearance cairn the landscape seem to have been transient and intermittent at excavated on Arran there was a small cache of stone tools the locations where the fragments of evidence survive. At (Barber 1997: 29–30). Whatever rituals were being perAchany Glen the carbonised seed assemblages indicate that formed, these sorts of deposits provided direct links between cereals were consumed in the hut-circles, and what were the occupants of the land and the land itself. Some of these probably locally-grown cereal grains were incorporated into links were covert and could only have been known by those turves used in the walls (McCullagh and Tipping 1998: 168– who witnessed them or were told about them, but others 69). So it is reasonable to suppose that at any one moment in were more overt statements, such as the burials placed in time, each farming unit comprised a core of cultivated plots cairns with kerbs. Even if you did not know who had built and occupied houses, while the absence of enclosures indi- this monument, you would have known it was not you and cates that there were outlying pastures. Evidently movement that someone else had a claim to the surrounding land. was a key element in the year, with an annual cycle in which While the settlements to which later shielings belonged at sowing in the spring the beasts were driven onto the pas- were relatively permanent and long-lived, the hut-circle settures, where they were presumably tended by herds through tlements seem anything but (Halliday 2007). They were the summer months, only returning after harvest. This cycle apparently moving intermittently, and while the shifts may carries another faint echo of traditional farming practice, in have taken place within what we perceive as single hut-circle which sheep and cattle went to the shieling grounds for the groups, the low level of impact of 800 years of activity in the summer (see Dixon 2018). The medieval and later shieling-­ peat column adjacent to the hut-circles at Achany Glen system was based around complex traditional rights and (above) is perhaps a sign they were more radical, that occuobligations set out in the leases of the townships (e.g. see Bil pation of each group was intermittent. The evidence of occu1990; Harrison 2016), and it would be a mistake to uncriti- pation, desertion and reconstruction at Tormore 10/1 thus cally transpose this system onto Bronze Age settlements. represents episodes in a longer-term cycle, in which the set-

96

tlement may have shifted a kilometre or so every 10–15 years, only returning at longer intervals. In such a system, of course, none of these locations was actually abandoned, so much as the intensity and character of the use changed. Each was receiving nutrients transferred in various ways out of the surrounding landscape, and, in effect, each was being improved, so even when it was not occupied, it still offered the best grazing. This is probably why the majority of the agricultural indicators in the palaeoenvironmental record from Achany Glen are about pasture, and trees do not regenerate on the cleared ground. The areas previously occupied in this longer cycle are the very pastures that the beasts were being sent to during the summer months, and grazed in turn until harvest. This is the context within which the burnt mounds operated. Furthermore, in a mirror of tathing practice, the daily cycle of driving the beasts out more widely by day and bringing them back by night saw further transfer of nutrients onto these locations. This is why so many post-medieval shieling grounds appear relatively green and verdant spots to this day, and why others later became the sites of permanent farms. Is it so far-fetched to think about the Bronze Age landscape in this sort of framework? An in extenso system in which settlement territories in both upland and lowland are relatively large, and at its maximum extent the pattern of Middle Bronze Age settlement appears so widespread. Figure 7.7 attempts to sketch out an arbitrary series of steps Fig. 7.7  Schematic depiction of a farming territory in the landscape, in which the black arrows represent the long term movement of the settlement around different locations within its territory, and the white arrows represent the short term annual movements around its grazings

S. Halliday

representing such a territory, in which the occupied hut-­ circles move schematically at intervals around the most favourable locations in the landscape. At each step the previously cultivated segment reverts to a pasture grazed in an annual cycle and lies fallow while accumulating further nutrients. There is certainly nothing in the evidence to contradict it, and if the intervals are as short as proposed here, it explains why they remain undetectable in either the radiocarbon chronologies or the sampling resolution of most pollen diagrams. Continuity was invested in the totality of the landscape, not in any particular activity at any one spot. In the diagram the number of hut-circles is shown increasing in the course of the cycle to represent an overall increase in the population. This, of course, cannot be demonstrated, but the remains of the Middle Bronze Age landscape are the most extensive of any period before or since, including the permanently settled landscape at the high point of the Highland population in the late eighteenth and early nineteenth centuries. Only the post-medieval shielings go out beyond the hut-­ circles. And the hut-circle landscape extends well beyond the distribution of large Early Bronze Age and Neolithic cairns, so some expansion in the population is implicit. As depicted, the schematic model is fixed in its six arbitrary steps, but the improvement of areas of pasture through grazing practice, followed by occupation and cultivation, was a tried and tested process in the expansion of later settlement. In the

7  Fields and Farming-Systems in Bronze Age Scotland

model, we should anticipate that any increase in the numbers of mouths to feed would have equated with increased production, and that this would have entailed the taking in and improvement of new pastures and the addition of further steps in the cycle. But while the first steps would have exploited the most favourable locations, additional steps would increasingly occupy poorer ground. Here perhaps are the seeds of the final desertion at places like Achany Glen towards the end of the second millennium BC, though the continuation of grazing at much the same level of intensity as before indicates reorganisation of the system rather than catastrophe for the population. If this is the case, in some areas it might herald truly sedentary settlements occupying their sites for hundreds of years. Rather than moving round the landscape, the parent settlement occupied the most favourable location in its territory and invested all its efforts in staying there. Thus at Cladh Hallan on South Uist, settlement of the machair is sandwiched between the sea and the impoverished peatlands of the interior; in stark contrast to any mainland hut-circle group, the sequence of hut-circles here is thought to be continuous spanning some 700 years 1100–400 cal BC (Parker Pearson et al. 2004: 70), essentially representing the period following on from the major desertion of settlement seen at Achany Glen on the mainland. But on the opposite side of the country at Kintore, there is not a hint of this sort of intense superimposed occupation; if anything the very opposite (Cook and Dunbar 2008: 346, Fig. 196). The same desertion visited upon Achany Glen is apparently attested by a statistical break in the radiocarbon dates, but a series of the round-houses also date from the Late Bronze Age and Iron Age. The same case can be made at Carn Dubh, above 300 m OD in the uplands of north-east Perthshire, where the distinctive double-walled hut-circles are Late Bronze Age, and a date of 745–385 cal BC was returned from a burnt threshold timber in a single-walled hut-circle (Rideout 1995). Further afield in East Lothian, two successive unenclosed round-houses that preceded the fort at Broxmouth are incorporated into a general phase for which the Bayesian analysis indicates a range of 715–550 cal BC to 515–415 cal BC; both structures revealed evidence of maintenance and modification, indicating longer life-spans than most, but the decay and superimposition of the structures within a relatively short chronology of 105–190 years is redolent of the Middle Bronze Age sequences elsewhere. At the very least the houses and a palisaded enclosure here represent three separate occupations rather than the single phase to which they have been attributed (Armit and McKenzie 2013: 25–38). Survey in the uplands of south-east Scotland routinely identifies sequences between similar unenclosed houses and palisades, which if only for their materials must have been relatively short-lived.

97

Whatever changed towards the end of the second millennium BC, the regional settlement records that characterise Iron Age Scotland only seem to radically diverge about the middle of the first millennium BC. The broch sequences of the far north, for example, seem to emerge out of the so-­ called Hallstatt plateau, while if Broxmouth is at all representative of the concentration of fortified enclosures in south-eastern Scotland and Northumberland, the Bayesian statistics place its start at 515–415  cal  BC.  In Eastern Scotland, from the Moray Firth southwards to at least the Tay estuary, however, the emphasis remains firmly on clusters of unenclosed hut-circles throughout the first millennium, and their occupations appear as intermittent as their Middle Bronze Age predecessors. Changes were certainly happening by the Late Bronze Age (discussed by Bradley 2007: 178–225), but they were neither universal nor synchronous, and the events recorded at Achany Glen, Cladh Hallan and Kintore are only representative of their own regions. A more radical change may not have been visited on some of these regions until the fifth or perhaps even the fourth century BC.

7.7

Conclusion

Detailed analysis of a few well-preserved round-houses in Scotland challenges the long use and permanence of many prehistoric settlements (Halliday 1999, 2007; Barber and Crone 2001), which in turn challenges current understanding of land-use systems in their contemporary landscapes. Admittedly, little of the evidence is clear-cut and conclusive at the landscape scale, and though discontinuities in a series of structural sequences can be demonstrated, we can no more measure their duration than we can the length of the episodes of occupation; furthermore, we cannot show that the scattered buildings of any hut-circle group are truly contemporary, or indeed successive. Likewise, excavation and proxy records can demonstrate the complexity of elements of the agricultural systems in operation, but they cannot map the true extent of the fields and grazings at any one point in time, nor any territories to which they may belong. In many respects the argument still comes back to our individual perceptions of how farming landscapes work. If we are beguiled by comparisons with the historic environment, unwittingly or otherwise (e.g. Pryor 2006: 10), with its ancient patterns of walls and hedges, there is a danger of conferring an unwarranted permanence on the settlements woven into the scenery; the discontinuities fade into the background and are lost in the mixture of derelict, disused and occupied buildings that you might find around any farmyard today. However, the intermixing of the Bronze Age remains is surely a problem for this sort of interpretation. Monuments like burnt mounds seem to be the signature of activities that more typically belong in pastoral environments, both in the uplands and the

98

lowlands, and yet they are found cheek-by-jowl with settlements and fields. The adoption of an alternative hypothesis, in which the hut-circles and the fields are themselves transient fixtures in a dynamic settlement landscape, representing a series of short-term occupations spread at intervals, allow all these activities to mesh together, and it is only the comparatively poor temporal resolution of radiocarbon chronologies that creates the impression that they all take place together at the same time. But is this particular to Scotland, or is it telling us something about the enclosed Bronze Age landscapes elsewhere? Is enclosure really any more of a manifestation of intensification of production than, for example, the made soils of the Northern Isles, or the round-houses ploughed-out in the Middle Bronze Age at Achany Glen? Not unless the boundaries were being used to manage the land with a view to increasing production (see Yates 2007: 120–22). And herein lies the problem, we do not really understand how these fields in southern England were being used and have no independent measures of intensification that can be applied. If there is a time when cereal production is probably increasing in southern England, it is during the Iron Age rather than the Middle Bronze Age, and thus unrelated to these enclosed landscapes (see discussion in Sharples 2010: 41–43); and yet the lynchets on the chalk downland are born of ploughing. This is a conundrum that cannot be solved by simply shifting the emphasis of the interpretation onto the intensive production of sheep and cattle, even though careful management of the fields at Butser showed that the traditional stocking ratios of unfenced downland grazing could be more than doubled (information from the late Peter Reynolds). The ploughing that created the lynchetted field-systems on the downland, however, need not have been any more intensive than that found in Achany Glen, nor any more continuous. The real difference may simply be that it took place within fixed boundaries, which themselves do not necessarily mean that the grazings were intensively managed; large areas of enclosed fields may simply have been thrown open to the beasts during long fallow breaks. The key question is why the fixed boundaries, and if the answer to that lies in social relationships between people and land rather than the practical expressions of intensification in improved farming systems (Sharples 2010: 42), then the perceived differences in the physical remains between north and south are misleading. This line of argument can also be extended to the remains of the settlements, where there is no clear relationship between the pattern of the field-systems and the settlements lying within them (Bradley 2007: 192), and most seem to have been equally short-lived (Brück 1999). The suspicion would be that southern English landscapes have much more in common with the chaotic multi-period fragments of the Scottish experience than is generally recognised. As the survey of the Salisbury Plain Training Area brings out (McOmish

S. Halliday

et al. 2002: 152–54), the evolution of the enclosed downland landscape is much more complex than the regimented lynchets at first suggest; a similar case can be made for some of the patterns of enclosures inherent in the systems of reaves on Dartmoor (Johnston 2005) and indeed the systems of ditches in The Fens. This is also a theme that Bradley has embraced (2007: 187–92). On the downs it is clear that some settlements were inserted into existing field-systems, such as at South Lodge (Barrett et al. 1991: 181), and we should perhaps anticipate, as in Scotland, that Middle Bronze Age round-houses were typically constructed on spots that had been cultivated previously; they fell out of use after relatively short occupations; the sites were deserted for varying lengths of time; and subsequently new ones were often erected at or near the same spots. The same process of churning witnessed at Achany Glen was played out on the downs and delivered a similar result; Early Bronze Age settlement is invisible, while the visible remains of the Middle Bronze Age landscape appear the most extensive having been differentially preserved by subsequent withdrawal of settlement into other places. The evidence that allows us to separate the occupations with periods of desertion in Scotland is perhaps a facet of differential preservation operating at a national scale rather than a material difference in settlement and land-­ use histories.

References Alexander, D. (2000). Excavation of Neolithic pits, later prehistoric structures and a Roman temporary camp along the line of the A96 Kintore and Blackburn Bypass, Aberdeenshire. Proceedings of the Society of Antiquaries of Scotland, 130, 11–75. Armit, I., & McKenzie, J. (2013). An inherited place: Broxmouth hillfort and the South-East Scottish Iron Age. Edinburgh: Society of Antiquaries of Scotland. Banks, I., Duffy, P., & MacGregor, G. 2008. Archaeology of landscape change in south-west Scotland, 6000 BC – AD 1400: Excavation at William Grant & Sons Distllery, Girvan. Scottish Archaeological Internet Reports (SAIR), 32. Available at: http://socantscot.org/ index.php/sair/issue/archive. Accessed 4 Dec 2019. Barber, J. W. (1997). The archaeological investigation of a prehistoric landscape: Excavations on Arran 1978–81 (STAR monograph 2). Loanhead: Scottish Trust for Archaeological Research. Barber, J. W., & Crone, B. A. (2001). The duration of structures, settlements and sites: Some evidence from Scotland. In B.  Raftery & J.  Hickey (Eds.), Recent developments in wetland research (Seandálaíocht: Mon 2, Department of Archaeology, UCD, and WARP (Wetland Archaeology Research Project occasional paper 14)) (pp. 69–86). Dublin: UCD & WARP. Barclay, G.  J. (1985). Excavations at Upper Suisgill, Sutherland. Proceedings of the Society of Antiquaries of Scotland, 115, 158–198. Barrett, J.  C., Bradley, R., & Green, M. (1991). Landscape, monuments and society: The prehistory of Cranborne Chase. Cambridge: Cambridge University Press. Bell, M. (2013). The Bronze Age in the Severn Estuary. In CBA research report 172. York: Council for British Archaeology. Bil, A. (1990). The shieling 1600–1840: The case of the Central Scottish highlands. Edinburgh: John Donald.

7  Fields and Farming-Systems in Bronze Age Scotland Bradley, R. (1989). Herbert Toms – A pioneer of analytical field survey. In M. Bowden, D. Mackay, & P. Topping (Eds.), From Cornwall to Caithness. Some aspects of British field archaeology. Papers presented to Norman V. Quinell (British archaeological reports, British series 209) (pp. 29–47). Oxford: Archaeopress. Bradley, R. (2007). The prehistory of Britain and Ireland (Cambridge World Archaeology). Cambridge: Cambridge University Press. Britton, K., & Müldner, G. (2013). Stable isotope evidence of saltmarsh economies. In M. Bell (Ed.), The Bronze Age in the Severn Estuary (CBA research report 172) (pp.  263–273). Council for British Archaeology: York. Brück, J. (1999). Houses, life-cycles and deposition on Middle Bronze Age settlements in Southern England. Proceedings of the Prehistoric Society, 65, 145–166. Burgess, C. (1985). Population, climate and upland settlement. In D.  Spratt & C.  Burgess (Eds.), Upland settlement in Britain: The second millennium B.C. and after (British Archaeological Reports, British series 143) (pp. 195–230). Oxford: Archaeopress. Burgess, C. (1989). Volcanoes, catastrophe and the global crisis of the 2nd millennium BC. Current Archaeology, 117, 325–329. Carter, S., & Dalland, M. (2005). Bronze Age banks at the Black Crofts, North Connel, Argyll, a synthesis and re-assessment. Proceedings of the Society of Antiquaries of Scotland, 135, 191–212. Cook, M., & Dunbar, L. (2008). Rituals, roundhouses and romans: Excavations at Kintore, Aberdeenshire 2000–2006, Volume 1 Forest Road (STAR monograph 8). Loanhead: Scottish Trust for Archaeological Research. Cowley, D.  C. (2011). Remote sensing for archaeology and heritage management  – Site discovery, interpretation and registration. In D.  C. Cowley (Ed.), Remote sensing for archaeological heritage management: Proceedings of the 11th EAC seritage management symposium, Reykjavik, Iceland, 25–27 March 2010 (EAC ­occasional paper no. 5 and occasional publication of the Aerial Archaeology Research Group no. 3) (pp. 43–55). Archaeolingua: Hungary. Cressey, M., & Anderson, S. (2011). A later prehistoric settlement and metalworking site at Seafield West, near Inverness, Highland. Scottish Archaeological Internet Report (SAIR), 47. Available at: http://socantscot.org/index.php/sair/issue/archive. Accessed 4 Dec 2019. Cressey, M., & Strachan, R. (2003). The excavation of two burnt mounds and a wooden trough near Beechwood Farm, Inshes, Inverness 1999. Proceedings of the Society of Antiquaries of Scotland, 133, 191–203. Discovery & Excavation in Scotland (DES). (n.d.). Annual publication originally of Scottish Regional Group of the Council for British Archaeology and latterly by Archaeology Scotland. Dixon, P. (2011). Crops and livestock in the pre-improvement era. In A. Fenton & K. Veitch (Eds.), Farming and the land (pp. 229–243). John Donald: Edinburgh. Dixon, P. (2016). Mukked and folded land: The evidence of field data for medieval cultivation techniques in Scotland. In J. Klapste (Ed.), Agrarian technology in the medieval landscape (RURALIA 10) (pp. 107–123). Turnhout: Brepols Publishers. Dixon, P. (2018). What do we really know about transhumance in medieval Scotland: A review. In E. Costello & E. Svensson (Eds.), Historical archaeologies of transhumance across Europe (Themes in Contemporary Archaeology 6, EAA) (pp. 59–74). London/New York: Routledge. Dockrill, S.  J. (2007). Investigations in Sanday, Orkney. Vol 2: Tofts Ness, Sanday; An island landscape through 3000 years of prehistory. Orcadian Ltd: Kirkwall. Dodgshon, R. (2011a). Highland Touns before the clearances. In A. Fenton & K. Veitch (Eds.), Farming and the land (pp. 111–132). John Donald: Edinburgh.

99 Dodgshon, R. (2011b). Livestock farming in the highlands and islands before and after the clearances. In A.  Fenton & K.  Veitch (Eds.), Farming and the land (pp. 545–565). John Donald: Edinburgh. Donnelly, M., & MacGregor, G. (2005). The excavation of Mesolithic activity, Neolithic and Bronze Age burnt mounds and RomanoBritish ring groove houses at Gallow Hill, Girvan. Scottish Archaeological Journal, 27(1), 31–69. Dunbar, L. (2007). Fluctuating settlement patterns in Bronze Age Sutherland: The excavation of a roundhouse at Navidale, Helmsdale. Proceedings of the Society of Antiquaries of Scotland, 137, 137–167. Evans, C. (2009). Overviewing Fengate; the field-system triumvirate – A dialogue (with Francis Pryor, Andrew Fleming & Richard Bradley). In C. Evans, E. Beadsmoore, M. Brudenell, & G. Lucas (Eds.), Fengate revisited: Further Fen-edge excavations, Bronze Age fieldsystems & settlement and the Wyman Abbott/Leeds archives (pp. 239–267). Cambridge: Cambridge Archaeological Unit. Feachem, R. W. (1961). Unenclosed platform settlements. Proceedings of the Society of Antiquaries of Scotland, 94, 79–85. Feachem, R. W. (1973). Ancient agriculture in the highland of Britain. Proceedings of the Prehistoric Society, 39, 332–353. Fleming, A. (2008). The Dartmoor reaves: Investigating prehistoric land divisions (2nd ed.). Oxford: Windgather Press, Oxbow Books. Fowler, P. J. (1981). Later prehistory. In S. Piggott (Ed.), The agrarian history of England and Wales: Volume 1.1 Prehistory (pp. 61–298). Cambridge: Cambridge University Press. Guttmann, E.  B. A., Dockrill, S.  J., & Simpson, I.  A. (2004). Arable agriculture in prehistory: New evidence from soils in the Northern Isles. Proceedings of the Society of Antiquaries of Scotland, 134, 53–64. Haggarty, A. (1991). Machrie Moor, Arran: Recent excavations at two stone circles. Proceedings of the Society of Antiquaries of Scotland, 121, 51–94. Halliday, S.  P. (1990). Patterns of fieldwork and the distribution of burnt mounds in Scotland. In V.  Buckley (Ed.), Burnt offerings  – International contributions to burnt mound archaeology (pp. 60–61). Wordwell: Dublin. Halliday, S. P. (1993). Marginal agriculture in Scotland. In T. C. Smout (Ed.), Scotland since prehistory: Natural change & human impact (pp. 64–78). Aberdeen: Scottish Cultural Press. Halliday, S.P. (1999). Hut-circle settlements in the Scottish landscape. In P.  Frodsham, P.  Topping, & D.  C. Cowley (Eds.) ‘We were always chasing time.’ Papers presented to Keith Blood. Northern Archaeology, 17/18 (special ed.): 49–65. Halliday, S.  P. (2007). Unenclosed round-houses in Scotland: Occupation, abandonment, and the character of settlement. In C.  Burgess, P.  Topping, & F.  Lynch (Eds.), Beyond Stonehenge: Essays on the Bronze Age in honour of Colin Burgess (pp. 49–56). Oxford: Oxbow Books. Harrison, J. G. (2016). The documentary evidence, in chapter 9, exploiting the margins. In J. A. Atkinson (Ed.), Ben Lawers: An archaeological landscape in time (Scottish Archaeological Internet Reports (SAIR), 62) (pp. 210–213) Available at: http://socantscot.org/index. php/sair/issue/archive. Accessed 4 Dec 2019. Hunter, F. (2006). Excavations at Birnie, Moray, 2005. Interim Report. Dept of Archaeology, National Museums of Scotland, Edinburgh. Hunter, F. (2007). Excavations at Birnie, Moray, 2006. Interim Report. Dept of Archaeology, National Museums of Scotland, Edinburgh. Jobey, G. (1980). Green Knowe unenclosed platform settlement and Harehope cairn, Peeblesshire. Proceedings of the Society of Antiquaries of Scotland, 110, 72–113. Johnston, R. (2005). Pattern without a plan; rethinking the Bronze Age coaxial field systems on Dartmoor, South-West England. Oxford Journal of Archaeology, 24(1), 1–21. Kristiansen, K. (1987). Centre and periphery in Bronze Age Scandinavia. In M. Rowlands, M. Larsen, & K. Kristiansen (Eds.),

100 Centre and periphery in the ancient world (pp. 74–85). Cambridge: Cambridge University Press. Lelong, O., & MacGregor, G. (2007). The lands of ancient Lothian: Interpreting the archaeology of the A1. Edinburgh: Society of Antiquaries of Scotland. Maynard, D. (1993). Burnt mounds around a pipeline in Dumfries and Galloway. Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, Series, 3(79), 33–52. McCullagh, R.  P. J., & Tipping, R. (Eds.). (1998). The Lairg project 1988–96: The evolution of an archaeological landscape in Northern Scotland (STAR monograph) (Vol. 3). Scottish Trust for Archaeological Research: Loanhead. McIntyre, A. (1998). Survey and excavation at Kilearnan Hill, Sutherland, 1982–3. Proceedings of the Society of Antiquaries of Scotland, 128, 167–201. McOmish, D., Field, D., & Brown, G. (2002). The Field archaeology of the Salisbury Plain Training Area. Swindon: English Heritage. Mercer, R.  J. (1996). The excavation of a succession of prehistoric round-houses at Cnoc Stanger, Reay, Caithness, Highland, 1981–2. Proceedings of the Society of Antiquaries of Scotland, 126, 157–189. Moore, H., & Wilson, G. (1999). Food for thought: A survey of burnt mounds of Shetland and excavations at Tangwick. Proceedings of the Society of Antiquaries of Scotland, 129, 203–237. Murray, H. K., Murray, J. C., Shepherd, A. N., & Shepherd, I. A. G. (1992). Evidence of agricultural activity of the later second millennium BC at Rattray, Aberdeenshire. Proceedings of the Society of Antiquaries of Scotland, 122, 113–125. Parker Pearson, M. (2012). The machair survey. In M. Parker Pearson (Ed.), From machair to mountains: Archaeological survey and excavation in South Uist (pp. 12–73). Oxford: Oxbow Books. Parker Pearson, M., Sharples, N., & Symonds, J. (2004). South Uist: Archaeology and history of a Hebridean island. Stroud: Tempus. Phillips, T., & Bradley, R. (2004). Developer-funded fieldwork in Scotland, 1990–2003: An overview of the prehistoric evidence. Proceedings of the Society of Antiquaries of Scotland, 134, 17–51. Piggott, S. (1958). Native economies and the Roman occupation of North Britain. In I. A. Richmond (Ed.), Roman and native in North Britain (pp. 1–27). Nelson: Edinburgh. Pryor, F. (2006). Farmers in prehistoric Britain (2nd ed.). Stroud: Tempus. RCAHMS (Royal Commission on the Ancient and Historical Monuments of Scotland). (1990). North-east Perth: an archaeological landscape. Edinburgh: HMSO. RCAHMS (Royal Commission on the Ancient and Historical Monuments of Scotland). (1993). Strath of Kildonan: An archaeological survey. Edinburgh: RCAHMS. RCAHMS (Royal Commission on the Ancient and Historical Monuments of Scotland). (1994). Glenesslin, Nithsdale: An archaeological survey. Edinburgh: RCAHMS. RCAHMS (Royal Commission on the Ancient and Historical Monuments of Scotland). (1997). Eastern Dumfriesshire: an archaeological landscape. Edinburgh: HMSO. RCAHMS (Royal Commission on the Ancient and Historical Monuments of Scotland). (2008). ‘Well Shelterd & Watered’: Menstrie Glen, a farming landscape near Stirling. Edinburgh: RCAHMS. Rees, A.  R. (2002). A first millennium AD cemetery, rectangular Bronze Age structure and late prehistoric settlement at Thornybank, Midlothian. Proceedings of the Society of Antiquaries of Scotland, 132, 313–355. Rideout, J. (1995). Carn Dubh, Moulin, Perthshire: Survey and excavation of an archaeological landscape 1987–90. Proceedings of the Society of Antiquaries of Scotland, 125, 139–195.

S. Halliday Ritchie, A. (2005). Kilellan Farm, Ardnave, Islay: Excavations of a prehistoric to early medieval site by Colin Burgess and others 1954– 1976. Edinburgh: Society of Antiquaries of Scotland. Ritchie, J.  N. G., & Welfare, H. (1983). Excavations at Ardnave, Islay. Proceedings of the Society of Antiquaries of Scotland, 113, 302–366. Ronan, D., & Higgins, J. (2005). Bronze Age settlement at Ross Bay, Kirkcudbright. Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, Series, 3(79), 47–70. Sharples, N. (2010). Social relations in later prehistory. Oxford: Oxford University Press. Sharples, N. (2012). The beaker-period and early Bronze Age settlement at Sligecnach, Cill Donnain. In M. Parker Pearson (Ed.), From machair to mountains: Archaeological survey and excavation in South Uist (pp. 215–258). Oxford: Oxbow Books. Shepherd, I. A. G., & Tuckwell, A. N. (1977). Traces of Beaker-period cultivation at Rosinish, Benbecula. Proceedings of the Society of Antiquaries of Scotland, 108, 108–113. Sheridan, A. (2012). Neolithic Shetland: A view from the “mainland”. In D. L. Mahler (Ed.), The border of farming and the cultural markers (Short papers from the network meeting in Lerwick, Shetland, September 5th–9th, 2011) (pp.  6–36). Copenhagen: National Museum of Denmark. Available at: http://nordligeverdener.natmus.dk/forskningsinitiativer/samlet_projektoversigt/shetlandsoeerne_landbrug_paa_graensen_4000_3000_fvt/. Accessed 5 Dec 2019. Stevenson, J.  B. (1984). Excavation of a hut-circle at Cùl a’Bhaile, Jura. Proceedings of the Society of Antiquaries of Scotland, 114, 127–160. Strachan, R., & Dunwell, A. (2003). Excavations of Neolithic and Bronze Age sites near Peterhead, Aberdeenshire, 1998. Proceedings of the Society of Antiquaries of Scotland, 113, 137–171. Strachan, R., Ralston, I., & Finlayson, W. (1998). Neolithic and later prehistoric structures, and early medieval metal-working at Blairhall Burn, Amisfield, Dumfriesshire. Proceedings of the Society of Antiquaries of Scotland, 128, 55–94. Tipping, R. (2010). Bowmont: An environmental history of the Bowmont Valley and the northern Cheviot Hills, 10,000 BC–AD 2000. Edinburgh: Society of Antiquaries of Scotland. Turner, V. (2011). From homestead enclosure to farm? Field development in Shetland in the Neolithic period. In: D.  L. Mahler & C. Andersen, (Eds.) Farming on the edge: Cultural landscapes of the North. Some features of the Neolithic of Shetland. Short papers from the network meeting in Lerwick, Shetland, September 7th– 10th 2010. National Museum of Denmark: Copenhagen, pp. 19–31. Available at: http://nordligeverdener.natmus.dk/forskningsinitiativer/samlet_projektoversigt/shetlandsoeerne_landbrug_paa_ graensen_4000_3000_fvt/. Accessed 5 Dec 2019. White, R., & Richardson, P. (2010). The excavation of Bronze Age roundhouses at Oldmeldrum, Aberdeenshire. Scottish Archaeological Internet Reports (SAIR), 43. Available at: http:// socantscot.org/index.php/sair/issue/archive. Accessed 4 Dec 2019. Whittle, A., Keith-Lucas, M., Noddle, B., Rees, S., & Romans, J.  C. C. (1986). Scord of Brouster, an early agricultural settlement on Shetland, excavations 1977–79. Oxford: Oxford University Committee for Archaeology. Wickstead, H. (2007) Land division and identity in later prehistoric Dartmoor. South-West Britain: Translocating Tenure. Submitted for the Degree of PhD University College London. Available at: https:// discovery.ucl.ac.uk/id/eprint/1445203/. Accessed 5 Dec 2019. Yates, D. T. (2007). Land, power and prestige: Bronze Age field systems in Southern England. Oxford: Oxbow Books.

8

Terraced Fields, Farming, and Farmers at the Settlements of Kalamianos and Stiri, Greece Lynne A. Kvapil

Abstract

This chapter focuses on the adoption of agricultural terracing as a technological enhancement that resulted in changes to cultivation practices and distribution and organization of farm labour. The macroscopic investigation and systematic documentation of terraces in and around the settlements of Kalamianos and Stiri in the south-­eastern Corinthia in southern Greece revealed that several large systems of terraces were likely contemporary to these Late Bronze Age settlements in the productive hinterlands of nearby Argolid palace-centres like Mycenae. It is suggested here that the construction of several systems of agricultural terraces, likely achieved with some palatial support, prompted the implementation of diverse cultivation techniques that would have altered labour needs throughout the agricultural calendar. Garden terraces within the settlement of Kalamianos emphasize the variety of growing environments enabled by terraced farming and raise questions regarding the identity of farm workers and the creation of gendered agricultural spaces. Keywords

Field systems · Terraces · Ancient farming · Mycenaean · Bronze Age · Women’s work

8.1

Introduction

Field systems, used to organize and demarcate agricultural space or encourage animal and plant husbandry, go hand in hand with several innovations that define agricultural prac-

tices in Europe during the Bronze Age. The most notable of these innovations is perhaps the use of the ard plow and traction animals (Sherratt 1983; Pullen 1992), but it also includes the diversification of crop assemblages customized to suit climate, topography, and social needs as well as subsistence needs (e.g. Dal Corso et  al. 2012; Halstead 2004; Buxó et al. 1997; van Zeist et al. 1991). For farmers practicing plant husbandry in vertical landscapes, the use of agricultural terraces would have offered many benefits  – the creation of level land for cultivation, soil retention despite deep plowing, increased exposure of plants to the sun, and the prevention of erosion (e.g. Flood and Soles 2014; Bevan and Conolly 2004; Frederick and Krahtopoulou 2000; Treacy and Denevan 1994; Rackham and Moody 1992). The process of transformation in the landscape from natural and undefined to managed and controlled presupposes a transformation of farming practices most likely accompanied by social changes, including shifts in the allocation of labour and composition of the agricultural labour force. This paper examines how these changes may have played out at two settlements in the hinterland of the Mycenaean palaces of the Argolid in Greece, where several systems of Late Bronze Age agricultural terraces were discovered. Survey work around the settlements of Kalamianos and Stiri, located in the south-eastern Corinthia near the modern village of Korphos found that the surrounding landscape was heavily terraced at various points in the periodic settlement history of the region (Table 8.1). Although systems of agricultural terraces are sometimes associated with single farmsteads (Kunen 2001; Whitelaw 1991, 1994), the terraces in the Korphos region were in fact monumental in scope, and their scale is indicative of changes to farming practices in Late Bronze Age Greece.

L. A. Kvapil (*) Butler University, Indianapolis, IN, USA e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 S. Arnoldussen et al. (eds.), Europe’s Early Fieldscapes, Themes in Contemporary Archaeology, https://doi.org/10.1007/978-3-030-71652-3_8

101

102

L. A. Kvapil

Table 8.1  Periods of occupation in the Kalamianos region and their Approximate datesa Periods of occupation at Kalamianos FN – EBA LBA a Classical – Hellenistic Late Roman – Byzantine Early modern a Modern a

Dates 4500–2000 BC 1400–1200 BC 480–100 BC 300–1400 AD 1400–1914 AD 1914 AD – Present day

Source: Kvapil 2012 a Dated periods associated with terrace construction

8.2

Terraces in the Bronze Age Aegean

The use of agricultural terraces in the Aegean is often associated as much with the scarcity of water as with the need to create space for farming in a largely vertical landscape. Some of the earliest terraces discovered in this area were check dam or cross-channel terraces. On the island of Pseira on Crete check dams have been dated to the Middle Bronze Age (Hope Simpson et al. 2005),1 and on Kythera to the end of the Middle Bronze Age and beginning of the Late Bronze Age (Krahtopoulou and Frederick 2008).2 The check dams were created through the construction of a low wall across a seasonal stream bed or drainage channel, likely partially filled. As the wall collected sediment, the terrace tread built up over time. Because this type of terrace is positioned in the path of naturally flowing water, soil is seasonally infused with moisture. The structure of the terrace, which combines built stone and sedimentary accumulation, is conducive to retaining moisture trapped in open spaces between stones and held in soil behind the riser wall. A level space for cultivation also forms in the process. Check dams were just as useful as stand-alone agricultural installations or as parts of broader water management or landscape management ­programs. Although the construction of this type of terrace implies an awareness of the dangers of landscape destabilization, little evidence has ever shown that terracing was employed broadly enough to play a substantial role in soil conservation (e.g. Weiberg 2014; Krahtopoulou and Frederick 2008). Contour (or hillslope) terraces, so named because they run parallel to the natural contour of a slope, can be built and can function similarly to check dams. However, some examples suggest that contour terraces enhanced cultivation in multiple other ways. At Gournia contour terraces thought to have been built in the Middle and Late Bronze Age, conThe Middle Bronze Age on the island of Crete, which includes the Middle Minoan IA to IIIB periods, lasts from c. 2100/2050 to 1700/1675  BC.  The absolute chronology used here follows Manning (2010). 2  The terraces on Kythera are dated to the Neopalatial period on Crete, which lasts from Middle Minoan IIIA to Late Minoan IB, c. 1750/1700–1470/1460 BC. 1 

sisted of riser walls tall enough to act as windbreaks aimed at possibly protecting vineyards from strong coastal winds (Watrous 2012).3 As part of a system of agricultural installations, Late Bronze Age contour terraces on the island of Pseira off Crete (Betancourt 2006; Betancourt et  al. 2005; Hope Simpson et  al. 2005) and at Berbati, east of the palatial centre of Mycenae on the Greek mainland, were used to organize the landscape into productive agriculture spaces.4 The latter may also have been associated with the construction of the road system stemming from the palatial centre at Mycenae (Schallin 1996). Certainly, both contour terraces and the Mycenaean roads shared many of the same principles of construction. The best visible sections of the Mycenaean roads that survive are those that consist of monumental retaining terraces. Roads and contour terraces are jointly beneficial to farmers as they inevitably increase accessibility to farmland (Jansen 1994) especially if traction animals are used to pull ploughs or transport agricultural goods for processing. Terrace construction in these contexts is also related to a broader emphasis, apparent at Mycenaean centres, on water management (e.g. Lane et  al. 2016; Knauss 1987, 2001), including numerous post-and-lintel culverts for drainage that were built into monumental terraces supporting portions of the Mycenae road system (Schallin 1996; Wells et al. 1990). Another basic type of agricultural terrace common to the Mediterranean is the pocket or tree-crop buttress terrace (Moody and Grove 1990; Spencer and Hale 1961), which consists of a short, curved or circular section of retaining wall meant to contain one or two trees. While these terraces are common today, no examples that the author knows of have been found that are thought to date to the prehistoric Aegean. Finally, box terraces utilize riser walls on several sides as a means of demarcating space and containing soil (Spencer and Hale 1961). The space created is suitable for vineyards or gardens. Although they are not typically included in terrace typologies relating to the Aegean, they are included here and will be discussed further below. These terraces are tentatively dated to the Minoan Proto-palatial and Neopalatial periods (Middle Minoan IB to Late Minoan IB, c. 1925/1900–1470/1460  BC) based on masonry style and associated sherds, but, as is frequently the case with terraces encountered during surface surveys, the chronology of terrace construction is based on circumstantial evidence that awaits corroboration by other means. “Sherds on the surface” (Betancourt et al. 2005), however, are not always the most reliable means of establishing a secure date, especially if they are the only factor considered, as opposed to the study presented here which examines multiple lines of evidence. 4  Two terraces were excavated on the island of Pseira. Terrace G 2 is dated to the Middle Bronze Age (Middle Minoan period, c. 2100/2050– 1700/1675 BC); terrace Q 21, the terrace referred to here, is dated to the Late Bronze Age (Late Minoan I, c. 1700/1675–1470/1460 BC). Q 21 appears to have been a contour terrace, although there is some confusion about its classification in the original publication (Hope Simpson et al. 2005) as well as references to it in other scholarship (Krahtopoulou and Frederick 2008). 3 

8  Terraced Fields, Farming, and Farmers at the Settlements of Kalamianos and Stiri, Greece

8.3

Field Methodology and Results

Between 2009 and 2011, fieldwork under the aegis of the Saronic Harbors Archaeological Research Project (SHARP) explored the terraced landscape around the settlements of Kalamianos and Stiri in the south-eastern Corinthia (Fig. 8.1; Tartaron et  al. 2011). SHARP is an archaeological survey project focused on exploring the region surrounding the Late Bronze Age settlement at Kalamianos. This settlement was initially identified when the extensive survey project, EKAS, the Eastern Korinthia Archaeological Survey (Tartaron et al. 2006), discovered the basal remains of numerous stone-built structures on a small cape south-east of the resort village of Korphos on the coast of the Saronic Gulf. Data from EKAS and SHARP indicated that the settlement at Kalamianos dated to the Late Bronze Age (c. 1400–1200  BC; see Table 8.1). The results of SHARP so far indicate that Kalamianos was most active in the thirteenth century BC, the Late Helladic IIIB period on the Greek mainland (Tartaron et al. 2011). Kalamianos was the site of approximately 50 ­structures including numerous large building complexes as well as a series of roads and a threshing floor (Pullen 2015; Kvapil 2012; Tartaron et al. 2011). A nearly complete circuit

Fig. 8.1  Map of southern Greece including Mycenaean palace centres and sites mentioned in the text with insert showing the region surrounding the Mediterranean By courtesy of the author

103

wall also encircled the 3.5  ha settlement. Underwater and geomorphological surveys confirmed the existence of a deep-­water harbour (Tartaron et  al. 2011; Tartaron et  al. 2003). The intensive pedestrian survey of the region revealed an upland settlement, Stiri, in addition to the coastal site. Stiri, a contemporary but smaller settlement (ca. 1.4 ha), also featured several building complexes and appears to have connected this region to the interior of the Peloponnese. Throughout the entire survey area, SHARP identified hundreds of agricultural terrace walls, which were recorded and systematically documented as part of the survey (Kvapil 2012; Kvapil forthcoming). Using field systems, including agricultural terraces, as a form of archaeological data does not come without some problems. Dry stone masonry and earth works are notoriously difficult to date with certainty, although new technologies are making great headway in this area. While remote sensing technologies like lidar (Light Detection and Ranging) have made it possible to map systems of terraces with improved accuracy (Chase et al. 2014), the remote study of terraces can solve the chronology problem. When possible, terraces can be excavated and layers of sediment contained within the tread fill or palaeosols preserved beneath the terrace can be roughly dated if identifiable material remains are

104

L. A. Kvapil

Fig. 8.2  Locations of terrace documentation carried out during SHARP By courtesy of the author

unearthed (Krahtopoulou and Frederick 2008; Hope Simpson et al. 2005; French and Whitelaw 1999). There are problems inherent in this approach as well, as sherds or other remains can at best provide a terminus post quem for the construction of terrace or placement of tread fill but do not provide dates for terrace use or reuse over time. On the other hand, OSL (Optically Stimulated Luminescence) testing has been used to date terraces in the Judean Highlands with some certainty by pinpointing the last time soils containing quartz sediments were exposed to light (Davidovich et al. 2012). While OSL techniques are useful in terms of dating the use of terraces, it does not account for the reuse of terraces, and OSL still requires that features be excavated, something which is not always possible, particularly when large field systems are, or have been, studied as part of a non-invasive surface survey project. Whether excavation is possible or not, it is necessary to consider terrace remains on a variety of scales – as individual terraces, as part of larger terracing or field systems, and in conjunction with the suite of other data that contextualizes them. This was the aim of the methodology used for the sys-

tematic documentation of terraces around Kalamianos and Stiri, which sought not only to determine a relative sequence of terrace use but also to identify features of construction styles, the distribution of terraces throughout the landscape, and spatial relationships between terraces and other structures. Over the course of this investigation, 145 individual terraces in ten locations were documented (Fig.  8.2). Documentation consisted of an assessment of macroscopically observed architectural attributes and signs of preservation (Table 8.2) that were used to determine a sequence of terrace phases in each segment of terrace wall and to contextualize perceived stratigraphic relationships. Stone size, shape, and coursing style were key variables in distinguishing distinct phases. Variations in the sizes or shapes of stones chosen for terrace riser walls can reflect conformity to a particular standard or style of construction as well as the method by which materials were acquired. A quarried stone might look much different than a repurposed field stone, for example. Whether or not walls were built with regular courses and the distribution of stones within those courses signifies qual-

8  Terraced Fields, Farming, and Farmers at the Settlements of Kalamianos and Stiri, Greece Table 8.2  Attributes indicating stratigraphic sequences of terraces in the Kalamianos region Architectural and Preservation Attributes Description Characteristics of Wall Variation and consistency in stone construction size and shape Coursing or stone bonding Placement and distribution of stones in wall courses Surface character of Dressing or shaping individual stones Chemical weathering Lichen growth Source: By courtesy of the author

ity of construction and the forethought applied prior to construction. A very large stone in the upper courses of a terrace riser wall might have been better suited to a basal course and thus was likely rolled into place later in the process of construction. The presence and alignment of rillenkarren (Tartaron et  al. 2006), a type of chemical weathering that produces vertical ridges in stones over time, and lichen were used to confirm that individual stones might have been drawn from earlier terrace walls or nearby structures and reused. Mapping of the layout of individual terraces was also instrumental in understanding the arrangement of terraces in the landscape (Kunen 2001). Continuous or discontinuous lines of terraces, signs of agglutinative construction, the interrelationship of terraces, and their proximity to each other and nearby structures were essential – not only for distinguishing the presence of multiple phases but also for understanding how each phase of terraces changed the geomorphology of each slope and how subsequent phases were constructed in reaction to those alterations of the landscape. In some places, terrace walls were superimposed directly on earlier terrace walls. But, in several places later walls were built above and several centimetres behind or even in front of pre-existing walls (Fig.  8.3). In some areas, the close ­investigation of terrace arrangement revealed the attempted removal of terrace walls. Low ridges, visible on the surface despite being overgrown, were created by the unremoved remains of cobblestone packing used to reinforce the interior of the terrace walls (Fig.  8.4). These features would have been missed without detailed surface mapping. While each of the three identified terrace phases was distinguishable according to their construction attributes, proximity to various areas of occupation was also critical for determining periods of terrace construction in a survey context. Whitelaw’s investigation of nineteenth century terraces in northwestern Keos, for instance, found that the distribution of agricultural installations in the rural landscape correlated to the organization of land holdings and modes of agricultural production. Installations associated with subsistence production were located near farmsteads while those

105

needed for processing exported goods were proximal to a settlement (Whitelaw 1991, 1994). Similar results were produced from an investigation of the spatial arrangement of agricultural installations in the Mayan lowlands of Belize (Kunen 2001). It was critical, therefore, to relate the position of systems of terrace walls to the wide range of landscape features documented by SHARP, including anthropogenic features like farmsteads, settlements, and enclosure walls and geomorphological features such as the arable basins characteristic of the karstic landscape of this region. Assessment of all these features was then used to create relative sequences of terracing, which were then correlated with patterns of land use established by SHARP (Table 8.1) with the aim of understanding the most likely periods for terrace construction. In most areas surveyed, three phases of agricultural terrace construction were identifiable. A total of 86 per cent of documented terraces were able to be aligned to a specific chronological period, of which 53 per cent were identified as belonging to the earliest phase of terraces based on the assemblage of attributes, stylistic uniformity, and stratigraphic relationship to later phases. The periodic settlement history of this region is extremely helpful in correlating phases of terrace construction with likely periods during which cultivators might have farmed terraced land. Occupation in the region was most active in the Early Bronze Age, Late Bronze Age, Early Modern, and Modern periods (Table  8.1). Evidence for activity in other periods, such as Classical and Hellenistic periods and the Late Roman-Byzantine era, was scant or isolated to specific locations, such as activity during the Roman period indicated by  pottery found near a lime kiln at Kalamianos (Tartaron et al. 2011). If this area was largely unoccupied during those periods, it seems unlikely that terrace construction would have been carried out at those times. The periods when there is strong evidence for human activity seem much more likely candidates for terracing. In connecting terrace phases to the periods of occupation at Kalamianos, it is useful to work from the most recent period to the earliest. Of the three phases, the most recent can be connected with agricultural activity that occurred in the Modern period, mostly likely during the late twentieth century. These terraces were primarily pocket terraces consisting of piled-up rings of stone, frequently built into earlier structures and sometimes including modern materials like chicken wire. These stone rings were haphazardly built up around individual olive trees, they were located in close proximity to currently inhabited houses, including one house directly adjacent to the ruins of the settlement of Kalamianos. The next most recent phase of terraces appears to have been built in the Early Modern period. These terraces, which were either of the contour type or were long pocket terraces designed to contain two to three trees, had well-constructed retaining walls but with stones of various sizes inconsistently

106 Fig. 8.3  Terrace Wall 0114 on the north slope of Stiri. The photograph shows the basal course of a LBA terrace wall with trace remains of rubble packing running east to west. This wall lies at a lower elevation and to the north of an Early Modern terrace wall, shown in the background, which also runs east to west but stands nearly 2.00 m high. The line of the LBA terrace wall is partially interrupted by an Early Modern pocket terrace, marked in the photograph by the black and white meter stick By courtesy of the author

Fig. 8.4  Relict or ghost terrace where the retaining wall has been removed. The presence of the terrace is indicated by the change in elevation and loose rubble packing along the line of the contour By courtesy of the author

L. A. Kvapil

8  Terraced Fields, Farming, and Farmers at the Settlements of Kalamianos and Stiri, Greece

107

Fig. 8.5  Terrace Wall 0043 within the settlement of Kalamianos, which preserves a terrace wall with a corner and has been classified by the author as a box terrace. Traces of rubble, used to counterweigh the facing stones of the terrace wall are also visible. By courtesy of the author

placed in rough and irregular courses. Stones appear to have been gathered from the vicinity but not intentionally acquired. In many places within the settlement of Kalamianos, the later terraces describe above were built into these terraces, indicating that they predated the most recent terraces. Both the recent phases of terraces at Kalamianos reused stones  and the basal courses from the earliest terraces as well as structure walls from the Late Bronze Age settlement. Contour terraces near Stiri were also built from reused stones and incorporated foundation courses  of earlier walls. An Early Modern date for these terraces makes sense based on patterns of reuse as well as their arrangement in relation to other Early Modern features, including several farmsteads and an Early Modern church and monastery. Threshing floors were also found near the monastery and at least one of the farmsteads. This pattern of settlement and agricultural installations suggests that agriculture was organized in a decentralized way through kin groups and the church. The earliest phase of walls was frequently found stratigraphically below the two later phases, and, although this phase could be associated with either the Early or the Late Bronze Age, a preponderance of indirect evidence suggests that the earliest phase of agricultural terraces most likely dates to the Late Bronze Age. The position of most of the earliest phase terraces, in proximity to the settlements at Kalamianos and Stiri, suggests that they were associated with the Late Bronze Age settlements. Not only were cohesive systems of terraces constructed so as to radiate out from these settlements, a large terraced area was located within the settlement at Kalamianos. The terraces within Kalamianos were arranged so that they both respected and aligned with the architecture of the settle-

ment. Terrace walls are adjacent to the walls of the building complexes but never cross them. The construction style of the terrace walls, which feature lightly hammer dressed stones in regular courses counterweighted over a well-set basal course, largely matches that used for the building complexes in each settlement as well. The most significant difference is the use of the interior rubble packing, which was placed so as to counterweigh the wall stones and hold their pyramidal interior ends in place (Fig.  8.5). In addition, far more Late Bronze Age ceramics were found in and around this phase of terraces, while  the many of the  same areas, where the earliest phase of terraces were found, yielded little to no Early Bronze Age material culture. One slope at Stiri, where some of the earliest phases terraces wall were located, did yield Early Bronze Age material. These walls, upon first investigation, appeared to have been built with two-faced retaining walls. Agricultural terraces are usually built with a single, exterior facing wall that is often supported on the interior with a rubble core, so terrace walls with two faces were unusual. The intensive survey of this slope by SHARP determined that it was likely the location of an Early Bronze Age settlement that was abandoned and converted to a terraced slope for cultivation in the Late Bronze Age, and further investigation of these unusual walls revealed that the Late Bronze Age contour terrace walls were built directly in front of short stretches of Early Bronze Age architectural terrace walls in such a way that they utilized the pre-­existing system of retaining walls for additional support wherever possible. Finally, the Early Bronze Age landscape, based on findings from SHARP, appears to have been exploited much differently and in a much less uniform way than it was in the

108

L. A. Kvapil

Late Bronze Age. Approximately twenty-five cairns and twenty fortified elliptical stone enclosures, dated to this period by associated artefacts and documented as part of the survey, have been interpreted as signs that the landscape was organized in a fragmented way that may reflect competition between family or kin groups, including contests over agricultural land (Tartaron 2013, 2015). In such a scenario, agricultural land proximal to the cairns and enclosures would likely have been used for small-scale subsistence production. Moreover, no earliest phase terraces were found to have been arranged in relation to these cairns or enclosures. As stated above, the earliest phase terraces were arranged in association with the settlements, with groups of terraces positioned within the settlement of Kalamianos, as well as adjacent to it, and near the settlement at Stiri. The threshing floor  at Kalamianos, mentioned above, seems also to have been part of the Late Bronze Age settlement. Rather than being part of a farmstead, like those of the Early Modern period, this threshing floor was positioned at the settlement’s northern edge, consequently appearing to have been built so that it would have been widely accessible. The proximity and accessibility from the settlement raises the possibility that various aspects of farm work were organized at the community level and utilized shared labour, whereas the threshing floors near the Early Modern farmsteads or the monastery would reflect the organization of labour from those points. Although there is no certain evidence for a construction date, the accumulated data suggest that these agricultural installations, the terrace walls and the threshing floor, belong to the Late Bronze Age settlements at Kalamianos and Stiri. If this was the case, then it seems as if the incorporation of terraced agriculture would mark a shift in the organization of production at that time away from small-scale primarily subsistence production carried out at a family or kinship level (a system suggested by the Early Bronze Age features in the regional landscape) to large-scale production featuring ­infrastructure aimed at producing beyond the subsistence level and implying changes to the organization of labour.

8.4

Cultivation at Kalamianos and Stiri

Changes brought about by the adoption of terraced farming begin with an increase in land available for cultivation, and it is tempting to view the result as merely increasing the magnitude of production since more land implies greater yields. Halstead (1992a, b, 1999) has proposed as much for agriculture organized by the Mycenaean palaces, suggesting that these centres, in order to support necessary functions of their elite society and dependent labourers, increased production by bringing more land under cultivation. In order for this to be manageable, he proposed that there was an accompanying decrease in the amount of human labour expended to culti-

vate this land, through the adoption of sometimes wasteful but labour-saving techniques, like broadcasting seed, and the use of draught animals, practices he defined as extensification (Halstead 1992a, b, 1999). For a Mycenaean palace, extensification may have been a useful strategy. The specialized cultivation of grains like wheat and barley, paired with the adoption of extensive cultivation techniques, would have been an efficient use of labour that would presumably have produced more staple goods. But, while this model provides a starting point for understanding cultivation practices, it cannot expatiate a view of agricultural production at hinterland settlements. Because of the focus on palatial economy, Halstead’s model also cannot account for the full range of plants that may have been consumed at a Mycenaean palace as indicated by physical evidence or the handful of texts that mention non-­ subsistence goods. Linear B texts from Mycenae (Bennett and Chadwick 1958; Chadwick et  al. 1963; Palmer 1999) and botanical remains from Tiryns (Kroll 1982) suggest instead that many different types of plants may have been grown utilizing a spectrum of approaches to cultivation (see below), from intensive, or heavily reliant on human labour, to the extensive approach described above. The evidence for a diverse range of cultivars accords well with the types of growing environments that would have been available at Kalamianos and Stiri. Because of the preserved field systems and the unmodified arable land, the surrounding region would have provided a variety of environments that might have been adapted to the cultivation of various of crops. The land most suitable for cultivation in proximity to the settlements exists in flat basins between limestone ridges that are characteristic of the region’s karstic landscape. These depressions filled, over time, with eroded soils that have been cultivated in the Modern period with olives and wheat without need for any sort of landscape modification. Ceramics dating to the Late Bronze Age, which were recovered by SHARP (Tartaron et  al. 2011), suggest that the basins closest to the settlements may have been fertilized with material from middens at the settlements (Alcock et  al. 1994; Bintliff and Snodgrass 1988; Murray and Kardulias 1986).5 While it is difficult to say with certainty how these basins were farmed, it is clear that a mix of labour-­ intensive practices, such as manuring and crop rotation, combined with extensive techniques, such as the use of traction animals and the broadcasting of seed, would have been well-suited to land of this type and would have the potential Analysis of the distribution of moveable finds indicate that in some cases erosional processes moved Late Bronze Age material into these basins, but not always. Early Bronze Age material, for example, was found in high density in down slope areas of a basin north of the settlement of Stiri from the settlement but no finds from that period were collected in the flat areas of the basin, whereas Late Bronze Age material was also found within the basin (Kvapil 2012).

5 

8  Terraced Fields, Farming, and Farmers at the Settlements of Kalamianos and Stiri, Greece

to boost production. Although we cannot know without palaeobotanical evidence what exactly was grown in the basins, or anywhere else in the region for that matter, the basins would have been ideal for the cultivation with grains and pulses. Most slopes with Late Bronze Age terraces were located in areas outside the settlements and within relatively close proximity to them. Some of these terraced fields also produced sherd scatters suggesting that  contemporary  cultivation practices might have included manuring (Kvapil 2012). These terraced fields, like the basins, would therefore have been suitable for a combination of intensive and extensive farming practices. Contour terraces, as opposed to pocket terraces, would have made the sloped hillsides accessible to traction animals, and they could have been planted with either grains and pulses or orchard trees such as olives or figs. The terraces found within the settlement of Kalamianos represent the most chronologically secure sign of the use of much more labour-intensive techniques, including fertilization with manure and irrigation. The terraces yielded ceramic evidence that is both closely associated with the Late Bronze Age architecture of the settlement and in a proximity that is typical for the application of manure from middens containing domestic refuse (Bintliff and Snodgrass 1988). In addition, these terraces were located close to fissures in the bedrock that provide access to groundwater – a common feature of karstic topography – that would have supplied water for irrigation (Kvapil 2012). The morphology of some of the terraces within Kalamianos is suggestive of their function. Because of their width, several can be classified as contour box terraces rather than strictly as contour terraces (Fig. 8.5). Instead of creating a narrow strip of land along the contour of the slope, these terraces created broad square plots that followed the gently sloped incline. The shape and signs of intensive farming techniques suggest the use of these terraces for horticulture, which would have required intensive human labour. The attention needed for gardening would have been readily available within a settlement, and the growing environment created by the boxes would also have been perfect for any number of garden plants, including herbs and aromatics similar to plants that may have been used medicinally as well as for cuisine and for the production of perfumed oils (e.g. Palmer 1999; Shelmerdine 1985; Wylock 1972). This condensed picture of agricultural production at Kalamianos and Stiri offers a sense of the shift in approach to farming that is also suggested by the adoption of terraces. Although the creation of many types of growing environments suggests an increased diversity in cultivars, the types of crops grown may have been specialized depending on the type of field. Unaltered land would have lent itself to extensive cultivation, while slopes organized into terraced plots

109

would have been well-suited for either extensive cultivation of grains and pulses or the intensive and specialized cultivation of orchard crops. This sort of model is equally well-­ suited to a context in which diversity might have been aimed at self-sufficiency or one that would have served subsistence, surplus, and social purposes (Pullen 2019a, b).

8.5

Changed Land, Changed Labour

Changes in the scale and modes of cultivation, brought on in part by the adoption of terraced farming, would have necessitated significant changes in the organization of agricultural labour. One change associated with extensified farming would have been the replacement of human with animal labour, which had the potential to decrease labour requirements at some stages of the production process, like field preparation and sowing, while increasing the need for workers at harvest and for processing (Halstead 1987a, 1995, 2001). Orchard farming and horticulture, on the other hand, would have operated on much different agricultural calendars that vacillated between seasonality and steady attendance. As stated above, the basins, as well as the contour terraces near the settlements, would have been ideal locations for the extensive cultivation of grains and pulses carried out with the use of traction animals for ploughing and labour-saving sowing methods like the broadcasting of seed. Field preparation, during which earth, weeds, and stubble are turned prior to sowing, and the broadcasting of seed, ideally after a second ploughing, could have been completed with one person behind an animal-driven plough (Halstead 1995, 1987b). Without the use of animal labour, the amount of land that could be turned and sown by hand, by a single person or two, would have been restricted. Cultivation by hand makes much more sense for small plots. Ploughing without traction animals is labour intensive and slow, and broadcasting, like spending money to make money, works well on a large-scale but is detrimentally wasteful if not enough land is sown to make a return on the investment. So, if the increase in the amount of land was coupled with the use of animal labour, the amount of labour for these preparatory stages would have remained roughly the same or may have decreased, while there would have been an increased need at the end of the season for harvesting and threshing. If a status quo of hand farming was attempted on expansive fields, pooling labour would have been one possible way to complete ploughing and sowing in a timely manner, in which case labour needs during the agricultural year would have come in waves with spikes early in the season for ploughing and sowing and then again at harvest. Using shared labour at harvest is often necessary due to the limited time available for reaping and removing crops

110

from fields (Whittaker 2000; Forbes 1982; Miller 1946). In an extensive farming regime, traction animals might have been used to transport the harvest from the fields to centrally-­ located threshing floors, like the one found at Kalamianos, where they may also have been used in the threshing process.6 Although the amount of material harvested would be bulkier since more of the grain stalk had to be harvested to feed working animals, traction animals could have pulled their weight, so to speak, by transporting the more massive harvest and working to thresh it as well (Halstead 1987b, 1995). Harvesting the crops, which was surely done by hand, would require human labour and a lot of it during a short period of time, making shared labour a much more efficient option especially with the extra effort required to winnow grain from stalks. Since harvest was already the most labour-­ intensive task and one not  relegated entirely to animals (Halstead and Jones 1989), the need for human labourers at this stage may not have drastically changed even with larger yields; instead the proportion of workers needed for each task changed with the addition of animals. The deployment of extensive methods that may have increased efficiency of field preparation and sowing despite the increase in land being cultivated, had the potential of leaving varying numbers of farm labourers, once needed at these stages, unemployed or underemployed for much of the agricultural season prior to harvest. Some of this labour might have been partly returned to other agricultural tasks, such as construction and maintenance of terraces or the planting and tending of terraced orchards that may have been part of the diversified approach to production at Kalamianos and Stiri. Others may not have returned to agricultural labour at all. Of those who may have continued to be occupied with agricultural work, some might have found employment in the terraced gardens within the settlement of Kalamianos. Produce cultivated in the box terraces might have added variety to the subsistence diet of those in the settlements, but the gardens might also have been used for the production of luxury plants that would have been valued by palatial or local elites emulating the haute cuisine of palatial society. Both textual and physical evidence indicate that horticulture was practiced. Linear B texts found just outside the palace at Mycenae mention eleven identifiable aromatic plants, often referred to as condiments (Palmer 1999; Wylock 1972; Bennett and Chadwick 1958). The systematic methods of recording these aromatic plants on the texts suggest that, while some may have been gathered wild, many will have been deliberately cultivated, such as cumin, sesame, safflower, and coriander (Kvapil et  al. 2019). Macrobotanical Palaima (2015), for example, connects the use of pairs of working oxen, loaned by the palace on Crete, to the harvest of approximately a million litres of grain, as recorded on tablet KN F(2) 852. 6 

L. A. Kvapil

remains of melon and celeries found during excavations at the Mycenaean palace at Tiryns also confirm the presence of garden crops (Kroll 1982). The box terraces within the settlement of Kalamianos would have provided the perfect growing environment for garden plants. As stated above, these terraces yielded evidence for fertilization from middens, which must have been located nearby. At least one fissure giving access to groundwater was in close proximity to the terraced area and thus may have provided water for irrigation, which is essential to fertilization with manure. The terraces would have helped soil retain both water and nutrients. Manuring, the application of water, and weeding in addition to hand sowing and harvesting mean that the cultivation of these plots might have supported continuous cropping throughout the year, with assemblages of cultivars changing with the seasons. Thus, horticulture in the box terraces of Kalamianos would have required constant labour.

8.6

Women’s Farm Work

The box terraces of Kalamianos also offer an opportunity to consider agricultural labour through a gendered lens. Might these spaces might have been inhabited by women labourers doing agricultural work? Linear B texts suggest that women might have  served in some capacity  as farm workers, although there is by no means an abundance of data from these sources, the texts present intriguing information about the sexual  division of labour and how that it might have extended to agricultural-related work. Based on her analysis of the roles of women in the workforce using texts found at the palaces at Pylos and Knossos, Olsen (2014) suggests that women and men tended to be assigned ‘gender-specific tasks’, and so, at least in the context of the Mycenaean palaces, some tasks were considered feminine and others masculine while only a few occupations seem to have been accessible to either sex. She also found that the types of jobs assigned to groups of women were usually associated with food or domestic duties and tended to require continual labour, work she describes as ‘menial, repetitive, and potentially unending’. The me-re-ti-ri-ja, a work group thought to be women grain-grinders, are a good example of domestic work relating to agriculture (Olsen 2014; Billigmeier and Turner 1981). Although not directly involved with cultivation, food processing of this type – hand-grinding grain into flour for baking – is a useful example of the sort of agricultural work that is also viewed as unskilled, endless domestic work allotted to women. An agricultural calendar composed of  men’s farm work and consisting of intermittent and definable tasks hardly counts as ‘unending’ work, and there is little about the cultivation of grain or even orchard crops that could be con-

8  Terraced Fields, Farming, and Farmers at the Settlements of Kalamianos and Stiri, Greece

sidered as repetitive or endless in the same manner as the constant grinding of grain. The box terraces at Kalamianos, however, if they were used for horticulture, present an environment that would have necessitated both seasonal and constant farm work, especially considering the good chance that these gardens were cultivated year-round and were fertilized and irrigated. Because the box terraces were in the middle of the settlement, their close proximity to the domestic space also would have made them ideal places of employment for women who might have had other domestic duties, including child rearing (Olsen 1998). That women may have been associated with irrigation is also implied by the mention of another group of women labourers identified at Pylos. This group of women, called the re-wo-to-ro-ko-wo has been interpreted by some as bath-­ attendants. Olsen notes, however, that Chadwick, commenting on the large number of women in this group, described their duty as ‘the carrying of all the water required in the household’ (Olsen 2014). In a palatial context, women occupied in this way might be charged with moving bath-water or provisioning water for cooking or crafting. Thus, at Pylos, the task of constantly managing the palatial water supply was gendered female. The re-wo-to-ro-ko-wo at Pylos raise the possibility that overseeing irrigation in the gardens of Kalamianos, because the work involved the continual supply of water for farming in the domestic sphere, might also have been work associated with women. Women managing water in the context of a Mycenaean palace may, for some, call to mind women hydria-bearers from images and texts of historical Greek periods that are distinctive of the urban context of a Greek polis. Gendered notions of water and watering can be shared across urban and rural contexts, but the actualization of those occupations would surely have been different in each setting, while the underlying idea remained the same – providing the household at large with water. Perhaps, too, these Bronze Age connections between women and water acquisition survived over time, materializing in the  later iconography of tasks considered quintessentially ‘female’.

8.7

 errace Construction as Agricultural T Labour

Deployment of large-scale landscape modification, such as the construction of large systems of terrace walls, would have affected the range of necessary agricultural work needed for farming and the timing needs for labour (Lansing 1991). The clearance of land for building, quarrying or gathering stones for the wall and the packing, and the transport of this stone from varying distances depending on availability all had to precede construction (Treacy 1987), although if stone for terraces was quarried close to the site of terracing,

111

which has been proposed for the architectural stone used for structures at Kalamianos (Tartaron et al. 2011) then transport would not be particularly laborious nor would draught animals in addition to human labour be needed. Since the arrangement of terraces around Kalamianos and Stiri suggest that they were constructed as terracing systems, they would have required significant planning in advance and regular maintenance. The amount and nature of the work suggests that more than one individual was needed for this job. Thus, both the amount of labour that had to be allotted to field preparation in advance of cultivation would have increased as would the number of labourers allocated to the task. It is unclear whether a typical farmer would have had the knowledge required to construct an agricultural terrace. Yet, the knowledge used to build the terraces at Kalamianos and Stiri had long been available to anyone who had been involved in large-scale architectural construction. Careful macroscopic observation of features of terrace construction showed that the risers were built on basal courses so securely placed that they were reused by both later phases of terraces in many places where the two later phases were identified (Kvapil 2012; Kvapil forthcoming). This suggests that whoever built the Late Bronze Age terraces understood the construction principles needed to build terraces sturdy enough to last. Structures that use similar principles of earth retaining can be found in abundance in the environs of the palace at Mycenae, for example, where elite architecture including retaining walls dates back to the end of the sixteenth century BC.  The corbelled walls of the chambers of early tholos tombs consisted of retaining walls built from stones that were positioned so that pointed ends position to the interior were counterweighed by an earth covered rubble packing. The same type of counterweighing was mirrored in the terraces at Kalamianos and Stiri. Earth retention technology of this sort had thus been in use for centuries, although, during the Late Bronze Age, it was most frequently deployed for palatial monuments. It is possible, therefore, that the labourers with the architectural knowledge were in the employ of the palace. As part of the palatial workforce, these labourers may have been deployed regionally and even loaned out for palatially-sponsored projects. In the case of Kalamianos and Stiri, this may have been done with the promise that a portion of the resulting agricultural surplus would be returned to the palace or in return for access to the harbour (Kvapil 2012). The practice of palaces making loans to groups or individuals is not unparalleled. Palatial loans of metal farm tools and pairs of oxen are referenced in Linear B texts from Pylos and Knossos, respectively (Palaima 2015; Halstead 1999; Killen 1993). In cases in which palaces were interested in economic operations, they appear to have been willing to support those endeavours. Perhaps this also was the case with skilled

112

L. A. Kvapil

labourers, who may even have come with their own equipment (Palaima 2015). The labour requirements for the large-scale construction of terraces would certainly have had some effect on labour usually expended on farming at least while the terraces were being built. It is possible that farmers participated in architectural construction, working construction during slow parts of the agricultural calendar. And, although single terraces could possibly have been constructed by a single farmer, building large-scale systems of terraces would have likely required multiple people and certainly would have been completed more efficiently by groups of workers (Kvapil 2012).

8.8

Conclusions and More Questions

This concise review illustrates how the addition of terraced fields including urban terraced gardens to the regional cultivation regime at Late Bronze Age Kalamianos and Stiri would likely have resulted in a shift in labour needs throughout the agricultural calendar. Particularly notable is the possibility of an increased reliance on seasonal, communal labour that arises when farming technology changes. Although the use of animals minimized the need for human workers early on in the season, groups of people would be needed to harvest and initially process the larger harvests. This shift in labour needs and replacement of humans with animals may have left some agricultural workers, especially those without access to their own land, unemployed for much of the year. Some of this labour could have been redirected to other tasks such as terrace construction, this sort of temporary work would not be a sustaining occupation, except ­perhaps in the case of horticultural employment. Owning oxen was expensive, but palatial loans of oxen in addition to biennial fallowing that would have provided grazing for animals and changes to harvesting methods to provide them with feed would have reduced the expense. Thus, the fate of those un- and underemployed as a result of these changes remains cloudy. The identity of these workers also remains obscured. Are we discussing primarily men who typically laboured in the fields outside the settlements and were thus displaced by changes to agricultural regimes? Or did these changes affect men, women, and even children? Linear B texts suggest that occupations, at least in the eyes of the palaces, were strictly gendered, but did these gendered designations extend to palatial hinterlands, and, if so, what were the mechanisms behind the transfer of these cultural ideas about work? There are more questions than answers regarding these topics, but the questions raised emphasize the impact that terraces and

technological change may have had on the people of the Late Bronze Age Aegean. Acknowledgements  I would like to thank Tom Tartaron and Daniel Pullen for allowing and encouraging my fieldwork as part of SHARP, and the editors of this volume for their insightful comments on the text.

References Alcock, S. E., Cherry, J. F., & Davis, J. L. (1994). Intensive survey, agricultural practice and the classical landscape of Greece. In I. Morris (Ed.), Classical Greece: Ancient histories and modern archaeologies (pp. 137–170). Cambridge: Cambridge University Press. Bennett, E.  L., & Chadwick, J. (1958). The Mycenae tablets II. Transactions of the American Philological Society, 48(1), 1–122. Betancourt, P. P. (2006). Survey conclusions. In P. P. Betancourt (Ed.), The Chrysokamino metallurgy workshop and its territory (Hesperia supplements 36) (pp.  257–278). Princeton: American School of Classical Studies at Athens. Betancourt, P. P., Davaras, K., & Hope Simpson, R. (2005). Pseira IX: The archaeological survey of Pseira Island part 2. The intensive surface survey. Philadelphia: INSTAP Academic Press. Bevan, A., & Conolly, J. (2004). GIS, archaeological survey, and landscape archaeology on the island of Kythera, Greece. Journal of Field Archaeology, 29(1/2), 123–138. Billigmeier, J.-C., & Turner, J. A. (1981). The socio-economic roles of women in Mycenaean Greece: A brief survey from evidence of the Linear B tablets. Women’s Studies, 8, 3–20. Bintliff, J. L., & Snodgrass, A. M. (1988). Off-site pottery distributions: A regional and interregional perspective. Current Anthropology, 29(3), 506–513. Buxó, i., Capdevila, R., Alonso, N., Canal, D., Echave, C., & González, I. (1997). Archaeobotanical remains of hulled and naked cereals in the Iberian Peninsula. Vegetation History and Archaeobotany, 6(1), 15–23. Chadwick, J., Bennett, E. L., French, E. B., Taylour, W., Verdelis, N. M., & Williams, C. K., II. (1963). The Mycenae tablets III. Transactions of the American Philological Society, 52(7), 1–76. Chase, A. F., Chase, D. Z., Awe, J. J., Weishampel, J. F., Iannone, G., Moyes, H., Yaeger, J., & Brown, M. K. (2014). The use of LiDAR in understanding the ancient Maya landscape. Caracol and Western Belize. Advances in Archaeological Practice, 2(3), 208–221. Dal Corso, M., Marchesini, M., Leonardi, G., & Kirleis, W. (2012). Environmental changes and human impact during the Bronze Age in northern Italy: On-site palynological investigation at Fondo Paviani, Verona. In J.  Kneisel, W.  Kirleis, M.  Dal Corso, N.  Taylor, & V. Tiedke (Eds.), Collapse or continuity? Environment and development of Bronze Age human landscapes. Proceedings of the international workshop socio-environmental dynamics over the last 12,000 years: The creation of landscapes II (14th–18th march 2011) in Kiel (pp. 71–83). Bonn: Verlag Dr. Rudolf Habelt GmbH. Davidovich, U., Porat, N., Gadot, Y., Avni, Y., & Lipschits, O. (2012). Archaeological investigations and OSL dating of terraces at Ramat Rahel, Israel. Journal of Field Archaeology, 37(3), 192–208. Flood, J.  M., & Soles, J.  S. (2014). Water management in Neopalatial Crete and the development of the Mediterranean dry-­ season. In G.  Touchais, R.  Laffineur, & F.  Rougemont (Eds.), PHYSIS. L’environnement naturel et al relation homme-milieu dans le monde égéen protohistorique. Actes de la 14e Recontre égéenne internationale, Paris, Institut National d’Histoire de l’Art (INHA) 11–14 décembre 2012 (pp. 79–84). Leuven - Liege: Peeters.

8  Terraced Fields, Farming, and Farmers at the Settlements of Kalamianos and Stiri, Greece Forbes, H.  A. (1982). Strategies and soils: Technology, production and environment in the Peninsula of Methana, Greece. PhD thesis, University of Pennsylvania. Frederick, C., & Krahtopoulou, A. (2000). Deconstructing agricultural terraces: Examining the influence of construction method on stratigraphy, dating and archaeological visibility. In P. Halstead & C. Frederick (Eds.), Landscape and land use in postglacial Greece (pp. 79–94). Sheffield: Sheffield Academic Press. French, C. A. I., & Whitelaw, T. (1999). Soil erosion, agricultural terracing and site formation processes at Markiani, Amorgos, Greece: The micromorphological perspective. Geoarchaeology, 14(2), 151–189. Halstead, P. (1987a). Man and other animals in later Greek prehistory. The Annual of the British School at Athens, 82, 71–83. Halstead, P. (1987b). Traditional and ancient rural economy in Mediterranean Europe: Plus ça change? The Journal of Hellenic Studies, 107, 77–87. Halstead, P. (1992a). Agriculture in the Bronze Age Aegean: Towards a model of palatial economy. In B. Wells (Ed.), Agriculture in Ancient Greece (pp. 104–117). Swedish Institute in Athens: Stockholm. Halstead, P. (1992b). The Mycenaean palatial economy: Making the most of the gaps in the evidence. Proceedings of the Cambridge Philological Society, 38, 57–86. Halstead, P. (1995). Plough and power: The economic and social significance of cultivation with the ox-drawn Ard in the Mediterranean. Bulletin on Sumerian Agriculture, 8, 11–22. Halstead, P. (1999). Mycenaean agriculture: The nature of palatial intervention. Bulletin of the Institute of Classical Studies, 43, 211–212. Halstead, P. (2001). Mycenaean wheat, flax and sheep: Palatial intervention in farming and its implications for rural society. In S. Voutsaki & J. T. Killen (Eds.), Economy and politics in the Mycenaean palace states. Proceedings of a conference held on 1–3 July 1999 in the Faculty of Classics, Cambridge (pp.  38–50). Cambridge: Cambridge Philological Society. Halstead, P. (2004). Life after Mediterranean polyculture: The subsistence subsystem and the emergence of civilisation revisited. In J. C. Barrett & P. Halstead (Eds.), The emergence of civilisation revisited (pp. 189–206). Oxford: Oxbow Books. Halstead, P., & Jones, G. (1989). Agrarian ecology in the Greek Islands: Time stress, scale and risk. The Journal of Hellenic Studies, 109, 41–55. Hope Simpson, R., Goldberg, P., & Clark, J. A. (2005). The two agricultural terraces chosen for excavation. In P. P. Betancourt, K. Davaras, & R. Hope Simpson (Eds.), Pseira IX: The archaeological survey of Pseira Island, part 2: The intensive surface survey (pp. 251–256). Philadelphia: INSTAP Academic Press. Jansen, A. (1994). Stations along the roads in the area of Mycenae: An analysis of the Mycenaean road system and its relation to the Mycenaean State. PhD thesis, University of Pennsylvania. Killen, J.  T. (1993). The oxen’s names on the Knossos Ch tablets. Minos, 27(28), 101–107. Knauss, J. (1987). Die Melioration des Kopaisbeckens durch die Minyer im 2 Jh.v.Chr. Kopais 2 - Wasserbau und Siedlungsbedingungen im Alterum. Munich: Obeinach Technische Universität München. Knauss, J. (2001). Späthelladische Wasserbauten: Erkundungen zu wasserwirtschaftlichen Infrastrukturen der mykenischen Welt. Munich: Technische Universität München. Krahtopoulou, A., & Frederick, C. (2008). The stratigraphic implications of long-term terrace agriculture in dynamic landscapes: Polycyclic terracing from Kythera Island, Greece. Geoarchaeology, 23(4), 550–585. Kroll, H. (1982). Kulturpflanzen von Tiryns. Archäologischer Anzeiger, 1982, 467–485. Kunen, J.  L. (2001). Ancient Maya agricultural installations and the development of intensive agriculture in NW Belize. Journal of Field Archaeology, 28(3/4), 325–346.

113

Kvapil, L.A. (2012). The Agricultural Terraces of Korphos-Kalamianos: A Case Study of the Dynamic Relationship Between Land Use and Socio-Political Organization in Prehistoric Greece. Cincinnati: PhD thesis, University of Cincinnati. Kvapil, L.A. (forthcoming). Bronze Age terracing systems. In: T.F.  Tartaron & D.J.  Pullen, eds. The Saronic Harbors Archaeological Research Project: Engaging land and sea in the Saronic Gulf, Greece. Kvapil, L.  A., Meier, J.  S., Price, G.  C., & Shelton, K.  S. (2019). Beyond city and country at Mycenae: Urban and rural practices in a subsistence landscape. In D. Garcia, R. Orgeolet, M. Pomadère, & J. Zurbach (Eds.), Country in the city. Agricultural functions in protohistoric urban settlements (Aegean and Western Mediterranean) (pp. 122–136). Oxford: Archaeopress. Lane, M.  F., Horsely, T.  J., Charami, A., & Bittner, W.  S. (2016). Archaeological geophysics of a Bronze Age agricultural landscape: The Aroura Project, central mainland Greece. Journal of Field Archaeology, 41(3), 271–296. Lansing, J. S. (1991). Priests and programmers: Technologies of power in the engineered landscape of Bali. Princeton: Princeton University Press. Manning, S.  W. (2010). Chronology and terminology. In E.  H. Cline (Ed.), The Oxford handbook of the Bronze Age Aegean (pp. 11–28). Oxford: Oxford University Press. Miller, W.  M. (1946). A threshing ring in southern Ohio. Hoosier Folklore, 5(1), 3–13. Moody, J. A., & Grove, A. T. (1990). Terraces and enclosure walls in the Cretan landscape. In S. Bottema, G. Entjes-Nieborg, & W. van Zeist (Eds.), Man’s role in the shaping of the eastern Mediterranean landscape: Proceedings of the INQUA/BAI symposium on the impact of ancient man on the landscape of the eastern Mediterranean region and the Near East, Groningen, Netherlands, 6–9 March (Vol. 1989, pp. 183–194). Rotterdam: A.A. Balkema. Murray, P., & Kardulias, P.  N. (1986). A modern-site survey in the southern Argolid, Greece. Journal of Field Archaeology, 13(1), 21–41. Olsen, B. A. (1998). Women, children and the family in the late Aegean Bronze Age: Differences in Minoan and Mycenaean constructions of gender. World Archaeology, 29(3), 380–392. Olsen, B. A. (2014). Women in Mycenaean Greece. The Linear B tablets from Pylos and Knossos. London/New York: Routledge. Palaima, T. (2015). The Mycenaean mobilization of labor in agriculture and building projects: Institution, individuals, compensation and status in the Linear B tablets. In P. Steinkeller & M. Hudson (Eds.), Labor in the ancient world (Vol. 5, pp.  617–647). ISLET-Verlag: Dresden. Palmer, R. (1999). Perishable goods in Mycenaean texts. In S. Deger-­ Jalkotzy, S. Hiller, & O. Panagl (Eds.), Floreant Studia Mycenaea. Akten des X.  Internationalen Mykenologischen Colloquiums in Salzburg vom 1.-5. Mai 1995. Band I (pp. 463–485). Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Pullen, D.  J. (1992). Ox and plow in the early Bronze Age Aegean. American Journal of Archaeology, 96(1), 45–54. Pullen, D.  J. (2015). How to build a Mycenaean town: The architecture of Kalamianos. In A.-L.  Schallin & I.  Tournavitou (Eds.), Mycenaeans up to date. The archaeology of the northeastern Peloponnese - current concepts and new directions (pp. 377–390). Stockholm: Svenska Institutet i Athen. Pullen, D.  J. (2019a). Agricultural self-sufficiency and Mycenaean Kalamianos on the Saronic Gulf. In D.  Garcia, R.  Orgeolet, M. Pomadére and J. Zurbach (Eds.), Country in the city. Agricultural functions in protohistoric urban settlements (Aegean and Western Mediterranean) (pp. 137–151). Archaeopress: Oxford. Pullen, D. J. (2019b). If you build it, will they come? Will they Stay? The Mycenaean port town of Kalamianos. In A.  Gyucha (Ed.), Coming together. Comparative approaches to population aggre-

114 gation and early urbanization (pp.  215–238). State University of New York Press: Albany. Rackham, O., & Moody, J.  A. (1992). Terraces. In B.  Wells (Ed.), Agriculture in ancient Greece. Proceedings of the seventh international symposium at the Swedish Archaeological Institute at Athens, 16–17 May 1990 (pp. 123–133). Svenska Institutet i Athen: Stockholm. Schallin, A.-L. (1996). The Berbati-Limnes archaeological survey. The late Helladic period. In B.  Wells (Ed.), The Berbati-Limnes archaeological survey 1988–1990 (pp. 123–175). Svenska Institutet i Athen: Stockholm. Shelmerdine, C. W. (1985). The perfume industry of Mycenaean Pylos. Göteborg: P. Åströms Förlag. Sherratt, A. (1983). The secondary exploitation of animals in the Old World. World Archaeology, 15(1), 90–104. Spencer, J., & Hale, G. (1961). The origin, nature, and distribution of agricultural terracing. Pacific Viewpoint, 2(1), 1–40. Tartaron, T.  F. (2013). Maritime networks in the Mycenaean world. Cambridge: Cambridge University Press. Tartaron, T. F. (2015). Late Bronze Age architecture and regional dynamics at Korphos in the Corinthia. In A.-L. Schallin & I. Tournavitou (Eds.), Mycenaeans up to date. The archaeology of the northeastern Peloponnese - current concepts and new directions (pp. 391–401). Stockholm: Svenska Institutet i Athen. Tartaron, T. F., Rothaus, R. M., & Pullen, D. J. (2003). Searching for prehistoric Aegean harbors with GIS, geomorphology, and archaeology. Athena Review, 3(4), 27–36. Tartaron, T. F., Gregory, T. E., Pullen, D. J., Noller, J. S., Rothaus, R. M., Rife, J.  L., Tzortopoulou-Gregory, L., Schon, R., Caraher, W.  R., Pettegrew, D.  K., & Nakassis, D. (2006). The eastern Korinthia archaeological survey: Integrated methods for a dynamic landscape. Hesperia, 75, 453–523. Tartaron, T.  F., Pullen, D.  J., Dunn, R.  K., Tzortopoulou-Gregory, L., Boyce, J. E., & Dill, A. (2011). The Saronic harbors archaeological research project (SHARP): Investigations at Mycenaean Kalamianos, 2007-2009. Hesperia, 80, 559–634.

L. A. Kvapil Treacy, J.  M. (1987). Building and rebuilding agricultural terraces in the Colca Valley of Peru. Yearbook, Conference of Latin American Geographers, 13, 51–57. Treacy, J.  M., & Denevan, W.  M. (1994). The creation of cultivable land through terracing. In N.  F. Miller & K.  L. Gleason (Eds.), The archaeology of garden and field (pp.  91–110). Philadelphia: University of Pennsylvania Press. van Zeist, W., Wasylikowa, K., & Behre, K.-E. (Eds.). (1991). Progress in Old World palaeoethnobotany. A retrospective view on the occasion of 20 years of the international work Group for Palaeoethnobotany. Rotterdam: A. A. Balkema. Watrous, V.  L. (2012). The harbor complex of the Minoan town at Gournia. American Journal of Archaeology, 116, 521–541. Weiberg, E. (2014). Timing, perception and response. Human dimensions of Erosion and sedimentations in the Greek Bronze Age. In G.  Touchais, R.  Laffineur, & F.  Rougemont (Eds.), PHYSIS. L’environnement naturel et al relation homme-milieu dans le monde égéen protohistorique. Actes de la 14e Recontre égéenne internationale, Paris, Institut National d’Histoire de l’Art (INHA) 11–14 décembre 2012 (pp. 41–48). Leuven - Liege: Peeters. Wells, B., Runnels, C. N., & Zangger, E. (1990). The Berbati-Limnes archaeological survey. The 1988 season. Opuscula Atheniensius, 18, 207–238. Whitelaw, T. (1991). The ethnoarchaeology of recent rural settlement and land use in northwest Keos. In J.  F. Cherry, J.  L. Davis, & E.  Mantzourani (Eds.), Landscape archaeology as long-term history: Northern Keos in the Cycladic Islands from earliest settlement until modern times (Monumenta Archaeologica 5, 16) (pp.  403– 454). Los Angeles: UCLA Institute of Archaeology. Whitelaw, T. (1994). An ethnoarchaeological study of rural land-use in north-west Keos: Insights and implications for the study of past Aegean landscapes. In P.  N. Doukellis & L.  G. Mendoni (Eds.), Structures Rurales et Sociétés Antiques: Actes du Colloque de Corfou (14–16 1992) (pp. 163–186). Paris: Les Belles Lettres. Whittaker, J. C. (2000). Alonia and Dhoukanes: The ethnoarchaeology of threshing in Cyprus. Near Eastern Archaeology, 63(2), 62–69. Wylock, M. (1972). Les Aromates dans les Tablettes Ge de Mycènes. Studi Micenei ed Egeo-Anatolici, 15, 105–146.

9

The Changing Fieldscapes of Loughcrew: New Insights from Airborne Lidar Corinne Roughley, Elizabeth Shee Twohig, Colin Shell, and Gillian Swanton

Abstract

Loughcrew (Co. Meath, Ireland) is famous for its passage tombs, but it is also a landscape with well-preserved multi-period relict field systems. Over 150 km of ‘field’ boundaries have been revealed by the Loughcrew Landscape Project’s airborne lidar survey. This complex of overlapping boundaries can be disentangled by considering the relationships between boundaries and also their relationships with prehistoric monuments and early medieval sites. Through such an approach, a sequence emerges which suggests that the earliest phase(s) of field system are most likely prehistoric in date. Later phases are likely to be medieval or more recent, but dating of such features remains challenging in general. A striking observation is that the Loughcrew fieldscape has been repeatedly, and radically, altered: in some places the shape, orientation, size and regularity of the boundaries have been completely reorganised three or four times. Some boundaries such as the linear feature which appears to predate the Loughcrew “Cursus” are long and winding and do not enclose a defined space; such boundaries are not archetypical field demarcations. Integration of the lidar evidence with historic mapping (from the late eighteenth century to the present) has allowed the impact of more recent landscape changes to be evaluated. Although the twentieth century saw a radiC. Roughley (*) Hughes Hall, University of Cambridge, Cambridge, UK e-mail: [email protected] E. Shee Twohig Department of Archaeology, University College Cork, Cork City, Ireland C. Shell McDonald Institute for Archaeological Research, University of Cambridge, Cambridge, UK G. Swanton Kynet Consultancy, Marlborough, Wiltshire, UK

cal reorganisation of land tenure and the fieldscape, it is the developments of the first half of the nineteenth century which have affected the preservation of earlier boundaries most dramatically. The extensive nature of the relict field systems challenges monument-based management strategies. The area has seen widespread development in recent years with many individual new houses being built scattered across the landscape in addition to a small number of larger developments. Many of the dispersed new buildings are in areas which include relict field systems. It is not proportionate to register all 150 km of features recorded as being instances of the monument class “field system”. However, the ongoing piecemeal development does have a potentially significant cumulative impact on the prehistoric fieldscape and this needs to be considered carefully.

Keywords

LiDAR · Historic mapping · Loughcrew · Ireland · Field boundaries

9.1

Introduction

Fields are an integral part of the Irish prehistoric landscape and, while monuments have dominated publications, there is a long tradition of research considering the role of fields in periods from the Neolithic onwards (e.g. Herity 1971: 258–65; Reeves-Smyth and Hammonds 1983; Cooney 1991: 123–39). This paper examines the archaeological evidence for field boundaries found in the Loughcrew landscape and discusses the interpretive and practical challenges the data pose. The sheer extent and complexity of the surviving features itself creates difficulties for the recording and manage-

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 S. Arnoldussen et al. (eds.), Europe’s Early Fieldscapes, Themes in Contemporary Archaeology, https://doi.org/10.1007/978-3-030-71652-3_9

115

116

ment of a landscape which is increasingly being developed, with many features being lost to new housing. Although none of the field boundaries are dated, the successive alteration of field orientation and shape also challenge our understanding of change in and through prehistoric and protohistoric fieldscapes. Furthermore, not all the linear banks appear to have necessarily enclosed an agricultural space, so the concept of a “field” may not invariably be appropriate. The presence in Ireland of extensive and substantial prehistoric field boundaries is well known from sites in the west of the country (Herity 1971: 258–65; Cooney 2000: 46–47; Cooney et al. 2011: 562–669). On the coast of northern Co. Mayo, there are cohesive prehistoric fields systems at Céide Fields and Belderg Beg (Caulfield 1978: 137–43; Caulfield 1983: 195–215; Verrill and Tipping 2010a: 1011–21; Verrill and Tipping 2010b: 1214–25). This has led to a very different discourse in Ireland surrounding Neolithic agriculture and land use (Cooney 1997: 23–31; Smyth 2011: 6–10). Unlike in British prehistory, settled agriculture in the Neolithic had been broadly accepted (Bradley 2007: 43ff). But recently, reconsideration of the dating evidence by Whitefield (2017) has seriously questioned the proposed Neolithic date for the Co. Mayo field systems and called for a “suspension of the paradigm that defines the coaxial field system on Céide Hill as Neolithic” (ibid., 18). Whitefield (2017: 1–23) has argued convincingly that both the original dating proposed for the field system (Caulfield 1978: 137–43; Caulfield 1983: 195–215) and subsequent Bayesian modelling (Cooney et al. 2011: 562–669) should be treated with caution. And although settlement evidence increases in the Late Neolithic (McLaughlin et  al. 2016: 117–53), it is not until the Chalcolithic that extensive field systems become more common. Well-preserved field walls, dated using differential bedrock weathering to the Chalcolithic and Early Bronze Age, have been recorded in the Burren (Jones et  al. 2010: 35–58) and elsewhere (see O’Connell and Molloy 2001: 122). Bronze Age cultivation ridges have been found at Carrownaglog (O’Connell 1986: 117–76) and Belderg Beg (Verrill and Tipping 2010b: 1214–25). However, archaeobotanical evidence does show that agriculture rapidly became important from c. 3750 cal BC and also suggests that it is plausible that a fixed plot system, rather than shifting cultivation, was in use in the Early Neolithic in Ireland (Whitehouse et  al. 2014: 181–205; McClatchie et  al. 2014: 206–15, 2016: 302–18). Recent developer-funded excavation has also greatly increased the number of settlement sites which are reliably dated to the early Neolithic (Smyth  2011: 1–31; Smyth 2014), adding weight to this picture of a settled landscape. In contrast, the middle Neolithic sees more ephemeral evidence for houses (ibid.) and agriculture (Whitehouse et  al. 2014: 181–205; McClatchie et al. 2016: 302–18) and a concomitant increase

C. Roughley et al.

in monument construction (McLaughlin et  al. 2016: 117–53). It is thus not a simple diachronic trajectory of increasing agriculture and enclosure but rather a more complex one with phases of reduced evidence for farming and settlement as well. The most famous prehistoric field systems from the west of Ireland are located in landscapes which are now marginal and were only occupied for relatively short periods of time. For example, the fields on Roughan Hill were probably built and used within a single millennium (Jones 2015: 77–100). The Loughcrew landscape was not covered by blanket bog (unlike the field systems in Co. Mayo) nor significantly degraded by over-cultivation so it has had the potential to be occupied over a long timescale. As the current Loughcrew landscape is dominated by pasture, a wealth of boundaries survive as upstanding features. There is thus more chance to investigate long-term change, but concomitantly it is more difficult to unravel the complex field boundaries into distinct use-phases and the dating of individual boundary features remains problematic.

9.2

The Loughcrew Landscape

Loughcrew is situated at the northern edge of the Irish midlands (Fig.  9.1). Newgrange and the Brú na Bóinne World Heritage Site lie about 50  km to the east. The Slieve na Calliagh hills rise to 250  m and sit across the watershed between the Inny and Boyne basins: the Upper Inny drains into the River Shannon and the Blackwater river is a major tributary of the Boyne (Fig. 9.1b).1 The soils are mostly deep, well-drained and mineral derived, with similar but shallower soil on the hills. There are smaller areas of poorly drained soils, peaty soils, and also alluvial and lacustrine soils. In the west of the area and to the south of the hills, the soils are largely derived from non-acidic parent materials (although there are also large erratics), whilst the soils to the east and the hills themselves are from acidic parent materials. The subsoil is predominantly Lower Palaeozoic sandstone and shale till in the north and east, and Carboniferous limestone till to the west, and south of the hills.2 The area is best known for the passage tomb complex which sits on top of the Slieve na Calliagh hills. These tombs were first recorded and investigated in the 1860s by Conwell (1864, 1866), and in 1943 one of the passage tombs was systematically excavated (Raftery 2009). It is plausible that the passage tombs are of similar date to other nearby passage tombs where recent research has provided date ranges of c. 1  Information from the Environmental Protection Agency’s ENVision project. http://gis.epa.ie/GetData/Download Accessed 8th August 2012. 2  Soils and subsoils data generated by Teagasc with the co-operation of the Forest Service, EPA and GSI. Project completed May 2006. http:// gis.epa.ie/GetData/Download Accessed 8th August 2012.

9  The Changing Fieldscapes of Loughcrew: New Insights from Airborne Lidar

Fig. 9.1  Location of study area and main prehistoric sites and extent of lidar survey (for transcription see Fig. 9.2)

117

118

C. Roughley et al.

Fig. 9.2  Transcription of lidar data (for location see Fig. 9.1c)

Linear features Enclosures Possible cursus Passage tombs Other mounds Townland boundaries 0

3350–3000  cal  BC (O’Sullivan 2005; Lynch et  al. 2014: 13–82). Although the passage tombs are the best known components of the prehistoric landscape at Loughcrew, they are only one constituent of a much more complex landscape (Fig.  9.1c; Shee Twohig 2001: 113–24). This includes significant rock art, standing stones, a stone circle, Early Bronze Age cists, cairns and barrows and an enclosure which may be a henge (Moore 1987; Shee Twohig 2001: 113–24; Shee Twohig et  al. 2010: 1–28).3 In addition, Newman (1995: 19–21) proposed that the linear earthwork found at Ballinvally townland might be a cursus. Prehistoric monuments are particularly concentrated in the area immediately to the north of the hills, in Ballinvally and Drumsawry townlands. The standing stones and linear earthworks may mark an ordering of the landscape along a north-south axis (Newman 1995: 19–21), and possibly create a formal approach to the passage tombs (Cooney 2000; Shee Twohig 2001: 113–24; Shell 2005: 1–3). This is not just a “ritual landscape” however, as Shee Twohig (2001: 1–28) noted. In addition to the monuments, there are also several fulachtaí fia – mounds of burnt stone. The interpretation of these mounds is still disputed (Brindley et  al. 1989–90: 25–33; Hawkes 2015: 44–77): domestic cooking and/or feasting have been frequently proposed, but alternatives include washing, dying, leather working, and saunas. Whatever their function, they have been considered “highly important indicators of settlement” of the Later Bronze Age (Brindley et  al. 1989–90: 32), though more recent dates also indicate sites of Neolithic and Early Bronze This is not a henge in the sense of a circular bank with internal ditch, rather it consists of a subcircular bank with upright stones set in it – for a further discussion of the forms of Irish henges see Condit and Simpson 2010.

3 

0.5

1 Km

Age date (McLaughlin et al. 2016: 117–53). There are also a large number of enclosures, many of which are likely to be early medieval ringforts (Moore 1987).

9.3

The Loughcrew Landscape Project

The use of lidar for revealing subtle topographic features is now well known. The Loughcrew dataset was one of the earliest surveys commissioned for archaeological research (Shell and Roughley 2004: 20–23). It covers a relatively small area (30 km2) with points at a ground sample distance of less than a metre and was focussed on the area with most known sites (see Fig.  9.1c). Most of the landscape is currently pasture which makes it ideal for lidar survey. In all, nearly 200  km of archaeological features have been transcribed (Fig.  9.2; Shell and Roughley 2004: 20–23) and areas with preserved cultivation ridges total nearly 300  ha (over 725 acres). The survey revealed several new enclosures and barrows, further strengthening the sense of an intensively used prehistoric landscape (Shell and Roughley 2004: 20–23). It also helped to clarify the form of the possible cursus monument in Ballinvally townland, which is difficult to discern on the ground (Newman 1995: 19–21; Shee Twohig 2001: 113–24). The interpretation of features was supported by small-scale geophysics and targeted ground-truthing. However, there are no direct dates as yet for any of the features revealed in the lidar survey. Subsequent developments in visualisation techniques (e.g. Hesse 2010: 67–72) have also made identification and transcription of features easier, but have not significantly changed the results. In addition to the lidar survey, the project also reconsidered aerial photographs taken by JK St Joseph in the 1960s and commissioned a new vertical photographic survey for a 10 by 12 km area. The vertical photographs were orthorecti-

9  The Changing Fieldscapes of Loughcrew: New Insights from Airborne Lidar

fied and merged into an orthophotomap to assist in interpretation of the lidar data and a photogrammetric Digital Surface Model (DSM) was created from them for the larger area they cover. Digital copies of historic mapping have then been georeferenced as far as possible to the lidar and orthophotomap. Ordnance Survey mapping for the study area exists from 1836. Mapping was much less consistent prior to the Ordnance Survey: Larkin’s 1812 map does provide useful detail on roads and settlements but does not include land divisions (Horner 2007). There are also estate maps for the Loughcrew estate. The earliest of these maps were drawn up following the division of the estate in 1776 (Connell 2004) and cover approximately half of the core study area. They provide detailed surveys for Loughcrew Demesne and Oldcastle (Hanly 1965: 249–252), even showing the cultivation ridges in use for fields adjacent to Oldcastle itself but unfortunately contain much less detail elsewhere, with only the major land holdings recorded. Finally, the 1655–1658 Down Survey maps4 provide the earliest mapping with useful detail for the study area.

9.4

Linear Features and Boundaries

Approximately 150 km of the features transcribed from the lidar data are linear in form and would appear to be field boundaries (Fig. 9.2). There is great morphological diversity, and in many places there are overlapping boundaries which imply successive field systems. Although some field boundaries were visible in St Joseph’s aerial photographs, the complexity and extent of the linear features were unexpected and thus posed a significant challenge for both the initial transcription phase and also for subsequent interpretation. Although not as heavily ploughed in recent centuries as Brú na Bóinne (Stout 2002), the study area does sit in a landscape which has been heavily used at a range of points prior to and during the nineteenth century. This generates two very different problems for further unravelling the field boundaries. Firstly, much of the landscape was cultivated by cottiers and tenant farmers in the eighteenth and nineteenth centuries, which means that many “relict” boundaries could be relatively recent and yet unmapped. The Ordnance Survey’s recording of field boundaries improved over time, with the Meath sheets considered to be “rather more perfect” than other midland counties (Andrews 1975: 105), but sod banks and lines of mearing stones used to mark ownership boundaries were never included (Andrews 1975: 104). Although these were themselves very slight features, such divisions could still lead to discernible boundaries in the lidar data. The second problem is that boundaries can persist, with ancient boundaries incorporated into subsequent land allotAvailable from http://downsurvey.tcd.ie/

4 

119

ment patterns and therefore becoming relatively hidden. Fowler (2002: 133) poses the question “what is the date of a field – its original creation, its use, its reuse, its alternative use, its abandonment, its absorption into a different type of field system or land-use?”. Going further, Chadwick (2008a: 6–7) calls for us to abandon both the term “relict” and with it the concept of time slices. Field boundaries which are also administrative boundaries are particularly likely to persist as features in the landscape, though this should not be over-­ emphasised (see Williamson 1986: 241–48). In Ireland, the smallest administrative unit is the townland (Duffy 2000: 187–218); in the study area, the average townland size is about 150  ha. Many townlands have boundaries which are clearly recognisable in the Down Survey maps, albeit in a simplified form. The persistence of these old townland boundaries has resulted today in several fields with three straight sides and a fourth one which is very irregular. Although it is not possible to date the construction of these boundaries (see also McErlean 1983: 315–40), features which underlie these irregular townland boundaries are very unlikely to be recent. Using map regression (Rippon 2004; Williamson 1987: 419–31), it has been possible to date the creation of some of the boundaries. Over 15 km of features could be determined to have been created since 1778. However, whilst comparison with extant mapping has been helpful, it still leaves 120 km of boundaries to be explained. As discussed above, it is not safe to assume that any boundary is modern just because it appears on a later map or, indeed, continues in use today. The lidar survey has also revealed extensive areas of cultivation ridges, but dating these is no less difficult (McCormick et al. 2011: 26). Bell and Watson (1984; 2009: 115–16) researched historic cultivation methods and have undertaken experimental work investigating the variety of techniques by which cultivation ridges can be made. Their work highlights the difficulties inherent in implying method or date from the forms of the resulting ridges. Although prehistoric examples of cultivation ridges do exist elsewhere in Ireland, it is unlikely that any of the cultivations ridges found in the study area are prehistoric in date. There is also no clear evidence for medieval cultivation ridges. The discovery of reversed-S shaped “aratral curves” preserved in the strip fields of Fethard (O’Keeffe 1999: 24), and medieval ridge-­ and-­furrow which is not curved in the Porchfield of Trim (Kelly 2005: 23–43) show that a medieval origin can no longer automatically be excluded in Ireland. However, by comparing the field evidence with historic mapping, it is clear that some of the cultivation ridges were created in the eighteenth and nineteenth centuries and it is likely that this is the case for the majority of ridges. Thus, although there are several overlapping field systems predating the cultivation

120

ridges, the ridges provide only a very recent terminus ante quem for the field systems. In addition to studying historic mapping, targeted fieldwork has been used help to unravel relationships between features in some instances. Although field visits are clearly an important part of the study of field boundaries (see lament in Chadwick 2008b: 207), in many cases it was unfortunately not possible to ascertain relationships from visual inspection. Some intersections are too denuded (cf. Herity 1988: 69); in other cases the overlying cultivation ridges are relatively substantial when compared with the slight nature of the upstanding linear features of interest and obscure crucial intersections which would otherwise provide a relative chronology. Furthermore, the scale of the task also means that it would be a significant undertaking to investigate all intersections.

9.4.1 Categorising Boundaries Given the complex and overlapping nature of the features revealed, it is necessary to apply some sort of categorisation process in order to disentangle the features. One approach might be an initial classification of fields based on morphology. Early work on field systems suggested a contrast between hand-dug irregular fields and ploughed rectilinear ones (Buchanan 1973: 580–618; Bradley 1978: 267), and characteristic examples of different field boundaries were proposed (Aalen 1983: 357–78). It is now accepted that the relationship between tillage technology and field shape is not straightforward (see Lewis 2012: 39–43). Indeed, Fowler (2002: 133) has commented “there is not much correlation between field-shape and time, process or product”. Whilst there are some forms of fields which may be recognisable with a degree of confidence, (e.g. ladder farms; Aalen et al. 2011: 206) these are limited to the relatively recent past. Although there is a very high density of features, relatively few complete “fields” could be identified with certainty. There are many “boundaries” but few fully defined enclosed spaces. This relates in part to differential preservation. In some areas it is clear that a field or even a field system stops at a major recent land holding boundary. There are also a significant number of short straight boundaries which appear to be one or two sides of a rectilinear field which has lost its other boundaries. But it is also possible that there are linear features which are not “field boundaries” at all. In order to try to start to pull out patterns from the data, the approach used was to categorise features rather than fields. Clearly this is not an approach without problems: a single field may have boundaries of very different character. Furthermore, applying morphological criteria becomes harder as the boundaries move across changing terrain, soils, and where they are differentially affected by subsequent land

C. Roughley et al.

uses. The basis for the initial categorisation was shape, with features being described as irregular, curvilinear or straight. Where possible, the form of a feature (bank, ditch, lynchet etc) was included as a further characteristic. The relative alignment of straight features to current and historic field boundaries was also noted. But the process has been an iterative one, particularly in very complex areas, with some boundaries changing category or being subdivided as the interpretation has been refined. Other boundaries have so far remained uncategorised, for example where different visualisations of the lidar data give diverse impressions as to the form of the feature. In spite of the problems, the approach has allowed some coherent sets of features to be identified.

9.4.2 Irregular Banks Over 16  km of bank have an irregular or winding form (Fig.  9.3). These have a range of forms: some enclosing small irregular fields whilst others appear to enclose larger spaces. There are also a number of banks which do not directly enclose space. Some of the banks can be seen to be relatively recent, as they postdate historic features (e.g. the Anglo-Norman ringwork). However, three sets of boundaries with possibly prehistoric origins have been identified: irregular fields in Stonefield (Fig. 9.4), smaller irregular fields in Drumsawry (Figs. 9.5, 9.6, and 9.7); and long winding banks in western Ballinvally (Fig. 9.7). A relatively distinct set of irregular banks appears to form an irregular field system consisting of fields between 3 and 8 ha in area. The clearest examples are in Stonefield townland (Fig.  9.4). There, a straight feature abuts one of the irregular banks, indicating that the irregular bank predates the straight one. There are also cairns situated on both the irregular banks and the straight bank, which might initially suggest a prehistoric date. However, these are clearance cairns and may be recent; many stones are indicated in this area on the nineteenth century field sheets for the first edition Geological Survey mapping which suggests that these cairns may be the result of relatively recent agricultural improvements. To the east of these cairns, an irregular bank appears to avoid a barrow (Fig. 9.4). This suggests that the bank here postdates the barrow, but the barrow itself is alas undated. The remains of an irregular field system in Drumsawry was first documented by St Joseph in 1963, when it was still well preserved (Figs. 9.5 and 9.6; Aalen 1970: 211; Moore 1987: 122).5 It consists of stone banks enclosing areas of 0.5–1.5 ha in a wide variety of shapes (Fig. 9.7). There is a large circular enclosure with a funnel-shaped entrance feature, described by Moore (1987: 125) as an ‘avenue’ [quotaSee also AHN038–039, AHN049–051, APE048–051, AJZ081, previews available from http://www.cambridgeairphotos.com/ 5 

9  The Changing Fieldscapes of Loughcrew: New Insights from Airborne Lidar

121

Fig. 9.3  Irregular banks in central part of study area; extent of Figs. 9.4 and 9.7

Irregular banks Cairns

Alignment conforms to 1836 field system Other rectilinear boundary Other features

Fig. 9.4  Irregular banks in Stonefield townland (a) Transcription of lidar (excluding modern drainage features)

(b) Lidar shaded from 122.5° at 20° azimuth. Note the complexity of features visible, only those which appear in more than one visualisation have been transcribed

122

C. Roughley et al.

Cairn

Enclosure

Townland boundary

Large enclosure

Fig. 9.5 CUCAP AHN52. Photograph taken looking north over Drumsawry field system with large enclosure in the foreground. The partially robbed cairn and enclosure are visible in the top right. The irregular stone wall running from foreground to the road follows the 1778 townland boundary. More recent boundaries cross cut the field system. (By permission of the Cambridge University Collection of Aerial Photography (c) Copyright reserved)

Large enclosure

Long straight bank

Cairn Enclosure

Fig. 9.6  CUCAP APE051. Photograph taken looking south-west over the Drumsawry field system. Enclosure and partially robbed cairn in foreground, with long straight bank running under the straight wall to the left. Large enclosure with entrance ‘avenue’ visible in the left of photograph, adjacent to irregular stone wall. (By permission of the Cambridge University Collection of Aerial Photography (c) Copyright reserved)

tion marks original] sitting within it and a small enclosure

which has been assumed to be an early medieval cashel (Aalen 1970: 211–12; Moore 1987: 71).6 The banks are also changed by an ongoing process of field clearance, through which boundaries have been created, modified and remade. This continues to the present, making it extremely difficult to date or even sequence the stony banks. The narrow cultivation ridges appear to be a recent phase, as they are even found within the long entrance of the large enclosure – surely not part of its original function. Towards the northern edge of this field system are several curvilinear banks which run east-west along the slope. They appear to be somewhat different in form from the rest of the field system. They run approximately parallel to each other, branching and re-joining to create spaces which are long and narrow. Some of these banks appear to run under the stone wall which divides Drumsawsry from Ballinvally townland (Figs.  9.5 and 9.6). This is most obvious on the northern slope in the vicinity of an enclosure and a cairn which had the remains of a drystone-built wall in the centre (Shee Twohig 2001: 113–124). Here it is clear that at least some of the banks predate this townland boundary (which remains as mapped in 1778). The relationships on the Ballinvally (east) side are a little hard to discern as the banks are much more denuded on this side of the townland boundary and there is also a steep slope and geological boundary. However, it would appear from both aerial photographs and lidar survey that one of the curvilinear banks extends up to the cairn, and it is possible that the cairn is sat on top of this bank. If this is the case, it would indicate a much earlier date for the at least this bank than the early medieval date previously proposed for the field system (e.g. Aalen 1970: 211–12). Further south in Ballinvally townland is an irregular bank which has an intersection with the earthworks that Newman (1995) suggested might be a cursus monument (Fig.  9.7). This bank was therefore identified early on in the project as being of particular interest. Its form gives the impression that the bank was constructed in small sections – whilst it has an overall NE-SW direction there are sections with wiggles and dog-legs which appear to relate to neither the topography nor other upstanding features. Some of this irregularity might be the result of subsequent use of the area, in particular the creation of cultivation ridges and later reorganisation of the field boundaries. Nevertheless, the general winding character appears intrinsic to the feature. It can be seen from the lidar data that the irregular bank is cut by the possible cursus (Fig. 9.7); this was confirmed in the field through visual inspection and geophysical survey. This monumental feature therefore provides both a relative terminus ante quem and indicates a clear local reorientation and reorganisation of the landscape. If the Ballinvally earth-

Stone ringfort.

6 

9  The Changing Fieldscapes of Loughcrew: New Insights from Airborne Lidar

123

Fig. 9.7  Eastern Drumsawry and western Ballinvally (a) Selected features transcribed from lidar and monuments

(b) Local relief model created from lidar data (pale locations are higher than their surroundings and dark locations are lower than their surroundings, see Hesse 2010: 67–72 for methodology)

works are a cursus, then it would imply that the irregular bank is Neolithic. Some doubt does remain surrounding the form and date of the “cursus” which means that the date of the meandering bank remains in question. Careful reconsideration by Newman (2007: 415–38) of the possible cursus at Tara (see Condit 1995: 16–18; Newman 1997: 150–53) has suggested that the latter site is unlikely to be a Neolithic cursus. The Ballinvally cursus is not a classic cursus monument, as it has a central bank, narrowly spaced ditches and an angle mid-­ way along its length (Fig. 9.7). But there is great diversity in cursus monuments. Recent excavations at Newgrange Farm (Leigh et al. 2018) have revealed a complex hybrid structure with double ditches and an internal bank as well as pits/postholes. Other known cursus monuments with central banks include Cleaven Dyke in Tayside (Barclay and Maxwell 1999: 98–106) and Scorton in Yorkshire (Topping 1982: 7–21). There are also several sites in East Anglia which are narrow with angles (e.g. Fornham all Saints, Maxey and possibly Stonea; Last 1999: 7–115). The Ballinvally earthwork

is certainly not recent as two sets of cultivation ridges overlie it, though these are undated. It is also unlikely that this earthwork is post-medieval in date as the feature runs under a stone wall that is a townland boundary and appears as a major landholding boundary in all records from the seventeenth century Down Survey onwards. The banks incorporate standing stones and although caution is essential it is still plausible that it is a Neolithic or Early Bronze Age feature. The irregular bank may be part of a wider set of irregular boundaries (Fig. 9.7), though it is hard to see relationships to other similar banks because this area appears to have had a particularly complex history with many overlapping banks and cultivation ridges. A bank of similar general form and appearance is present on approximately the same alignment 110–120 m to the north-west. To the north-east there are two further banks which are on a similar alignment and approximately 150 m apart. One of these appears to incorporate the bank of a possible henge monument (see Shee Twohig 2001: 113–124) and would thus postdate the henge. This bank

124

appears to be incorporated into a trackway which turns an abrupt corner at this point and continues in a similar direction. These irregular banks also run approximately parallel to a small stream 120–150  m to the south-west which flows from south-west to north-east.7 It is therefore possible that there is a field system which is later than the henge and earlier than linear earthwork. However, there are no plausible cross-banks associated with these boundaries. This may be because they were originally of a different construction (e.g. hedges rather than walls) or they may have been obscured by the overlying cultivation ridges. Alternatively, this may relate to a real absence. The irregular banks are more extensive, diverse and complex than has been previously considered, particularly in the area to the north of the Slieve na Calliagh hills. In addition, there are further irregular field systems in other parts of the survey area which cannot yet be related to other features. The field boundaries are more rarely part of overlapping sequences in areas away from the core area of prehistoric monuments. This is in part because they are more fragmentary or denuded, and is likely to relate to differential preservation.

C. Roughley et al.

They are very straight, deviating little even when crossing undulating terrain. Some of the boundaries are visible on photographs taken by St Joseph in 1963 (e.g. Figs. 9.5 and 9.6) and though they are slight can be located on the ground. The form of the straight banks initially suggested a relatively recent origin but further consideration suggests that this is not the case. Most of the straight banks are overlain with wide (>6  m) cultivation ridges which are likely to be late eighteenth or early nineteenth century, ruling out a very recent date. The proximity of a major straight boundary to the irregular townland boundary makes little sense practically, and thus suggests that these boundaries predate it, though such assumptions should always be verified. However, to the north-east of the nearby enclosure, a N-S boundary very clearly overlaps an irregular bank (Fig. 9.7). As this would suggest that these straight banks postdate at least some of the irregular banks in the area, they would appear to be part of a wider reorganisation of the landscape. In the western part of the field system is an enclosure and a large cairn (the latter unfortunately partially destroyed in 1961: Moore 1987: 21; Shee Twohig 2001: 113–24; Figs. 9.5 and 9.6). A short bank extends at right angles from the W-E bank to the enclosure. Just to the east of this junction, there also appears to be an entrance through the W-E bank. Although a little difficult to see now, this gap is confirmed by 9.4.3 Rectilinear Boundaries St Joseph’s 1960s photography (Fig. 9.5). There are two N-S Around half of the boundaries categorised are straight or boundaries which about the W-E boundary. A further short nearly straight. Approximately half of these straight features bank also connects one of the N-S boundaries to the encloconform to the 1836 field system (Fig. 9.8). There are two sure again, suggesting a very definite relationship between possible interpretations for these  – they are either subdivi- the linear features and the enclosure. However, the linear feasions of the 1836 fields which have since gone out of use, or ture within the enclosure is harder to define and does not the larger fields were defined by amalgamating these pre-­ show clearly either from visual inspection or in the geophysexisting smaller units. It is not possible to be certain which ical survey. Although many of the linear banks are likely to be relaoccurred, nor at what point this might have taken place, but it is almost certain that these are among the most recent fea- tively recent, the low straight banks in northern Ballinvally tures recorded. Those straight boundaries which do not con- seem to form a distinct set of features and are potentially form to the 1836 boundaries appear to belong not in just one earlier. They are located in an area with a concentration of previous system but several. Many of these are very frag- prehistoric monuments and which also has complex irregular mentary, with few full fields being defined. In places it is also banks. This part of the landscape appears to have a particuimpossible to be sure which boundaries may have belonged larly complex history. together as they are separated by significant gaps. In the northern part of Ballinvally townland, towards the edge of the raised area which extends north from the base of 9.4.4 Boundary Changes the Slieve na Calliagh hills, there are relatively long straight features (up to 280 m in length) which are 6–10 m wide but Although it is not possible to assign dates to any of the above only a few centimetres high (Fig. 9.7). These features seem field systems, the significant alterations which must have to be unrelated to any mapped boundaries. They appear at taken place themselves pose interesting questions. Because first sight to be largely coherent and planned, though that they are upstanding features, rather than ditches which could assumption may be questionable (see Johnston 2005: 1–21). have silted up over time, the question as to why the landscape was altered becomes particularly prominent. Although now very slight, it is likely that such boundaries would have 7  Although now constrained by modern farming, some sections of this stream appear from their unstraightened form to be approximately in been more visible when new field systems were created. their natural location.

9  The Changing Fieldscapes of Loughcrew: New Insights from Airborne Lidar

125

Fig. 9.8 Rectilinear boundaries transcribed from the lidar data and information from Ordnance Survey 1836 maps

Fig. 5

Features in lidar

Possible cursus

Roads_1836

Conforms to 1836 field system

Trackway

Rectilinear but do not conform

Other features

Fields_1836

Low linear banks discussed in text

An early transformation appears to be from a landscape demarcated by irregular banks running NE-SW to one with a monumental route running N-S.  Based on the distribution map, it is still reasonable to consider that the standing stones might mark a prehistoric routeway running north-south across the landscape as Shee Twohig (2001: 1–28) proposed. This routeway potentially became increasingly structured through the construction of the cursus which incorporated two of the standing stones. If the irregular banks formed a coherent field system predating the “cursus”, then this axis of movement was created across a landscape which had previously been divided along an approximately north-east-­ south-west axis. Instead of the “cursus” and standing stones formalising existing pathways, this route may instead have been created to cross-cut a previously defined structure. However, several elements of this scenario are very speculative at this stage, not least the uncertainty regarding the point at which the line of standing stones running north was erected. Subsequent transformation from one fieldscape to another would have been far from straightforward. While prehistorians have considered the question “why enclose” in considerable detail (cf. Cooney 1991: 123–139), there is less literature looking at why fieldscapes might change. Fowler’s (2002: 137ff) consideration of early medieval field boundaries helpfully distinguishes several different possible situations. Although his concern is with the relationship between pre-­ existing prehistoric field systems and subsequent early medieval field systems, the distinctions he makes are valuable

0

200 m

Streams (1836)

over a range of timespans and situations. Fowler (ibid.) notes that adaptation of a pre-existing field system can take place both when that earlier system is still in use or after a period of abandonment. The irregular field system in Drumsawry townland may well have been adapted and reconfigured several times. The creation of a completely new field system on top of an existing field system is a very different scenario, particularly when the preceding fields are still in use rather than after a discontinuity. Although the rectilinear boundaries in Ballinvally area appear to relate in some way to the enclosure and cairn in the north-west corner of the townland, they cross-cut the earlier field boundaries and appear to impose a new structure onto the landscape. Changes in cultivation technology have been proposed as drivers of change. For example, a shift from hand-dug irregular fields to rectilinear ploughed fields was once widely accepted (cf. Bradley 1978: 265–80). An increase or reduction in population might make it more expedient to alter the fieldscape, as might changes in practice, perhaps as a result of soil degradation or climate change. However, it is important to remember that tenure and land holding practices might also lead to a reconsideration of appropriate field shapes and sizes (Fowler 2002). Evans (quoted in Chadwick 2008a: 14–15) has highlighted the diversity of reasons for forming field boundaries and the diverse roles they can play in society: fields are not simply about subsistence, and cross-­ cut any attempt to divide landscape features into categories of functional versus symbolic and prosaic versus ritual (Chadwick 2008b: 210; Lewis 2008: 239). The concept of a

126

C. Roughley et al.

Fig. 9.9  Impact of nineteenth and twentieth century farming on the preservation of upstanding features Boolies

Drumsawry Mountain

Features from lidar

Small landholdings in Boolies Drumsawry_road

Field boundary Road

9.5

 he Effect of the more Recent Land T Use

As the landscape has been occupied relatively intensively over a significant period of time, it is inevitable that many of the upstanding features have been affected by subsequent land use. The effects of early nineteenth century agricultural change can be seen clearly in two areas: Drumsawry and Boolies (Fig. 9.9). In Drumsawry, a new road was built by the Naper estate sometime between 1812 and 1836 along which a string of farms developed. This area had previously been relatively uncultivated, indeed is denoted “Drumsawry

500 m

Subdivisions from 1778 map

Land Comission

“ritual landscape” has been widely used by prehistorians (Robb 1998). It can help us situate monuments within their context rather than seeing them as isolated sites. However, the term can lead to a privileging of the “ritual” and “monument” focus over “field boundaries”. The interaction between “monuments” and “field boundaries” needs to be more fully explored. At this stage, it is not possible to move beyond speculation without significant further research as there are too many missing elements – not least the need for high resolution dating. Nevertheless, the successive reorganisations of boundaries and monuments in this landscape suggest this will be a fruitful direction for further research.

0

Townland boundaries

Mountain” in the estate rent rolls 1733–40 (National Library of Ireland MS 3031). This development resulted in intensive use of the land and the eradication of earlier features. Approximately half of Boolies townland was used for small land holdings held by tenants of an absentee landlord (Griffith’s valuation).8 Although less is known about this area (it is outside the Loughcrew estate) the land was presumably used extremely intensively and very little remains under these farms in contrast with the rest of the townland which was farmed more extensively. Any interpretation of the prehistoric landscape has to take into consideration the effect of subsequent activities. In the first half of the twentieth century, the work of the Land Commission in the 1920s and 1930s had a very dramatic impact on the landscape in many areas of Ireland (Aalen et al. 2011: 104–05). In the area north of the Slieve na Calliagh hills, a new road was built in 1924 (O’Keeffe 2010) and new farms were created on a regular plan (Fig. 9.9). The new landholding system cut across earlier field systems. Field boundaries were constructed with little regard to pre-­ existing features, for example bisecting an early medieval rath (earthen ringfort) in Ballinvally townland. However, although some monuments have been very significantly damaged, there are as many features present in the lidar data Available from http://www.askaboutireland.ie/griffith-valuation/

8 

9  The Changing Fieldscapes of Loughcrew: New Insights from Airborne Lidar

in areas affected by the Land Commission as elsewhere (Fig. 9.9). Very significant damage has occurred during the second half of the twentieth century to the field system in Drumsawry. In the 1960s, this field system covered approximately 25 ha but by this point it was already damaged. As discussed above, there are continuities through to the eastern side of the Townland boundary with Ballinvally where the banks are much denuded but still visible. It is not possible to ascertain the original extent due to the limited coverage of the 1960s photography. By 1995, many of the low stone banks which were upstanding features in 1963 had been cleared.9 More recently, a significant number of new houses have been built, especially between 1995 and 2008. Many were constructed as single dwellings on plots adjacent to existing houses. However, the potential damage to archaeological features is more considerable than might be expected given the size of the footprint  – many are bungalows with large landscaped driveways and garden areas. In itself, the sheer quantity of field boundaries throughout the Loughcrew landscape as now recorded poses significant problems for their effective management.

9.6

Conclusions

This paper has demonstrated that there is hitherto unexpected potential for research into field systems in the Loughcrew landscape. The degree of preservation and the overlapping nature of the field systems provide the opportunity for temporal changes to be considered. The number and complexity of features revealed has proved challenging to record and interpret. However, through applying an albeit imperfect classification based on the form of unmapped field boundaries, it has been possible to highlight some of the diversity and extent of field systems which are preserved as upstanding features in this landscape. The relationships between field boundaries and monuments suggest that some of these field systems may be prehistoric, possibly Neolithic. Given current debate surrounding the dating of Céide Fields (Whitefield 2017: 1–23), determining the date of the earliest boundaries through excavation and high-precision dating is perhaps even more crucial now than when the survey was undertaken. The presence of both field boundaries and monuments in the same area provides the opportunity to further understand the relationships between field and monuments in this important landscape, and questions the validity of the term “ritual landscape”. Although currently poorly understood in chronological terms, the Loughcrew field boundaries are clearly a key part Ordnance Survey Ireland orthophotography available from https:// www.osi.ie/services/mapgenie/ 9 

127

of the landscape and need to be considered carefully in the future management of the landscape. Acknowledgements The lidar and vertical photographic surveys, ground-truthing fieldwork and geophysical investigation were funded by Irish Heritage Council Archaeology Scheme grants (12334 in 2004 and 14052 in 2005). Fieldwork was greatly helped by Ciaran O’Reilly and the support of current users of the landscape. The encouragement of the conference organisers and the helpful comments of the referees have also significantly improved this paper.

References Aalen, F. H. A. (1970). The origin of enclosures in eastern Ireland. In N. Stephens & R. E. Glasscock (Eds.), Irish geographical studies in honour of Estyn Evans (pp. 209–223). Belfast: Queen’s University Belfast. Aalen, F. H. A. (1983). Perspectives on the Irish landscape in prehistory and history. In T.  Reeves-Smyth & F.  Hamond (Eds.), Landscape archaeology in Ireland (British archaeological reports, British series, 116) (pp. 357–378). Oxford: Oxbow. Aalen, F. H. A., Whelan, K., & Stout, M. (Eds.). (2011). Atlas of the Irish rural landscape (2nd ed.). Cork: Cork University Press. Andrews, J.  H. (1975). A paper landscape: The Ordnance Survey in nineteenth century Ireland. Dublin: Four Courts Press. Barclay, G.  J., & Maxwell, G.  S. (1999). The Cleaven Dyke project: A summary account. In A. Barclay & J. Harding (Eds.), Pathways and ceremonies: The cursus monuments of Britain and Ireland (pp. 98–106). Oxford: Oxbow Books. Bell, J. (1984). A contribution to the study of cultivation ridges in Ireland. The Journal of the Royal Society of Antiquaries of Ireland, 11, 80–97. Bell, J., & Watson, M. (2009). A history of Irish farming, 1750–1950. Dublin: Four Courts Press. Bradley, R. (1978). Prehistoric field systems in Britain and north-west Europe: A review of some recent work. World Archaeology, 9(3), 265–280. Bradley, R. (2002). The past in prehistoric societies. London: Routledge. Bradley, R. (2007). The prehistory of Britain and Ireland (Cambridge world archaeology). Cambridge: Cambridge University Press. Brindley, A.  L., Lanting, J.  N., & Mook, W.  G. (1989). Radiocarbon dates from Irish Fulchat Fiadh and other burnt mounds. The Journal of Irish Archaeology, 5, 25–33. Buchanan, R. H. (1973). Field systems in Ireland. In A. R. H. Baker & R. A. Butlin (Eds.), Studies of field systems in the British Isles (pp. 580–618). Cambridge: Cambridge University Press. Caulfield, S. (1978). Neolithic fields: The Irish evidence. In H.  C. Bowen & P.  J. Fowler (Eds.), Early land allotment in the British Isles (British archaeological reports 48) (pp.  137–143). Oxford: Archaeopress. Caulfield, S. (1983). The Neolithic settlement of North Connaught. In T. Reeves-Smyth & F. Hamond (Eds.), Landscape archaeology in Ireland (British archaeological reports, British series 116) (pp. 195– 215). Oxford: Oxbow. Chadwick, A. (2008a). Introduction. In A.  Chadwick (Ed.), Recent approaches to the archaeology of land allotment (British archaeological reports international series 1875) (pp.  1–24). Oxford: Archaeopress. Chadwick, A. (2008b). Fields of discourse? Towards more self-­critical, theoretical and interpretative approaches to the archaeology of field systems and land allotment. In A.  Chadwick (Ed.), Recent approaches to the archaeology of land allotment (British archaeo-

128 logical reports international series 1875) (pp.  205–237). Oxford: Archaeopress. Condit, T. (1995). Avenues for research. Archaeology Ireland, 9(4), 16–18. Condit, T., & Simpson, D. (2010). Irish hengiform enclosures and related monuments: A review. In A. Gibson & D. Simpson (Eds.), Prehistoric ritual and religion. Sutton: Thrupp. Connell, P. (2004). The land and people of County Meath, 1750–1850. Dublin: Four Courts Press. Conwell, E.  A. (1864). On the ancient remains hitherto undescribed in the County of Meath. Proceedings of the Royal Irish Academy, (series 1) 9, 42–50. Conwell, E.  A. (1866). Examination of the ancient sepulchral cairns on the Loughcrew Hills, Co. Meath. Proceedings of the Royal Irish Academy, (ser. 1) 9, 355–379. Cooney, G. (2000). Landscapes of Neolithic Ireland. London: Routledge. Cooney, G. (1997). Images of settlement and the landscape in the Neolithic. In P. Topping (Ed.), Neolithic landscapes (Neolithic studies group seminar papers, volume: 2) (pp. 23–31). Oxford: Oxbow books. Cooney, G. (1991). Irish Neolithic landscapes and land use systems: The implications of field systems. Rural History, 2(2), 123–139. Cooney, G., Bayliss, A., Healy, F., Whittle, A., Danaher, E., Cagney, L., Mallory, J., Smyth, J., Kador, T., & O’Sullivan, M. (2011). Ireland. In A. Whittle, F. Healy, & A. Bayliss (Eds.), Gathering time: Dating the Early Neolithic enclosures of southern Britain and Ireland (pp. 562–669). Oxford: Oxbow Books. Duffy, P. J. (2000). Heritage and history: Exploring landscape and place in County Meath. Ríocht na Mídhe, 11, 187–218. Fowler, P. (2002). Farming in the first millennium AD: British agriculture between Julius Caesar and William the Conqueror. Cambridge: Cambridge University Press. Hanly, J. (1965). Oldcastle and Loughcrew 1778. Ríocht na Mídhe, 3, 249–252. Hawkes, A. (2015). Fulachtaí fia and Bronze Age cooking in Ireland: Reappraising the evidence. Proceedings of the Royal Irish Academy, 115C, 47–77. Herity, M. (1971). Prehistoric fields in Ireland. Irish University Review, 1(2), 258–265. Herity, M. (1988). A survey of the Royal Site of Cruachain in Connacht IV. Ancient field systems at Rathcroghan and Carnfree. The Journal of the Royal Society of Antiquaries of Ireland, 118, 67–84. Hesse, R. (2010). LiDAR-derived local relief models – A new tool for archaeological prospection. Archaeological Prospection, 17, 67–72. Horner, A. (2007). Mapping Meath in the early nineteenth century: With an atlas of William Larkin’s map of County Meath 1812. Bray: Wordwell. Johnston, R. (2005). Pattern without a plan: Rethinking the Bronze Age coaxial field systems on Dartmoor, South-West England. Oxford Journal of Archaeology, 24(1), 1–21. Jones, C., Carey, O., & Hennigar, C. (2010). Domestic production and the political economy in prehistory: Evidence from the Burren, Co. Clare. Proceedings of the Royal Irish Academy, 111C, 35–58. Jones, C. (2015). Dating ancient field walls in karst landscapes using differential bedrock lowering. Geoarchaeology, 31(2), 77–100. Kelly, D. (2005). The Porchfield of Trim: A medieval “open-field”. Irish Geography, 38(1), 23–43. Last, J. (1999). Out of line: Cursuses and monument typology in eastern England. In A.  Barclay & J.  Harding (Eds.), Pathways and ­ceremonies: the cursus monuments of Britain and Ireland (Neolithic studies group seminar paper 4) (pp. 107–115). Oxford: Oxbow. Leigh, J., Stout, G., & Stout, M. (2018). A pathway to the cosmos at Newgrange farm. Archaeology Ireland, 32(4), 25–29. Lewis, H. (2008). “The pleasant land of counterpane”: Linking site-­ specific archaeological land use to the landscape of prehistoric field

C. Roughley et al. systems. In A. Chadwick (Ed.), Recent approaches to the archaeology of land allotment (British archaeological reports international series 1875) (pp. 239–250). Oxford: Archaeopress. Lewis, H. (2012). Investigating ancient tillage: An experimental and soil micromorphological study (British archaeological reports international series, 1875, 2388). Oxford: Archaeopress. Lynch, A., McCormick, F., Shee Twohig, E., McClatchie, M., Molloy, K., Schulting, R., OCarroll, E., & Sternke, F. (2014). Newgrange revisited: New insights from excavations at the back of the mound in 1984–8. Journal of Irish Archaeology, 23, 13–82. McCormick, F., Kerr, T., McClatchie, M., & O’Sullivan, A.. (2011). The archaeology of livestock and cereal production in early medieval Ireland, AD 400–1100. EMAP Report 5.1 [online]. Available at http://www.emap.ie/documents/EMAP_Report_5_Archaeology_ of_Livestock_and_Cereal_Production_WEB.pdf. Accessed 23 Mar 2017. McErlean, T. (1983). The Irish townland system of landscape organisation. In T. Reeves-Smyth & F. Hamond (Eds.), Landscape archaeology in Ireland (British archaeological reports, British series 116) (pp. 315–340). Oxford: Oxbow. McClatchie, M., Bogaard, A., Colledge, S., Whitehouse, N.  J., Schulting, R. J., Barratt, P., & McLaughlin, T. R. (2014). Neolithic farming in North-Western Europe: Archaeobotanical evidence from Ireland. Journal of Archaeological Science, 51, 206–215. McClatchie, M., Bogaard, A., Colledge, S., Whitehouse, N.  J., Schulting, R. J., Barratt, P., & McLaughlin, T. R. (2016). Farming and foraging in Neolithic Ireland: An archaeobotanical perspective. Antiquity, 90(issue 350), 302–318. McLaughlin, T., Whitehouse, N., Schulting, R. J., McClatchie, M., & Barratt, P. (2016). The changing face of Neolithic and Bronze Age Ireland: A big data approach to the settlement and burial records. Journal of World Prehistory, 29(issue 2), 117–153. Moore, M.  J. (1987). Archaeological inventory of County Meath. Dublin: The Stationery Office. Newman, C. (1995). A cursus at Loughcrew, Co. Meath. Archaeology Ireland, 9(4), 19–21. Newman, C. (2007). Procession and symbolism at Tara: Analysis of Tech Midchúarta (the “banqueting hall”) in the context of the sacral campus. Oxford Journal of Archaeology, 26(4), 415–438. Newman, C. (1997). Tara: An archaeological survey (Discovery programme reports 5 (Monograph 2)). Dublin: Royal Irish Academy. O’Connell, M. (1986). Reconstruction of local landscape development in the post-Atlantic based on palaeoecological investigations at Carrownaglogh prehistoric field system, County Mayo, Ireland. Review of Palaeobotany and Palynology, 49(117), 176. O’Connell, M., & Molloy, K. (2001). Farming and woodland dynamics in Ireland during the Neolithic. Proceedings of the Royal Irish Academy, 101B, 99–128. O’Keeffe, M. (2010). Michael Tobin, Oldcastle, Born 1918. Irish Life and Lore County Meath Collection CD 14[online]. Available at http://hdl.handle.net/10599/8428. Accessed 10 Dec 2013. O’Keeffe, T. (1999). Townscape as text: The topography of social interaction in Fethard, Co. Tipperary, AD 1300–1700. Irish Geography, 32(1), 9–25. O’Sullivan, M. (2005). Duma na nGiall: The mound of the hostages, Tara. Dublin: Wordwell/University College Dublin. Raftery, J. (2009). Newtown, Loughcrew, Oldcastle, County Meath. Cairn H, August 5  – November 10, 1943. Draft report published posthumously. In G.  Cooney (Ed.), Relics of old decency: Archaeological studies in later prehistory. Festschrift for Barry Raftery (pp. 529–540). Dublin: Wordwell. Reeves-Smyth, T., & Hamond, F. (Eds.). (1983). Landscape archaeology in Ireland (British archaeological reports, British series 116). Oxford: Oxbow.

9  The Changing Fieldscapes of Loughcrew: New Insights from Airborne Lidar Rippon, S. (2004). Historic landscape analysis: Deciphering the countryside (Practical handbooks in archaeology 16). York: Council for British Archaeology. Robb, J. G. (1998). The ‘ritual landscape’ concept in archaeology: A heritage construction. Landscape Research, 23(2), 159–174. Shee Twohig, E. (2001). Change and continuity: Post passage tomb ceremonial near Loughcrew, Co. Meath. Révue Archéologique de l’Ouest, Supplement, 9: 113–124. Shee Twohig, E., Roughley, C., Shell, C., O’Reilly, C., Clarke, P., & Swanton, G. (2010). Open air rock art at Loughcrew. Journal of Irish Archaeology, 19, 1–28. Shell, C. (2005). The Loughcrew landscape project. PAST, 51, 1–3. Shell, C. A., & Roughley, C. (2004). Exploring the Loughcrew landscape: A new approach with airborne Lidar. Archaeology Ireland, 18(2), 20–23. Smyth, J. (2011). The house and group identity in the Irish Neolithic. Proceedings of the Royal Irish Academy. Section C: Archaeology, Celtic Studies, History, Linguistics, Literature, Special Issue: Domestic life in Ireland, 111C, 1–31. Smyth, J. (2014). Settlement in the Irish Neolithic: New discoveries at the edge of Europe (Prehistoric Society research paper 6). London: Oxbow Books.

129

Stout, G. (2002). Newgrange and the bend of the Boyne. Cork: Cork University Press. Topping, P. (1982). Excavation at the cursus at Scorton, North Yorkshire 1978. Yorkshire Archaeological Journal, 54, 7–21. Verrill, L., & Tipping, R. (2010a). Use and abandonment of a Neolithic field system at Belderrig, Co. Mayo, Ireland: Evidence for economic marginality. The Holocene, 20(7), 1011–1021. Verrill, L., & Tipping, R. (2010b). A palynological and geoarchaeological investigation into Bronze Age farming at Belderg Beg, Co. Mayo, Ireland. Journal of Archaeological Science, 37, 1214–1225. Williamson, T. (1986). Parish boundaries and early fields: Continuity and discontinuity. Journal of Historical Geography, 12, 241–248. Williamson, T. (1987). Early co-axial field systems on the East Anglian Boulder clays. Proceedings of the Prehistoric Society, 53, 419–431. Whitefield, A. (2017). Neolithic ‘Celtic’ fields? A reinterpretation of the chronological evidence from Céide fields in North-Western Ireland. European Journal of Archaeology, 20(2), 1–23. Whitehouse, N.  J., Schulting, R.  J., McClatchie, M., Barratt, P., McLaughlin, T.  R., Bogaard, A., Colledge, S., Marchant, R., Gaffrey, J., & Bunting, M.  J. (2014). Neolithic agriculture on the European western frontier: The boom and bust of early farming in Ireland. Journal of Archaeological Science, 51, 181–205.

My Home Is My Castle! Field Systems and Farms: Rhythm and Land Appropriation During the Bronze Age in North-West France (2300–800 BCE)

10

Cyril Marcigny and Rebecca Peake

Abstract

Keywords

With the first studies dating from the 1990s, France could be seen as being somewhat behind in its research on ancient field systems compared to other countries. Post-­ war agriculture of north-west France has totally obliterated ancient fossilised landscapes; levelling micro-reliefs and so erasing the smaller irregularities in the landscape such as field systems that in certain cases were laid down more than 4000  years ago. Only since the wide scale development of preventive archaeology during the last 20  years and with its more extensive excavations, has research into prehistoric field systems been possible. This has resulted in the present overview of agricultural boundaries and their chronology from the laying down of the first field systems in the third millennium BCE to their abandonment during the Late Bronze Age. Our aim is not to reconstruct ancient landscapes over large areas  – as excavations rarely cover more than 5 ha at a time –but to understand how and why communities took possession of the landscape at the end of the Neolithic and how these planimetric features developed over the following thousand years. This focus is of obvious historic significance, as it targets the monopolisation and the management of the landscape and opens the way to a new reading of Bronze Age societies.

North-West France · Field systems · Neolithic · Bronze Age · Land management

C. Marcigny (*) Inrap, UMR 6566-CReeAAH, Institut national de recherches archéologiques préventives, Centre archéologique Inrap de Normandie, Bourguébus, France e-mail: [email protected] R. Peake Inrap, UMR 6298-ArTeHiS, Institut national de recherches archéologiques préventives, Centre archéologique Inrap de Passy, Passy, France e-mail: [email protected]

10.1 Introduction The first enclosed field systems emerge at the end of the third millennium BCE in north-west France. These new features that permanently mark the landscape were used to organize large areas for the increased demand of usable land, but beyond this structural role, they are also tangible evidence of a fundamental change in land tenurial regimes at the beginning of the Bronze Age. The change stems from a notably higher density of population than elsewhere in France in the areas around the Channel (or Manche-Mer du Nord) as well as in the south of Britain and in Normandy, during the Early Bronze Age and Middle Bronze Age I (1700–1500  BCE; Bradley et al. 2016: 20; 151–70; Marcigny et al. 2018). This has led to choice areas being given over to agricultural development and which were consequently structured by the first field systems (Fig. 10.1). However, these systems do not appear to be extensive, yet they form small, very organized, zones amidst vast areas of what seems to be uncultivated land. Within these organized zones, settlements were linked by pathways that facilitated movement from farm to field, which provides the ­opportunity to study these areas as an integral system (Béguier et  al. 2011; Marcigny and Ghesquière 2008). The extensive excavations carried out in France within the framework of preventive archaeology over the last 20 years have enabled us to draw up a timeline of domestic contexts and settlements. Following the various changes over the course of a millennium has shed light on how land management has evolved with time. This has led to the drawing up of a working hypothesis based on the assumption that agricultural features reflect forms of social change with the drawing up of (made in the sociological sense of a sustainable transformation of

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 S. Arnoldussen et al. (eds.), Europe’s Early Fieldscapes, Themes in Contemporary Archaeology, https://doi.org/10.1007/978-3-030-71652-3_10

131

132

C. Marcigny and R. Peake

Fig. 10.1  Cartography of the field systems identified in the south of Britain and in northwest France. (© C. Marcigny, Inrap: after Yates 2007 for Great Britain and by courtesy of S. Blanchet, Inrap, unpublished data for Brittany)

Hibaldstow

Welland

Marlborough Down

Tendring Peninsula Blackwater

Fyfield and Overton Down Exmoor Salisbury Plain Cranborne The Weald Chase New Forest Sussex Romney Dartmoor Marsh

Bodmin Moor

Dorset Downs

Lannion

Isle of Thanet

Hengistbury Head

Omonville Cosqueville Montfarville Goury Saint-Vigor Cherbourg Réville/Tatihou Bernières Langrune Bayeux Saint-Aubin-d’A Cairon Biéville-Beuville Fontenay-le-M Goulet

Cléder St-Méen-le-G Plouedern

the social organization or culture (Boudon 1984: 18, 93–110; Foster 1962: 1–9). This work was pivotal in developing a historical scenario that retraces the economic and socio-­ political evolution of communities during the Bronze Age from the late third millennium BC to the dawn of the first millennium BC. In this paper, we will not use regional chronologies, but will refer to the periods in centuries BCE.  It will however be useful as an introduction to detail – as shown in Fig. 10.2 – the chronological framework used in Western Europe whilst linking it to fluctuations in climate. In Western France, even though many academics use French or German chronological systems, Dutch and British timescales are more adapted to illustrating the major evolutionary steps of the Bronze Age in this area.

10.2 T  he Third Millennium: Twenty-First – Eighteenth Century BCE The first field systems of Normandy and southern Britain dating to the Early Bronze Age I (2300–2000 BCE; Marcigny 2012a, b; Evans et  al. 2016) are amongst the earliest in North-west Europe. In the North-west of France, most of the known field systems have been detected along the Calvados coastline in Normandy at Bernières-sur-Mer (Marcigny and Ghesquière 2003a), Bayeux, Cairon (Flotté et al. 2012) and Saint-Aubin-d’Arquenay (Ghesquière 2014: fig. 1). However, in the Cotes d’Armor in Brittany more recent archaeological investigations with extensive stripping of the

top soil over several hectares have revealed ditches dating to this early period (unpublished, S. Blanchet, Inrap) as part of the research project “Elements for a new approach to the Bronze Age in France – The chronological framework and forms of settlement“. The land plots cover several tens of hectares and are delimited by 1.8–2.5 m wide and 1.5–2 m deep ditches. However no site has been studied in its entirety. Their great depth having no apparent functional explanation other than to send a strong message of ownership by marking the limits as permanently as possible (Marcigny and Ghesquière 2008). The site plans vary, most are coaxial (Tatihou, Cairon or Bernières-sur-Mer, for example; Marcigny 2012b) others are more irregular (like Saint-­ Aubin-­d’Arquenay, Ghesquière 2014). At the same time, enclosing plots with ditches materialises ownership in a collective and individual sense, but also raises the question of how the land management with all of its social and economic implications (Marcigny 2012a, b; Brun and Marcigny 2012) was organised. Other features dating to the Early Bronze Age II (2000– 1650/1600  BCE) and thus contemporaneous to these first field systems are settlements themselves. Two main types, enclosed and open settlements, can be distinguished and they are linked to social status (the open settlements are “poorer” than the enclosed settlements; Marcigny 2012a). Indeed, there is no topographical relationship between the field systems and the enclosed settlements, and these show evidence of being of a higher status than the open settlements. These high status enclosed settlements are found in Lannion (Côtes

10  My Home Is My Castle! Field Systems and Farms: Rhythm and Land Appropriation During the Bronze Age in North-West… C résiduel

10

-10

14

133

2500

1500

Subatlantique

2000

500 BC

Subboréal

1000

Fig. 10.2  Synchronised chronology with measurements of residual 14C in the atmosphere (used as an indicator of climate); The French chronological system is used in this paper. (© C. Marcigny, Inrap)

d’Armor; Escats et  al. 2011), Ancretiéville-Saint-Victor (Seine-Maritime; Ghesquière and Marcigny 1996), Mondeville (Calvados, Chancerel et al. 2006a) or Luc-sur-­ Mer (Calvados, Flotté et al. 2012). The enclosure at Luc-sur-­ Mer is particularly emblematic. Dating to around 1900–1800 BCE, the enclosing ditches were found to contain special deposits of Armorican arrowheads, fine ware pottery and evidence of metalworking was detected in the settlement by the presence of tuyères and mould fragments (Flotté et al. 2012). In contrast to the enclosed settlements, open settlements are more common in this area and these farms are contained within an older established field system. Their buildings, grain stores and small outbuildings, and their finds, whetstones, flint ploughing and harvesting blades, relate ­exclusively to agricultural activities. The most studied example of this type of farm is the Tatihou site (Manche; Marcigny and Ghesquière 2003b), the first phase of which is characterised by a circular dwelling with grains stores and fire pits (Fig. 10.3).

Main communication routes, which also contribute to structuring the landscape, are also established during EBA II and are clearly visible within the field systems. Routes such as the “chemin Saulnier“(a right of way in use until the modern era) can be followed over several kilometres. Excavations at Banneville-la-Campagne (Calvados; Béguier et al. 2011) and the recent discovery of an enclosure at Hérouvillette (Calvados, Besnard-Vauterin et al. 2015) that opens directly onto the “chemin Saulnier“date its first construction phase to the Early Bronze Age. The location of the Hérouvillette enclosure is exceptionally not within an established field system.

10.2.1 A Historical Interpretation From the end of the third millennium BCE to the beginning of the second millennium BCE, the first field systems in North-west France appear simultaneously with the development of a strong social hierarchy underlined by funerary

134

C. Marcigny and R. Peake

Fig. 10.3  Phasing of the Early and Middle Bronze Age remains on the île Tatihou. A single and open agricultural settlement is replaced by three neighbouring farmsteads the finds of which attest to their contemporaneity. (© C. Marcigny, Inrap)

UEA

Phase 1 Early Bronze Age

0

100 m

UEA 3

UEA 1

UEA 2

Phase 2 Middle Bronze Age 0

contexts in southern England (Wessex Culture; Burgess 1980) and western France (Armorican Tumulus culture; Briard 1984). The power of the new elite was expressed in their control of agricultural land as underlined in the unpublished PhD dissertation of C.  Nicolas (2013), in which he pinpoints the strong association between the most fertile agricultural land and the richest Armorican burial mounds. In this context, the link with the agrarian planimetry is as obvious as it is today as land with a high agricultural potential is locked into large land plots (Dubreuil 1992). The analysis of the material culture found in the enclosed settlements of this period as well as the configuration of the sites themselves (enclosures defined by deep ditches) illustrate the high social status of their occupants – perhaps the seat of the ruling families of the area. This type of residence is generally located on a highly visible topographic feature, and systematically on the edge of the field systems (Marcigny 2012a). In fact, only open settlements, interpreted as farms are situated amidst agricultural land. This observation is particularly interesting and can be considered a good indicator of the socio-economic ties between the two types of site. Using a simple historical analogy, it is possible that the social

100 m

organisation at the beginning of the Bronze Age was similar to the Middle Ages (Bloch 1931: 185–98) or France‘s Ancien Regime (Bloch 1931: 98–216). In this context, land was under tenancy or involved sharecropping, controlled by the residents of high-status enclosed settlements but entrusted to smallholders (a sort of tythe, the exact terms of which during the Bronze Age can only be left to the imagination of the archaeologist). The smallholders would have absolutely no right to the land or at best, the right was shared between farmers and elites, which would result in the absence of enclosing systems.

10.3 The Seventeenth–Sixteenth Centuries BCE Important transformations occur in settlement type at the end of the Early Bronze Age and during the Middle Bronze Age (1650–1500  cal. BC). The enclosed settlements disappear from the landscape and the farms located within the field systems change (Fig. 10.4). Small units that were previously open settlements become enclosed, delimited by wide and

10  My Home Is My Castle! Field Systems and Farms: Rhythm and Land Appropriation During the Bronze Age in North-West…

135

EBA I (2300 to 2000 BCE)

EBA II (2000 to 1650/1600 BCE)

vegetation path field system crop or prairie

plain enclosure

tumulus or ring ditch

open settlement

grave

enclosed settlement

cemetery

Fig. 10.4  Modelling of land occupation in Normandy during the Early Bronze Age I and II, top left of the image, theoretical section of the field. (© C. Marcigny, Inrap)

deep ditches and these farms are found within the old field systems such as Cairon or Tatihou (Flotté et  al. 2012; Marcigny and Ghesquière 2003b) or in areas which were hitherto undeveloped (Marcigny 2012b). In the first case, the general pattern of the field systems is not or only slightly modified and farms are placed within an existing plot. The only change observed is that of the digging

of a deep ditch around the farm that marks the landscape more permanently while superimposing itself on the pre-­ existing field system. The Cairon site is a good example, as the field system founded at the beginning of the Early Bronze Age II around a tumulus containing a Bell Beaker grave is in part included in the new land organisation. Dating to the Early/Middle

136

C. Marcigny and R. Peake

Fig. 10.5  Cairon (Calvados). Plan and detail of the excavation at Hauts du Manoir 2. (© E. Ghesquière/C. Marcigny, Inrap. After C.C. Besnard-Vauterin, D. Giazzon & E. Ghesquière, Inrap)

"Rue des Ecureuils"

TR. 5 22

TR. 7

2 23 24 4

2 22

TR. 6

Tr.1

5

Tr.3

2

1

EBA I and/or II field system

21

7 8

3 4

9

10

6

57 56

14

53

11 12 13

55

Tr.4

15

77

17

22

16

20 21

18 19

49

48 47

31

46

30 23

22

66

32

33

34

64 62

61 59 58

65

63

104

43 41 60 3937 42 4038 36

102

44 101 45

79

Tr.2

103

ring ditch (funerary)

35

74 73

68 67 70 69 90

Tr.7 106

107 89

88

105

Tr.5

87b is

87 86 85

83 81 84 80 82

27

23

51 50 75 93 91

94 92

24 25 26 28

54

52

76

95 96

Bronze Age fire pit

C

29

100

78 98 97

99

Tr.6

EBA II/MBA I enclosure

187 7

"Les Hauts du Manoir 2" TR. 27 TR. 26

1 hectare 127 153

TR. 23 TR. 22

128 8

TR. 21

35

TR. 9

34

49 obu obus

TR. 10

104

48

50 103

51 1 105

TR. 11

102

C

TR. 12

"Les Hauts du Manoir 1"

78

85

TR. 14

64

sH

aut

75

sd

62

63

"Le

uM

ano

ir 2

"

4

15

5

14

"Rue Mac Mahon/rue on n u ue des es Frênes" e

enclosure

T EBA II / MBA I EBA II EBA I T

Bell Beaker grave

Bronze Age it sees the foundation of an enclosed farm delimited by deep ditches (Fig. 10.5). This can be compared to a similar situation at Tatihou and at Lannion in Brittany (excavation by S. Blanchet, Inrap) but also across the Channel at Ormesby St. Michael in Norfolk (Gilmour et al. 2014). When the areas under study are sufficiently large as is the case at Tatihou (the study covers nearly 20  ha, Marcigny and Ghesquière 2003b), it is not uncommon to observe a much more intensive human presence within the field systems with the establishment of several farms where during the Early Bronze Age there was only one (Fig. 10.3). This “collective” use of the field systems has also been observed on Dartmoor (Fleming 1986: 165–68, 1987: 195–97). In the second case, the field systems are founded during the Middle Bronze I in areas that were previously undeveloped (Fig. 10.5). The boundary ditches of these new plots do not have the same characteristics as those dating to the Early

Bronze, as ditches are shallower and often less regular in plan (for example at Saint-Vigor-d’Ymonville, Seine-­ Maritime; Clément-Sauleau et al. 2002). These new agrarian landscapes do not seem to have the same meaning as before. It appears that the attention is no longer focused on “marking” the landscape, but more on its strict functional use (drainage system, delineation, et cetera).

10.3.1 A Historical Interpretation From the seventeenth century BCE, when the social elites in place since the Early Bronze Age are no longer visible through the prism of funerary contexts, Bronze Age II field systems are largely taken over by enclosed farms, organised within a network of small hamlets. The general framework of the ancient field systems does not change. It seems that the

10  My Home Is My Castle! Field Systems and Farms: Rhythm and Land Appropriation During the Bronze Age in North-West…

137

factors used to identify and structure this particular form of agricultural land management are still upheld; ensuring its sustainability beyond the disappearance of the elite class. These factors contribute to stabilising the system, its identification and its permanence. Ditches are dug and maintained by farmers and the banks and hedges rebuilt as at Tatihou (Marcigny and Ghesquière 2003a, b: 163–64). The land plots are used over several generations without any significant change in their overall plan. It becomes not just a tool to manage the land and its production, but also a base for the establishment of a community, which will ensure its social reproduction and its territorial identity via a landscape fixed over several centuries. During the same phase, new land is occupied and developed according to different methods (shallow and discontinuous ditches), that contributes to the long-term stability of the whole system of land management. We propose as a hypothesis, therefore, at the end of the Early Bronze Age and during the Middle Bronze Age I, farmers seem to have taken hold of the land management that can now be interpreted as “cooperative”, similar to the model developed for Dartmoor by Fleming (1986). Each farmer has “the right to enclose,” allowing property to be delimited by the digging of deep ditches (earthen bank/hedge) and thereby asserting the farmer’s right to individual propriety. Land is shared between farmers and landowners giving them full ownership and joint exploitation rights (and using the historical metaphor, this is considered a communal property).

tural practices, provoking changes in land management. The permanent nature of land occupation is no longer ensured and the farming system established during the previous centuries collapses. The field systems in Normandy and Brittany are abandoned, ditches are even filled in and the hedges are burnt as for example at Digulleville, (Marcigny, study in progress). Settlements are delocalised. Environmental data (pollen and carpology) also indicates these changes around the fourteenth and thirteenth century BCE, in a large area that covers most of Northern Europe as suggested by A. Harding (1989). In Western France, this is underlined by the decease of cereals during the Middle/Late Bronze Age transition and the rise in the number of heliophytes (David 2014). This entire transition between the Middle and Late Bronze Age is marked by agricultural instability and a probable economic stress linked to the loss of agricultural resources. At the end of the period, this economic stress can be linked to the development of defensive and offensive weapons and the role of warriors in society (Uckelmann 2008) that bear witness to the violence that affected society during the thirteenth century BCE (Harding 1999). Traces of violence are not easy to identify in the archaeological record, however several sites have provided such evidence. An enclosure ditch at Tormarton (UK) contained the bodies of five young men who died in battle (Osgood 2006), a battlefield was identified at Tollense in Germany (Jantzen et  al. 2011) and the graves of Nord-Trøndelag in Norway (Fyllingen 2003) are a case in point.

10.4 The Fifteenth–Thirteenth Centuries BCE

10.5 Thirteenth–Twelfth Centuries BCE

The intensification and densification of occupation slows considerably during the Middle Bronze Age II and Late Bronze Age I and there are no new settlements and field systems (Marcigny et al. 2018, Fig. 10.6), perhaps in connection with the deterioration of the climate (Fig.  10.2). The new areas developed during this period include only remote farms distant from each other by a few hundred metres and connected by routes as illustrated at Nonant and Vaux-sur-­ Seulles (Calvados, Marcigny 2005). Some older field systems are still in use and maintained (on Tatihou for example), but most agricultural plots are no longer delimited by ditches. Most of the settlements remain enclosed even if open settlements emerge. The same phenomenon is observed in the south of Britain (Burgess 1980).

10.4.1 A Historical Interpretation The climate deterioration observed during the Middle Bronze Age II (Fig. 10.2) has a probably profound effect on agricul-

The observations of the previous phase (the abandonment of field systems and the delocalisation of settlements) are confirmed during the thirteenth and twelfth centuries BCE and the changes outlined at the turn of the thirteenth century are consolidated. This leads to a new form of occupation that initiates a new cycle of land management during the middle phase of the Late Bronze Age (Fig. 10.6). Some farms are enclosed (like those at Mondeville and Grentheville, Besnard-Vauterin et  al. 2006; Chancerel et  al. 2006b), yet there are no visible field systems (they either exist but are superficial or they completely disappear). It is interesting to note that this widespread abandonment of land plots delimited by ditches finds a parallel across the Channel with the apparent abandonment of Dartmoor. This phenomenon is linked to the climatic deterioration of 1395–1155  BCE (Fig. 10.2) and illustrated by the accumulation of layers of peat that cover the ditches of the ancient field systems (Fyfe et al. 2008). It is during this same period that fortified hilltop sites emerge in western France (30 sites have been investigated; Delrieu 2013). Most of these sites accommodate relatively

138 Fig. 10.6  Modelling of land occupation in Normandy during the Middle and Late Bronze Age. (© C. Marcigny, Inrap)

C. Marcigny and R. Peake MBA I (1650/1600 to 1500 BCE)

MBA II and LBA I (1500 to 1250 BCE)

LBA II to IIIb (1250 to 800 BCE)

small settlements, perhaps reserved for the high status members of society. The quantity and quality of the finds of these sites suggest a material culture made up of imports, metalworking, large quantities of bronze objects. This is the case for the sites of Basly (Calvados; San Juan et al. 2012), Port-­ en-­Bessin, Calvados (Lefort and Marcigny 2013) or that of Mauron (Morbihan, Tinevez et al. 2011) being the only sites to have been fully excavated in this area. The farms are no longer enclosed and adopt a new open plan without any form of delimitation whatsoever. This is especially the case in the second half of the twelfth century when these farms become the norm (e.g. Guichen, Ille-et-­ Vilaine; Hinguant et al. 1999; Guichainville, Eure; Marcigny and Carpentier 2006; Mont-Saint-Aignan, Seine-Maritime; unpublished B.  Aubry, Inrap, to name but a few French examples). They can be grouped together to form a village that lasts several generations, as seen in Britain (Reading, Berkshire; Brossler et  al. 2004), Ireland (Corrstown, Ginn

and Rathbone 2012; this site is however founded during the late fourteenth century BCE) or in France at Malleville-sur-­ le-Bec (Eure; Mare 2005) to name a few recent excavations (Marcigny 2012c; Rathbone 2013). The site of Malleville-­ sur-­le-Bec also includes a ring fort, a new type of structure also identified at Cagny (Calvados), Lamballe (Côtes d’Armor, Blanchet 2011) or across the Channel at Springfield Lyons (Essex; Brown and Medlycott 2013), Mucking (Essex; Evans et  al. 2015) and Navan (Ireland; Mallory and Lynn 2002). Most of these ring forts are founded around the late thirteenth century and twelfth century BCE and are subsequently maintained until the tenth–ninth century BCE.

10.5.1 A Historical Interpretation From the thirteenth century onwards field systems are abandoned and settlements are either open or enclosed (perhaps

10  My Home Is My Castle! Field Systems and Farms: Rhythm and Land Appropriation During the Bronze Age in North-West…

depending on their relative hierarchical status) and have generally a much shorter occupation period than previously. New fortified settlements emerge either located on hilltops or as ring forts, with some becoming the seat of the new social elite that takes control of certain production such as metalworking (Fort-Harrouard; Mohen 1989 or Springfield Lyons; Brown and Medlycott 2013: 47–74). This period is also marked by a greater polarization of settlements leading to the creation of villages with no administrative or commercial function and whose economic activity is based on agriculture. This is illustrated by agricultural features such as storage buildings and pits, ovens or the material finds discovered at Cahagnes, Malleville-sur-le-Bec (Normandie; Marcigny 2012a, c) or Caudan (Bretagne; Levan 2015). Evidence of craft production is absent on these sites and similar observations have been made in Britain (Rathbone 2013). These villages are organised over one or two generations until the ninth century BCE. This evolution can be compared to a similar process observed today in West Africa or in South America, where the emergence of a new elite of powerful landowners has led to the eviction of small-scale farmers who gradually group together to form farming villages (such as the “agro-towns” in West Africa; Curtis 2012). If we use this comparison for what happened in Europe during late prehistory, these “landless” farmers could sell their labour to the landowners/resident elites, who during the Bronze Age lived probably in the fortified settlements or ring forts. This new configuration of society was put into place between the end of the thirteenth century and the twelfth century BCE (during the Bronze final IIa) as a result of the socio-economic upheavals of the previous phase. This period of instability ended in the second half of the twelfth century with a reorganisation of society during the last phases of the Late Bronze Age as is suggested on a European scale by P. Brun’s research (Brun 1993).

10.6 Conclusion The objectives of this short essay on agrarian structures of the Bronze Age were to evaluate the various factors of societal change between the late third millennium and the first half of the twelfth century BC.  Our aim was to map their evolution throughout this time period with a focus on the transition from the Middle Bronze Age to the Late Bronze Age. These are, of course, only proposals where in many respects the historical analogies proposed can only hint at the more complex processes in place during the Bronze Age (exchange then commerce for example). However, the relationship between settlement and land management can contribute to rendering explicit extrinsic and intrinsic factors that have profoundly influenced the large-scale changes that affected Bronze Age society. The extrinsic factors such as

139

new agricultural practices and demographic factors have contributed to the development of the first field systems however the intrinsic factors through time and by conflict have also changed society‘s relationship with the land. Four main phases chart these changes in Western Europe and they are built around three crucial periods impacted by various societal factors. The first around the eighteenth and seventeenth centuries BCE sees the end of the social elite of the Early Bronze Age and the likely takeover of agricultural land by farmers whose social status changes (they thus acquire property rights). The second focuses on the early fifteenth century BCE and could be linked to an environmental (and therefore economic) crisis linked to the climate variations of the Middle Bronze Age II.  The thirteenth century BCE and the early twelfth century BCE mark the end of this crisis, when communities rebuild themselves around a new elite with ideals that announce the changes of the Iron Age.

References Béguier, I., Germain-Vallée, C., & Hincker, V. (2011). Etude micromorphologique du Chemin Saulnier: apport de l’archéologie des Paysages de la Plaine de Caen (Calvados, Basse-Normandie). Norois, 220(3), 121–142. Besnard-Vauterin, C.-C., Chancerel, A., & Marcigny, C. (2006). L’enceinte de l’âge du Bronze moyen de l’Etoile (Mondeville). In A.  Chancerel, C.  Marcigny, & E.  Ghesquière (Eds.), Le plateau de Mondeville (Calvados), du Néolithique à l’âge du Bronze (Documents d’Archéologie Française 99) (pp.  112–138). Paris: Éditions de la Maison des sciences de l’homme. Besnard-Vauterin, C.-C., Auxiette, G., Besnard, M., Deloze, V., Fiant, C., Giazzon, S., Le Puil-Texier, M., & Sehier, E. (2015). L’occupation d’un micro-terroir de la Protohistoire à l’Antiquité: le site d’Hérouvillette « Les Pérelles » (Calvados). Revue Archéologique de l’Ouest, 32, 129–176. Blanchet, S. (2011). Des grands travaux à l’âge du Bronze… L’enceinte bretonne de la Tourelle à Lamballe. Archéopages, 33, 18–27. Bloch, M. (1931). Les caractères originaux de l’histoire rurale française. Paris: Société d’édition « Les Belles Lettres ». Boudon, R. (1984). La place du désordre. Critique des théories du changement social. Paris: Presses Universitaires de France. Bradley, R., Haselgrove, C., Vander Linden, M., & Webley, L. (Eds.). (2016). The later prehistory of North-West Europe: The evidence of development-led fieldwork. Oxford: Oxford University Press. Briard, J. (1984). Les tumulus d’Armorique (Coll. L’Âge du Bronze en France 3). Paris: Picard. Brossler, A., Early, R., & Allen, C. (2004). Green Park (Reading Business Park), Phase 2 excavations 1995-Neolithic and Bronze Age sites (Thames Valley landscapes monograph 19). Oxford: Oxford Archaeology. Brown, N., & Medlycott, M. (2013). The Neolithic and Bronze Age enclosures at Springfield Lyons, Essex: excavations 1981–1991 (East Anglian Archaeology monograph 149). Norwich: East Anglian Archaeology. Brun, P. (1993). La complexification sociale en Europe moyenne pendant l’âge du Fer: essai de modélisation. In A.  Daubigny (Ed.), Fonctionnement social de l’âge du Fer: opérateurs et hypothèses pour la France. Actes de la Table ronde de Lons-le-Saunier, 24–26 Oct. 1990 (pp. 275–290). Lons-le-Saulnier: Cercle Girardot.

140 Brun, P., & Marcigny, C. (2012). Une connaissance de l’âge du bronze transfigurée par l’archéologie preventive. In D.  Garcia (Ed.), Nouveaux champs de la recherche archéologique (Archéopages, hors-série 10 ans) (pp.  132–139). Paris: Institut National de Recherches Archéologiques Préventives. Burgess, C. (1980). The age of Stonehenge. London: J.M. Dent & Sons. Chancerel, A., Ghesquière, E., & Marcigny, C. (2006a). L’enclos du Néolithique récent/Bronze ancien du MIR (Mondeville). In A.  Chancerel, C.  Marcigny, & E.  Ghesquière (Eds.), Le plateau de Mondeville (Calvados), du Néolithique à l’âge du Bronze (Documents d’Archéologie Française 99) (pp.  100–110). Paris: Maison des sciences de l’Homme. Chancerel, A., Marcigny, C., & San Juan, G. (2006b). La double enceinte de l’âge du Bronze moyen de la ZI sud (Mondeville, Grentheville). In A.  Chancerel, C.  Marcigny, & E.  Ghesquière (Eds.), Le plateau de Mondeville (Calvados), du Néolithique à l’âge du Bronze (Documents d’Archéologie Française 99) (pp. 140–172). Paris: Maison des sciences de l’Homme. Clément-Sauleau, S., Ghesquière, E. & Marcigny, C. (2002). Du Ve millénaire au début de l’Antiquité, les occupations de Saint-Vigor-­ d’Ymonville (76). Haute-Normandie Archéologique, numéro spécial: 12–17. Curtis, D.  R. (2012). Pre-industrial societies and strategies for the exploitation of resources. A theoretical framework for understanding why some settlements are resilient and some settlements are vulnerable to crisis. PhD thesis. Utrecht: Utrecht University. David, R. (2014). Modélisation de la végétation holocène du Nord-Ouest de la France: Reconstruction de la chronologie et de l’évolution du couvert végétal du Bassin parisien et du Massif armoricain. PhD thesis. Rennes: Université de Rennes 1. Delrieu, F. (2013). Chronologie et statut des sites fortifiés de hauteur au Bronze final et au 1er Fer ancien dans le Nord-Ouest de la France (Haute-Normandie, Basse-Normandie et Bretagne). In S.  Krausz, A. Colin, K. Gruel, I. Ralston, & T. Dechezleprêtre (Eds.), L’âge du Fer en Europe, Mélanges offerts à Olivier Buchsenschutz (Mémoires 32) (pp. 131–146). Ausonius éd: Bordeaux. Dubreuil, V. (1992). Typologie des paysages ruraux de l’Ouest de la France à partir de classifications d’images du satellite NOAA. Norois, 155, 283–296. Escats, Y., Blanchet, S., & Nicolas, T. (2011). Une enceinte et une nécropole protohistoriques à Lannion: présentation luminaire. Bulletin de l’Association pour la Promotion des Recherches sur l’Âge du Bronze, 8, 77–79. Evans, C., Appleby, G., & Lucy, S. (2015). Lives in land—Mucking excavations by Margaret and Tom Jones 1965–78. Oxford/ Havertown: Oxbow. Evans, T., Tabor, J., & Vander Linden, M. (2016). Twice-crossed river. Prehistoric and palaeoenvironmental investigations at Barleycroft Farm/Over, Cambridgeshire (Landscape archives series). Cambridge: Cambridge Archaeological Unit. Fleming, A. (1986). Les délimitations territoriales de l’Age du Bronze dans la région de Dartmoor (Grande-Bretagne). In A.  Ferdière & E.  Zadora-Rio (Eds.), La prospection archéologique: paysages et peuplement (pp.  161–168). Paris: Éditions de la Maison des sciences de l’homme. Fleming, A. (1987). Coaxial field systems: Some questions of time and space. Antiquity, 61, 188–202. Flotté, D., Ghesquière, E., Giazzon, D., Marcigny, C., Ménager, L., & Vipard, L. (2012). Nouveaux parcellaires et habitats du Bronze ancien/moyen en Basse-Normandie: les fouilles de 2010. Bulletin de l’Association pour la Promotion des Recherches sur l’Âge du Bronze, 10, 82–90. Foster, G. (1962). Traditional cultures: The impact of technological change. New York: Harper & Brothers. Fyfe, R.  M., Brück, J., Johnston, R., Lewis, H., Roland, T.  P., & Wickstead, H. (2008). Historical context and chronology of Bronze

C. Marcigny and R. Peake Age land enclosure on Dartmoor, UK. Journal of Archaeological Science, 35(8), 2250–2261. Fyllingen, H. (2003). Society and violence in the Early Bronze Age: An analysis of human skeletons from Nord-Trøndelag, Norway. Norwegian Archaeological Review, 36(1), 27–43. Ghesquière, E. (2014). Parcellaires de l’âge du Bronze ancien à Saint-­ Aubin-­ d’Arquenay « Eco-quartier des Vignettes » (Calvados). Revue Archéologique de l’Ouest, 32, 57–82. Ghesquière, E., & Marcigny, C. (1996). L’enceinte Chalcolithique de la « Plaine d’Ectot » à Ancretiéville-Saint-Victor (Seine-Maritime). Bulletin de la Société Préhistorique Française, 93(1), 97–102. Gilmour, N., Horlock, S., Mortimer, R., & Tremlett, S. (2014). Middle Bronze Age enclosures in the Norfolk Broads: A case study at Ormesby St Michael, England. Proceedings of the Prehistoric Society, 80, 141–158. Ginn, V., & Rathbone, S. (2012). Corrstown: A coastal community. Excavations of a Bronze Age Village in Northern Ireland. Oxford: Oxbow. Harding, A. F. (1989). Interpreting the evidence for agricultural change in the Late Bronze Age in northern Europe. In H.-Å. Nordström & A. Knape (Eds.), Bronze Age Studies. Transactions of the British-­ Scandinavian Colloquium in Stockholm, May 10–11, 1985 (Studies 6) (pp. 173–181). Stockholm: The Museum of National Antiquities. Harding, A.  F. (1999). Warfare: A defining characteristic of Bronze Age Europe? In J. Carman & A. F. Harding (Eds.), Ancient warfare (pp. 157–173). Stroud: Sutton. Hinguant, S., Le Goff, E., & Maguer, P. (1999). Une occupation de l’âge du Bronze en Moyenne Vallée de Vilaine, La Basse Bouexière (Guichen, Ille-et-Vilaine). Revue Archéologique de l’Ouest, 16, 67–83. Jantzen, D., Brinker, U., Orschiedt, J., Heinemeier, J., Piek, J., Hauenstein, K., Krüger, J., Lidke, G., Lübke, H., Lampe, R., Lorenz, S., Schult, M., & Terberger, T. (2011). A Bronze Age battlefield? Weapons and trauma in the Tollense Valley, North-Eastern Germany. Antiquity, 85, 417–433. Lefort, A., & Marcigny, C. (2013). L’oppidum du Mont-Castel (Port-en-Bessin-Huppain, Calvados), Premiers résultats. Revue Archéologique de l’Ouest, 29, 107–131. Levan, M. (2015). Premier hameau de l’âge du Bronze final en Bretagne. Archéopages, 40, 48–49. Mallory, J. P., & Lynn, C. J. (2002). Recent excavations and speculations on the Navan complex. Antiquity, 76(292), 532–541. Marcigny, C. (2005). Une ferme de l’âge du Bronze à Nonant (Calvados). In C. Marcigny, C. Colonna, E. Ghesquière, & G. Verron (Eds.), La Normandie à l’aube de l’histoire, les découvertes archéologiques de l’âge du Bronze 2300-800 av. J.-C (pp. 48–49). Paris: Somogy éditions d’art. Marcigny, C. (2012a). Au bord de la mer. Rythmes et natures des occupations protohistoriques en Normandie (IIIe millénaire - fin de l’âge du Fer). In M. Honegger & C. Mordant (Eds.), Au bord de l’eau, Archéologie des zones littorales du Néolithique à la Protohistoire (pp. 365–384). Cahier d’Archéologie Romande: Lausanne. Marcigny, C. (2012b). Les paysages ruraux de l’âge du Bronze, structures agraires et organisations sociales dans l’Ouest de la France. In V. Carpentier & C. Marcigny (Eds.), Des Hommes aux Champs, Pour une archéologie des espaces ruraux du Néolithique au Moyen Age (pp. 71–80). Rennes: Presses Universitaires de Rennes. Marcigny, C. (2012c). L’habitat de l’âge du Bronze. In A.  Lehoërff (Ed.), Par-delà l’horizon, Sociétés en Manche et mer du Nord, il y a 3500 ans (pp. 123–125). Paris: Somogy éditions d’art. Marcigny, C., & Carpentier, V. (2006). Les fouilles extensives du Long Buisson, entre Evreux et le Vieil-Evreux (Eure). In: P. Brun, C. Marcigny and J. Vanmoerkerke, eds. Une archéologie des réseaux locaux. Quelles surfaces étudier pour quelle représentativité? Les Nouvelles de l’Archéologie, 104(105), 57–60.

10  My Home Is My Castle! Field Systems and Farms: Rhythm and Land Appropriation During the Bronze Age in North-West… Marcigny, C., & Ghesquière, E. (2003a). Parcellaire et nécropoles de l’âge du Bronze ancien à Bernières-sur-Mer (Calvados). Bulletin de la Société Préhistorique Française, 100(1), 117–134. Marcigny, C., & Ghesquière, E. (2003b). L’Île de Tatihou à l’âge du Bronze (Manche). Habitats et occupation du sol (Documents d’archéologie française 96). Paris: Maison des sciences de l’Homme. Marcigny, C., & Ghesquière, E. (2008). Espace rural et systèmes agraires dans l’ouest de la France à l’âge du Bronze: quelques exemples normands. In J. Guilaine (Ed.), Villes, villages, campagnes de l’Âge du Bronze, séminaires du Collège de France (pp.  56–278). Paris: Editions Errance. Marcigny, C., Néré, E., Peake, R., Riquier, V., & Le Denmat, G. (2018). Rythme et nature des occupations du IIIe millénaire à l’aube de l’âge du Fer en France septentrionale. In O.  Lemercier, I. Sénépart, M. Besse, & C. Mordant (Eds.), Habitations et habitat du Néolithique à l’âge du Bronze en France et ses marges (pp. 525– 536). Archives de l’Ecologie Préhistorique: Toulouse. Mare, E. (2005). Le village de Malleville-sur-le-Bec (Eure). In C.  Marcigny, C.  Colonna, E.  Ghesquière, & G.  Verron (Eds.), La Normandie à l’aube de l’histoire, les découvertes archéologiques de l’âge du Bronze 2300-800 av. J.-C (pp. 52–53). Paris: Somogy éditions d’art. Mohen, J. P. (1989). Ateliers métallurgiques dans l’habitat protohistorique du Fort-Harrouard. Bulletin de la Société préhistorique française, 10–12, 404–408. Nicolas, C. (2013). Symboles de pouvoir au temps de Stonehenge: les productions d’armatures de prestige de la Bretagne au Danemark

141

(2500–1700 av. J.-C.). PhD thesis. Paris: Université de Paris 1 Panthéon-Sorbonne. Osgood, H. (2006). The dead of Tormarton: Bronze Age combat victims? In T.  Otto, H.  Thrane, & H.  Vandkilde (Eds.), Warfare and society: Archaeological and social anthropological perspectives (pp. 331–340). Aarhus: Aarhus University Press. Rathbone, S. (2013). A consideration of villages in Neolithic and Bronze Age Britain and Ireland. Proceedings of the Prehistoric Society, 79, 39–60. San Juan, G., Dron, J.-L., Fromon, N., Lepaumier, H., & Besnard, M. (2012). L’éperon fortifié de « La Campagne » à Basly (Calvados). Données nouvelles concernant la phase Bronze final. Bulletin de l’Association pour la Promotion des Recherches sur l’Âge du Bronze, 10, 91–94. Tinevez, J.  Y., Quesnel, L., Marcoux, N., Donnart, K., Bardel, V., Gautier, M., Bernard, V., Fontugne, M., Van der Plicht, J., & Oberlin, C. (2011). Le site de La Rochette à Mauron (Morbihan): les multiples occupations d’un promontoire. Revue Archéologique de l’Ouest, 28, 71–148. Uckelmann, M. (2008). Irland oder Iberien  – Überlegungen zum Ursprung einer Ornamentform der Bronzezeit. In F.  Verse, B.  Knoche, J.  Graefe, M.  Hohlbein, K.  Schierhold, S.  Siemann, M.  Uckelmann, & G.  Woltermann (Eds.), Durch die Zeiten … Festschrift für Albrecht Jockenhövel, zum 65, Geburtstag (pp. 259– 268). Marie Leidorf: Rahden. Yates, D. T. (2007). Land, power and prestige. Bronze Age field systems in Southern England. Oxford: Oxbow Books.

Tracing the Remains of a Late Bronze Age Field System in Central Mainland Greece

11

Michael F. Lane and Vassilis L. Aravantinos

Abstract

The assumption that the Late Bronze Age (‘Late Helladic’ or LH) texts in the Linear B script, used on Crete and mainland Greece and dating to about 1390 to 1190 BC, can be used in conjunction with ethno-archaeological data to model the topography of economic transactions was recently investigated with  fieldwork in central Greece. Previously calculated  parameters of  the dimensions and organisation of land plots recorded in the Linear B archives informed a methodology for discovering a comparable system of fields under suitable conditions in the Aegean. The AROURA project, that formed the basis for this article, identified the Kopaic Basin in northern Boiotia as a suitable study region. Since the end of the last ice age, it has contained a shallow lake with seasonally fluctuating wetland margins, except in  the modern era, when it was thoroughly drained, and LH, when it was partially drained, and dykes protected polders of dry land from floods. One of these polders encloses the colossal fortress of Glas, whose storehouses contained thousands of metric tonnes of wheat, indicating extensive cultivation nearby. The polder also displays expansive traces of a premodern landscape beneath the present plough soil. AROURA executed a program of magnetometry, ground-­ truthed with soil profiling, over 60  ha of land selected from inside the polder around Glas. This fieldwork revealed evidence of canals connected to rivers that were diverted during the LH.  The canals  appear to feed a network of irrigated fields, particularly to the west of Glas near the polder’s dyke. These plots are evidently demarcated by low levees made of lake sediment and parallel M. F. Lane (*) Ancient Studies Department, University of Maryland, Baltimore County, Baltimore, MD, USA V. L. Aravantinos 9th Ephorate of Prehistoric and Classical Antiquities, Thebes, Greece

ditches. Their size and configuration conform to the topographical model, and radiocarbon and luminescence dates corroborate their already circumstantially probable attribution to the LH. These discoveries raise new questions concerning the process of agricultural intensification and state formation in LH Greece, particularly concerning the origin and history of relevant hydraulic technologies. Keywords

Bronze Age · Mycenaean · Hydraulic engineering · Agriculture · Agricultural strategy · Socio-economic complexity

11.1 G  las and the Late Bronze Age Drainage of the Kopais The fortress of Glas (Γλας) sits in an ancient polder, or tract of land claimed from lake waters, in the Kopaic Basin (or ‘Kopais’) in northern Boiotia Department, central Greece (Fig.  11.1). The construction and primary inhabitation of Glas has been dated by interregional synchronism of ceramic pottery series to the period from c. 1300 until c. 1190 BC, that is to the Late Bronze Age, or ‘Late Helladic’ in Aegean terms, and specifically to the Late Helladic IIIB period (Iakovidis 1998: 188–91, 2001: 142–45; Table  11.1). The fortress has a boulder-built Cyclopean outer wall about 2.5  km in circumference, the longest of its kind in the Aegean. It is 8 m thick in some places (Iakovidis 1992). It seems both logical and logistically plausible that the surrounding wetlands were drained before Glas was built, comparanda from other areas in later times notwithstanding. This inference means that the terminus ante quem of the creation of the polder is about the beginning of the thirteenth century BC.  The polder was created  by channelling nearly 25 km of major tributary rivers between stone walls and into karstic sinkholes on the edges of the former lake, erection of

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 S. Arnoldussen et al. (eds.), Europe’s Early Fieldscapes, Themes in Contemporary Archaeology, https://doi.org/10.1007/978-3-030-71652-3_11

143

144

M. F. Lane and V. L. Aravantinos

Fig. 11.1  Location of Glas and the Kopaic Basin in mainland Greece, showing reconstruction of drainage system. (After Knauss 1984, with permission)

dams to divert the spate of seasonal streams from the surrounding mountains, and construction of kilometres of stone-and-earth dykes to protect land from fluctuating water levels in the remaining parts of the lake (Knauss 1984, 1987). Widely separated excavated sections of the Cyclopean retaining walls of the rivers provide a terminus post quem for their construction that ranges from the Mycenaean ‘transitional’ period of the Middle Helladic (MH) III through Late Helladic (LH) IIA (c. 1700–1500  BC) until the LH IIIA2 (c. 1400–1300 BC; Aravantinos et al. 2006; Kountouri et al. 2012). Hence it is possible, though not certain, that the terminus post quem date of whole drainage and flood control system, as well as the polder, (which is one of several identified in the Basin; Knauss 1984, 1987: 145–225, 2001: 26–42), is just a matter of decades earlier than the building of Glas (LH IIIA2–B1 transition). Moreover,  remains of the fortress include a precisely planned multi-storey residential or administrative suite, embellished in the style of the regional palaces (Iakovidis 1998, 2001), and two storehouses

nearly 150  m long that evidently contained thousands of tonnes of einkorn wheat (Triticum monococcum), as well as scores of litres of olive oil or wine or both (Iakovidis 1998: 15–23; 175; Jones 1995). The proxies of charred remains of a single grain species and ceramic vessels are entirely consistent with contemporary textual linguistic evidence of a palace-administered extensive agricultural regime (Halstead 1992). This means maintenance of a strategy of low human labour input for a modest, though reliable, yield of a few subsistence cultivars, complemented by more labour-­ intensive, non-subsistence arboriculture and viticulture (Halstead 2014).  Glas burnt down for the last time around 1190  BC (LH IIIB2/C1 Early; Iakovidis 2001: 145; 156), and the drainage system failed at some point thereafter, since by the time of the reappearance of written records some centuries later (Hom.Il.5.709; Ar.Ach.860–80; Thrphr.H.P.4.10), the Kopais is known only as a lake (Kalcyk 1984). Indeed, it was not permanently drained again until the

11  Tracing the Remains of a Late Bronze Age Field System in Central Mainland Greece Table 11.1  Mainland Greek Bronze Age chronology Period/sub-period name Early Helladic (EH) I EH IIA EH IIB EH III Middle Helladic (MH) I MH II MH III Late Helladic (LH) I LH IIA LH IIB LH IIIA1 LH IIIA2 LH IIIB1 LH IIIB2 LH IIIC Early

Approx. years (BC), high 3100–2700

Traditional low chronology

145

Table 11.2  (A) Hypothetical areas of land measured in seed grain, using values of Mycenaean dry measurement in Ventris and Chadwick (1973) vs Those in Chadwick (1966) and Palmer (1989) (after Lane 2009). (B) Dimensions of hypothetical plots based on sowing density (enframed above) and ploughing rates (after Lane 2009) (A)

2700–2400 2400–2200 2200–2000 2000–1900 1900–1750 1750–1690 1690–1610

1900–1700 1700–1590 1590–1500

1610–1500 1500–1440 1440–1400 1400–1300 1300–1230 1230–1190 1190–1170

1500–1430 1430–1390 1390–1360 1360–1300

System Ventris and Chadwick (equivalent areas)

Unit Z V T GRA GRA 10

Chadwick/ Palmer (equivalent areas)

Z V T GRA GRA 10

60 litres/ha .0083 ha 0.03 ha 0.20 ha 2.00 ha 20.0 ha

40 litres/ha 0.01 ha 0.04 ha 0.24 ha 2.40 ha 24.0 ha

60 litres/ha 0.007 ha 0.027 ha 0.16 ha 1.60 ha 16.0 ha

50 litres/ha 0.008 ha 0.032 ha 0.192 ha 1.92 ha 19.2 ha

(B)

Sources: Betancourt (2007), Manning et  al. (2006), Shelmerdine (2008), and Warren and Hankey (1989)

twentieth century of our era (Dean 1937; Durand-Claye 1888; Papadopoulos 1997). By virtue of lying in a polje, or alluvial plain in a karstic graben, the former Kopaic Lake was a basin of net sedimentary deposition (Higgins and Higgins 1996: 76–78). It thus  seems a nearly-perfect place for the discovery of traces of the constituent practices of a palace-administered agricultural regime, rather than simply their products’ proxies, such as seeds and vessels. Therefore, between 2007 and 2009, a topographic model of typical LH IIIB grain estates was developed. The foundations of the model were the quantitative parameters on the size and subdivision of plots and allotments in contemporary, Mycenaean Greek landholding texts written in the Linear B script. These were further interpreted, with respect to the organization and dimensions of land plots, by historical, ethnographic, and agronomic accounts of broadcast sowing densities and animal-drawn ploughing rates (Lane 2009, 2012; Table 11.2) (Fig. 11.2). The authors then embarked in 2010 on a geophysical survey titled ‘Archaeological Reconnaissance of Un-investigated Remains of Agriculture’ (AROURA), sampling  systematically and adaptively about 1,000  ha of the polder around Glas. By 2012, AROURA had sampled about 60 ha (c. 6%) of this study area with magnetometry at one-­eighth metre interval on metre-wide traverses, within transects 90 m wide and up to 360 long (Fig. 11.2). The goal was to detect elements in a hierarchised list of expected land plot  qualities linked to administrative and agricultural practices (Table  11.3). Magnetometry results were verified through comparing sediment and soil profile descriptions and measuring volume-specific magnetic susceptibility at verti-

Sowing density 40 litres/ha 50 litres/ha 0.01 ha 0.0125 ha 0.04 ha 0.05 ha 0.24 ha 0.30 ha 2.40 ha 3.00 ha 24.0 ha 30.0 ha

Area T GRA GRA 10

Square dimensions Sown (40–50 Ploughed l/ha) c. 49 × 40 m c. 52 × 52 m c. c. 196 × 98 m 209 × 105 m c. c. 490 × 490 m 522 × 522 m

Oblong dimensions Sown (40–50 l/ha) Ploughed c. 40 × 60 m c. 42 × 65 m c. c. 200 × 120 m 210 × 130 m c. c. 400 × 600 m 420 × 650 m

cal intervals. Analyses of permutations of satellite spectral data (‘supervised reclassification’) permitted magnetic anomalies to be traced into areas  not sampled geophysically (Lane et al. 2016; Stussi 2011). Collection of portable objects from the ground surface was conducted to determine if there was any correlation of material and chronological types with features discovered with geophysics (Aravantinos and Lane 2010; Lane 2010; Lane and Aravantinos 2012).

11.2 Geophysical Results 11.2.1 Connecting Glas with the Drainage System The geophysical results showed that the most important predicted elements were present. Together with the surface collections, they raised significant questions about the nature and extent of the LH mechanisms of hydraulic engineering in the Kopais, as well as about the date and duration of their construction and use. These questions have important implications for theories of regional development of social complexity. Magnetometry detected segments of several linear anomalies, as well as a couple of gently arcing anomalies, which

146

M. F. Lane and V. L. Aravantinos

Fig. 11.2  AROURA survey area, showing transects sampled with magnetometry, 2010–2012

directly or indirectly connect major components of the Kopaic drainage system with Glas. While some of these correspond to features previous researchers had observed on the ground surface (Knauss 1984; Lauffer 1979), others were detected for the first time and traced further through the landscape in satellite imagery. Those linear anomalies making a perpendicular connection with the base of the outcropping of Glas were typified as ‘joining’. Those making an indirect connection, via a joining anomaly, were described as ‘bounding’, because they correspond to features that seem to separate Glas from the singularly interesting ‘reticulate’ patterns of anomalies and corresponding features detailed below. Joining anomalies are essentially radial from Glas, bounding anomalies are essentially perimetric. Chief among the anomalies corresponding to already identified features was the so-called ‘Revetted Canal’ (Fig. 11.3). It showed up clearly to the south of Glas, running south southwest from the eastern tip of the rock outcrop on which the fortress sits to the scarp of Mount Fteliá, along which the Peripheral Canal, described below, flowed. It appears as two parallel anomalies, negative with respect to background (lighter shading in figures), each a couple of metres wide, separated by about five metres, such as one would expect of a

channel between stone walls. This interpretation is supported by previous investigators’ written and photographic records of fieldstone, though they initially thought the canal a dam (Fig. 11.1; Knauss 1984: 213–27, cf. 1987: 207–18; Lauffer 1979: 452–53; Threpsiadis 1965). Concentrations of stone were also observed in modern irrigation ditches that dissect the anomalies and by probing above the anomalies. Previous investigators had recorded a north-eastern stretch of the Revetted Canal, running between the promontory of Nisí and the eastern tip of Glas. It was presumed to connect with the Peripheral Canal at its northern end, both canals fed by water flowing south from the ancient artificial channel that combined the waters of the Melas and Kephissos rivers. While the corresponding crop and soil mark (hereafter ‘field mark’) shows up in historical aerial photographs, it is difficult to detect in recent satellite data because it runs nearly parallel with modern field boundaries and concentric ploughing furrows. It may have been intersected in the western corner of one magnetometric sampling transect, but if so, its corresponding magnetic anomaly is obscured by modern farming and irrigation features. The Peripheral Canal is so called because it describes a shallow arc from the promontory of Nisí to the north scarp of

11  Tracing the Remains of a Late Bronze Age Field System in Central Mainland Greece

147

Table 11.3  Hierarchised list of features AROURA expected to be able to detect Feature Rectilinear or near-­rectilinear field partitions, built or excavated Ditches for drainage or irrigation, or both

Plough scars in the subsoil from repeated ploughing in one direction Pits for vines and tree crops, particularly in peripheral areas Scattered remains of built or excavated outbuildings (shelter/ protection, crop processing, etc.)

Expected geomagnetic signature Linear magnetically positive or negative anomalies (or both) relative to background, each vertex tens or hundreds of metres long and several metres wide Linear magnetically positive anomalies (reflecting in-filling with iron-oxide rich topsoil) internal to the field partitions, probably narrower than them Palimpsest of shorter quasi-linear (less regular) positive anomalies approximately parallel or perpendicular to each other, no more than half a metre wide Positive anomalies spaced at regular intervals, between about a metre and a few metres wide Negative anomalies (reflecting stone or mud-brick construction, or tamped earth; cf. built field partitions) a few to tens of metres wide on the edges or in the interstices of field divisions

Fig. 11.3  Revetted Canal, magnetometry results (greyscale) and interpretation (Transect D1; cf. Figs. 11.2 and 11.18)

Mount Fteliá (Fig.  11.4). Knauss initially identified its surface traces as those of a dam protecting the polder from influx of a seasonal stream descending from the east. Later, he conjectured that, along with the Revetted Canal, it was a channel bearing waters from the LH artificial river channel to Fteliá and thence around the peak’s eastern end to the Vrystiká Sinkhole (Knauss 1984: 195–204; 1987: 207–18). A field mark appearing to correspond to the Peripheral Canal is seen in satellite data near the north end of Nisí, between 400 and 500  m from the eminence of the Aghía Marína Pýrghos (AMP) settlement site (Fig.  11.5). Magnetometry made clear the differences in character between the Revetted Canal and Peripheral Canal and therefore clarified their pos-

Fig. 11.4  Peripheral Canal, magnetometry results (greyscale) and interpretation (Transect C1; cf. Figs. 11.2 and 11.18)

148

M. F. Lane and V. L. Aravantinos

Fig. 11.5  Field mark of Peripheral Canal, near Aghía Marína Pýrghos, or ‘AMP’ (AROURA/Worldview-2 red–green–blue–NIR composite image; cf. Figs. 11.2 and 11.18)

sible functions. The remains of the Cyclopean walls of the canalized rivers rise as high as two metres above the current surrounding plain (Kountouri et al. 2012; Lane, field notes). Hence the ‘revetments’ of the Revetted Canal were, like the Cyclopean retaining walls, evidently meant to channel a stream at or above the level of the polder, thus also supplying Glas with fresh water. In contrast, the Peripheral Canal appears to be dug deeply into the subsoil and to have a levee along its western edge, akin to Knauss’s ‘dam’. Therefore, it could have served both as a deep overflow channel for the canalized rivers or Revetted Canal, especially if floodgates existed at the northern end, and as a sump for seasonal waters flowing from the valley to the east. Both it and the Revetted Canal nonetheless could have joined to empty in the Vrystiká Sinkhole. The most prominent joining anomaly discovered during AROURA is visible in satellite and magnetometry data running from the embankment of the crushed-stone road around Glas to the present course of the Melas River on a bearing of about 13° east of north (Fig.  11.6). The embankment may hide its terminus nearer the scarp of the outcrop on which the fortress sits. Post-Helladic down-­cutting, not to mention certain modern field features, such as hedgerows and plough scars, could account for the disappearance of any trace of the anomaly in available data between the modern Melas and the ancient joined, canalized rivers, a span of about 100 m. In the magnetometry data, the anomaly consists of two components, each properly speaking a separate anomaly. One of these components is negative with respect to the magnetic background (light shading in Fig.  11.7), and about five

metres wide. It is joined along its eastern edge by a narrower, parallel positive component (dark in Fig. 11.7). Such combination of anomalies is typical of a zone of re-deposited subsoil or compacted or eroded topsoil, beside which is a strip of more magnetically susceptible topsoil or the topsoil fill of a cut (Aspinall et al. 2008). In other words, the pattern is consistent with a levee paralleled with an adjoining ditch. Less clear in the magnetometry data than the above, although nearly as clear in satellite images, is a joining anomaly running from about 120 m north of Glas’ west gate for about 900  m to a point about 300  m short of the dyke enclosing the polder to the west. There it joins a reticulate pattern of field marks perceptible in satellite data (Fig. 11.8). Before merging with this reticulate pattern, a linear anomaly of the bounding type intersects it at nearly a right angle. This perpendicular anomaly bifurcates near the point of intersection, the north-western branch curving gently to a point just north of the hill of the village of Kástro (see Figs. 11.2 and 11.18), the site of Kopai in the Classical Period (c. 480–300 BC). The LH canalization of the Melas and Kephissos runs around the north side of the hill, so this curving branch could connect with it. In satellite data, it continues for about 250 m before it disappears beneath the footprint of a modern farm. The northern, straight branch continues on a bearing of about four degrees west of north until it too becomes obscure in the vicinity of intersecting field marks, whose corresponding magnetic anomalies, in two magnetometry transects north of Glas, suggest paleochannels. All these linear anomalies appear to be between three and four metres wide. In the mag-

11  Tracing the Remains of a Late Bronze Age Field System in Central Mainland Greece

149

Fig. 11.6  Joining anomaly and field mark between Glas and LH canalized rivers (blue–green–NIR image; cf. Figs. 11.2 and 11.18)

netometry data, at least the major north–south branch clearly consists of a negative component joined to the east by a parallel positive component. South of the aforementioned intersection, the straight branch of the bounding anomaly continues south on the same bearing. It appears both in magnetometry data and field marks to extend some 600 m, after which it abruptly changes bearing to about 30° east of south toward the scarp of the western tip of Fteliá. There it presumably merges with the remains of the Peripheral Canal (Fig. 11.9). It perhaps represents a feature that channelled further water into the Vrystiká Sinkhole. While nowhere along the southern extent of this major bounding anomaly is the magnetic contrast so strong as it is to the north, a broad parallel negative component may still just be visible on its east side. Taken together, the joining and bounding anomalies link Glas to the LH drainage works, strengthening the circumstantial topographic case for their being part of the same system. It may be further argued that they are connected through specific functions. It is now clear that the Revetted Canal and the Peripheral Canal served respectively as a secondary drainage conduit and overflow channel or sump. The joining and bounding anomalies bear all the characteristics of levees

built up of sediment excavated from adjacent ditches. Connected with the canalized rivers, as at least the majority of them appears to be, an obvious hydraulic hypothesis is that they represent feeder canals for a drainage or irrigation system inside the polder. As much as elements of water control, the levees may also have been causeways between Glas and nearby settlements, e.g. Aghía Marína Pýrghos and Kopai’s LH predecessor (Fossey 1988: 277–90, Farinetti 2011: 127–35), and between the canalized rivers and the drainage sinkholes.

11.2.2 Evidence of an Irrigated Agricultural Field System The most astonishing magnetometric results, in terms of both scale and intricacy, are those from the area between the polder dyke and the approximately north–south bounding anomaly to its east. They comprise two distinct network patterns, a northern one (Reticulate 1) whose linear elements are parallel and perpendicular to the dyke and southern one (Reticulate 2), some of whose linear elements are parallel to the dyke. The latter’s  other elements, running from north-

150

Fig. 11.7  Joining anomaly (Transect A1), magnetometry results and interpretation (cf. Figs. 11.2 and 11.18)

M. F. Lane and V. L. Aravantinos

east to south-west, are more closely perpendicular to the southern, south-east-tending extent of the bounding anomaly. The linear anomalies in Reticulate 1 form nearly square quadrilaterals, each with a perimeter of almost 120 m, their sides appearing not to vary in length, given the resolution of the data, more than 3% from 29 to 30 m (Fig. 11.10). The linear anomalies in Reticulate 2 to the south form rhomboids with perimeters of between 104 and 106 m. The anomalies perpendicular to the bounding anomaly are again separated by about 30  m, but those parallel to the polder dyke are between 21 and 22 m apart. They meet at an angle of 60° (or 120°) (Fig. 11.11). In both patterns, the major component of each linear anomaly is between two and three metres wide and has a negative magnetic character, indicating compacted topsoil or built-up subsoil. There are faint indications in places of an adjoining parallel positive magnetic component, like a narrow filled-in ditch. Some further details of the reticulate patterns point to the aliquot character of the system and possibly to more than one phase of construction. The northernmost sampled part of Reticulate 1 displays a ‘herring-bone’ pattern. The lines perpendicular to the dyke, approximately west–east, seem to be spaced every 15  m, while the intersecting, approximately north–south lines are spaced about every 30 m. This observation suggests either a deliberate division of square units into

Fig. 11.8  Joining and bounding anomalies west of Glas (Google Earth, 2008 data). Note bifurcation in Area K

Fig. 11.9  Bounding and other anomalies west of Glas (yellow–NIR image; cf. Fig. 11.2)

Fig. 11.10  Reticulate 1, magnetometry results and interpretation

152

M. F. Lane and V. L. Aravantinos

Fig. 11.11  Reticulate 2, magnetometry results and interpretation

halves or a palimpsest of two patterns, one offset 15 m to the north or south from the other. Reticulate 2 also conforms well to a simple metrology. Rhomboids of 20  m by 30  m enclose an area two thirds of that defined by 30-m2. Knauss’s speculative interpretation of the field marks, which he knew only from aerial photographs, as settlement traces (‘Siedlungsspuren’) does not  conform to the evidence. No other anomaly in the patterns described suggests residential inhabitation, even temporary; there is no trace of building foundations, storage constructions, hearths, or other typical features. Furthermore, Knauss was partial to the old, loosely topographical, identification of Glas with the Homeric town of Arne, which does not stand up to scholarly scrutiny (De Ridder 1894; Fossey 1988: 408–24).

11.3 ‘Ground-Truthing’ the Character and Nature of the Anomalies Three methods were employed independently to confirm the character and nature of the magnetic anomalies: augering into points above anomalies and adjacent background areas, cleaning and profiling modern ditch sections where they dissect anomalies, and measuring volume-specific magnetic susceptibility at intervals down augered soil cores or ditch sections.

A hand-driven auger fitted with a mud (‘Dutch’) bit removed cores of sediment in stratigraphic sequence. Cores were successfully removed from outside and inside the Revetted Canal, and from above the joining anomaly that connects the canalized rivers with Glas, the northern branch of the bounding anomaly west of Glas, and linear elements of Reticulate 1, as well as from within background areas adjacent to each of these anomalies. Augering generally continued no deeper than two metres below present ground level or until an impenetrable coarse-clast layer or groundwater was encountered, reaching into subsoil in any  case. Ditches were about two to two and a half metres deep. Depths were measured with a plumb-line and tape measure from a levelled horizontal line spanning the trench. A threemetre-wide section was drawn, so as to comprise all or most of the anticipated features. This width also permitted greater horizontal stratigraphic detail than augering. Both stratigraphically arranged core segments and strata (soil horizons) in the ditch sections were recorded as soil profiles (according to the US standards found in Soil Survey Division 1993: 59–196). There was a clear difference between every test (anomaly) and control (background) soil profile in a pair, there being in the former  a layer of silty sediment up to 40  cm thick above the subsoil horizons and generally at the base of the plough zone. The underlying subsoil, judging from its

11  Tracing the Remains of a Late Bronze Age Field System in Central Mainland Greece

153

Fig. 11.12  Soil profiles from Reticulate 1, indicating feature corresponding to anomaly vs background (where χ indicates magnetic susceptibility)

gastropod shell content and, in one case, gleyish lowest horizon, represented ancient lakebed deposits (Fig. 11.12). The horizons corresponding to the linear anomalies were light grey to light greyish brown, occasionally with faint mottles of darker sediment. The major components of these horizons resembled those of the underlying sediment, albeit without the characteristic structures and mottling of subsoil (B) horizons. Eluviated (E) horizons were attenuated or missing. These observations about the anomaly horizons, or ‘features’, as they became known, were confirmed by a greyish lens of like thickness in the ditch profiles. Furthermore, in at least one profile, there was an abutting dark brown stratum on the feature’s eastern edge (Figs.  11.13 and 11.14). The presence of limestone cobbles and boulders on the edges of modern fields, as well as of cobbles in one ditch profile, confirm previous investigators’ hypothesis that certain features in the plain were shored up with stone (Fig. 11.15; Knauss 1984: 213–27, 1987: 207–18; Lauffer 1979). In short, magnetometry appears to have revealed linear deposits of lakebed sediment, sometimes combined with cul-

tural or non-cultural material, which may have been paralleled, at least in places, with a fill of darker (more topsoil-rich) sediment, such as could have arisen from erosion into an abandoned adjacent ditch. Measurement of magnetic susceptibility down the profiles not only confirmed the negative or positive nature of the anomalies’ corresponding features relative to background or overlying and underlying horizons but also how subtle the differences in susceptibility were, proving that the extraordinarily weak contrast in the magnetometry data was not an issue of operator or instrument error. The profiles inside and outside of the Revetted Canal (not illustrated) confirmed that no sediment correlating to a fill existed, but rather that the bottom of the plough zone (c.  50–65  cm deep), at the level of the truncated retaining walls, was the subsoil into which their foundations must have been built. The presence of water flowing between them could have resulted in the different subsoil profiles. That inside the canal has a distinctive light yellowish brown horizon, appearing eluviated, from about 50 to 80  cm, which could be the remains of the raised channel’s bottom. In other

154

M. F. Lane and V. L. Aravantinos

Fig. 11.13  Ditch profile from Reticulate 1, feature corresponding to negative anomaly labelled ‘2’ and possible ancient infilled ditch, corresponding to adjacent positive anomaly, labelled ‘3’

Fig. 11.14  Ditch profile from bounding anomaly west of Glas, corresponding feature labelled ‘2’ and possible ancient infilled ditch or ditches, corresponding to adjacent positive anomaly, labelled ‘3’ and ‘4’

words, in contrast with the magnetometric evidence of the Peripheral Canal, the course of the Revetted Canal was evidently above the level of the ancient lakebed, not cut into it. Unfortunately for detailed investigation of the Peripheral Canal, augering into both it and its background area was inhibited by the high concentration of gravel and oblong cobbles in both the topsoil and upper subsoil in areas where the Canal had been detected with magnetometry. The profiles (not illustrated) nonetheless contrasted with each other, that above the Canal deeper and with more gravel than that

above the embankment to its west. It is worth noting that this coarse-clast content could be due to periodic flooding from the seasonal stream whose alluvial fan is just to the east of the Peripheral Canal, between Nisí and Fteliá (Fig. 11.2) – the very flooding that this canal was probably excavated to abate. The rough edges and apparently uneven depth of the Peripheral Canal in plan in the magnetometry data suggest that it was dredged or repaired in a series of pits (Fig. 11.4), perhaps reflecting a persistent problem for LH engineers, if it was not excavated in this manner in the first place.

11  Tracing the Remains of a Late Bronze Age Field System in Central Mainland Greece

155

Fig. 11.15  Ditch profile from Reticulate 1, feature labelled ‘2’ corresponding to anomaly (possible displaced paving stone in stratum ‘1’ above it)

11.4 Age of the Features Crucial to our argument is that the radiometric analyses of the features confirm their LH date (Table 11.4). Samples of sediment were taken for accelerator mass spectrometry (AMS) dating of their organic carbon-isotope component from both features in soil cores and overlying, underlying, or adjacent horizons. Optically stimulated luminescence (OSL) dating, with complementary thermoluminescence (TL) dating, was also carried out on the sediments of which the features in ditch profiles were composed. The working assumption was that radiocarbon dates would reflect the terminus post quem of the features, because they consist of recently culturally deposited nearby lakebed material, and that the luminescence dates would reflect the terminus ante quem of the features’ construction, because the last ‘zeroing event’ (bleaching out of inherent radiation) happened with the burial of material during deposition. In the case of one linear feature making up part of Reticulate 1, the median radiocarbon date is calibrated BC 1708 and 1855 (95% probability, or ‘2 sigma’) or calibrated BC 1786 and 1837 (68% probability, or ‘1 sigma’). Luminescence dates of the same in one of the ditch profiles (LUM 316/13 and LUM 318/13), measured in years before 2013 (1 sigma) are 1637 ± 330 or 1687 ± 320 BC (OSL) and 1277  ±  200 or 1257  ±  200  BC (TL). The analyst remarked that the TL dates of calcite are typically 15–35% lower than calendar age. Thus the mean TL dates could be in the range of 1724–1468 or 1697–1446 BC. Moreover, in the case of extreme fluctuation in groundwater since mod-

ern drainage and irrigation began, these luminescence dates should be lowered by 10–30  years per millennium (N.  Zacharias, Uni Peloponnese, pers. comm. 2013). Hence the OSL dates could correspond to calendar years c. 1600–1528 or 1650–1578 and the TL dates to between 1714/1687 and 1430/1416  BC.  In any case, not only are the OSL and TL dates of this feature closely consistent with each other, but they are also consistent with the premised chronological bracketing, meaning that construction could have taken place between the last few decades of the eighteenth century BC and the first few of the seventeenth century. The OSL dates from the same linear feature, though on the opposite bank of the ditch, are curiously higher, in or around the twenty-third century BC. Unfortunately, no corresponding TL dates are available. As noted above, there is no such anomalous radiocarbon date from this feature, so it is quite possible that the test sample included older material underlying the feature. Two OSL dates from a single profile of the northern branch of the bounding anomaly west of Glas are nearer to the LH era, although they are separated from each other by several centuries  – 1837  ±  280 or 1126 ± 200 BC – the latter of which would be some 60 years into the post-palatial period, according to conventional chronology. The calcite TL date is mysteriously low, although bringing to bear the typical adjustments given above, the date range 1043–882 BC is close to the lower OSL date from this profile. One possible explanation is that there is intrusive contamination of part of the profile (e.g. evidence of root turbation in Fig. 11.14). Another is that post-palatial repair

156

M. F. Lane and V. L. Aravantinos

Table 11.4  AMS Radiocarbon and luminescence dates from selected features and backgrounds Sample name/loc. Reticulate pattern * 2011I2-01

Lab. no.

Conv. 14C

Cal. date (1 σ)

Cal. date (2 σ)

Beta-­331307

3440 ± 40 BP

BC 1880–1660 BC 1650–1640 Intercept BC 1740

2013I2-01

Beta-­371126

3530 ± 30 BP

BC 1860–1850 BC 1770–1720 BC 1720–1690 Intercept BC 1740 BC 1900–1880 BC 1840–1820 BC 1800–1780 Intercept BC 1880

Pit-like Anomaly, Transect B1 2010B1-01

Beta-­301995

6470 ± 30 BP

BC 5480–5460 BC 5400–5390 Intercept BC 5470

BC 5480–5370 Intercept BC 5470

Vrystika Sinkhole 2011VK-01

Beta-­331308

3340 ± 30 BP

BC 1682–1610 Intercept BC 1620

BC 1720–1720 BC 1690–1530 Intercept BC 1620

190 ± 30 BP

AD 1660–1680 AD 1740–1760 AD 1760–1800 AD 1940–1950+ Intercept AD 1670 AD 1780 AD 1800 AD 1940 AD 1950

AD 1650–1690 AD 1730–1810 AD 1920–1950+ Intercept AD 1670 AD 1780 AD 1800 AD 1940 AD 1950

*

Control 1 (A2 horizon, above feature) 2013J1-01 Beta-­371125

BC 1940–1770 Intercept BC 1880

Sample name/loc. Reticulate pattern 1 * 2011I2-P01

U.o.P. lab code

Geological dose

Dose rate

Age BP (1 σ)

LUM 316/13

2011I2-P01

LUM 317/13

Quartz 3.47 ± 0.30 Calcite (TL) 2.605 ± 0.13 Quartz 3.55 ± 0.30 Calcite (TL) 2.65 ± 0.12 Quartz 4.12 ± 0.37 Quartz 4.15 ± 0.45

0.95 ± 0.07 0.80 ± 0.07 0.96 ± 0.07 0.81 ± 0.07 0.96 ± 0.07 0.96 ± 0.07

3650 ± 330 (1637 ± BC) 3290 ± 200 (1277 ± BC) 3700 ± 320 (1687 ± BC) 3270 ± 200 (1257 ± BC) 4250 ± 380 (2237 ± BC) 4333 ± 470 (2317 ± BC)

1.05 ± 0.08 0.90 ± 0.07 0.96 ± 0.07

3850 ± 280 (1837 ± BC) 2780 ± 250 (767 ± BC) 3140 ± 200 (1127 ± BC)

*

2011I2-P02 LUM 318/13 2011I2-P02 LUM 319/13 Bounding anomaly running from area J into area K * 2011J1-P01 LUM 320/13

Quartz 4.04 ± 0.17 Calcite (TL) 2.50 ± 0.20 2011J1-P01 LUM 321/13 Quartz 3.01 ± 0.25 * Date overlapping with present definition of the Late Helladic Period

and reuse of part of this feature took place, whether for its original purpose or not. We now can attempt to bracket the sequence of events around Glas during the last few centuries of the Bronze Age. The construction and primary inhabitation of Glas has been cross-dated by ceramic seriation – a robust and precise system for the Eastern Mediterranean in the Late Bronze Age – to the LH IIIB, c. 1300–1190 BC. This chronology is supported at its lower end by a radiocarbon date (Beta-­ 412470) lately obtained from the burnt grain of Glas’ storeroom H1 (G.  Jones, Uni Sheffield, pers. comm. 2015): calibrated BC 1225–1045 (2σ, 1210–1110, 1σ; 2940  ±  30 BP) with the earliest calibration intercept at 1185 (others at 1155, 1145, and 1125). This span means that any possible

repair and reuse of the feature corresponding to the bounding anomaly, perhaps to direct water to a system of cultivated fields, would have post-dated the demise of Glas’ administrative centre. However, Glas would also appear to have been built nearly 400 years later than the earliest features. Large stretches of the extant drainage system would likewise seem to be later. Archaeologists in the early twentieth century noted but did not collect ‘Minyan Ware’ pottery at Glas (Wace and Thompson 1912: 193) typical of the period from the MH III (and earlier) through the LH I (sometimes into the LH II; Sarri 2010), i.e. c. 1750/1700–1610/1500  BC (see Table 11.1). Hence it is not out of the question that the irrigation and drainage system, from its earliest to its latest phase, spans the period from LH I through LH IIIB.

11  Tracing the Remains of a Late Bronze Age Field System in Central Mainland Greece

157

11.5 Other AROURA Investigations

previously recognised. There are also specimens of coarse wares for storage, which, judging from the fabrics, could be AROURA conducted systematic field walking for collection either of EH or MH date (K. Sarri, pers. comm., 2014). Also of portable archaeological finds from the ground surface and represented, in much smaller quantities, are potsherds of an intensive surface collection at the aforementioned settle- Late Archaic or Classical (c. 600–300 BC) and Late Classical ment site of AMP. The former was carried out on two-­metre-­ or Hellenistic (c. 400–150 BC) date, as well as a few pieces wide traverses across 30-m magnetometry grid squares of Medieval or early modern lead-glazed and tin-glazed above and adjacent to features, while the latter was realized wares, possibly Catalonian/Frankish, Late Byzantine, or in two-metre units within three grid squares mapped onto Ottoman (Ch. Loizou, pers. comm. 2015). Only one sherd AMP (225 units per 30-m2). Field walking above and beside can positively be identified to the Geometric Period features in the plain recovered little at all from mown and (c. 1050–720  BC). Significantly, the LH material includes ploughed fields (Fig. 11.16a, b), relative to other surface sur- decorative motifs typical of the LH IIIA1/2 phase veys in Greece, and even less of plausibly LH date: 43 finds (c. 1400/1360  BC; Fig.  11.17), immediately preceding the in 16,200 m2 around the Revetted Canal, or 0.14 find per tra- conventional date of Glas’ construction and inhabitation, verse on average; 839 finds in 9000 m on the lake-­ward side and the LH IIIB2/C Early (c. 1190  BC) transition to the of the polder dyke, or 5.6 finds per traverse on average (cf. post-palatial phase, as well as possibly of the  LH IIB figures in Bintliff et al. 2007: 183–312). This dearth is con- (c. 1500/1430–1440/1390  BC) early palatial phase sistent with the hypothesised extensive agricultural regime (S. Vitale, pers. comm. 2014). of the era, which would not have entailed intensive artificial Ceramic evidence thus  indicates that the most intensive manuring or middening (Pettegrew 2001; Bintliff et  al. phase of occupation of AMP is that from the MH through the 2002). Moreover, even if the ancient topsoil is entirely bur- LH, consistent with the radiometric dates of the features in ied, ground-truthing has turned up no artefacts that suggest the polder. The presence of robbed-out stone-lined cists at these practices. the east end of and on the saddle below AMP, constructed in AMP had previously been identified, through cursory a manner, so far as one can tell from extant remains, typical inspection of its surface, as a site with MH and LH phases, as of adult graves of the MH–LH transition (Nordquist 1990, well as possibly Ceramic Neolithic (c. 6100–3100  BC), Whittaker 2014: 82–92), further corroborates the equation of Early Helladic (EH) I–II (3100–2200), Geometric (c. 950– intensive inhabitation with the creation and development of 720), Archaic (c. 720–480  BC), and Classical (c. 480– the polder. 330  BC) Period phases (Fossey 1988: 277–90; Farinetti 2011: 127–35, 305–13). Intensive surface collection clarified this assessment. About 22% of all the potsherds collected 11.6 Reconstruction of the Hydraulic there (280 of 1255) could be identified by decoration, shape, Systems, and Bronze Age or paste and fabric to the Helladic Age, particularly to the Comparanda MH and LH. This amount accounted for about 94% of the tentatively datable pottery. In addition to specimens of LH 11.6.1 Hypothetical Reconstruction (Mycenaean palatial) decorated ware, there were specimens of fine MH brown burnished (‘Minyan’), Matte-painted, The various lines of engineering and chronometric evidence and Mainland Polychrome wares. Indeed, the MH indicate that the drainage system of the Kopais, whose major component may be larger and more important than components have been treated thoroughly elsewhere (Knauss

Fig. 11.16  Field walking and intensive surface collection areas in AROURA survey area: (a) Area H (left) and Areas D and E (right, ESRI/ DigitalGlobe image); (b) AMP (right, AROURA/Worldview-2 image)

158

Fig. 11.17  Reconstruction drawing of a LH IIIA1/2 stemmed goblet (kylix) from AMP

M. F. Lane and V. L. Aravantinos

1984, 1987, 1990: 168–225), was designed to claim land from the Kopaic Lake’s margins for agricultural ends, rather than for residential settlement (Fig.  11.18). The gigantic food stores inside the Cyclopean walls of Glas support such a reconstruction. AROURA both contributed further detail to the water level control mechanisms, an issue that previous investigators had not resolved to their satisfaction, and it revealed what is almost certainly a system of irrigated fields, more likely than ‘settlement traces’ à la Knauss, if architectural residential habitation is understood. By the LH IIIB, at least, the Melas and Kephissos flowed through Cyclopean channels raised some two metres above the present plain (Kountouri et al. 2012; Mamassis et al. 2016), which is taken as approximately their elevation above the ancient lakebed. Once the rivers had been diverted away from the centre of the lake and dykes were built, their waters could be permitted into the polder with floodgates – an object of future archaeological field research  –  both to irrigate crops (or prevent over-drainage) and to provide fresh water for Glas. The seasonal floods that once would have made the area of the polder a marshland (Aronis 1963: 7–14; Griffiths et al. 2002; cf. Paus. 9.38.6), wetter still in longer cycles (Durand-­Claye 1888; cf. Thrphr. C.P. 5.12.3, H.P. 4.11.2), could have been controlled through the Revetted Canal and possibly the sump

Fig. 11.18  Provisional reconstruction of LH drainage and irrigation system around Glas (Google Earth 2014 base image)

11  Tracing the Remains of a Late Bronze Age Field System in Central Mainland Greece

of the Peripheral Canal. The bounding and joining features, appearing to consist of a basic cut-and-fill levee plus ditch, evidently topped or shored up in places with stone, would have been the primary and secondary feeders of the reticulate pattern of fields, easily serviced from atop the parallel, likely reinforced, causeways. The rectilinear field system itself seems to be subdivided with similar features, and it is possible that simple floodgates could further have regulated irrigation inside this network. Should water levels have become too high in the feeder canals, they could have been reduced through connection with the conjoined Peripheral and Revetted Canals alongside Fteliá, emptying into the Vrystiká Sinkhole, as the bounding anomaly west of Glas suggests. They could even have emptied through gates in the dyke – another object for future fieldwork. Even if the Cyclopean extents of the canalization of the Melas–Kephissos were not finished until the LH IIIB, it is still very possible that a less elaborate arrangement of earthen dykes, such as some investigators have conceived (Knauss 1990: 145–67; Mamassis et  al. 2016), diverted the waters toward the sinkholes in the north-eastern bay of the Kopais and supplied water to feeder canals within the polder (or, at least, to seasonally dry land). This would account for the radiometric dates several centuries earlier than the LH IIIB fortified administrative complex and storage centre of Glas. As remarked above, the nearby settlement site of AMP was inhabited most intensively during the later MH and throughout the LH, and Glas may yet prove to have a more substantial early Mycenaean phase than previously recognised. It is not out of the question that LH features were restored and reused in later periods, but there is no evidence that the basic system is anything but Helladic. Knauss believes he has observed Archaic Period walls crossing certain inlets of the north-eastern bay, though the polygonal masonry described could also be Hellenistic (c. 330–150 BC; Knauss 1987: 243–78, 1990: 236–42). The ancient geographer Strabon and subsequent historians record that Alexander the Great (356–23 BC) hired a certain Krates of nearby Khalkis to dig a trench across (diataphreusai) the Kopaic Basin to drain it, so that peripheral cities could benefit from claimed land, but that the drainage project was unfinished and came to naught because of political infighting after the death of Alexander among the beneficiaries to be (Str.9.2.16–18, D.L.4.23). The two nearest member cities of the erstwhile Boiotian League were Kopai and Akraiphia (Bakhuizen 1994), the latter about five kilometres due south-east of Glas, with a view over the easternmost bay of the Kopais. Previous investigators have persuasively argued that the wide canal made of straight segments many kilometres long, visible in aerial photographs and satellite data running from near Ághios Dhimítrios in the western Basin to the Vrystiká Sinkhole is Krates’ work (Knauss 1984: 217; 233–36; Lauffer 1979), as could be the 16 vertical shafts spaced every

159

few hundred metres from the Bínia Sinkhole northeast of Glas into the Kefalári Pass, though not completing the traverse to the cave’s outlet in the Bay of Larymna (Kambanis 1892–1893; Philippson 1894). Epigraphical evidence exists of Caligulan or Neronian era restoration of a dyke of uncertain date across the bay south of Akraiphia. It seems to have fallen quickly into disuse and does not hint at anything so long-lived and systematic as the LH constructions (IG 7.2712–13; Boatwright 2000: 112–17). The second-century AD  Emperor Hadrian financed the building of aqueducts and irrigation works for the cities on the west side of the Kopais. However, as in Alexander’s time, these fell quickly into disuse after Hadrian’s death (AD 138), particularly because of questions of fiscal responsibility for their upkeep (SEG 35.405; Boatwright 2000: 112–17; Oliver 1989: 253–72). The antiquarian and travel writer Pausanias, who sailed from Akraiphia to Kopai at about this time, makes no mention of dykes, canals, or irrigated fields in use (Paus.9.24.1). No archaeological or historical evidence can be found of any drainage or irrigation undertakings in later periods. Wet rice cultivation may have taken place in marshland near the town of Livadheiá to the west of the Basin in Ottoman times (fifteenth–seventeenth century AD), nothing in the existing Ottoman taxation records indicates any such cultivation in the vicinity of Kopai, then ‘Topólia’ (Kiel 1997). At the opposite chronological pole, Diodoros Sikeliotes, Strabon, and subsequent historians preserve the tradition of the city of Orkhomenos having extensive land, that the ‘place now occupied by Lake Kopais is said formerly to have been dry land [anepsukhthai “to have been exposed to fresh air”] and was cultivated in every sort of way, being subject to the Orkhomenians, who lived close by; and indeed this is adduced as evidence of their wealth’ (Str.9.2.40), until the Theban champion Herakles plugged up the caverns in order to exact revenge against the Orkhomenians (Apollod.2.4.11, Diod.Sic.4.18.6, Paus.9.38.6– 8, Polyaen.1.3.5). Orkhomenos’ wealth in agriculture is adumbrated in Homer’s comparison of the riches that flow toward it to those that flow toward Thebes in Upper Egypt (Iliad 9.381–82). As remarked above, the reticulate patterns of features are consistent with Mycenaean metrology (see Table  11.2 and relevant text). Linear B records show that the areas of cultivated fields were measured equivalent to their seed corn (L.R.  Palmer 1963: 96–99; Ventris and Chadwick 1973: 393–94). The largest unit is GRA, probably a measure of between 96 and 120  litres of grain (Chadwick 1966; R. Palmer 1989; Ventris and Chadwick 1973: 55–56), but a much more common measure, which Mycenologists denote as T, is one tenth of this amount, equivalent to perhaps 1600– 3000 m2 (0.16–0.30 ha) of land, per previous, independent estimates of sowing densities (in other words, a quadrilateral 40 to 55 metres on each side; Lane 2009; Ventris and

160

Chadwick 1973: 393–94). While these dimensions are greater than those of the largest such rectangles in the reticulate patterns, they are of the right order of magnitude for an area that can be ploughed in a day, representing perhaps double the adduced sowing density (or half the inferred ploughing rate) expected of a field whose dimensions can be described as 30-metre-wide square  in our model. An area 30  m (or 100  ft) on a side is very close to the size of a Classical plethron or modern Greek strémma (Burford 1993: 113, Psychoghios 1995: 24–25), comparable also to the Roman actus of 35 by 35 metres, according to certain descriptions (Col.Rust.5.16). The fractions of Mycenaean T are denoted V = 1/6 and Z = 1/24 (i.e. V ¼). Hence the ‘herring-bone’ pattern in the northern part of Reticulate 1 could represent area V 3 (= T ½), and the rhomboids in Reticulate 2 area V 4 (= T 2/3). Alternatively, in keeping with the expected sowing and ploughing rates, each square in Reticulate 1 could represent V 3, just as two Roman actūs make up a iūgerum, the ‘herring-bone’ Z 6 = V 1½ = T ¼, and the rhomboids Z 8 = V 2 = T 1/3.

M. F. Lane and V. L. Aravantinos

reverse of his vengeful act against Orkhomenos. Knauss, with the support of certain philologists, thinks that the association of the place name Orkhomenos with two separate LH engineered landscapes is no coincidence but rather refers directly to the protection of the settlement from flooding or the improvement of the land for human purposes (Knauss 1991; Lane 2016; Lauffer 1974). Feats of hydraulic engineering are known also from Middle and Late Bronze Age Crete (Middle–Late ‘Minoan’), though they are of a different scale and character than those evident on the mainland. The most famous of these is the system of stone and terracotta aqueducts in and around the Minoan courtyard complexes (‘palaces’), such as Knossos (Angelakis et  al. 2007; Gorokhovich et  al. 2011). More similar in principle to the hydraulic engineering in the Kopaic Basin and elsewhere on the mainland is the series of stream-­ diverting, anti-erosion dams on the island of Pseira off Crete’s north coast (Betancourt and McCoy 2012). The only place suitable for a drainage system of mainland scale is the endorheic plateau of Lasíthi in eastern central Crete, where, so far, no evidence of such has turned up (Watrous 1982). There are also the soil retention terraces around such Minoan 11.7 Cases for Comparison outposts as that on Kythera off the coast of the Peloponnese, some of which date back to the Bronze Age (Bevan et  al. Knauss and Zangger, separately and together, have identified 2013; Krahtopoulou and Frederick 2008). and endeavoured to explain the technology of similar and The notable  differences between mainland and Cretan relevant feats of LH engineering in mainland Greece (Knauss engineering, in light of well attested Minoan influence on 1984, 1987, 1990, 2001; Zangger 1994; Zangger et al. 1997). many aspects of Mycenaean building, invites comparison These include dykes, polders, and Cyclopean channels in with adjacent regions of the Eastern Mediterranean during other karst poljes, stream redirection and a possible artificial the Middle and Late Bronze Age, if not earlier, to find formal port below the palace at Pylos, and the already well known parallels or cultural connections and to answer questions of flood protection dam by the palace at Tiryns. Knauss technological development: Does the technology emerge especially devoted over two decades of his academic career from a long indigenous tradition, or is it a calque on some to studying the evidence of LH drainage and flood-control other tradition or technological development? The Hittites engineering. He found systematic parallels to the drainage and their neighbours in Anatolia built with a method works around Glas and in the north-eastern bay, just west of reminiscent of Cyclopean masonry and also undertook major Glas, in an apparent polder between Strovíki and Kástro works of hydraulic engineering. Hittite defensive and (Knauss 1984: 222–24, 1987: 178–82), as well as in semi-­ retaining walls famously have a sloping glacis, at least in closed hydric environments (mainly other poljes) in southern their lower courses, sometimes reinforced with earth Boiotia, in Phokis to the west, and in Arkadia in the (MacQueen 1975: 103–04; Nossov 2008: 55), which is not a Peloponnese (Knauss 1990: 221–22, 2001). Probably the pronounced feature of mainland Greek constructions (Loader most convincing of his reconstructions are those below 1998). They are also typically casemate-built, and the walls Arkadian Orkhomenos and in the Pheneos Valley polje, of Troy Phase VI (eighteenth–thirteenth century BC) are at which the authors  have examined, all of which involve least skeuomorphs of this method (Nossov 2008: 9; 14; redirection of streams into sinkholes and the building of Sagona and Zimansky 2009: 235; 267; Wright 2005: 192). polder dykes to prevent periodic re-flooding (Knauss 2001: Helladic Cyclopean walls, in contrast, are generally built in a 15–22). In both cases, the method of construction of the ‘pseudo-casemate’ fashion, consisting of abutting open-­ dykes or dams is similar, including Cyclopean stones, and is ended sections (Loader 1998; Wright 2005). These things not dated to any post-Bronze Age period by ancient authors noted, excavated and exposed sections of retaining walls in or modern archaeologists (Paus.8.13.4, 8.14.2; Salavoura the Kopaic system have a slight slope away from vertical, at 2015: 96–99). Curiously, the Roman Republican Period poet least the lower courses of which are shored up with loam and Catullus (Carmen 68b) alludes to a tale (source unknown) of sealed with clay (Knauss 2001: 36–40; Kountouri et al. 2012; Herakles channelling the Pheneos streams into sinkholes, the Mamassis et al. 2016). Moreover, one section of the retaining

11  Tracing the Remains of a Late Bronze Age Field System in Central Mainland Greece

wall of the Melas–Kephissos, as it nears Strovíki, whose terminus post quem is LH IIIA2, is clearly built in a casemate style (Aravantinos et al. 2006; Kountouri et al. 2012). One wonders whether this is due to local technological tradition or the exigencies of building in a wetland, including preventing seepage into the rubble core of the wall sections. One salient difference, given the current state of archaeological evidence, between Anatolian hydraulic engineering of the Middle and Late Bronze Age and that most clearly attested in mainland Greece in the same period is that the former seems mainly aimed at creating cisterns and reservoirs (Çınaroğlu and Çelik 2010: 342–52; MacQueen 1975: 80; 110; 189), gathering seasonal rains, perennial streams, and groundwater, while the latter seems to be designed mainly to create polders (Knauss 2001). The Helladic systems therefore more closely resemble those of the Middle Bronze Age (Middle Kingdom) and later Egypt, especially in the Fayyum Depression (Lake Moiris), where dykes protected settlements such as el-Lahun and constrained the shape of the lake, and floodgates regulated both the flow of water from the Nile into the closed system and the level of water in the irrigation canals (Flinders Petrie 1890; Kemp 2006: 211–21; Quirke 2005: 40–55). The Middle Kingdom engineering in the Fayyum, expanded and extended in the Late Kingdom (LBA), would have provided a precedent in the East Mediterranean Basin no later than the end of the twentieth century BC (Callender 2000: 152–54; 157; David 1986: 40–41), at least a century earlier than the radiocarbon terminus post quem of the features around Glas in the Kopaic Basin. The hoard of Minoan silverware at el-Tod in Upper Egypt indicates substantial contact with the Aegean in this period (Bisson de la Roque 1953), suggesting communication of technologies was possible. The reservoirs at the Hittite capital of Hattusa in central Anatolia can date to no earlier than the mid-sixteenth century (Erkul et al. 2007), a century later than the earliest luminescence terminus ante quem for the creation of the field system around Glas. The Hittite king Tudhaliya IV recorded that he had several dams built or repaired during his reign in the last third of the thirteenth century BC, including the well known extant example at Alacahöyük north of Hattusa (Bryce 2005: 323– 25; Çınaroğlu and Çelik 2010: 342–52). Given the current state of relevant evidence and its chronology, a plausible hypothesis is that the drainage systems of Helladic Period Greece could represent both an indigenous local solution to a drainage problem, emerging over several centuries, and, especially in its last phases, a representation of a regional architectural and engineering koine, much as are grey metal-­ skeuomorph burnished fine wares (e.g. Minyan and Anatolian Grey Ware) around the Northern Aegean from the middle of the MBA to the beginning of the LBA (Pavúk 2010). This hypothesis nevertheless begs a fundamental explanatory question.

161

11.8 Why the Hydraulic Engineering? It seems to us that there are four likely explanations for the development of the drainage and irrigation systems in the region, which are not necessarily mutually exclusive. Two of these would be contingent on a longer chronology than is generally conceived. The first of these is that the land was claimed for subsistence from the later MH onwards in response to environmental constraints on demographic growth in the lower Kephissos Valley, near Orkhomenos. Recent settlement pattern data indicate that Boiotia did not suffer the depopulation from the end of the Early Helladic II to the middle of the MH II (c. 2200–1800 BC) that southern mainland Greece did (Bintliff 2010; Farinetti 2011). Unlike its later rival Thebes, 35 kilometres away, Orkhomenos had only a narrow arable hinterland that presumably would eventually have encountered territory claimed by MH settlements upriver, such as Kápraina and Ághios Vlásios in the Kephissos Valley (Fossey 1986: 63–67; 1988: 375–82). Hence seasonal reclamation of the lakebed margins, perhaps initially used as summer pasture, as the Sarakatsani did in modern times (Dasios 2010: 81–130), took place first, accompanied piecemeal by dykes or dams for diverting water and canals for drainage and irrigation (Halstead 2014: 277–81). Littoral settlements without any immediate arable hinterland, such as AMP north-east of Glas, could thus be sustained. As the engineering became more systematic – that is elaborate and integrated between farming communities – a central administration could have developed to assure its construction, repair, and maintenance, as well as the distribution of foodstuffs among littoral settlements. Such a scenario raises serious questions about the development and singular trajectory of state-level institutions in ‘Mycenaean’ Greece from the LH IIB (discussion in Bintliff 2010; Galaty and Parkinson 2007). Given the amount of human and animal labour required, especially in the latest phases of the drainage works, one might also hypothesise that more political– economic motives existed from the end of the MH. Establishing a regime of extensive cultivation, albeit one sustained by technological intensification (including animal labour), could have ‘freed’ human labour for other economic activities. As the classic southern mainland Mycenaean states emerged, vying with each other for a place in the Eastern Mediterranean network of diplomatic gift exchange and fealty through prestige goods (Galaty and Parkinson 2007), northern Boiotian communities may have felt the need to compete in order to stay independent, devoting more labour to full-time or seasonal crafting and seasonal repairs of hydraulic works. They may eventually have tried to corner the market in the grain trade from Greece, grain that we know by the end of the thirteenth century is sorely needed in parts of the Hittite Empire (Bryce 2002: 254–56; 2005: 322).

162

M. F. Lane and V. L. Aravantinos

Table 11.5  Hypothetical annual round of deployment of labour for LH extensive agriculture and hydraulic maintenance (cf. du Boulay 1974: 274–77) Bold: Human labour-saving (employment of animal labour) Italic: Labour-intensive, redeployment of work gangs () JANUARY Pruning vineyards  Dispensation of olive oil? APRIL Tapping pines for resin  Springtime construction and maintenance of hydraulic works (with traction animals)  JULY Harvesting grain 

OCTOBER Animals return from summer pasture (e.g. sheep and goats) Ploughing (with traction animals) Dispensation of grain, dried figs, and currants or young wine for winter?

FEBRUARY Pruning vineyards  Hoeing vineyards  MAY Planting tree seedlings  Propagating vines  AUGUST Late summer construction and maintenance of hydraulic works (with traction animals)  Processing grain (threshing with traction animals) NOVEMBER Olive harvest  Sowing grain (broadcast)

A still more ideological–political hypothesis may plausibly be advanced. Surplus labour relative to claims to land could have existed already by the end of the MH, be it the result of vagaries of local climate, poor agricultural strategic decisions, coercive means, or any of the other factors of inequality that have been posited for the Final Neolithic (c. 4500  BC) onward (Halstead 1995, 2014: 57–60, 318–19, 351–53; Halstead and O’Shea 1982). If this was the case, then it may have been important to the wealthy and well connected to devise projects that would both mitigate the effects of inequality of means and ensure social solidarity. Whatever labour could not be engaged in household or palace manufactories and directly compensated with subsistence goods (see Melena 1983; Killen 1999, 2001; Nosch 2006), could instead be put both to the execution of engineering projects in season and to sharing in the land and crops through their concerted efforts (Table  11.5). While this hypothesis need not preclude the use of surpluses for long-­distance exchange too, one would expect greater evidence of consumption in nearby settlements of the narrow range of cultivars represented at Glas than if these settlements were sustained by a more diversified and intensive  agro-pastoral regime. The hypothesis also suggests that precedent had been set for claiming land from waters, perhaps the sort of gradual approach described above but more ancient still, to which the powers that  there  were could have pointed to justify their projects. Finally, assuming the shortest timeline, is the hypothesis that the major phase of the drainage and irrigation systems was a response (evidently failed by the twelfth century) to the effects of climatic or ecological change. These effects could be local, regional, or supra-regional. It is conceivable that the limited good arable soils in northern Boiotia had been over-exploited – either lacking sufficient fertilisation or

MARCH Tapping pines for resin  Hoe vineyards  JUNE Harvesting grain  Some animals to summer pasture (e.g. sheep and goats) SEPTEMBER Late construction and maintenance of hydraulic works (with traction animals)  Fig harvest  Grape harvest  DECEMBER Olive harvest 

having suffered considerable erosion into the Kopaic, or both – resulting in the secondary effect of silting up the river courses and karstic sinkholes, augmenting the lake and reducing available farmland, while hill farming may not have been adequately managed  – for example, with terracing (Kvapil, this volume)  – resulting in some of the same disastrous consequences. The jury is still out on whether cooling and drying of the climate was a factor in the devolution of the palace economies of the Late Bronze Age in the Aegean. However, the end of the fourteenth or beginning of the thirteenth century BC, the latest phase of construction of the drainage works and the fortress of Glas, would be quite early for significant effects in any model (Moody 2005; Kaniewski et al. 2015; Knapp and Manning 2016; Rohling et al. 2009). Economically disruptive droughts in more distant places, such as Anatolia or Cyprus, in the later thirteenth century (Kaniewski et al. 2013) could have provided an inducement to expand and intensify the system so as to supply food for export to allies. Intensification could have increased the risk of depletion and mineralization of soils, decreasing the productivity of the land. Until the details of the drainage and irrigation system are better understood, it is impossible to know if subtler problems of over-irrigation or, inversely, desiccation and aeolian erosion took place in the polder (Bureau of Reclamation 1993: 26–27; Lane, field notes; Soil Survey Division 1993: 192–96). There is certainly no evidence yet of catastrophic failure, such as the breaching of a dyke. Hence the required detailed and integrated archaeological research plan for the Helladic Kopais is already sketched out, involving further chronometry, palaeo-­environmental and subsistence studies, geomorphology and sedimentology, and surface survey. In summary, investment in drainage and irrigation may have accompanied the population expansion right through

11  Tracing the Remains of a Late Bronze Age Field System in Central Mainland Greece

the Bronze Age, as more land was required for subsistence. Two further factors of this population expansion could be pre-existing political territorial claims and relatively permanent unequal distribution of landed property. These might have provided political incentives for the few to maintain their economic advantages while enhancing social solidarity, with respect to internal and external rivals, through subsistence redistribution tied to collective labour projects. The population would also have had to respond, ultimately inadequately, to the ecological effects of their agricultural intensification through land acquisition. In any case, whether these activities and events transpired over the longer or shorter term, they comprise processes that contrast with those commonly described for the emergence political economic complexity and the Mycenaean states to the south. Only further research can tell. Acknowledgements The authors thank Dr Alexandra Charami, Superintendent of Antiquities of Boiotia during the AROURA study period 2012–2014, Prof Emeritus J. Knauss of the Technical University of Munich, who contributed original notes and plans from his research 1984–2001, Dr T. Horsley of Northern Illinois University, AROURA’s Principal Geophysical Investigator, assisted by Ms A. Cuneo, American Schools of Oriental Research, Mr W.S. Bittner, AROURA Information Technologies Specialist, and, not least, the Institute for Aegean Prehistory for its generous grants from 2010 through 2014.

References Angelakis, A.  N., Savvakis, Y.  M., & Charalampakis, G. (2007). Aqueducts during the Minoan era. Water Science and Technology: Water Supply, 7(1), 95–101. Aravantinos, V. L., Kountouri, E., & Fappas, Y. (2006). Το μυκηναϊκό αποστραγγιστικό σύστημα της Κωπαΐδας: νέα δεδομένα και πρώτες εκτιμήσεις. In Αρχαία Ελληνική Τεχνολογία: 2ο Διεθνές Συνέδριο/Ancient Greek Technology: 2nd International Conference (pp. 557–564). Athens: Τεχνική Επιμελητήριο Ελλάδας. Aravantinos, V.  L., & Lane, M.  F. (2010). Συνεργασία της ΑΣΚΣΑ με τη Θ′ ΕΠΚΑ για Γεωφυσική Έρευνα σε Τμήμα της Κωπαῒδας (Πρόγραμμα AROURA). Αρχαιολογικόν Δελτίον, Μέρος Β′ 2 Χρονικά, 65, 995. Aronis, G. (1963). Notes for a field study tour of karst groundwater investigation of areas of Greece, 21–25 March 1963. Athens: U.N. Food and Agriculture Organisation. Aspinall, A., Gaffney, C., & Schmidt, A. (2008). Magnetometry for archaeologists. Landham: Altamira Press. Bakhuizen, S. C. (1994). Thebes and Boeotia in the fourth century B.C. Phoenix, 48(4), 307–330. Betancourt, P.  P. (2007). Introduction to Aegean art. Philadelphia: INSTAP Academic Press. Betancourt, P. P., & McCoy, F. W. (2012). Dams and water management systems of Minoan Pseira. Philadelphia: INSTAP Academic Press. Bevan, A., Conolly, J., Colledge, S., Frederick, C., Palmer, C., Siddall, R., & Stellatou, A. (2013). The long-term ecology of agricultural terraces and enclosed fields from Antikythera, Greece. Human Ecology, 41, 255–272. Bintliff, J. L. (2010). The middle Bronze Age through the surface survey record of the Greek mainland: Demographic and sociopolitical

163

insights. In A. Philippa-Touchais, G. Touchais, S. Voustaki, & J. C. Wright (Eds.), Mesohelladika: The Greek mainland in the middle Bronze Age (pp. 755–763). École Française d’Athènes: Athens. Bintliff, J. L., Howard, P., Snodgrass, A. M., & Dickinson, O. T. P. K. (2007). Testing the hinterland: The work of the Boeotia survey (1989–1991) in the southern approaches to the city of Thespiai. Cambridge: McDonald Institute of Archaeological Research. Bintliff, J.  L., Farinetti, E., Howard, P., Sarri, K., & Sbonias, K. (2002). Classical farms, hidden prehistoric landscapes and Greek rural survey: A response and an update. Journal of Mediterranean Archaeology, 15(2), 259–265. Bisson de la Roque, F. (1953). Le Trésor de Tôd. Cairo: Institut Français d’Archéologie Orientale. Boatwright, M. T. (2000). Hadrian and the cities of the Roman empire. Princeton: Princeton University Press. du Boulay, J. (1974). Portrait of a Greek mountain village. Oxford: Clarendon Press. Bryce, T. (2002). Life and society in the Hittite world. New  York: Oxford University Press. Bryce, T. (2005). The kingdom of the Hittites. New  York: Oxford University Press. Bureau of Reclamation Staff. (1993). Drainage manual: A water resources technical publication. Washington, DC: U.S. Department of the Interior. Burford, A. (1993). Land and labor in the ancient Greek world. Baltimore: Johns Hopkins University Press. Callender, G. (2000). The Middle Kingdom renaissance (c. 2055– 1650 BC). In I.  Shaw (Ed.), The Oxford history of ancient Egypt (pp. 137–171). Oxford: Oxford University Press. Chadwick, J. (1966). The olive oil tablets of Knossos. In J. Chadwick & L. R. Palmer (Eds.), Proceedings of the Cambridge colloquium on Mycenaean studies (pp. 26–32). Cambridge: Cambridge University Press. Çınaroğlu, A., & Çelik, D. (2010). Atatürk & Alacahöyük. Ankara: Ekici Form Ofset. Dasios, P.  Dh. (2010). Η «στράτα» των νομάδων Σαρακατσαναίων Αγραφιωτών και το πολιτισμικό ίχνος της. PhD dissertation, Panteion University of Social and Political Sciences, Athens. David, A. R. (1986). The pyramid builders of ancient Egypt: A modern investigation of pharaoh’s workforce. New York: Routledge. De Ridder, A. (1894). Arné? Bulletin de Correspondance Hellénique, 18(1), 446–452. Dean, A. J. (1937). The Lake Copais, Boeotia, Greece: Its drainage and development. Journal of the Institution of Civil Engineers, 5(5), 287–304. Durand-Claye, M.  A. (1888). Rapport sur le desséchement du lac Copaïs Grèce. Paris: Imprimerie Nationale. Erkul, E., Hüser, A., Stümpel, H., & Wunderlich, T. (2007). Combined geophysical survey of an ancient Hittite dam: New and old high-­ tech. In A. Posluschny, K. Lambers, & I. Herzog (Eds.), Layers of Perception. Proceedings of the 35th International Conference on Computer Applications and Quantitative Methods in Archaeology (CAA), Berlin, 2.–6. April 2007 (pp. 1–5). Bonn: Dr. Rudolf Habelt GmbH. Farinetti, E. (2011). Boeotian landscapes: A GIS-based study for the reconstruction and interpretation of the archaeological datasets of ancient Boeotia. Oxford: Archaeopress. Flinders Petrie, W.  M. (1890). Kahun, Gurob, and Hawara. London: Kegan Paul, Trench, Trübner and Co. Fossey, J.  M. (1986). The ancient topography of Eastern Phokis. Amsterdam: J.C. Gieben. Fossey, J.  M. (1988). Topography and population of ancient Boiotia. Chicago: Ares Publishers. Galaty, M.  L., & Parkinson, W.  A. (2007). Introduction: Mycenaean palaces rethought. In M.  L. Galaty & W.  A. Parkinson (Eds.), Rethinking Mycenaean palaces II: Revised and expanded second

164 edition (pp.  1–17). Los Angeles: Cotsen Institute of Archaeology, University of California, Los Angeles. Gorokhovich, Y., Mays, L., & Lee, L. (2011). A survey of ancient Minoan water technologies. Water Science and Technology: Water Supply, 11, 388–399. Griffiths, S.  J., Street-Perrott, F.  A., Holmes, J.  A., Leng, M.  J., & Tzedakis, C. (2002). Chemical and isotopic composition of modern water bodies in the Lake Kopais Basin, Central Greece: Analogues for the interpretation of the lacustrine sedimentary sequence. Sedimentary Geology, 148, 79–103. Halstead, P. (1992). Agriculture in the Bronze Age Aegean: Towards a model of palatial economy. In B. Wells (Ed.), Agriculture in Ancient Greece (pp. 105–117). Svenska Institutet i Athen: Stockholm. Halstead, P. (1995). From sharing to hoarding: The Neolithic foundations of Aegean Bronze Age society? In R.  Laffineur & W.-D.  Niemeier (Eds.), Politeia: Society and state in the Aegean Bronze Age. Proceedings of the 5th international Aegean conference, University of Heidelberg, Archäologisches Institut, 10–13 April 1994 (pp. 11–20). Liège: Université de Liège. Halstead, P. (2014). Two oxen ahead: Pre-mechanized farming in the Mediterranean. Malden: Wiley/Blackwell. Halstead, P., & O’Shea, J. (1982). A friend in need is a friend indeed. In C. Renfrew & S. Shennan (Eds.), Ranking, resource and exchange: Aspects of the archaeology of early European society (pp. 92–99). Cambridge: Cambridge University Press. Higgins, M.  D., & Higgins, R. (1996). A geological companion to Greece and the Aegean. London: Duckworth Publishers. Iakovidis, S. E. (1992). The Mycenaean Fortress of Gla. In J. T. Killen & J.-P. Olivier (Eds.), Mykenaïka: Actes du IXe Colloque international sur les textes mycéniens et égéens organisé par le Centre de l’antiquité grecque et romaine de la Fondation hellènique des recherches scientifiques et l’École française d’Athènes (pp.  365– 380). Paris: l’École française d’Athènes. Iakovidis, S. E. (1998). Γλας ΙΙ: Η Ανασκαφή, 1981–1991. Athens: Εν Αθήναις Αρχαιολογικής Εταιρείας. Iakovidis, S. E. (2001). Gla and the Kopais in the 13th Century B.C. Athens: Εν Αθήναις Αρχαιολογικής Εταιρείας. Jones, G. (1995). Charred grain from the Late Bronze Age fortress of Gla, Boiotia. Annual of the British School at Athens, 90, 235–238. Kalcyk, H. (1984). Die Kopais in historischen and archäologischen Quellen. In J.  Knauss, B.  Heinrich, & H.  Kalcyk (Eds.), Wasserbauten der Minyer in der Kopais: die älteste Flußregulierung Europas (pp.  13–70). Munich/Obernach: Technische Universität München. Kambanis, M. L. (1892–1893). Le dessèchement du lac Copais par les anciens. Bulletin de Correspondance Hellénique, 16–17(121–37), 322–342. Kaniewski, D., Guiot, J., & Van Campo, E. (2015). Drought and societal collapse 3200 years ago in the Eastern Mediterranean: A review. WIREs Climate Change, 6(4), 369–382. Kaniewski, D., Van Campo, E., Guiot, J., Le Burel, S., Otto, T., & Baeteman, C. (2013). Environmental roots of the Late Bronze Age crisis. PLoS One, 8(8), e71004. Kemp, B.  J. (2006). Ancient Egypt: Anatomy of a civilization. New York: Routledge. Kiel, M. (1997). The rise and decline of Turkish Boeotia, 15th–19th century. In J. L. Bintliff (Ed.), Recent developments in the history and archaeology of Central Greece: Proceedings of the 6th international Boeotian conference (pp. 315–358). Oxford: Archaeopress. Killen, J. T. (1999). Mycenaean o-pa. In S. Deger-Jalkotzy, O. Panagle, & S. Hiller (Eds.), Floreant Studia Mycenaea: Akten des X. internazionalen mykenologischen Colloquiums in Salzburg vom 1.–5 Mai 1995 (pp.  325–341). Vienna: Österreichische Akademie der Wissenschaften. Killen, J.  T. (2001). Some thoughts on ta-ra-si-ja. In S.  Voustaki & J.  T. Killen (Eds.), Economy and politics in the Mycenaean pal-

M. F. Lane and V. L. Aravantinos ace states: Proceedings of a conference held on 1–3 July 1999 in the Faculty of Classics, Cambridge (pp.  161–180). Cambridge: Cambridge Philological Society. Knapp, A.  B., & Manning, S.  W. (2016). Crisis in context: The end of the Late Bronze Age in the Eastern Mediterranean. American Journal of Archaeology, 120(1), 99–149. Knauss, J. (1984). Systemanalyse der antiken Wasserbauten. In J. Knauss, B. Heinrich, & H. Kalcyk (Eds.), Die Wasserbauten der Minyer in der Kopais: die älteste Flußregulierung Europas (pp. 167– 242). Munich/Obernach: Technische Universität München. Knauss, J. (1987). Die Melioration des Kopaisbeckens durch die Minyer im 2. Jt. v. Chr. Munich: Technische Universität München. Knauss, J. (1990). Wasserbau und Geschichte: Minysche Epoche– Bayerische Zeit. Munich: Technische Universität München. Knauss, J. (1991). Arkadian and Boiotian Orchomenos, Centres of Mycenaean hydraulic engineering. Irrigation and Drainage Systems, 5, 363–381. Knauss, J. (2001). Späthelladisch Wasserbautern: Erkundungen zu wasserwirtschaftlichen Infrastrukturen der mykenischen Welt. Munich: Technische Universität München. Kountouri, E., Petrochilos, N., Koutsoyiannis, D., Mamassis, N., Zarkadoulas, A., Vött, A., Hadler, H. & Willerhäuser, T. (2012). A new project of surface survey, geophysical and excavation research of the Mycenaean drainage works of the North Kopais: The first study season. In: Editorial Staff, IWA Specialized Conference on Water and Wastewater, 22–24 March 2012 (pp. 467–476). London: IWA Publishing. Krahtopoulou, A., & Frederick, C. (2008). Landscapes: Polycyclic terracing from Kythera Island, Greece. Geoarchaeology, 23(4), 550–585. Lane, M. F. (2009). From da-mo to δῆμος: Survival of a Mycenaean land allocation tradition in the classical era? In C.  Bachhuber & G.  Roberts (Eds.), Forces of transformation: The end of the Bronze Age in the Eastern Mediterranean (pp.  109–116). Oxford: British Association for Near Eastern Archaeology and Oxbow Books. Lane, M.  F. (2010). Πρόγραμμα AROURA. Αρχαιολογικόν Δελτίον, Μέρος Β′ 2 Χρονικά, 65, 996–1004. Lane, M. F. (2012). Landholding at Pa-ki-ja-na: Toward spatial modeling of Mycenaean agricultural estates. Pasiphae, 6, 59–115. Lane, M. F. (2016). Four probable terms in Mycenaean Greek pertaining to water management, some possible references, and broader implications. Kadmos, 55(1/2), 83–129. Lane, M.  F., & Aravantinos, V.  L. (2012). Archaeological reconnaissance of unexplored remains of agriculture (AROURA): Interim report, 2010–2011 campaigns. Teiresias, 42(1), 2–26. Lane, M.  F., Horsley, T.  J., Charami, A., & Bittner, W.  S. (2016). Archaeological geophysics of a Bronze Age agricultural landscape: The AROURA project, central mainland Greece. Journal of Field Archaeology, 41(3), 271–296. Lauffer, S. (1974). Orchomenos. Realencyklopädie der classischen Altertumswissenschaft, RE Supplementum (p.  14). Stuttgart: Metzler. Lauffer, S. (1979). Topographische Untersuchungen. Αρχαιολογικόν Δελτίον, Μέρος Β′ 2 Χρονικά, 29(1973–1974), 449–454. Loader, N.  C. (1998). Building in cyclopean masonry: With special reference to the Mycenaean fortifications on mainland Greece. Jonsered: P. Åström. MacQueen, J. G. (1975). The Hittites and their contemporaries in Asia minor. Boulder: Westview Press. Mamassis, N., Moutsakas, S., & Zarkadoulas, N. (2016). The operation of ancient reclamation works at Lake Copais in Greece. Water History, 7(3), 271–287. Manning, S. W., Ramsey, C. B., Kutschera, W., Higham, T., Kromer, B., Steiner, P., & Wild, E. M. (2006). Chronology for the Aegean Late Bronze Age 1700–1400 B.C. Science, 312, 565–569.

11  Tracing the Remains of a Late Bronze Age Field System in Central Mainland Greece Melena, J.  L. (1983). Further thoughts on Mycenaean o-pa. In A.  Heubeck & G.  Neumann (Eds.), Res Mycenaeae: Akten des VII.  Mykenologischen Colloquiums in Nürburg vom 6–10 April 1981 (pp. 258–286). Göttingen: Vandenhoeck & Ruprecht. Moody, J. (2005). Drought and the decline of Mycenae, updated. In A. Dakouri-Hild & S. Sherratt (Eds.), Autochthon: Papers presented to O.T.P.K. Dickinson on the occasion of his retirement (pp. 126– 133). Oxford: Archaeopress. Nordquist, G. C. (1990). Middle Helladic burial rites: Some speculations. In R. Hägg & G. C. Norquist (Eds.), Celebrations of death and divinity in the Bronze Age Argolid. Proceedings of the sixth international symposium at the Swedish Institute at Athens, 11–13 June 1988 (pp. 35–41). Svenska Institutet i Athen: Stockholm. Nosch, M.-L. B. (2006). More thoughts on the Mycenaean ta-ra-si-ja system. In M. Perna (Ed.), Fiscality in Mycenaean and Near Eastern archives: Proceedings of a conference held at Sopraintendenza Archivistica per la Campania, Naples, 21–23 October 2004 (pp. 161–182). Paris: De Boccard. Nossov, K. S. (2008). Hittite fortifications. Oxford: Osprey Publishing. Oliver, J.  H. (1989). Greek constitutions of early Roman emperors from inscriptions and papyri. Philadelphia: American Philosophical Society. Palmer, L. R. (1963). Interpretation of Mycenaean Greek texts. Oxford: Clarendon Press. Palmer, R. (1989). Subsistence rations and Pylos and Knossos. Minos, 24, 89–124. Papadopoulos, A. (1997). The socio-economic effects of the drainage of Lake Copais. In J.  L. Bintliff (Ed.), Recent developments in the history and archaeology of Central Greece: Proceedings of the 6th international Boeotian conference (pp. 365–377). Oxford: Archaeopress. Pavúk, P. (2010). Minyan or not? The second millennium Grey Ware in Western Anatolia and its relation to mainland Greece. In A.  Philippa-Touchais, G.  Touchais, S.  Voustaki, & J.  C. Wright (Eds.), Mesohelladika: The Greek mainland in the Middle Bronze Age (pp. 931–943). École française d’Athènes: Athens. Pettegrew, D.  K. (2001). Chasing the classical farmstead: Assessing the formation and signature of rural settlements in Greek landscape archaeology. Journal of Mediterranean Archaeology, 14(2), 189–209. Philippson, A. (1894). Der Kopais-See in Griechenland und seine Umbegung. Zeitschrift für Erdkunde zu Berlin, 29, 1–90. Psychoghios, D.  K. (1995). Προίκες, φόροι, σταφίδες και ψωμί: οικονομία και οιογένεια στην αγροτική Ελλάδα του 19ου αιώνα. Athens: Εθνικό Κέντρο Κοινωνικών Ερευνών. Quirke, S. (2005). Lahun: A town in Egypt 1800 BC, and the history of its landscape. London: Golden House Publications.

165

Rohling, E., Hayes, A., Mayewski, P.  A., & Kucera, M. (2009). Holocene climate variability in the eastern Mediterranean, and the end of the Late Bronze Age. In C. Bachhuber & G. Roberts (Eds.), Forces of transformation: The end of the Late Bronze Age in the Mediterranean (pp.  2–5). Oxford: British Association for Near Eastern Archaeology and Oxbow Books. Sagona, A., & Zimansky, P. (2009). Ancient Turkey. New  York: Routledge. Salavoura, E. (2015). Μυκηναϊκή Αρκαδία: Αρχαιολογική και Τοπογραφική Θεώρηση. Athens: ΗΟΡΟΣ. Sarri, K. (2010). Orchomenos IV: Orchomenos in der mittleren Bronzezeit. Munich: Verlag der Bayerischen Akademie der Wissenschaften/Verlag C.H. Beck. Soil Survey Division Staff. (1993). Soil survey manual. Washington, DC: U.S. Government Printing Office. Shelmerdine, C.  W. (2008). Background: Sources and methods. In C. W. Shelmerdine (Ed.), The Cambridge companion to the Aegean Bronze Age (pp. 1–18). New York: Cambridge University Press. Stussi, N. (2011). Pléiades Constellation: Capability overview. Presentation at the 2011 JACIE conference, Boulder, CO. https://calval.cr.usgs.gov/JACIE_files/JACIE11/Presentations/WedPM/235_ Stussi_JACIE_11.135.pdf Threpsiadis, I. (1965). Ανασκαφή Άρνης (Γκλα) Κωπαΐδος. Πρακτικά της εν Αθήnαις Αρχαιολογικής Εταιρeίας του Έτους, 1959, 21–25. Ventris, M., & Chadwick, J. (1973). Documents in Mycenaean Greek (2nd ed.). Cambridge: Cambridge University Press. Wace, A. J. B., & Thompson, M. S. (1912). Prehistoric Thessaly: Being some account of recent excavations and explorations in North-­ Eastern Greece from Lake Kopais to the Borders of Macedonia. Cambridge: Cambridge University Press. Warren, P. M., & Hankey, W. (1989). Aegean Bronze Age chronology. Bristol: Bristol Classical Press. Watrous, L. V. (1982). Lasithi: A history of settlement on a Highland plain in Crete. American School of Classical Studies: Princeton. Whittaker, H. (2014). Religion and society in Middle Bronze Age Greece. New York: Cambridge University Press. Wright, J. C. (2005). Offsets in Mycenaean architecture. In A. Dakouri-­ Hild & S.  Sherratt (Eds.), Autochthon: Papers presented to O.T.P.K. Dickinson on the occasion of his retirement (pp. 191–199). Oxford: Archaeopress. Zangger, E. (1994). Landscape changes around Tiryns during the Bronze Age. American Journal of Archaeology, 98, 189–212. Zangger, E., Thompson, M.  E., Yazvenko, S.  B., Kuhnke, F., & Knauss, J. (1997). The Pylos regional archaeological project, part II: Landscape evolution and site preservation. Hesperia, 66(4), 549–641.

Reconstructing Enclosed and Parcelled Out Landscapes from the First Millennium BC in Himmerland, Denmark: Arable Fields, Grazing Land and Settlement Patterns Examined in Three Micro-regions

12

Michael Vinter

Abstract

For the Himmerland area in north-­ eastern Jutland, Denmark, extensive later prehistoric field systems have been mapped from aerial photographs. Looking in detail at the archaeological remains preserved, historical maps and palaeo-­ ecological data for three micro-­ regions, it could be estimated that as much as 70% of the landscape was once covered by field systems comprising lynchets and banks. Palynological analyses indicate the expansion of open landscapes with extensive grassland and minor, scattered, stands of trees between 1200 and 300 BC, after which heather heathlands take over – hinting at a notable change in grazing strategies and a change from arable to outfield. Agricultural usage during the Late Bronze Age and Early Iron Age (1000–50 BC) was also indicated by increasing sediment deposition in peat deposits. Older barrows and contemporary Early Iron Age settlements were integrated into the vast systems of banks and lynchets. By the middle of the Pre-­Roman Iron Age (c. 300 BC), cultivation became concentrated in smaller areas, closer to the now nucleated settlements. Keywords

Denmark · Himmerland · Celtic fields · Land use

M. Vinter (*) Moesgaard Museum, Højbjerg, Denmark e-mail: [email protected]

12.1 Introduction Air photographs and plans of ancient fields are a unique and as yet unused source for knowledge about the development of the cultural landscape through the Iron Age. The developments of the last 25 years have made it ever more clear what huge archaeological resources still lie preserved in Denmark. Former Keeper of National Antiquities, Steen Hvass, in Digging into the Past (1993: 194)

Across North-­ western Europe traces of ancient enclosed field systems can be found on marginal lands either preserved as lynchets and low field banks or as soil marks on aerial photographs. These are often referred to as “Celtic fields”. They can cover several hundred hectares, but their original size is difficult to estimate. Furthermore, the genesis, development, dating of the field systems, their relationship to contemporary settlements and impact on the landscape are difficult to grasp. This chapter tries to answer some of these questions through an analysis of the archaeological record, historical maps and pollen analyses in three micro-­­ regions in Himmerland, Denmark in which prehistoric field systems have been mapped from aerial photographs. The study suggests that as much as 70% of the landscape was enclosed in the Late Bronze Age-­Early Iron Age, and that a marked change in land use took place c. 300  BC.  At that time, large parts of the field systems were left to give way to heathland and grassland. This coincides with a shift in the settlement pattern from wandering single farms to nucleated settlements. The field systems of the first millennium BC and their use have mainly been studied as a separate field from general settlement archaeology. This is due partly to the difficulties involved in determining their original extent, which for a long time was only based on Gudmund Hatt’s investigations and surveys undertaken in the 1930s. But it is also a consequence of something that has protected them for 2000 years,

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 S. Arnoldussen et al. (eds.), Europe’s Early Fieldscapes, Themes in Contemporary Archaeology, https://doi.org/10.1007/978-3-030-71652-3_12

167

168

namely their peripheral location in the landscape relative to present-­day villages, which means that they have rarely been subject to archaeological excavations. The archaeological excavations of the last 50 years have expanded our knowledge of the internal structure, size and position in the landscape of Early Iron Age settlements but have not equally provided much new information on their relations to the presumed field systems of the time. The remains of these field systems have been the subject of several groundbreaking studies since the 1950s. These studies are primarily represented by the work of Viggo Nielsen, who surveyed the field systems in the forests of eastern Denmark and recorded evidence of ploughing in the Store Vildmose bog (Nielsen 1984: 135–63, 1993, 2000b). His work demonstrates that the prehistoric field systems were not solely a Jutish phenomenon, but occurred all across Denmark, and that some were established already in the Late Bronze Age. Both archaeological and pedological investigations have subsequently been undertaken at several other localities with ancient field systems, such as Alstrup Krat (Bech 2003), Øster Lem Hede (Nielsen and Dalsgaard 2017) and recently three field systems in Eastern Jutland (Nielsen et al. 2019). In Alstrup Krat, elevated soil phosphate values were found as a consequence of manuring, and cultivation surfaces from the Bronze Age were discovered below the Iron Age field systems (Bech 2003). Even though several studies undertaken in Denmark, the Netherlands and Northwest Germany in the 1960s and 1970s demonstrated that the use of aerial photography, especially on sandy soils, could increase the number and extent of known prehistoric field systems considerably, this method has remained virtually unexploited in Denmark.1 Danish studies have generally been based on a single series of aerial photographs, with mapping of the observed soil marks being the exception rather than the norm, and only few surveys subsequently being published. A study of three prehistoric field systems in Himmerland shows that examination of multiple aerial photo series can generate a detailed picture of their morphology and extent. Traces of earthen banks and lynchets, which represented the divisions between the individual fields or plots, are still visible in many places despite 150 years of cultivation. But in most of those areas where cultivation has been practised since the twelfth century, these traces have been erased (Vinter 2011: 65–90). Judging from the dimensions of the field systems, they constituted a striking and dominating component of the Himmerland landscape in the first millen-

M. Vinter

nium BC, when they also incorporated the imposing monuments of preceding prehistoric periods, for example megaliths and Bronze Age barrows (Løvschal 2014a, 731). The field systems are particularly important for Danish archaeology as they constitute the only prehistoric example of a coherent cultivation system prior to the introduction of ridge-­and-­furrow cultivation at the beginning of the Viking Age. They therefore provide a unique opportunity to reconstruct a complete prehistoric landscape and the human influence on, and exploitation of, this landscape. Land use and the relationship between arable fields and pastures in the Early Iron Age has been the subject of several Danish studies (Ejstrud and Jensen 2000; Hvass 1985; Lewis 1985: 123–60; Nielsen 1998: 271–92). These have involved the mapping of meadows (i.e. grazing land) and potential arable land within a naturally delimited resource area on the basis of historical maps. Most often, these studies have focused on determining how much arable and grazing land each farmstead had access to. The starting point has always been a particular known settlement, with an analysis of the land use in the landscape being a secondary consideration. In the study of Himmerland reported in this chapter, the field systems and the exploitation of the landscape have been the primary focus, with the incorporation of the settlements into the landscape being addressed subsequently. No ideal object, i.e. a totally excavated settlement and a surveyed field system, was available for study. The analysis of the links between settlement and field system was therefore based on analogies to the archaeological record for Himmerland in general and on the settlements recorded in other field systems in Himmerland, Denmark and The Netherlands. The chapter starts with a short introduction to the data sources and methods applied, and is followed by an analysis of the land use based on pollen diagrams. Then the topography, settlement history and land use patterns are studied and evaluated in three micro-­regions and general trends are summarized. In the last section of the chapter, analogous data of the regional settlement structure is discussed and linked to land use history in the micro regions and compared with known examples of the relationship between field systems and settlement from Denmark and the Netherlands.

12.2 Sources and Methods

To investigate land use patterns and reconstruct the prehistoric agricultural landscape in Himmerland, three micro-­­ regions or resource areas were selected: Gundersted Hede, Skørbæk Hede and Store Binderup (Fig. 12.1). These areas 1  Denmark: Jeansson 1963; Newcomb 1971; Olesen 1981, 1983; were chosen due the presence of well-­preserved prehistoric Sørensen 1975, 1979, 1982, 1991; Stoumann Hansen anf Sørensen field systems that have been mapped in detail on the basis of 1984. The Netherlands: Brongers 1976. North-­ west Germany: Zimmermann 1979. Research overviews were produced by Müller-­ multiple aerial photographs, the availability of pollen data Wille 1965: 1979 and particularly thoroughly by Klamm 1993. from the immediate vicinity of the fields and because the

12  Reconstructing Enclosed and Parcelled Out Landscapes from the First Millennium BC in Himmerland, Denmark: Arable Fields…

169

Fig. 12.1  Location of the three micro-­regions and the two pollen diagrams. 1. Skørbæk Hede, 2. Gundersted Hede, 3. Store Binderup, 4. Navnsø pollen diagram, 5. Borremose pollen diagram. Bronze Age settlements (green squares) and Early Iron Age settlements (blue squares). Sandy soils (yellow), clay soils (brown), sandy deposits in wetlands and on raised seabed (green). (By the author)

fields can, to a reasonable extent, be defined topographically. The delimitation of the field systems was undertaken primarily on the basis of historical maps. The maps show the landscape 150–200  years ago, prior to the great drainage and reclamation projects gathering impetus in the late nineteenth century, and before which the areas of bog and meadow had probably remained more or less unchanged since the Bronze Age.2 The general development of the landscape and its vegetation history is revealed by two pollen diagrams from sites located within 5  km of the micro-­ regions: Navnsø and Borremose (Fig. 12.1, 4–5). The influence that the establishment of Early Iron Age field systems had on the landscape was investigated by examining the relative proportions of trees, grass and other vegetation through time in the pollen studies. This information was combined with the character and extent of the mapped field systems within delimited resource areas. Settlements in the micro-­regions and in the rest of Himmerland, were examined on the basis of data from the Danish national register of Sites and Monuments from 2008.3 This database is not yet fully updated with earlier records from the area, and is by no means error-­free, but it was considered adequate for a more general study such as this. It was, however, necessary to obtain supplementary information from aerial photographs with respect to unrecorded barrows and Iron Age settlements. Several different maps have been employed: A digitised version of the Royal Society maps from the late eighteenth century constitute the main foundation, but the first cadastral maps from around the same time and the first edition ordnance maps from c. 1880 can also contribute several details. They can also be used in a critical analysis of the field systems recorded on aerial photographs and of land use in historical times. 3  http://www.kulturarv.dk/fundogfortidsminder/ [accessed 2019]. 2 

Within the micro-­regions, the primary aim of the analysis of the archaeological record was to explore land use patterns in a long-­term perspective, putting the situation in the Early Iron Age into perspective, while a study of the settlement traces in other recorded field systems would give an idea of how the Bronze Age and Iron Age settlement related to the field systems. As these settlement traces are fragmentary, and therefore not representative of Early Iron Age settlement in Himmerland, it was important to conclude the study with a consideration of how well-­investigated localities from the same period in Denmark and north-­western Europe compare with and aid in the interpretation of the Early Iron Age landscape in the three micro-­regions.

12.3 T  he Vegetation History of Western Himmerland The pollen diagrams describe general trends in the composition of the tree component of the vegetation and the relationship between woodland and grassland within a given area over time. As the diagrams reveal trends, it is necessary to evaluate the human impact on the landscape within a given period in relation to the preceding and subsequent periods. The vegetation history of the Store Binderup micro-­region has been investigated in the Borremose bog, which lies approximately 2 km away. Several of the pollen cores were found to have been disturbed by peat cutting, but one was complete (Jensen 1977: 96–119). The core covers approximately a period from 5000 BC until 500 AD. Woodland is dominant until 2200 BC, and from c. 2200–500 BC, the forest clearance is continuous and grass and herb pollen increases. These developments culminate in the middle of the Pre-­Roman Iron Age and are subsequently maintained at

170

this level. The maxima for grass and herb pollen coincide with the highest values of ribwort plantain and a marked, brief peak in sheep’s sorrel, indicating an open, grazed landscape. Pollen values of both plants then fall, as heather expands massively. Coincident with the grass maximum, a sand component was identified in the peat, indicating an open landscape with an element of soil/sand drift. An interesting feature of this pollen diagram is the large amounts of heather pollen (Calluna vulgaris). Unfortunately, it is not possible to determine whether this comes from plants growing on the surface of the bog or from a wider area in the surroundings. The rise in Calluna pollen clearly takes place later than the increase in grass and herbs, at around 1000  BC.  Calluna pollen values show a marked peak at around 300 BC, after which they appear to decrease slightly, though they remain at a high level. It is clear that heather expands across an open and cultivated landscape with relatively few trees, but it is difficult to determine whether it forms part of a cultivation/grazing strategy due to the methodological/interpretational problems inherent in determining the source(s) of the heather pollen. This situation is resolved by the pollen diagram from Lake Navnsø, which lies 3–4 km south of the micro-­regions at Gundersted and Skørbæk Hede, and which therefore gives a picture of the vegetation history associated with the establishment, use and abandonment of these field systems (Fig.  12.2). The pollen diagram reveals the vegetation history for almost the entire Holocene: from 8000  BC to the

M. Vinter

present day (Odgaard 1999: 7–17). The development of the landscape around Lake Navnsø is almost identical to that evident at Borremose. Woodland dominates until c. 2400 BC, after which it declines, and grasses and herbs expand, reaching a maximum at c. 300 BC and maintaining this level until the present day. The occurrence of ribwort plantain is greatest in the Bronze Age (1800–500 BC), and this indicates the presence of an open landscape with extensive grassland and minor, scattered stands of trees. As was the case at Borremose, heather expands slowly from the beginning of the Late Bronze Age (c. 1000  BC) into a grazing landscape that endures for a millennium, before increasing rapidly around 300 BC. The relative proportions of grass/herbs, heather and woodland remain stable until c. AD 1000, when heather expands at the expense of woodland. In the pollen diagrams from Skånsø and Kragsø, which lie closer to the moraine, the development of the vegetation resembles more that evident in Himmerland. At first there was a landscape with open grassland, but in the course of the first millennium BC this was gradually replaced by heather heath. Heather therefore became a permanent element in the landscape of Himmerland from the Pre-­Roman Iron Age onwards. The question then is which factors were responsible for promoting this expansion? Part of the explanation can be found in pollen diagrams from three lakes in western Jutland: Solsø, Skånsø and Kragsø (Odgaard 1994: 137–48; 1998: 14–28). At Solsø, which was not ice-­covered during the last glaciation, heather heathland spread as the forest was

Fig. 12.2  Pollen diagram from Lake Navnsø. (After Odgaard 1999: 13, fig. 5)

12  Reconstructing Enclosed and Parcelled Out Landscapes from the First Millennium BC in Himmerland, Denmark: Arable Fields…

171

cleared from c. 3000 BC, with no preceding open-­grassland phase. Leaching of soil nutrients took place so rapidly that only grasses of low nutrient value could exist here, and these were little better as animal fodder than young heather plants. As heather can also be used for winter grazing, heather heath simply represented a better grazing resource than the grassland. For heather to form new shoots or for its seeds to germinate on dry surfaces requires burning or cutting. In other words, its rejuvenation is reliant on human intervention. Otherwise, after about 30  years, the plants will complete their life cycle and die, and the area will become colonised by woodland. After an episode of burning, grass will dominate for 3–5 years, while heather grows up below it. There is a strong correlation between heather pollen and microscopic charcoal, which indicates that the areas of heather were likely maintained by periodic burning (Odgaard 1994: 137– 48). Heather also has several other potential uses, which makes it a versatile resource. It can be used for winter grazing, fuel and building material and heather turf, mixed with animal dung, is used for soil improvement. The arrival and expansion of heather was perhaps, in the first instance, a response to the poorer grazing on parts of the commons, which can be attributed to a reduction in soil nutrients brought about by intense, long-­term grazing pressure, but this is unlikely to be the only cause. A marked change in cultivation strategy, or perhaps more correctly, grazing strategy, could have been crucial to the expansion of heather. This strategy involved reducing the area of cultivated land and an increased emphasis on, and maintenance of, the heather on the surrounding grazing areas. The common landscape was then, in around 300 BC, swiftly replaced by another kind of grazing landscape, in which heather played a major role. This land use appears to have continued up into historical times. The relative proportions of grassland and heather heath are difficult to estimate, as grassland areas are found within heather heathlands, and we know that the necessary burning of the heather results in a subsequent period with grass cover. The pollen diagrams cannot, unfortunately, tell us anything about the extent of the cultivated areas, as cereal crops, except rye, spread little pollen. Moreover, periods of grass cover would also have been part of the cultivation rotation sequence. For an indication of the extent of the cultivated areas, we therefore need to look at studies of sediment deposition.

sents the result of soil erosion and movement. No investigations have been undertaken of dust deposition in Himmerland, but this aspect has been examined in the Store Vildmose bog in northern Jutland, and Abkær bog in southern Jutland (Aaby 1990: 130–40; 1994: 25–40). At both localities, increasing sediment deposition is observed during the first half of the Pre-­Roman Iron Age (500–250 BC). But while it declines at the very end of the Pre-­Roman Iron Age (c. 100 BC) at Abkær Mose, this does not happen until around AD 500 at Store Vildmose. From AD 800–1050, and after 1500, increasing quantities of dust are seen deposited at both localities. This material must originate from bare, unvegetated surfaces, where there is nothing to retain the mineral grains. According to Bent Aaby, in these inland areas this phenomenon must indicate, first and foremost, the presence of ploughed fields. The increasing values for sediment movement and deposition during the Pre-­Roman Iron Age therefore suggest that larger areas were coming under the plough. Compared with the sediment deposition evident from AD 1500 onwards, levels in the Pre-­Roman Iron Age are significantly lower. The question is whether this deposition of sediments during the Pre-­Roman Iron Age could derive from heather heath and grassland which, as a result of burning and overgrazing, lay exposed to wind erosion. A decrease in sediment deposition is seen at both localities coincident with the regeneration of the woodland. At Abkær Mose, very high sediment deposition is accompanied from AD 1500 onwards by high values of grass and heather pollen, as well as evidence of almost total woodland clearance. The majority of the open areas in the landscape must have been grazing land, while arable fields were probably less significant. The arable fields would, in the most cases, have been sown with winter rye, which would have ameliorated the effects of wind erosion to some degree. The opening of the landscape in the Late Bronze Age and Early Iron Age (1000–50 BC) was also accompanied by increasing sediment deposition. The question is whether variations in the quantities of sediment deposition within the peat can be unequivocally related to increases and decreases in the extent of cultivated areas, or whether, to a greater degree, they reflect the relative proportions of woodland and grazing land? In order to understand how these general trends in the vegetation history and sediment disposition relate to the archaeological record on a more local scale, we will now zoom in on three micro regions.

12.4 C  ultivated land and Sediment Deposition

12.5 The Topography and the Settlement History of Skørbæk Hede

Investigations of soil/sand movement can tell us something about the extent and intensity of arable cultivation. As mentioned previously, sand was identified at Borremose in the peat from the Pre-­Roman Iron Age. This most likely repre-

The micro-­region centred on the field system at Skørbæk Hede lies on a peninsula that extends 7 km north-­south and c. 2.5 km east-­west. It is delimited to all sides by either river valleys or wet meadows (Fig. 12.3). The central part of the

172

Fig. 12.3  The micro-­region at Skørbæk showing the Early Iron Age field system mapped from aerial photographs. Meadow/bog, based on evidence from several historical maps (green), cultivated land on the Royal Society maps (yellow), heath on the Royal. Society maps (light purple), red circle = barrow on aerial photo and undated/Bronze Age barrow from the Sites and Monuments database, blue square = Iron Age settlement. Background topographical map from c. 1880. By author.

peninsula comprises a prominent ridge which dips evenly to the west and south. Large areas of the northern part are cut through by steep ravines, which presumably were not suited to cultivation, but would have been used for grazing. These areas are shown as heath-­covered on historical maps from 1880. The soils throughout the micro-­region consist of fine sand. Knowledge of Palaeolithic and Mesolithic settlement (11,000–4000  BC) is limited to two shell middens on the margins of river valleys. From the Neolithic (4000–1800 BC), there are five megaliths scattered across the southern part of the peninsula, which were in active use during most of the Neolithic. More than 70 barrows have been recorded from the area, and most of these probably date from the Bronze Age (1800–500 BC). They are found primarily on the north-­­ south ridge and on the western side of the peninsula, with the remaining areas remarkably empty of barrows. At least a further 30 barrows of various sizes are visible on aerial photo-

M. Vinter

graphs (Fig. 12.3), with the majority in the southern half of the peninsula, to the west of the ridge. Several of these occur in the boundary banks of the field system and could therefore represent clearance cairns. If the distribution of the barrows is taken as an indication of settlement in the later part of the Neolithic and into the Bronze Age, this appears to have been concentrated in the central and western parts of the peninsula. The same pattern is found repeated in the settlement traces from the Late Bronze Age and Early Iron Age, which all derive from work by the Danish National Museum in the 1920s and 1930s. Since then, no archaeological investigations have been undertaken in this area. The excavated Iron Age settlement at Skørbæk Hede is well-­known in Danish archaeology. It is one of the few examples of a settlement positioned centrally within a field system, where it can be said with certainty that the two are contemporaneous. When it was investigated in 1937, the area had apparently lain untouched for 2000  years, ever since it was abandoned around the birth of Christ (Hatt 1938). The latest phase of the settlement consisted of four houses with turf walls, which were still evident as low, elongated-­oval banks that clearly respected the field boundaries. During the excavation, traces were found of a further six houses, which represented an earlier settlement, with three houses and a few replacements during two main phases. Pottery dates the settlement to 100  BC–AD 100. The many buildings within a small area give the impression of a stable settlement, where it was unnecessary to cultivate the former house sites, with their nutrient-­ rich byre sections. Moreover, the many cobbled areas would have required considerable clearance work prior to a possible return to cultivation. Beneath one of the houses were two pits containing pottery dating from the end of the Late Bronze Age (700–500 BC). Two wells, located between 80 and 300 m north of the settlement, contained ceramic dating to The Early Iron Age (500–300  BC). The settlement traces inside the field system spans most of the first millennium BC. Skørbæk Hede was not occupied in isolation. Evidence suggests that the field system at Skærbæk Hede was cultivated by several settlements and it could well comprise several combined field systems. A rarely mentioned, coeval settlement and burial lie only 1  km to the northeast of the settlement on Skørbæk Hede. An excavation was undertaken at the beginning of the 1920s, in one meter squares. Only one house was investigated, but the settlement is likely to have been a good deal larger.4 This locality lies only 200 m from the area of the field system recorded from aerial photographs. The northern part of the micro-­region has only two recorded settlement traces: one is an unclassified record from 700– 500 BC, and the other is a house from 100 BC–AD 100 with

The file is in the archives of the Danish National Museum.

4 

12  Reconstructing Enclosed and Parcelled Out Landscapes from the First Millennium BC in Himmerland, Denmark: Arable Fields…

a sunken byre end, found on aerial photographs.5 Here too, traces of field banks were found on aerial photographs within a few hundred metres. There are no settlement traces in the area later than the end of the second century AD. The villages of Skørbæk and Ejdrup are first mentioned in written sources in the fourteenth and fifteenth centuries, but the Romanesque church in Ejdrup dates from the middle of the twelfth century (Trap and Nielsen 1961: 1170). There are several fourteenth century records of castle mounds in the nearby river valleys. The location of the villages to the east of the ridge suggests that between 400 and 1100  AD the settlement moved closer to the large area of meadowland in the river valley to the east.

12.6 Land Use at Skørbæk Hede On historical maps from c. 1780, about 750 ha, or approximately 45% of the peninsula is depicted as cultivated land, while the remaining 900 ha was heath. The cultivated areas are situated around the villages of Ejdrup and Skørbæk and by the manor at Halkær Hovedgård. The entire western and northern parts comprise extensive areas of heath and provided common land for the two villages, where they could obtain fuel, turves for soil improvement and winter fodder for the cattle. Comparing the map with the pollen diagram from Navnsø, the relative proportions of grass/herbs and heather heath appear to have remained stable from the Viking Age (750–1050 AD) to the eighteenth century. This historical agrarian landscape differed markedly from the grassland-­dominated landscape of the Bronze Age with its greater numbers of trees and absence of heather. There would have been a clear line of sight between the area’s many barrows, and it is likely that Bronze Age settlement was concentrated around the monuments on the central ridge and in the areas to the west; the arable fields probably also lay here. Around 300  BC, heather expanded at the expense of grassland, and this situation continued until the beginning of the twelfth century. As mentioned above, the nutrient content of the soil was a critical factor in whether an area was covered by grass or by heather. On this basis, one scenario is that persistent grazing pressure and cultivation through the Bronze Age depleted the reserves of soil nutrients on the grassland and brought about a shift from grass to heather. The transition appears, however, to have taken place too rapidly for soil nutrient depletion to have been the only cause. A more plausible explanation is that there was a fundamental change in the grazing and cultivation strategy, and the nature of this change may be elucidated from the evidence for preThe Danish National Museum, Danish Prehistory, archive no. 240/25; Vinter 2009.

5 

173

historic field systems. At Skørbæk, aerial photographs reveal traces of a field system covering 290 ha that was probably established in 700–300 BC (Fig. 12.3). If this field system was associated with a common land use in the landscape and that there was a continual alternation between grassland and arable crops, the heath would not have been able to expand because it has a poor resistance to cultivation. If, however, parts of the field system were taken out of cultivation, the heather would have been better able to expand, and could subsequently have been maintained by periodic burning. It has previously been demonstrated that the traces of field banks and lynchets are generally observed in areas not cultivated in historical times, so the field system must be presumed to have covered significant parts of the areas that came under later cultivation (Vinter 2011: 65–90). On Skørbæk Hede, this is very likely to be the case for the areas of high ground to the west of Skørbæk and a significant part of the historically cultivated areas around Ejdrup. It is also here that there is evidence of continuous cultivation from the Neolithic to the Early Roman Iron Age. Conversely, the heath areas on the northwestern part of the peninsula, together with those to the south and west of Ejdrup, were unsuitable for arable agriculture due to their topography and were probably not cultivated prior to the nineteenth century. There are no traces of either barrows, field banks or lynchets on the aerial photographs of these areas. The absence of barrows to the north, south and east of Skørbæk suggests that this area too was of marginal importance during both the Neolithic and Early Bronze Age. Any calculation of the relative proportions of heath, grassland and field system during later prehistory is rather speculative in nature. However, it is reasonably certain that secure traces of field banks and lynchets are present on 290  ha of the 1400  ha considered to be suitable for arable agriculture, corresponding to 20% of the total. The field system was undoubtedly larger than this. The areas outlined in the central and western parts of the peninsula comprised more than 50% of the potential arable area. Within this 700  ha large area, three presumably stable settlement sites with house structures from around the birth of Christ or the Early Roman Iron Age, have been recorded. These settlements were placed in a landscape, where substantial parts of the field systems had been abandoned, replaced by heath and cereal production concentrated in smaller areas around the settlements. The earlier settlements from Late Bronze Age and Early Pre-­Roman Iron Age (c. 700–300/250 BC) have so far only been documented as dispersed finds of wells and pits, located in a structured landscape divided into field plots dominated by grassland intermixed with cereal cultivation. In all probability, the house structures of this time moved around within the field system on regular intervals.

174

12.7 The Topography and Settlement History Around Gundersted The micro-­region around the field system at Gundersted abuts Skørbæk Hede to the east, being separated only by the Vidkjær Å watercourse, which forms the micro-­ region’s eastern and southern boundary. It is rather more difficult to define topographic boundaries to the west and north (Fig. 12.4). Today, this is Himmerland’s largest contiguous area of heath, comprising Gundersted Hede, Lundby Hede and Ajstrup Hede. To the west, and especially to the northeast, there is some undulating ground, but otherwise the heaths are relatively flat. They are situated on 900 ha of postglacial aeolian sand deposits, while the remainder of the area comprises fluvio-­ glacial sand. The heath areas have a remarkably sparse distribution of barrows, which suggests that they have always been marginal in terms of settlement history. Only in the northwestern and northeastern border regions of the area are barrow groups evident. The central part of Lundby Hede is cut through by several hundred-­ metre-­ wide systems of eroded or sunken trackways, known as holloways, which run north-­south and northwest-­­southeast. In several places on the heath, these holloways cut through areas with field banks and have erased the traces of them. To the southeast of the drift sand area, a low ridge extends through the landscape from northeast to southwest. It slopes evenly down towards the areas around Gunderstedgård and Borup, marked as cultivated land on historical maps. The area’s more than 40 barFig. 12.4  The micro-­region at Gundersted showing the Early Iron Age field system mapped from aerial photographs. Meadow/bog is marked in green, based on evidence from several historical maps. Yellow = cultivated land on the Royal Society maps, light purple = heath on the Royal Society maps, red circle = barrow on aerial photo and undated/Bronze Age barrow from the Sites and Monuments database, blue square = Iron Age settlement. Background topographical map from around 1880. (By author)

M. Vinter

rows are also concentrated on this ridge and are, in many cases, incorporated into the banks and lynchets of the field system. The same is true of several unrecorded barrows that have been found on aerial photographs. Almost all the barrows are undated, but most of them must be from the Late Neolithic/Bronze Age. From the Neolithic there is a damaged megalith, a stone cist and two wetland offerings of axes and amber. The only secure Bronze Age finds are a spearhead found in a stone cist and a stone bearing cup marks in the church dike. By the latter, a small burial cairn from c. 500 BC and a thirteenth century churchyard have also been investigated. Unlike Skørbæk Hede, the traces of Early Iron Age settlements here are restricted to a single locality. In 1989, in the southwestern part of the surveyed field system, two stone pavements were excavated. These measured 2 × 2 and 2 × 5 m, respectively, and were spaced 40 m apart. Between the stones lay pottery that could be dated no more precisely than to the Iron Age. Whether these pavements represent parts of a larger settlement or simply two single farmsteads are unclear, but it seems very likely that they belong to the Early Iron Age and are thereby coeval with the field system. As was also the case at Skørbøk Hede, the settlement traces lie close to a small group of barrows. The later historic settlements, Gunderstedgård and Borup, are mentioned for the first time in written sources in the fifteenth century but building materials of Romanesque character were incorporated into Gundersted church, which burned down at the end of the sixteenth century. The early medieval churchyard shows

12  Reconstructing Enclosed and Parcelled Out Landscapes from the First Millennium BC in Himmerland, Denmark: Arable Fields…

175

that the area was settled from the middle of the twelfth century (Trap and Nielsen 1961: 1245–46). There is also a castle mound at Gunderstedgård, which is probably coeval with the church and churchyard.

tain estimate would be that about 50% of the cultivable areas outside the drift sand area has been covered by the field system. A much more definite picture of the land use is evident 15 km to the south at Store Binderup.

12.8 Land Use at Gundersted

12.9 The Topography and Settlement History at Store Binderup

Land use in historical times at Gundersted and Borup, as derived from the 1780 maps, shows many similarities to the situation at Skørbæk and Ejdrup. The cultivated areas and historical settlements are also situated close to meadowland by the river, while to the west and north there are extensive areas of heathland which could have been a common grazing/resource area. On the heathland, as at Skørbæk and Ejdrup, there are large areas covered with traces of field banks and lynchets, which form rectangular field plots. Most of the field system is associated with the barrows along the ridge, suggesting a continuity of settlement between the Bronze Age and Early Iron Age. If this relationship between heathlands, barrows and field systems is transferred to Skørbæk, it supports the hypothesis that the ridge to the west of Skørbæk, which also has numerous barrows, was once also enclosed by later prehistoric field banks and lynchets. Moreover, there are vegetation marks that may be traces of several later prehistoric field boundaries in the southern part of the now scheduled heath area at Ajstrup and Lundby Hede. It is difficult to evaluate how much of the drift sand area has been cultivated, due to the present-­day vegetation and the extensive systems of holloways.6 The absence of barrows and the fragmentary traces of field banks suggest that this has always been a marginal area and the soils here would favour growth of heather. The ridge between the drift sand and the cultivated areas extends over c. 370 ha, and traces of coherent field systems are evident on aerial photographs on 250 ha, corresponding to c. 70%. In comparison, Hatt estimated that the two field systems at Gundersted covered an area of only 65  ha (Hatt 1949: 9). The field systems were undoubtedly larger than 250 ha in the past and would have extended into the areas that were cultivated in historical times. In contrast to Skørbæk, no coherent traces of field boundaries are evident running down to the Vidkjær River. There are, though, fragments to the south of Borup, including some fields on a small sandy promontory directly opposite the field system at Skørbæk Hede. It is difficult to evaluate the extent to which the historically cultivated areas were also under the plough during the Early Iron Age. On the one hand, the traces of field banks run directly up to these areas, but on the other, there are virtually no barrows here, and none whatsoever to the northeast of Borup. A very uncer-

The micro-­region at Store Binderup lies only 1  km to the south of Denmark’s best known Pre-­Roman Iron Age locality, namely the Borremose fortified village. The area covers 2 × 2 km and is delimited to all sides by wet meadowland around the Lerkenfeld Å watercourse (Fig. 12.5). In the middle of the area is a 1 km long and 200 m wide wetland and a few minor bogs also lie scattered across the area. The terrain is relatively flat and dips evenly down towards the meadows. The soils comprise a range of sandy deposits. On historical maps from c. 1780, the entire area is marked as heath, surrounded by cultivated land, and there are historical villages to the north, south and west. The area constituted common land for the inhabitants of one of these villages. In 1814, the micro-­region was unoccupied and still lay as heath, about half of it was cultivated by around 1880, and the remainder had come under the plough by 1937. The area was apparently not settled or cultivated in historical times. There is no evidence of activity in the Stone Age. The 16 recorded barrows, with a few exceptions, are all located in the north-­­ eastern part of the area. Several are of considerable size, and from one of these there are reports of the discovery of numerous urns, which must date from the Late Bronze Age (1000– 500 BC). Studies of aerial photographs and historical maps suggest the existence of at least 25 further barrows of varying sizes. These are more or less in the same area as those already recorded, with the exception of a group of barrows evident in the south-­eastern part of the area, which otherwise is without finds. Many of the barrows have been incorporated in the banks of the field system. In the middle lies the area’s one and only recorded settlement, dated to 100 BC–AD 100.7 Later settlement traces are limited to the twelfth-­century church in Kongens Tisted south of the micro region.

12.10 Land Use at Store Binderup The exploitation of the landscape at Store Binderup is significantly less complex than at Gundersted and Skørbæk, as the traces of Early Iron Age field systems have not been erased by historical cultivation but are thought to be present According to personal communication with Torben Trier Christensen, Nordjyllands Historiske Museum, there is only one find report in the file (archive no. 3182).

7 

Recently released LiDAR data of the scheduled heath areas shows that most of these are covered with field boundaries. 6 

176

M. Vinter

Fig. 12.5  The micro-­region at Store Binderup showing the Early Iron Age field system mapped from aerial photographs. Meadow/bog is marked in green, based on evidence from several historical maps. Yellow = cultivated land on the Royal Society maps, light purple = heath on the Royal Society maps, red circle = barrow on aerial photo and undated/Bronze Age barrow from the Sites and Monuments database, blue square = Iron Age settlement (upper left the Borremose fortified village and directly south of this a settlement from the Late Pre-­Roman Iron Age). Background topographical map from around 1880. (By author)

in almost their entire original extent. At the same time, the micro region is much smaller, and constitutes a more well-­­ defined topographic unit than the other two, and therefore provides a more valid estimation of the original extent of the field system. The first marked change in the landscape according to the pollen data took place in the Late Neolithic and continued up into the Early Bronze Age. The centre of gravity in this early settlement lay in the northern half of the area, judging from the occurrence of large, preserved barrows. The barrows stood clearly visible in an open common-­­ like landscape which, in its southern part, could have had denser tree cover. It would also be obvious to place the Early Bronze Age’s cultivated areas in the northern part of the area, around the barrows. The extensive parcelling out of the landscape that took place in the Late Bronze Age and Early Pre-­­ Roman Iron Age could consequently build on earlier boundaries of the Early Bronze Age. The total extent of the field system, which covers 190  ha, or 75%, of the micro-­­ region’s 250  ha of cultivable land shows a strongly structured and regulated landscape. Only the central southern part appears unstructured and unregulated. The field system was in continuous use, as there is no actual woodland regeneration in the Borremose pollen diagram. Large parts of the field system must therefore have been under grass, resulting in a picture of a cultivation system where the movement and grazing of cattle within the individual plots must have been regulated with the aid of fences or tethering to avoid damage to crops. In the later part of the Pre-­Roman Iron Age, parts of the system were abandoned, and heather was able to spread

and expand, but parts of the field system continued in use. Judging from the pollen evidence, no crucial changes took place in the relationship between trees, grass and heather in the area until the end of pollen diagram ends around 500 AD. On this basis, detailed mapping of the settlement in the micro-­region during both the Early and Late Iron Age could provide answers to the many questions about land use in the Bronze Age and Iron Age. In the case of the Early Iron Age, the main question relates to the size of the area which is required by each farmstead unit? While in the Late Iron Age, the question is whether the field systems continued under the plough at this time or whether the settlement relocated? If such large areas were included in the cultivation strategy, it can be difficult to conceive the function of the meadows, or whether they had any great significance. Direct access to extensive meadowlands does not appear to have been of crucial significance at either Gundersted or Skørbæk. These were available at Store Binderup, and had significant potential as a source of fodder, but we cannot know the extent to which they were included in agricultural practices and strategies.

12.11 L  and Use in the Micro-regions: A Summary The enclosure of land within field systems became a major element in the landscape during the Late Bronze Age and Early Iron Age. The fields may have covered 50–70% of the

12  Reconstructing Enclosed and Parcelled Out Landscapes from the First Millennium BC in Himmerland, Denmark: Arable Fields…

cultivable area, with their limits extending out to the natural boundaries in the terrain, and thereby creating a strongly structured and regulated landscape. Similar high exploitation percentages have been determined on the island of Bornholm (Nielsen 2000b: 178–179). There appears to have been a tendency for the field systems to be associated with barrows, which were incorporated into field banks and lynchets, indicating continuity in the exploitation of the landscape. Whether the field banks were generally based on much earlier boundaries cannot be determined with certainty, but this has been demonstrated at Alstrup Krat, where lynchets overlie cultivation layers from the Early Bronze Age (Bech 2003). Given that the laying out of the field systems took place primarily between 800 and 300  BC, they appear to have been established in an extensively exploited common-­­ like landscape with widespread grazing areas and few trees. This parcelled-­out grassland landscape with permanent fields may well have been created to control and concentrate the dung of farm animals on the arable land. The manuring of the fields has been well documented at localities such as Alstrup Krat (Kristiansen 2001; Bech 2003). Some of the fields were abandoned around 300 BC, after which heather was able to expand. As heather does not tolerate cultivation, it must be presumed that the areas it colonised were not part of the arable field system, but constituted outfield with a certain element of grass cover that was rarely or never cultivated. It is important to emphasise that the fields were cultivated regularly and did not lie fallow for 20–30  years. Both the formation of the lynchets, which results from erosion of the bare, vegetation-­free surfaces, and the fact that the forest was unable to regenerate, point in this direction. It is difficult to determine the length of time that a field lay fallow, but judging from historical sources, 6–10 years seems a reasonable estimate. This would be within the memory of the local population. It is also worth noting that when fields were laid fallow, 3–5 years would elapse before grass covered the entire surface (Mikkelsen and Nørbach 2003: 129–30 with references). Even though this latter reference relates to reports from the less nutrient-­rich soils of western Jutland, and the fact that the stubble is able to retain some of the soil, it seems likely that fallow fields in Himmerland were exposed to wind erosion for several years. This would result in the soil both being deposited by the field banks and, in the case of the lighter fractions, blowing further afield. The latter constitutes the sediment component found deposited in peat bogs, and in relatively large amounts in the Early Iron Age. The significance of the meadow areas is difficult to establish as the grass pollen evident in the pollen diagrams could derive from fallow fields and grazing land. Meadows were accessible but do not appear to have been a determining factor in the choice of location for the field systems. All the micro-­regions contain settlement from the Early Iron Age within or immediately adjacent to the surveyed

177

field systems. At Skørbæk there are, moreover, traces of earlier settlement in the form of pits and wells. Even though the settlement traces cease around the birth of Christ in all the micro-­regions, the pollen diagrams show no change in the exploitation of the landscape prior to AD 500. Given the substantial size of the areas covered by the field systems, it seems extremely unlikely that their cultivation did not continue into the Late Iron Age. The settlement traces at Skørbæk Hede suggest continuous settlement within the field system throughout the system’s presumed period of use. There is however still a very fragmentary picture of the links between settlement, field system and land use. To obtain a better understanding, it is therefore necessary to look more closely at the archaeological record for Iron Age settlement in Himmerland.

12.12 S  ettlements in Himmerland During the Bronze Age and Early Iron Age and Their Relation to Field Systems Settlement archaeology, and in particular the excavation of settlements from the Early Iron Age, has a long tradition in Himmerland, extending back to the 1920s. The resulting archaeological record has previously been considered to be representative of the Early Iron Age, though with the reservation that a more detailed chronological division is not possible (Nielsen 1998: 271). This must be said to still be the case. The overall settlement record for the Bronze Age and Iron Age is extensive. There are 511 known localities, evenly distributed across all of Himmerland (Fig. 12.1). If the records are allocated to sub-­periods, the spatial representativity is no longer pronounced, because 44% of them are dated broadly to the Early Iron Age (650 years) or Iron Age (1250 years). If only localities with house structures are examined, the broad dating frame is less pronounced. Of the 117 localities, 20% are dated to the Bronze Age, 25% to the Early Iron Age/Iron Age, 20% to the Pre-­Roman Iron Age and 20% to the Early Roman Iron Age. Remarkably, only three localities can be dated unequivocally to the Early Pre-­ Roman Iron Age. Due to these dating issues, it is rather difficult to draw any conclusions of local displacement of the settlement and exploitation of the landscape during the Early Iron Age. The problems of dating and representativity regarding the Early Iron Age settlement, and the location of this relative to the field systems, are further exacerbated by the broad and uncertain dating of the field systems in general. There are only few examples of how the settlements and field systems relate to each other, but these can give an idea of their overall appearance. If the records for houses of the Early Iron Age are cross-­referenced with the areas covered by field systems, five field systems can be identified to contain remains house

178

structures.8 Within the field system at Stenildvad, just south of Aars, the local museum has investigated a total of six houses and a pit complex from 500–250 BC.9 Examination of multiple series of aerial photographs revealed that the field system was only fragmentarily preserved due the impact of cultivation in historical times. A well-­preserved field system is evident on aerial photographs at Haverslev. The house recorded here consists solely of a stone pavement.10 In the well-­preserved field system at Stenild, the record refers to a hearth. Nine other sites with pits from the Early Iron Age have been recorded within the field system, and 800 m to the north, four houses from 700–250  BC and four longhouses from 200–600 AD were excavated in 2007.11 There is also a record of a house site within the large field system at Vindblæs/Kornum, but with no further information given.12 A couple of hundred metres to the southeast of these fields, six houses from 250–50  BC and several cremation graves from the period between 600 and 400 BC were excavated. The last example is Skørbæk Hede, which is the only instance of an excavated settlement with houses located within a mapped field system. Further to this is the site of Østerbølle, where remains were discovered of 16 houses, three small out-­buildings, a well and a burial ground from 50 BC–AD 150. Only three were excavated, while the others are now scheduled. The houses lie in two east-­west-­orientated rows. Two in the northern row are clearly incorporated into the field bank associated with a small fragment of preserved field system (Hatt 1938). This field system is unfortunately not visible on aerial photographs. Other settlement remains, which in most cases are different kinds of pits, do not appear to be frequent occurrences in the recorded field systems either. Only 18 of the 137 known field systems contain traces of settlement from the Bronze Age and Early Iron Age, and in half of these cases the remains are merely dated broadly to the Iron Age. Three of these refer to localities with a house site.13 Finally, mention should be made of the field system at Gundestrup, which contains numerous traces of settlement activities from both the Bronze Age and the Early Iron Age. The Bronze Age remains comprise a back-­filled well and several areas with cooking pits, pits and postholes. From the Early Iron Age there are two large settlements, where recent investigations have demonstrated the remains of parts of a The records of the field systems from the Early Iron Age in the Sites and Monuments database are based on P.H. Sørensen’s records, but it is not known who has drawn the polygons. They are though generally larger than the extent of the actual traces on the aerial photos. 9  Site nos. 120814–323, 324, 329 (cf. the Sites and Monuments database). 10  Site no. 120810–111. 11  Site nos. 120211–95, 120210–58. 12  Site no. 120703–25. 13  Site nos. 120511–27A, 120510–35, 120107–22. 8 

M. Vinter

house with a clay floor and a cobbled area, dated to 100 BC– AD 150.14 This account of settlement traces from the Bronze Age and Early Iron Age in the mapped field systems shows that there are actually a number of examples, even though they only occur in a minority of the field systems. There are houses, pits, cooking pits, wells and various forms of graves or burials. The finds from the Bronze Age clearly suggest that these areas have had a long settlement continuity. The information on the Iron Age settlements is often of an earlier date, and the dating frame is broad. Nevertheless, there are secure examples of houses from 500–250 BC at Stenildvad, and houses from around the birth of Christ at Gundestrup, Skørbæk and Østerbølle, which document settlement activities in the field systems in general throughout the entire Early Iron Age. The picture of the settlement and its position relative to the field systems is however so fragmentary, that it is necessary to look more closely at better investigated and structured settlements in order to investigate how the settlements/houses could have fitted into the field systems. It had previously been argued that the transition from the Late Bronze Age to the Early Pre-­ Roman Iron Age, around 500  BC, marked a tangible shift in the settlement pattern with the emergence of the village, as evident at Grøntoft in western Jutland (Becker 1972: 79–109). However, new studies show that the most striking shift took place later, in the middle of the Pre-­Roman Iron Age, around 250 BC (Webley 2008: 23). This shift is also evident in Himmerland, even though the area has some special features that are not seen elsewhere. Settlement in the Bronze Age consisted of single, isolated farmsteads that moved around within a resource area. New houses were rarely built in the same places as previous ones, they were not fenced in and house structures can rarely be shown to be contemporaneous. This wandering Bronze Age farmstead is now well-­documented in Denmark and was also to be found in other parts of Europe at the same time (Webley 2008: 23 with references). In recent years, it has also been demonstrated at several sites in Himmerland, by way of finds of single farmsteads dating from the Late Bronze Age. This settlement pattern continued into the Pre-­Roman Iron Age and even though secure sites from this period are rare in Himmerland, there is an example at Svenstrup Gårde, which shows a remarkably long settlement continuity. Within an area of 10 ha there are 100 buildings of various sizes, dating from the period between the Late Bronze Age and the Early Roman Iron Age. In part of the area, the buildings are interpreted as representing one or two mobile farmsteads dating from 500–250  BC, and three farmstead complexes from 250 BC–AD 150, which lay more or less in the same Site nos. 120814–273, 120814–248, 120805–153, 120805–122, 120814–205, 120805–152. 14 

12  Reconstructing Enclosed and Parcelled Out Landscapes from the First Millennium BC in Himmerland, Denmark: Arable Fields…

place throughout the entire period.15 Around the area are many other settlement traces from the Bronze Age in the form of pits, cultural layers, a large cooking pit complex extending over several hectares and a settlement from the Late Roman Iron Age (AD 175–375).16 On the historical map from 1880 the entire area is shown as being cultivated. Therefore, no soil marks of field banks can be observed on aerial photographs, but soil marks are clearly visible just beyond it. The same situation occurs at Højgård where, within a small area, there were eight houses dating from the period 800–50  BC and two Viking Age houses (AD 750– 1050).17 Svenstrup Gårde and Højgård show, first and foremost, long-­term, continuous settlement within a small area, extending from 800 BC until AD 150. But they also suggest that the Late Iron Age is represented in these same areas. Long-­term continuity of the settlement within a small area is also evident at Sejlflod, where there was a village in three phases, which moved around within an area of 6  ha during the period 500 BC–AD 150. In the Late Roman Iron Age, it moved again (Nielsen 1991: 111–25). Sejlflod must be considered as the only secure example of a wandering village in Himmerland, but both the Højgård and Svenstrup Gårde localities suggest that it is not unique. The tell sites of Nørre Tranders, Nørre Hedegårde, Trandersgård and Lykkensgård are a special northern Jutish settlement phenomenon with a long settlement continuity that began in 600 BC and continued until AD 200/400. The number of coeval farmsteads is given as between four and ten. In the early phases, the farmsteads lay scattered, whereas they became part of a stringent, permanent parcelled structure in the later phases.18 The tendency towards concentration and parcellation of the settlement appears to have been a general feature in the period from 250  BC until AD 150. At the localities of Tolstrup and Års Mark, 71 longhouses from this period were excavated within areas of, respectively, 2.8 and 5.5 ha.19 The distance between these two localities is only 800 m. The fortified village of Borremose must also be considered to be part of the same phenomenon. Single farmsteads were also very probably part of the settlement picture of the period, but it remains uncertain to what extent. Early Iron Age localities, discovered in recent years with the aid of aerial photographs, most often consist of five to ten houses visible within an area of 150 × 150 m. They are therefore more like Tolstrup and Års Mark and the tell sites than single farmsteads, which are only known from Site no. 120801–305, 308, 319, 333. Site no. 120801–64, 65, 228, 336, 340, 343. 17  Site no. 120804–54. 18  Site nos. 120113–24, 120113–32 120113–24. Nielsen and Haue 2002: 141; Nielsen (2000a). 19  Site nos. 120814–111, 120814–250. 15  16 

179

a few localities (Vinter 2009). The predominant settlement form in Himmerland during the Late Pre-­Roman and Early Roman Iron Age appears then to have been a village of four to ten coeval houses, concentrated within a small area of two to six ha. Late Bronze Age and Early Pre-­Roman Iron Age settlement is less well documented, but is, in most cases, found within the same areas. However, it also appears to be more ephemeral and scattered across larger areas. As mentioned above, the settlement traces in the micro-­­ regions are extremely fragmentary. The Skørbæk Hede settlement resembles a typical settlement in Himmerland around the birth of Christ. There are several coeval farmsteads and the individual farms appear to have occupied the same spot through several phases. It covered an area of only 50 × 50 m and was thereby smaller than other settlements but was probably originally considerably larger than the remains revealed within the investigated area. The distance to the nearest known coeval settlement is just less than 1 km, as in the case of Års Mark and Tolstrup. The most striking difference at the Skørbæk Hede settlement and other settlements in Himmerland is the connection between field banks and houses. The houses are clearly fitted into a parcelled-­out field system and appear to be subordinate to it, while the houses at several other settlements around the birth of Christ appear to have been locked into a different kind of parcelling system. Is this an expression of two different parcelling systems – an earlier system that divided up fields and was already on the way out around the birth of Christ – and a later system directed towards the parcelling-­out of farmsteads, with no visible markers for the field boundaries? Or are these actually two aspects of the same phenomenon? This question represents somewhat of a Gordian knot due to the difficulties in dating the periods of use of the field systems, poor knowledge of the settlement pattern in the Late Bronze Age/Early Pre-­Roman Iron Age and the nature of its relations to the field systems and the cultivation boundaries around the well-­­ documented settlements from around the birth of Christ onwards. These problems cannot be readily solved on the basis of the available evidence from Himmerland, but a closer dating of the field systems and the links to their respective settlement can be found elsewhere in the archaeological records of Denmark and the Netherlands. In terms of dating, the investigations at Alstrup Krat in East Jutland showed that the formation of a lynchet took place between 750 and 500 BC and that the latest cultivation layer could, based on pottery finds, be dated to 250–50 BC (Bech 2003). In eastern Denmark, the field systems appear to have been established during the period 750–340 BC, based on pottery and radiocarbon dates. Pollen studies here show a major opening up of the landscape during this same period (Nielsen 2000b, 2010). In Store Vildmose, in northern Jutland, short-­term cultivation took place between 400 and 200 BC (Nielsen 1991). The cultivation at Lodbjerg in north-

180

west Jutland began in the Late Bronze Age, but the most distinctive cultivation layer, with two field banks and a house placed immediately adjacent, is dated to 300 BC (Liversage et al. 1987: 55–84). In Hatt’s investigations (1925–50), the stratigraphic observations appear to date the establishment of the field banks to 500–250  BC, and at Nørre Fjand, a village from around the birth of Christ was established on top of field banks and drift sand deposits (Hatt 1949, 1957). Mention should of course also be made of the excavations at Grøntoft, where it proved possible to demonstrate an extremely dynamic alternation between field banks and houses during the Early Pre-­Roman Iron Age (Becker 1972: 79–109). The most recent example of the diachronic relationship between a field system and houses in the Early Pre-­Roman Iron Age comes from Brændgårds Hede in western Jutland. A solitary house, aligned with the field banks, must represent an early phase of the settlement, while the others represent the concentration of the settlement that took place late in the Pre-­­ Roman Iron Age (Fig. 12.6).20 In the Netherlands, there are several examples of excavated houses in field systems (Gerritsen 2003). In a minor excavation at Peelo, three single farmsteads from, respectively, the end of the Late Bronze Age, the Early Pre-­Roman Iron Age and the Late Pre-­Roman Iron Age, were found in the middle of a field plot and partially buried beneath very broad field banks (Kooi and de Langen 1987). Excavation of a small portion of a field system at Hijken (Fig.  12.7; Harsema 2005: 548) revealed the presence of houses from both the Bronze Age and the Early Pre-­Roman Iron Age. Remains of a fence which followed the surface soil marks of field banks clearly show that the houses had stood within well-­defined, parcelled-­out plots and that, contrary to the interpretation at Grøntoft, there was continuity in the field banks. Hijken is the only example to date of a fence found associated with field banks (Fig.  12.7). A house from the Early Roman Iron Age has also been demonstrated in the middle of a plot at Selingen-­Zuidveld (Van Giffen 1940). It later became overlain by a field bank, which subdivided the plot. Several researchers in the Netherlands have suggested that the use of the field systems established in the Early Iron Age contracted during the Late Pre-­Roman Iron Age, at the same time as wandering farmsteads moved towards a fixed and concentrated structure involving several farms together (but see Arnoldussen, this volume). In parts of the Netherlands, the settlement is thought to have moved on to clay soils, where cultivation was concentrated in smaller areas around the villages, while the former field systems on sandy soils formed pastures (Gerritsen 2003: 242–44; Webley 2008: 42 with references). Oral communication from Esben Schlosser Mauritsen, who is also thanked for permission to use this material here. 20 

M. Vinter

In Denmark, and presumably also in the Netherlands, the allotment of field systems and the formation of lynchets and field banks took place in the period c. 750–250 BC, and the primary boundaries appear to have had a permanent character. Settlements consisted of single farmsteads, which respected the field boundaries, and which moved around on the field plots that appeared over time until 300 BC. Subsequently, the settlements became concentrated within a few hectares and comprised several coeval farmsteads. The focus of the parcelling-­out process apparently shifted from being centred on the fields to also include the settlement to a greater extent. The cultivation of the field systems did not cease, but the cultivation strategy changed. Until 300 BC, there was extensive cultivation with continual relocation of the farmsteads such that these, over time, came to cover as much as 50–70% of the landscape. This strategy also led to cultivation of marginal areas such as the dune heaths at Lodbjerg and perhaps the aeolian drift sand deposits to the north of Gundersted. This system was, however, intensive enough for cultivation to result in the formation of lynchets. The pollen diagrams show that the relationship between grass/heather and trees remained stable from the Late Bronze Age until sometime in the Viking Age. The only marked change is the expansion of heather, which is evident from c. 300 BC onwards. A plausible explanation for this development is that the cultivation became concentrated in smaller areas, as also suggested by Webley (2008) as well as a series of other overarching changes in the general use and layout of the landscape (Løvschal 2014b). Large parts of the former combined cultivation and grazing areas then became colonised by heather, which was maintained by burning and grazing husbandry, and prevented reforestation. Heather is a diverse resource with many potential uses, such as winter grazing, fuel and as peat manure for soil improvement. The reasons for this shift in settlement structure, cultivation strategy and land use are difficult to ascertain. They could involve either natural factors, intentional human decisions, or a combination of both. For example, it is conceivable that leaching of nutrients resulting from long-­term grazing and cultivation of large areas was rectified by concentrating manuring activity on a smaller cultivated area, where the soil was improved using a mixture of heather turf and animal manure. Then there is easier access to winter fodder and perhaps a greater emphasis on sheep rearing, as shown by the animal bone assemblages from northern Jutland (Kveiborg 2008). The farms could then either be accommodated within parts of the existing parcelled-­out system, as seen at Skørbæk and Østerbølle, or be established on top of it, as has been documented at Nørre Fjand. In the micro-­regions, these settlements have been identified in two places on Skørbæk Hede and at a single locality at Store Binderup. Whether this development continued into the Late

12  Reconstructing Enclosed and Parcelled Out Landscapes from the First Millennium BC in Himmerland, Denmark: Arable Fields…

0

181

50 meter Skala: 1:2.139

Fig. 12.6  Excavation plan of Brændgårds Hede (yellow) superimposed on Hatt’s map of the field banks at the site (pink). The houses to the west are of uniform orientation which clearly deviates from the single house

Fig. 12.7  Excavation plan of the Pre-­Roman Iron Age settlement at Hijken. The broken lines mark the course of fences. The Bronze Age houses, which are oriented north-­south, are not shown. The north-­south-­oriented house in the northwestern corner is marked on other versions of the same plan as a Bronze Age house. (After Harsema 2005: 548 fig. 24.6)

(red circle) and the field bank to the east of it. The pit zone systems are evident as dark bands above the houses to the west. (By permission of Esben Schlosser Mauritsen, Arkvest)

182

Iron Age cannot be determined based on the present archaeological records for the micro-­regions.

12.13 Conclusion This study has shown how a combined analysis of pollen data, historical maps, aerial photographs and the existing archaeological record can generate a somewhat clearer picture of the landscape and its use in the Early Iron Age and adjacent periods, even though there remain substantial gaps in our knowledge of the field systems and their associated settlements. Within the micro-­regions, aerial photographs show that the field systems covered very large areas which, during the Late Bronze Age and Pre-­Roman Iron Age, appear to have taken on the character of permanent structures in a regulated landscape. The pollen diagrams reveal a marked change in the vegetation in the middle of the Pre-­Roman Iron Age, when heather heath became part of the grazing landscape. This took place at the same time as a nucleation of settlement and continuity in the locations of houses. Taking this shift in settlement organisation together with the changes in vegetation, the middle of the Pre-­Roman Iron Age (c. 300 BC) saw parts of the field systems colonised by heather as they were used for less intensive grazing and fodder, while cultivation became concentrated in smaller areas, close to the now nucleated settlements. The greatest problem relative to understanding the landscape and settlement in the first millennium BC is a lack of information on Early Pre-­Roman Iron Age settlement in general, and in particular its location in and chronological relationship to the regulated, enclosed landscape. Hopefully, this study provides an example of the uses to which the field systems mapped from aerial photographs, this unexploited source, so expectantly described by Steen Hvass in the introductory quote, can be put.

References Aaby, B. (1990). Pollen og støvnedfald fortæller om fortidens landbrug. Nationalmuseets Arbejdsmark, 1990, 130–140. Aaby, B. (1994). Landskabsudnyttelse, muldflugt og sandflugt i Nordjylland. In J.  Lund & J.  Ringtved (Eds.), Limfjordsprojektet. Rapport nr. 7. Sandflugt, dyrkning og bebyggelse i Limfjordsområdet (pp. 25–40). Aarhus: Aarhus University. Bech, J. (2003). Fra fortidsminder til kulturmiljø – hvad Alstrup Krat og Hohøj gemte. Copenhagen: Kulturministeriet, Kulturarvsstyrelsen, Miljøministeriet og Skov- og Naturstyrelsen. Becker, C.  J. (1972). Früheisenzeitliche Dörfer bei Grøntoft, Westjütland. 3. Vorbericht: Die Ausgrabungen 1967–68. Acta Archaeologica, XLII, 79–109. Brongers, J. A. (1976). Air photography and Celtic fields research in the Netherlands (Nederlandse Oudheden 6). Amersfoort: ROB.

M. Vinter Ejstrud, B., & Jensen, C. K. (2000). Vendehøj – landsby og gravplads. Kronologi, organisation, struktur og udvikling i en østjysk landsby fra 2.årh. f.Kr. til 2.årh. e.Kr. Højbjerg: Jysk Arkæologisk Selskab. Gerritsen, F. (2003). Local Identities. In Landscape and community in the Late Prehistoric Meuse-Demer Scheldt Region (Amsterdam Archaeological Studies 9). Amsterdam: Amsterdam University Press. Harsema, O. (2005). Farms amongst Celtic fields. Settlements on the northern sands. In L.  P. Louwe Kooijmans, P.  W. van den Broeke, H.  Fokkens, & A.  L. van Gijn (Eds.), The prehistory of the Netherlands (Vol. 2, pp.  543–556). Amsterdam: Amsterdam University Press. Hatt, G. (1938). Jernalderens bopladser i Himmerland. Aarbøger for Nordisk Oldkyndighed og Historie, 1938, 119–268. Hatt, G. (1949). Oldtidsagre (Arkæologisk-Kunsthistoriske Skrifter Bd. II no.1). Copenhagen: Det Kongelige Danske Videnskabernes Selskab. Hatt, G. (1957). Nørre Fjand, an Early Iron Age village site in West Jutland (Arkæologisk-Kunsthistoriske Skrifter Bd. II no.2). Copenhagen: Det Kongelige Danske Videnskabernes Selskab. Hvass, S. (1985). Hodde. Et vestjysk landsbysamfund fra ældre jernalder (Arkæologiske Studier, Vol. VII). Copenhagen: Akademisk Forlag. Hvass, S. (1993). Bebyggelsen. In S.  Hvass & B.  Storgaard (Eds.), Da klinger i muld…25 års arkæologi i Danmark (pp.  187–194). Copenhagen/Højbjerg: Det Kgl. Nordiske Oldskriftselskab/Jysk Arkæologisk Selskab. Jeansson, N. R. (1963). Fossila åkrar i Himmerland – En flygbildsinventering. Svensk Geografisk Årsbok, 39, 111–118. Jensen, A. (1977). Geologiske undersøgelser omkring Borremosebebyggelsen. Aarbøger for Nordisk Oldkyndighed og Historie, 1975, 96–119. Klamm, M. (1993). Aufbau und Entstehung Eisenzeitliche Ackerfluren (“celtic fields”). Stand der Forschung (Göttinger Bodenkundliche Berichte 102). Göttingen: Universität Göttingen. Kooi, P.  B., & de Langen, G.  J. (1987). Bewoning in de vroege ijzertijd op het Kleuvenveld te Peelo, gem. Assen. Nieuwe Drentse Volksalmanak, 104, 151–165. Kristiansen, S.  M. (2001). Present-day soil distribution explained by prehistoric land-use: Podzol-Arenosol variation in an ancient woodland in Denmark. Geoderma, 103, 273–289. Kveiborg, J. (2008). Fårehyrder, kvægbønder eller svineavlere. En revurdering af jernalderens dyrehold. Kuml, 2008, 59–100. Lewis, B. (1985). Overbygård og Nørre Fjand. En analyse af nogle jernalderlandsbyers tilliggender og økonomi. Kuml, 1985, 123–160. Liversage, D., Munro, M. A. R., Courty, M.-A., & Nørnberg, P. (1987). Studies of a buried Iron Age field. Acta Archaeologica, 56, 55–84. Løvschal, M. (2014a). Emerging boundaries: Social embedment of landscape and settlement divisions in northwestern Europe during the first millennium BC. Current Anthropology, 55(6), 725–750. Løvschal, M. (2014b). From neural synapses to culture-historical boundaries: An archaeological comment on the plastic mind. Journal of Cognition and Culture, 14, 415–434. Mikkelsen, P.  H., & Nørbach, L.  C. (2003). Drengsted. Bebyggelse, jernproduktion og agerbrug i yngre romersk og ældre germansk jernalder. Højbjerg: Jysk Arkæologisk Selskab. Müller-Wille, M. (1965). Eisenzeitliche Fluren in den festländischen Nordseegebieten. Münster: Westfalen. Müller-Wille, M. (1979). Flursysteme der Bronze- und Eisenzeit in den Nordseegebieten. In H.  Beck, D.  Denecke, & H.  Jankuhn (Eds.), Untersuchungen zur Eisenzeitlichen und Frühmittelalterlichen Flur in Mitteleuropa und ihrer Nutzung: Bericht über die Kolloquien der Kommission fûr die Altertumskunde Mittel- und Nordeuropas in den Jahren 1975 und 1976 (pp. 196–239). Vanderhoeck & Rupert: Göttingen.

12  Reconstructing Enclosed and Parcelled Out Landscapes from the First Millennium BC in Himmerland, Denmark: Arable Fields… Newcomb, R. (1971). Celtic fields in Himmerland, Denmark, as revealed by vertical photography at a scale of 1:25000. Photogrammetria, 27, 101–113. Nielsen, V. (1984). Prehistoric field boundaries in Eastern Denmark. Journal of Danish Archaeology, 3, 135–163. Nielsen, J.  N. (1991). Befolkningens størrelse i Sejlflodlandsbyen i Nordjylland fra 500 f.Kr. – 500 e.Kr. In C. Fabech & J. Ringtved (Eds.), Samfundsorganisation og regional variation. Norden i Romersk Jernalder og Folkevandringstid (Jysk Arkæologisk Selskabs Skrifter XXVII) (pp.  111–125). Højbjerg: Jysk Arkæologisk Selskab. Nielsen, V. (1993). Jernalderens pløjning – Store Vildmose. Hjørring: Vendsyssel Historiske Museum. Nielsen, J.  N. (1998). Ældre jernalders bebyggelse i det østlige Limfjordsområde. In J.  Lund & J.  Ringtved (Eds.), Variation og enhed omkring Limfjorden. Rapport nr. 8, II (pp. 271–292). Aarhus: Aarhus University. Nielsen, J.  N. (2000a). Landsbyen under klitten. In S.  Hvass (Ed.), Vor skjulte kulturarv. Arkæologien under overfladen (pp.  94–95). Copenhagen/Højbjerg: Det Kongelige Nordiske Oldskriftselskab & Jysk Arkæologisk Selskab. Nielsen, V. (2000b). Oldtidsagre i Danmark. Bornholm. Højbjerg: Jysk Arkæologisk Selskab. Nielsen, V. (2010). Oldtidsagre i Danmark. Sjælland, Møn og Lolland-­ Falster. Højbjerg: Jysk Arkæologisk Selskab. Nielsen, N.  H., & Dalsgaard, K. (2017). Dynamics of Celtic fields  – A geoarchaeological investigation of Øster Lem Hede, Western Jutland, Denmark. Geoarchaeology, 32(3), 414–434. Nielsen, J.  N., & Haue, N. (2002). Arkæologiske Udgravninger i Danmark, 2001, 141–142. Nielsen, N.  H., Kristiansen, S.  M., Ljungberg, T., Enevold, R., & Løvschal, M. (2019). Low and variable: Manuring intensity in Danish Celtic fields. Journal of Archaeological Science: Reports, 27, 101955. Odgaard, B. V. (1994). Postglacial vegetationsdynamik i hedeegne syd for Limfjorden. In J. Lund & J. Ringtved (Eds.), Limfjordsprojektet. Rapport nr. 7. Sandflugt, dyrkning og bebyggelse i Limfjordsområdet (pp. 137–145). Aarhus: Aarhus University. Odgaard, B.  V. (1998). Hedens vegetationshistorie. In Skov- og Naturstyrelsen (Ed.), Seminarrapport “Den danske hede” afholdt på Skarrildhus 21. september 1998 (pp. 14–28). Copenhagen: Skovog Naturstyrelsen.

183

Odgaard, B.  V. (1999). Fossil pollen as a record of past biodiversity. Journal of Biogeography, 26(1), 7–17. Olesen, L.  H. (1981). Jernalderagre ved Halgård Bæk. Holstebro Museum Årsskrift, 1981, 5–20. Olesen, L. H. (1983). Vestjyske agersystemer set fra luften og jorden. Antikvariske Studier, 6, 300–309. Sørensen, P.  H. (1975). Jysk Oldtidsagerbrug  – lokaliseret efter luftfotografier. Kulturgeografi, 120, 337–354. Sørensen, P.  H. (1979). Luftarkæologi: Hvilke oplysninger giver soil marks? In H. Thrane (Ed.), Fra jernalder til middelalder: beretning fra et symposium d. 17.-19. maj 1979 afholdt af Odense Universitet (pp. 30–40). Odense: Odense Universitet, Historisk Institut. Sørensen, P. H. (1982). The use of air photographs in celtic field studies. Journal of Danish Archaeology, 1, 77–86. Sørensen, P.  H. (1991). Jyske Oldtidsagre. Nordjyllands og Viborg Amter. Copenhagen: Skov- og Naturstyrelsen. Stoumann-Hansen, S., & Sørensen, P.  H. (1984). Tinghøj på Borre Hede. Et glimt af et århundredes mennesker og forskning omkring et ældre jernalders agersystem. Kuml, 1984, 191–211. Trap, J. P., & Nielsen, N. (1961). Trap Danmark (5th ed.). Copenhagen: Gads Forlag. van Giffen, A. E. (1940). De zgn. heidensche legerplaats te Zuidveld bij Sellingen, gem. Vlachtwedde. Verslag Museum van Oudheden Groningen, 1939, 86–93. Vinter, M. (2009). Luftfotografering og bebyggelse Himmerland  – Katalog over himmerlandske luftfotolokaliteter og en vurdering af deres beliggenhed i forhold til jordbund og historisk arealudnyttelse. Aarhus: Aarhus University. (Unpublished article). Vinter, M. (2011). Kortlægning af marksystemer fra jernalderen. En kildekritisk vurdering af luftfotografiernes anvendelighed. Kuml, 2010, 65–90. Webley, L. (2008). Iron Age households. Structure and practice in Western Denmark, 500 B.C.-200 A.D. Højbjerg: Jysk Arkæologisk Selskab. Zimmermann, H.  W. (1979). Untersuchungen zur Landwirtschaft während der Römischen Kaiserzeit in der Siedlungkammer Flögeln, Kr. Cuxhaven. In H.  Beck, D.  Denecke, & H.  Jankuhn (Eds.), Untersuchungen zur Eisenzeitlichen und Frühmittelalterlichen Flur in Mitteleuropa und ihrer Nutzzung: Bericht über die Kolloquien der Kommission fûr die Altertumskunde Mittel- und Nordeuropas in den Jahren 1975 und 1976 (pp. 240–249). Göttingen: Vanderhoeck & Rupert.

Understanding the Chronologies of England’s Field Systems

13

Robert Johnston, Rowan May, and David McOmish

Abstract

The chapter presents the outcomes from a project funded by Historic England to collate and analyse scientific dates for field systems in England. The project identified 393 scientific dates from more than 120 excavated field systems. The analysis provides a constructive critique of fieldscape chronologies and recommends methodological improvements and directions for future research. The results indicate that bounded fields appeared in the English landscape during the first quarter of the second millennium BC, and that subsequent enclosure was an ongoing, geographically discontinuous process. The active apportionment of land and maintenance of field systems was possibly interrupted during c. 1000–600 BC, as a hiatus was observed in the project’s data. Post-Roman and medieval field systems are rarely targeted for scientific dating, with the consequence that the continuity of field systems from the Roman period into later landscapes remains ill-defined. There are major biases in the current dataset of scientific dates. The project has illustrated the requirement for standardised reporting of scientific dates in grey literature reports, publications, and data archives. Geographical biases within the data occur due to regional concentrations in the application of development-led archaeology, and the impacts of regional priorities and research agendas. Future legislative protection, research frameworks and fieldwork strategies should prioritise understanding the long chronologies of field systems and R. Johnston (*) Department of Archaeology, University of Sheffield, Sheffield, UK e-mail: [email protected] R. May ArcHeritage, Sheffield, South Yorkshire, UK e-mail: [email protected] D. McOmish National Strategy Team, Historic England, London, UK e-mail: [email protected]

rebalance our current attention towards periodised perspectives. Keywords

England · Field systems · Dating · Chronology · Long-­ term developments

13.1 Introduction In this chapter, we report on a project that collated and reviewed the scientific dating of field systems in England. Bounded field systems began shaping England’s landscapes in the early-mid second millennium BC. Enclosure of land and its reorganisation continued with varying temporal and spatial intensities throughout the subsequent four thousand years. It is estimated that field systems now cover at least 70 per cent of England and are the most extensive form of heritage asset in the country (McOmish 2011; cf. Vinter, this volume). By extension, field systems offer an important source for understanding organisation, identity and change within landscapes at regional and national scales. A total of 393 scientific dates from 120 sites were recorded, derived from a wide variety of field and boundary types, including large-scale land division, and covering all periods from early Bronze Age to post-medieval (Johnston et al. 2020). The results show that the English landscape was enclosed during later prehistory and that this was an ongoing, though geographically discontinuous process. Based on the analysis of the scientific dates, large-scale apportionment of land into field systems began in the early centuries of the second millennium BC and became more commonplace after 1700 BC. The active apportionment of land and maintenance of field systems may have been interrupted during the period c. 1000–600 BC. Land enclosure re-emerged and extended into new regions during the middle and late first

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 S. Arnoldussen et al. (eds.), Europe’s Early Fieldscapes, Themes in Contemporary Archaeology, https://doi.org/10.1007/978-3-030-71652-3_13

185

186

millennium BC. There are few scientific dates from Roman and historic period field systems. There are a number of biases in the current dataset of scientific dates. The project identified a lack of a consistent recording method for scientific dates both regionally and nationally, severely limiting the dissemination of this information for research and comparative purposes. The nature of development-led archaeology means that most data arises from areas of modern economic activity, leading to an under-­ representation of uplands and heathlands, where dating is currently poorly understood. Additionally, there is a low number of scientific dates associated with Roman field systems across most regions, whilst post-Roman and medieval fields remain largely inaccessible through scientific dating. Our chapter introduces the project and the data, identifies the biases that affected the quality and completeness of the data, reviews the dating strategies, geographical distribution and chronological variation of the dates. The conclusion identifies broad patterns in the spatial and temporal distributions of field systems in England. It presents several recommendations for revising dating strategies, the data management of scientific dates and future research directions in the study of field systems.

13.2 Project Context and Aims The past and present patterns of England’s field systems have been surveyed systematically, if not always analytically, across many regions. This field and desk-based research has delivered convincing arguments for the chronological sequence of land enclosure within these regions, based on a combination of relative dating using the physical relationships of landscape features, morphology and typology, map regression (cf. Roughley, this volume), and incorporating dates from excavations and proxy sources (such as palaeoenvironmental sequences; cf. Vinter, this volume). For example, the current understanding of the relict field systems on Salisbury Plain reflects a balanced evaluation of a wide variety of sources combined with targeted analytical survey (McOmish et  al. 2002). By comparison, landscape characterisation offers the coarsest resolution and the most reliance on typologies, although it is also the most accessible and widely-used (measured in geographic terms) method for identifying the historic character of field systems (Turner 2006). While accepting that there are many strengths to the current understanding of field systems, there are also significant weaknesses. A key one is a lack of secure absolute dates for the formation, use, transformation, and abandonment of the fields and enclosure systems (cf. Arnoldussen, this volume). In part this is because until the advent of large-scale rescue projects (typified by Fengate and Mucking (Evans et  al.

R. Johnston et al.

2009: 12–14) in a lowland context, and Shaugh Moor in the uplands (Balaam et al. 1982)) the investigation of field systems was restricted to the localised trenching of a few or maybe only one boundary. Larger scale archaeological investigations, whether across open areas or in corridors, have vastly increased the opportunities for systematically sampling and dating field systems. Alongside this, scientific dating is more widely employed and with much stricter criteria applied to the selection and reporting of samples. Scientific dating offers an important method for understanding the chronologies of field systems. Scientific dates can be independent of and may be more precise than artefacts and relative dating, however, it is by no means straightforward. Samples for radiocarbon dating are rarely from ‘event contexts’, as field boundaries do not usually provide the locations where activities happened in situ. Field surfaces themselves were subjected to re-working, soil formation processes, erosion and truncation. Physical boundaries provided important ‘traps’ for datable material, whether sealed within or beneath banks or incorporated into ditch sediments. This has benefits but also brings challenges. The datable material in boundaries can derive from secondary formation processes such as manuring using midden deposits that have accumulated elsewhere, followed by erosion from a field surface into a boundary. Complex formation processes make it important that all scientific dates are critically evaluated against their contexts, in a technical and an agricultural sense. Optical Stimulated Luminescence (OSL) dating overcomes some of the problems with radiocarbon methods, yet thus far its use remains relatively limited. OSL has the advantage that it dates sediments and that it estimates an event (the time when the sediment was last exposed to light), which may in turn correlate with an archaeological context – the construction of a bank, for example. On Big Moor, Derbyshire, OSL was used to date the cultivation of the fields to the later second and early first millennium BC (Heath 2003). Gwithian in Cornwall, where sand overlies the cultivated horizons, is another (Nowakowski 2009). Additionally, there are important examples of OSL’s use in northwest European field systems (Arnoldussen, this volume). The ‘Scientific Dates for Field Systems in England’ project was completed during 2015–16 in a partnership between Historic England, the University of Sheffield and ArcHeritage, a commercial archaeological consultancy. The project’s aims were to: (a) produce a collated list of relevant scientific dates for field systems in England; (b) provide a constructive critique of fieldscape chronologies based on collated scientific dates; (c) recommend methodological enhancements and future research that arise from the project’s results. The project assessed the current understanding of the chronology of field systems of all periods in England. The focus of the work was on reviewing known dates for the

13  Understanding the Chronologies of England’s Field Systems

activity of enclosure for agricultural purposes and aimed to review existing dates for fields of all periods drawn from published and unpublished sources. The intention was that all scientific dates (radiocarbon and luminescence) should be included within the review, with a focus on dates that have been derived from field boundary features such as lynchets, banks and ditches, or similar.

13.3 Methodology 13.3.1 Sources The ‘Scientific Dates for Field Systems in England’ project recorded 393 scientific dates from 120 excavated sites (‘site’ refers to the archaeologically investigated area, however large or small) using information from unpublished and published sources. Our starting point was a search of England’s regional monument databases (HERs), the national monument database, and several key online sources, including the Council for British Archaeology’s database of radiocarbon dates and the Archaeology Data Service. We supplemented this primary search with a review of key bibliographic databases and reviews, such as the British and Irish Archaeological Bibliography, the Archaeological Investigations Project’s (AIP) database of grey literature reports, and the period surveys in regional research frameworks. A series of research projects relating to fields have been undertaken or were ongoing during the project. These included individual PhD research (e.g. Yates 2007; Chadwick 2008; English 2013; Løvschal 2014), as well as larger scale research projects undertaken by universities, local government authorities or archaeological consultancies (e.g. Cooper and Green 2016; Rippon et al. 2015; Bradley et al. 2015). Finally, we requested information on unreported sites from over 40 offices of major archaeological contracting units that operate in England. There were predictable challenges with accessing data from some of the project’s core sources. The regional monument databases, although immensely valuable, could not provide a comprehensive picture. Given the constraints on local government funding, the HERs have substantial backlogs of sites yet to be entered on their databases. Regional differences in staffing levels and staff time pressures as well as different database packages and the type of information recorded meant that most could not search specifically for excavated sites with associated scientific dates. The recording of scientific dating in the searchable fields of most of our sources was limited. For example, the AIP database only mentioned radiocarbon dating in the summary field for two of the nearly 50 records of projects with field systems.

187

13.3.2 Database Structure Information gathered during the project was collected in a relational database (Table  13.1; full dataset available: Johnston et al. 2020). We included locational information in the site table (national grid references and latitude-longitude) so that the distribution of sites could be presented and queried within a GIS (Fig. 13.1). Up to two chronological periods were recorded for each site as there were instances where field systems spanned multiple periods. The ‘primary’ period was defined by the earliest dates associated with field or boundary-related features. The database used Historic England’s thesaurus of monument types to categorise the field systems (FISH 2020). During analysis, the ‘main’ monument type was selected as the category most representative of the dated features. Table 13.1  The project’s database structure, comprising four related tables and their constituent fields 1. Site table: Site UID County Easting and northing Date of intervention

Archaeological period(s) × 2 (based on historic England’s periodisation) Site summary 2. Dating method table: Site UID 3. Radiocarbon date table: Date UID Laboratory code and unique sample number Type of material dated [defined list] Description of context 14 C age δ −13C (per cent) Calibrated date [to 2 sigma, 95% confidence] 4. OSL date table: Date UID Laboratory code and unique sample number Description of context De measurement and error Luminescence age Calendar date Comment/interpretation

Site name NGR Latitude and longitude Monument type(s) x 2 (based on historic England’s monuments thesaurus) Short reference

SMR/HER monument or event no. Dating method (e.g. radiocarbon, OSL) Site UID Project sample reference Description of material Objective for dating sample C error Measurement type [AMS/ radiometric] Comment/interpretation 14

Site UID Project’s sample no Objective for dating sample Total dose rate Age error Date of measurement

188

R. Johnston et al.

Fig. 13.1  Map showing the locations of the sites and the regions recorded in the database

13.3.3 Field System Categorisation Field systems form distinctive and readily observable spatial patterns that make them well suited to categorisation within morphological schemes. It is reasonably straightforward to distinguish field systems according to their relative sizes, the shape of the boundaries (whether straight or curvilinear, for instance), and the ways that they were constructed and used. While there is no single scheme for categorising prehistoric and historic field systems in England, there are recurring, widely accepted terms: e.g. aggregated, coaxial, Celtic, brickwork, assarts, open fields, or planned enclosure. Based on decades of archaeological survey and excavations, most of these categories are ascribed to broad time periods. For example, curvilinear aggregated fields in upland landscapes are predominately classified as later prehistoric. The regular patterns and straight boundaries of planned field systems that

continue in use in the present-day were largely created during the 18th and 19th centuries. These categories have been refined further to take account of specific regional and historical processes. An example of this can be found around some villages in Derbyshire and other Midland counties, where the post-medieval enclosure by agreement of open fields created narrow strip fields with distinctive curving boundaries, which preserve the layout and ploughing patterns of the medieval field systems (Hall 2012). The coaxial field systems on Dartmoor dating to the second millennium BC (termed reaves) are another, much earlier, instance of a field pattern that is a distinctive regional manifestation of a broader, national category (coaxial; Fleming 1978, 2008). The method of Historic Landscape Characterisation that has been applied across England during the last 15–20 years has played a major role in drawing out many of these regional schemes for classifying field systems, particularly medieval and later field systems (Turner 2006).

13  Understanding the Chronologies of England’s Field Systems Fig. 13.2  Examples of field system types commonly represented in the database: aggregated (Swillington Common), coaxial (Shaugh Moor/Saddlesborough), linear / pit alignment (Whitemoor Haye Quarry). (Plans adapted from Roberts et al. 2001: 48; Smith et al. 1981: 210; Coates 1999: Fig. 15)

189

Swillington Common West Yorkshire

N 100m

Whitemoor Haye Quarry, Staffordshire

Pit Ali gnme nt

Shaugh Moor, Devon

N 500m

While acknowledging the importance of the refined, regionally specific schemes that are available, especially for historic field systems, this project uses a more straightforward and higher-level framework for the data management and analysis. There are two key reasons for adopting this approach: maintaining a sufficient sample size within each category, and ensuring a degree of consistency between regional frameworks. There are too few scientific dates across England as a whole to support a more fine-grained categorical scheme. For example, distinguishing reaves on Dartmoor from other coaxial field systems or between curvilinear and rectilinear aggregated field systems would lead to lots of categories each with few dates. The variability in how different excavation reports, which form the project’s primary source of information, apply a variety of terminologies reflected regional and subjective preferences. The term ‘Celtic’, for example, is widely applied to later prehistoric field systems defined by lynchets on the chalk downland of southern England, which is the region where the label was first coined (Crawford 1923). Yet ‘Celtic fields’ also appear in descriptions of fields in County Durham (Bowes Moor field system – site 28; see Appendix for a list of sites) and Cornwall (Gwithian and Stencoose, sites 5 and 59).

N

100m

There are six types of site in the project’s database. They align closely with the types that David McOmish (2011) identifies in his introduction to field systems in England: coaxial/cohesive, regular and irregular accreted (treated as one category in the database  – ‘aggregated’), formal terraced, open, enclosed and parliamentary fields (Fig.  13.2). Field systems were (and are) not immutable structures in the landscape. They were lived in, cultivated, modified through decades and centuries, and they were shaped as much by natural processes as by social life. Any category applied to a field system will necessarily be reductive and mask variability through time and space. In many coaxial systems, for instance, the primary spinal boundaries or trackways create wide strips that are infilled with aggregated patterns of smaller fields. Given the potential variability within a field system and the large areas they cover, it can frequently be difficult to assign a category to an excavated field system. In some cases, the definition of the fields is limited by a lack of sufficient detail exposed in the excavated areas. In other examples, formation processes have affected the preservation of boundaries, such that only fragments of individual boundaries are visible. These are common limitations with the project’s dataset, and the types ‘field boundary’ and ‘uncategorised’ are, combined, the most common site type.

190

13.3.4 Dating Methods Three dating methods were recorded during the project, with the most commonly used being radiocarbon dating, with OSL occasionally used, and one instance of thermoluminescence (TL) dating. We calibrated all the radiocarbon measurements collated during the project using OxCal (version 4.2) and the IntCal 13 radiocarbon calibration curve (Bronk Ramsey 2009; Reimer et al. 2013). The dates are quoted at 2 sigma (95%) confidence, and to avoid a false perception of precision, the dates were rounded outwards to 10 years, or 5 years where the error is less than 25 years (following conventions described in Bayliss and Marshall in prep.). There are currently no set specifications for the reporting of luminescence ages (Duller 2008). We only included limited measurement data in the database: the De measurement and the luminescence age. Calendar years have been calculated from the luminescence age and error range and rounded out to the nearest 10 years. In many cases the measurement year was not recorded and this has been estimated based on publication date.

13.3.5 Biases There were biases affecting the completeness and representativeness of the database. A primary factor was the difference in responses from regional monument databases, their levels of recording scientific dating information and the extent to which their data was up-to-date. Our consultation with contracting units was intended to address HER bias, but only a small number of units responded positively to our requests. The poor response rate provided its own skewing factor. For example, the large number of sites recorded from Cornwall (southwest England) and Cambridgeshire (east) partly reflects the considerable assistance provided by the Cornwall and Cambridge Archaeological Units (although many excavations of field systems in these counties were also identified through searches of publications). We also found biases created by the uneven distribution of infrastructure and housing development and aggregate extraction. The rise in development-led excavations has greatly increased the available information on field systems, but it is driven by modern economic activity rather than research priorities. The corollary of this is that regions where modern development is limited or small-scale are poorly represented in the data. This includes protected areas such as national parks and Areas of Outstanding Natural Beauty, where most field systems have been recorded through survey and dated on morphological or stratigraphic grounds. The precise dating of land enclosure is not seen as a research priority in some regions, which results in further local and regional biases. We identified several projects

R. Johnston et al.

where excavated field boundaries were treated as the lowest priority for scientific dating, at least partly due to difficulties in obtaining datable samples. The availability of suitable samples also creates regional biases, with well-preserved, often waterlogged features in the southeast and east of England providing a greater number of secure dates than the more limited survival of remains on well-drained sands and gravels in the Midlands and north. Period bias is marked by an over-representation of later prehistoric dates (second and first millennia BC) and very few dates from Roman to post-­ medieval periods (AD 43-1900). This is often due to a reliance on artefactual material for dating later periods.

13.4 Dating Methods and Materials The majority of the scientific dates were derived from radiocarbon analyses (380), with a small number of OSL dates (12) and one instance of a TL measurement. Most of the 380 radiocarbon dates recorded during the project were dated by AMS measurement (77%). This reflects the appropriateness of AMS for dating small, single-entity, samples such as charcoal fragments, individual bones and plant macrofossils. Radiometric dates (21%) are more common for older projects, prior to the development of AMS dating in the 1980s, although radiometric dating was utilised by more recent projects for sediments such as peat or larger pieces of waterlogged wood. Projects chose a variety of materials for dating, with charcoal being the most common, followed by plant macrofossils (grains, seeds, nutshells and fruit stones), waterlogged wood and carbonised residue on the interior of pottery sherds. Animal and human bones were also dated, with the relatively low numbers perhaps reflecting the difficulties in establishing if the bone was residual or contemporary with the use of the boundary. Articulated bones should be more reliable as they are unlikely to have moved from other, older, deposits. That said, anomalies do occur, as at Lynton Way, Sawston, where bone from a partially complete dog skeleton from a pit cutting (and therefore younger than) an enclosure ditch returned a date older than the ditch (Weston et  al. 2007). Groups of similar bones, possibly from the same animal, were found in three deposits within the ditch. A minority of the dated material comprised organic sediments (peat) and two humic acid samples – the latter were both from waterholes investigated during the Heathrow Terminal 5 excavations and returned results that were significantly different from waterlogged wood in the same features (site 85; Appendix). Peat has primarily been used for dating stone or earthwork boundaries, reflecting the predominance of upstanding field boundaries in moorland areas (where peat is a common). The excavators interpreted the dates from peat underlying boundaries as providing terminus post quem

13  Understanding the Chronologies of England’s Field Systems

(TPQs) for the field boundaries. Occasionally, peat was sampled from within ditches, and it is generally assumed to provide a terminus ante quem (TAQ) for the boundary. The database includes four dates on sediment used to date soil horizons, including material accumulated within ard cultivation marks cut into a buried soil (Lafone St, Southwark – site 72; Appendix). In some cases, the soil horizons are difficult to relate to the field boundaries they are dating, as at the Coach and Horses Earthwork on the A66 excavations (site 29).

13.5 Dating Strategies The majority (85%) of the 120 sites provided small numbers (five or fewer) scientific dates for field boundaries. Of these, just over half provided 2 to 5 dates (46 per cent of total sites), with 36 per cent of the total sites recorded in the project providing only one date directly associated with a boundary. Some of the low figures represented small sites uncovered during corridor investigations, such as road and pipeline schemes, where only the limited extents of boundary features were excavated (e.g. the Asselby to Pannal pipeline in North and West Yorkshire, sites 9–17; Appendix for site details). Other factors affecting the low numbers of dates are a lack of suitable samples for dating, and excavators’ decisions to prioritise artefactual chronologies and only use scientific dating as an additional test. The proportions of sites with 6 to 10 (8 per cent) and 11 or more (7 per cent) dates represent the minority of scientifically-dated field systems. These are mainly large area excavations undertaken in advance of development or quarrying, such as Perry Oaks (Heathrow Terminal 5; site 85), Bestwall and Huntsman’s Quarries (sites 4 and 76) and Sewerby Cottage Farm (site 94). The only small-area project with over 11 dates is Gwithian (site 5), which is a long-term research and rescue excavation from the mid-twentieth century that was reassessed in the early 2000s. Many of the dates associated with field ditches derive from material from slow infilling or deliberate backfilling deposits, and can only be used to date the end of use of the field systems rather than their constructions or periods of use. In some cases, burials or cremations inserted into ditch fills have been dated, their locations indicating that the ditch was still a visible feature in the landscape, though possibly infilled to the extent that it no longer formed part of a working field system. Ditches may have had associated banks, fences and hedges, which have subsequently been lost through post-depositional erosion and destruction. At the Elliott site, Fengate (site 64), waterlogged hedging material and stakes dumped in a pit pre-dated any of the dates from the field and enclosure ditches. The wood was assumed to be material associated with field boundary hedges, and may

191

provide information on the early development of the field system, though the relationship remains unproven and the dates are significantly earlier than those from ditch fills (Evans et al. 2009). In many cases, ditches were cleaned and recut over a long period of time, which can add further complications to dating the origin of the field system. At Old Rydon Lane, Exeter (site 106), the field ditches did not appear to have been kept clear and were thought to have been dug for material to construct hedge banks running alongside, with a later episode of recutting perhaps associated with strengthening of the banks. In southern England, dating of waterholes has been increasingly undertaken to date associated field systems where the ditch fills contain low quantities of artefactual and ecofactual material. The waterholes may have direct association with field boundaries that cut, were cut by, respected or were respected by them, and waterlogging of the fills has the potential to preserve material suitable for single-entity sampling (for example, at Heathrow Terminal 5; site 85). Material such as wooden stakes from revetments can give accurate dates for the construction of the waterholes, although these may only provide a date for before or after an associated field ditch was dug. At Huntsman’s Quarry, Kemerton, dating of waterholes and settlement remains suggests that the Bronze Age fields may have been laid out c. 100 years before settlement was established within the field system (site 76). This has implications for the interpretation of other sites, since the physical proximity of field boundaries and domestic structures need not mean that they had contemporaneous origins. Scientific dating of earth or stone banks has been undertaken at 21 sites, representing 5 per cent of the sites recorded in the database. Earthwork field boundaries survive in landscape zones, such as upland and heathland, where there has not been historic cultivation and intensive grazing. These are areas largely outside zones of modern development and aggregates extraction and so rarely become the object of development-led excavation. Most earth and stone field boundaries were radiocarbon dated using charcoal or sediment samples, with only one instance of a charred grain and two dates on waterlogged wood from stakes forming part of the boundary feature. Radiocarbon dates from old land surfaces underlying stone and earth banks provide TPQs for a large proportion of the dates from sites. At Big Moor (site 2) dates from a pit sealed by a bank were unexpectedly early (Mesolithic), and therefore unhelpful in dating the bank. Some dates have been taken on sediment overlying or adjacent to earthworks, providing TAQs for their construction and use. The problem with TAQs is illustrated at Simy Folds (site 96), where samples were taken from a slag heap overlying a boundary wall and hearths within two buildings which were built on top of another boundary. These provided early medieval dates. On the basis of pottery sherds found during the excavations and

192

a radiocarbon date from another building, the excavators thought it possible that the boundaries were significantly earlier, possibly prehistoric, and reused to house a small settlement in the early medieval period. Most luminescence dates were associated with earth or stone boundary features (nine of the 13 dates, from three sites). These dates were on samples taken from buried surfaces underlying banks (at Shovel Down; site 45), deposits accumulated within a stony boundary (at Big Moor; site 2), or wind-blown sand deposited between periods of cultivation (at Gwithian, site 5). At Big Moor (site 2), two of the dates from the core of the bank and overlying accumulated colluvium were stratigraphically consistent, but two others from the later sediment were anomalously early and probably relate to the geomorphological formation of the sediment (Heath 2003). Both samples were low in the profile of the outer edge of the bank and close to the old land surface, and they may have been insufficiently bleached during d­ eposition through solifluction. At Gwithian (site 5), both radiocarbon and OSL dating were used to refine the chronology of a series of occupation and agricultural deposits, interleaved by wind-blown sand layers. The OSL dates for two sand deposits separated by a ploughsoil deposit were broadly contemporary with radiocarbon dates from carbonised residue on pottery sherds from occupation/cultivation horizons, and were consistent with each other within 1 sigma errors, but the lower deposit (layer 6, Aber-101 GWT-6 3360 ± 160a, 1515–1190 BC) returned a more recent date than the upper deposit (layer 4, UID 1379, Aber-101 GWT-4 3650 ± 160a, 1805–1480 BC). This anomaly, despite the relatively high precision of the dates, was interpreted as being due to rapid deposition of sand, with only a brief period of stabilisation in between, due to cultivation.

13.6 Geographic Variation While affected by some data collection biases, the distribution of sites recorded during the project reflects patterns of both excavated field systems and scientific dating across England. In order to give an indication of the reliability of the coverage of scientifically dated fields, a comparison was Fig. 13.3  Chart comparing the numbers of scientificallydated sites by regions

R. Johnston et al.

made between the collected scientific dates and the distribution of excavated or evaluated field system sites reported by AIP (1990–2010). For the thirteen counties with no scientific dates in the database, nine also have either no or fewer than five projects with associated field systems on AIP.  The remaining counties, which have five or more projects on AIP, are Nottinghamshire, Shropshire, Suffolk and Warwickshire. The spatial distribution shows a significant lack of scientifically dated field systems in the Midlands region, with only eight sites recorded (Fig. 13.3). Relatively few field-related projects are recorded in AIP for the western and central Midlands (Merseyside, Cheshire, Derbyshire, Staffordshire, Shropshire, Worcestershire). It is surprising that few scientifically dated sites have been recorded from Nottinghamshire and Lincolnshire, where ditched field systems are a common feature recorded during sand and gravel quarrying. Only one site with scientific dates was identified from Lincolnshire and no scientifically dated examples are known from Nottinghamshire. Ongoing quarry projects may be a factor here, with post-excavation analysis and full publication sometimes awaiting the end of staged phases of excavation. A backlog of publication of sites was previously noted in the East Midlands Resource Assessment (Cooper 2006). The low numbers of dates from Northamptonshire fields is also important, as the HER records at least 27 excavated sites with field-related features, though only two of these have associated scientific dates (sites 100–101). Local contracting units reported that dating of fields in this region is mainly achieved through artefactual material and stratigraphic relationships. The larger number of projects from Yorkshire and Humberside is partly a result of access to data, with West Yorkshire being one of the few HERs which has searchable scientific dates, and partly a consequence of large-scale linear excavation projects with several scientifically-dated field-systems (e.g. the M1-A1 link road and the Asselby to Pannal pipeline; Appendix sites 9–17). The region is well-­ known for its extensive later prehistoric and Roman field systems (Chadwick 2008). The lack of datable artefacts in many prehistoric to Roman field ditches in this period is also a factor in the number of scientific dates.

13  Understanding the Chronologies of England’s Field Systems

Scientifically dated sites in the East region are dominated by Cambridgeshire (including Peterborough), where good preservation of field boundaries and waterlogged deposits have combined with large-scale quarrying and development projects and a strong research agenda relating to Bronze Age field systems. This pattern is also reflected in the AIP data, where Cambridgeshire has 50 more relevant project reports than any other county. Further to the east there are few dated sites, with only one dated site recorded in Norfolk. This ­pattern was noted by Yates (2007, 81), who identified no sites with scientific (or artefactual) dates for Bronze Age fields from Norfolk, although information from Norfolk HER suggests that scientific dating of field system is being undertaken on at least two sites that are yet to be published. In the South East, 22 sites have been scientifically dated, particularly in the Thames Valley, reflecting the location of large-scale development and quarrying projects as well as good preservation of buried archaeology. This includes a cluster of sites to the west of London, in Berkshire. Within Greater London, three smaller sites with evidence for field systems or stock enclosures have associated scientific dates. Relatively few sites from Kent have been recorded. Although Kent HER were unable to provide any data, three sites were identified from local journals and publications, mainly concentrated on the east coast, near Sheppey. Along the southern coast, there are two clusters of dated sites around Bognor Regis and Southampton. Thirty-one sites have been recorded in the South West, particularly in Cornwall. In Devon, the majority are on Dartmoor, principally from the Shaugh Moor rescue project (Appendix: sites 40, 42–44), with a recent increase in scientifically dated ditched field systems in areas outside the moors. In Cornwall, the scientifically dated sites are spread relatively widely and many derive from recent development-­ led projects. Several sites have also been recorded on the Isles of Scilly  – generally research and rescue excavations associated with field walls exposed by coastal erosion. The only relevant sites identified in Wiltshire were from the Wessex Linear Ditches Project (sites 107–113), mainly datFig. 13.4  Chart comparing the numbers of scientificallydated sites by regions and the primary periods recorded in the database

193

ing large-scale land division, although a pit alignment and field boundaries pre-dating the linear ditches were also investigated.

13.7 Chronological Variation by Region There were clear regional differences in the main periods to which fields and linear boundaries have been scientifically dated (Fig.  13.4). In the South West, Midlands and East regions, Bronze Age dates (2300–700  BC) dominate the dataset, with sites which have the Bronze Age as the primary dated period representing between 61 per cent and 83 per cent of dated sites. The pattern in the South East is different, with Bronze Age and Iron Age sites each forming 41 per cent of the dataset. In the Yorkshire and Humber and North regions, there are much lower proportions of dated Bronze Age sites, between 4 per cent and 7 per cent. The Iron Age (700 BC–AD 43) is the most common primary period of field systems in the Yorkshire and Humber region (61 per cent), and forms a significant portion of the dates from all regions apart from the East, ranging from 27 per cent in the North to 41 per cent in the South East. The high proportion of Iron Age dates from Yorkshire in part reflects a general lack of artefacts from Iron Age field ditches in this area, with scientific dating providing the only means of establishing a chronology for the features. Yorkshire is also the region with the greatest proportion of sites with Roman dates (23 per cent), with no other regions having greater than 7 per cent of dates, and the Midlands and East regions having no scientifically dated sites from the Roman period (AD 43–450). Early medieval dates (AD 450–1066) are more common than medieval dates (AD 1066–1485) in all areas apart from the North, and generally range between 4 to 9 per cent, with no dates from any medieval period in the Midlands and South West. In the North, the medieval and early medieval periods combined make up just over half of the sites. The North and Yorkshire and Humber regions also have the wid-

194

est spread of dates from different periods, each having examples of all but one of the period categories. The only site where the primary dated period is post-medieval is from the Asselby to Pannal Pipeline, North Yorkshire, where two presumed Iron Age ditches gave consistent post-medieval dates and are thought to be associated with an eighteenthcentury enclosure (site 9).

13.8 Chronological Variation by Type 13.8.1 Aggregated Fields Aggregated field systems developed piecemeal, without adherence to an axial alignment or through the application of a predefined plan over a large area. They can be comprised of irregular, curving and sometimes discontinuous boundaries, or more regular, rectilinear arrangements of fields. The enclosed fields are often less than 0.5 ha, and a field system may only cover a relatively small area (a few hectares). Some of the most impressive examples of later prehistoric aggregated field systems can be found on the granite moorland of Devon and Cornwall: Bodmin Moor, Dartmoor and West Penwith (Johnson and Rose 1994; Newman 2011; Herring et  al. 2016). Processes of piecemeal, aggregated enclosure appear to have continued throughout much of the last 4000  years, with regular and irregular aggregated fields found alongside settlements of later prehistoric through to medieval date in many areas. The project database includes 104 scientific dates from 15 sites that are categorised as aggregated (Fig. 13.5). The earliest dates calibrated to the first half of the second millennium BC and are from both upland and lowland landscapes in southern England and the Midlands. At Eaglestone Flat (site 35), a group of cremation burials were excavated alongside the stone boundaries of a fragmentary field system. The TPQ for the boundaries (based on charcoal underlying the banks) is c. 1900–1600 BC, with a TAQ of 1700–1300 BC. The two Fig. 13.5  Chart comparing the numbers of sites and scientific dates within each category of field system

R. Johnston et al.

radiocarbon dates from peat below a field wall on Stannon Down (site 39) provided a TPQ for the boundary of c. 1700– 1500 BC. A stronger group of dates are available for the first phase field system at Bestwall Quarry (site 4). These are from carbonised residues on pottery, short-lived, identifiable charcoal and charred seeds. The dates provide a period of use for the ditches in 2000–1500 BC, with a more likely range of 1750–1500 – an early date from residue on a sherd of Beaker pottery has a very wide error range and may be residual in the ditch. The dates from the first phase field systems at Bestwall Quarry are similar to those from Gwithian (Cornwall), where the earliest phase of settlement with a terraced field system (layer 8) is dated to 1890– 1620 BC. However, this is based on one date and the other two dates from this phase were interpreted as too late, and the result of intrusive material. There were subsequent phases of aggregated fields at both Bestwall Quarry and Gwithian. At Bestwall Quarry, the majority of dates (12) fall within 1750–1300, with the second phase of field systems dated to 1500–1300  BC (three further dates are interpreted as later material deposited in the upper fills of still visible ditches). A similar range is covered by the dates from Big Moor (site 2) and the farmsteads at Terminal 5 Heathrow (site 85). At Gwithian, the latest phase of fields (layer 3) spans 1400–900 BC, making it one of the few field systems to provide scientific dates that span the second and first millennia BC. A second cluster of aggregated field systems are dated to the second half of the first millennium BC: North Rigton (site 14), Lockington Quarry (site 99) and Great Doddington (site 100). In all three cases, the TAQs for the field systems are pre-Roman. Only Lockington Quarry has material evidence of occupation continuing into the Roman period, although this is not visible in the radiocarbon chronology. Following this, there are isolated dates in the Roman period for aggregated field systems at Stencoose (site 59) and Ashville Trading Estate (site 48). At Bowes Moor (site 28), a sequence of four dates associated with an aggregated field

13  Understanding the Chronologies of England’s Field Systems

system span AD 30–1160. However, the dates are from organic-rich horizons within the field soil: a thin buried A-horizon and an overlying peat. These dates provide a TAQ for the construction of the boundaries.

13.8.2 Coaxial Fields Coaxial field systems have dominant alignments that structure the development of the fields. These axes may continue for hundreds of metres and in many cases can be traced across several kilometres of landscape. Andrew Fleming (1978) mapped a continuous coaxial field system covering 3000  ha and 6  km in length around Dartmeet, Dartmoor (Devon). Coaxial field systems are not attributable to a specific period in time. Coaxial field systems in southern England originate in the middle and late second millennium BC (Yates 2007). Iron Age and Roman coaxial landscapes are recognised in the East Midlands and northern England (Chadwick 2008), whilst coaxial boundaries have been shown to structure landscape development from later prehistory into the Middle Ages, for example in west Cambridgeshire (Oosthuizen 2003). There are 128 scientific dates from 33 sites categorised as ‘coaxial’ in the database. These include field systems that were categorised as cohesive, brickwork and rectilinear. In broad terms, the scientific dates support the wide chronology for coaxial field systems that is proposed in Fleming’s (2008, 159–86) review, as they span the second millennium BC to the first millennium AD. There were notable intensities and lacuna within this time span. The earliest dates for coaxial field systems are from Shaugh Moor and Holne Moor (sites 40–44), Elliott Site, Fengate (site 64), and Thanet (site 67); these pre-date 1700  BC and some have a calibrated range pre-dating 2000 BC. In most cases, there are reasons to question whether the dates provide a credible estimate of the period when the field boundaries were built and in use. With one exception, the dates from boundaries on Shaugh Moor and Holne Moor are all TPQs from bulk peat samples and charcoal. The exception to this is a wooden stake recovered from the boundary ditch on Shaugh Moor, which is interpreted as the remains of timber fence that formed the first phase of the boundary. The calibrated radiocarbon date is 1890–1430 BC (site 40). Of similar character, the earliest dates from Elliot Site, Fengate, came from worked wood in a pit with hedge material and stakes (site 64). They calibrate to the final centuries of the third millennium BC. The three dates from the field system at Monkton Road, Thanet (Kent, site 67), are tightly grouped within c. 1900–1700 BC, and appear to provide a reliable estimate for activity contemporary with the early infilling of the ditches. The majority of second millennium BC coaxial field systems fall after 1600 BC and before 1000 BC. These include

195

Clay Farm (site 78), Colne Fen (site 57), Castle Hill (site 89), Heathrow (site 85) and Kemerton (site 76). At Edgerley Drain Road, Fengate, the dates were on samples in pits in stratigraphic association with the field boundaries. The most reliable of these (1640–1440 BC – site 68) is on bone from a cow burial that post-dates the infilling of the boundary. The duration of the chronologies for the second millennium BC coaxial field systems is closely correlated with the numbers of scientific dates from each site. Clay Farm, Colne Fen, Tower’s Fen, and Castle Hill have relatively short chronologies based on the scientific dating  – a couple of centuries after 1500 BC – and these are based on only two dates from the same or associated features in most cases. By comparison, the chronologies for Newark Road (site 80), Cranford Lane (site 95) and Kemerton (site 76) are longer (up to 1700–1000 BC), and based on 11 to 14 scientific dates from each site. Of the 27 coaxial field systems, only two have chronological boundaries that lie predominantly within c.1000 and 500 BC: Cranford Lane (site 95) and Trenowah (site 50). At Cranford Lane, there are two dates from a well within the field system that are interpreted as the latest use of the field systems during 800–600  BC.  At Trenowah, there are two Early Iron Age dates from an infilled boundary ditch, c.800–400 BC. The chronology of the field systems at Trenowah spans the first millennium BC, and may continue into the Roman period. It is one of nine coaxial field systems that are dated within 500  BC–AD 900. With the exception of Trenowah and Tremough (site 49) and Wrotham Quarry (site 90) these are all in the Midlands and North, from Lincolnshire to County Durham, and with the majority in South and West Yorkshire, in an area recognised as rich in later Iron Age and Roman field systems (Chadwick 2008). The concentration of coaxial field systems of this date in the North and Midlands primarily relates to the lack of material culture, specifically ceramics, from excavations. This makes scientific dating the only tool for establishing chronological frameworks. One coaxial field system has provided post-Roman dates: Ravock (site 30). At this site, the medieval and post-medieval dates from a lens of soil in a boundary and from peat overlying another boundary do not correspond with the excavators’ interpretation of the boundaries as much older and possibly prehistoric.

13.8.3 Linear Land Division (Including Pit Alignments) The category of linear land division includes linear banks and ditches covering long distances, as well as pit alignments that do not appear to be associated with a ceremonial function (Wigley 2007). In the South West region, it has been noted that pit alignments and linear ditches combined to form larger

196

landscape divisions, possibly representing different ways of marking the same boundary perhaps in areas of different ground conditions and topography (Fitzpatrick 2007). There are 35 dates from 16 linear boundaries (including six pit alignments) within the database, spanning 1400 BC to AD 600. The linear banks and ditches dated by the Wessex Linear Ditch Project (sites 108 to 113) fall largely between 600–100  BC.  One date from animal bone at Weather Hill (1220–800 BC) is interpreted by the excavators as residual, whilst a late date from a deliberate deposition in one of the Sidbury Double Linear Ditch terminals appeared to indicate that the feature remained prominent over a long period. At Ardleigh, Essex, a series of dates on a pair of ditches span 1490–830 BC; these ditches were interpreted by the excavators as more characteristic of large-scale linear land division rather than field boundaries, although only a limited portion of each was recorded. The only post-Iron Age date comes from the Wether Hill site at Ingram Valley (site 117) but this was considered to be a TAQ, derived from peat formation in the base of the ditch, and cannot be considered reliable. A date from turf sealed by the associated bank dated to 400–40 BC. Of the six pit alignments included in the dataset, only two have multiple, stratigraphically related dates. At Holmfield, West Yorkshire, the pits were dug around c. 40 BC–AD 50, with subsequent human burials inserted during the first century AD (site 24). The dates from three pits at Redscar Wood (site 73) span the first to sixth centuries AD. The dates from within each pit were consistent with each other, and those between two pits were also similar (Roman), though the dates the third pit were fifth to sixth century AD. This was interpreted as representing either maintenance of the boundary over a long period of time, or two phases of boundary construction. Single dates from individual pits were taken from site WWBP on the A1[M] Darrington to Dishforth road scheme (site 22), South Elmsall (site 7), and Haxton Down (site 112). All of these fall into the range 400–150 BC, though the differences in dates from the sites with multiple determinations indicate that single dates cannot be taken as definitive dating of pit alignments.

13.8.4 Uncategorised Boundaries and Field Systems There are 45 scientific dates from uncategorised boundaries and field systems that were collected during the project. Considering the broad chronological distribution of these dates, 13 sites fall within the Middle Bronze Age, four in the Late Bronze Age, 14  in the Iron Age, six in the Roman period, two post-Roman, and six medieval or later. This distribution of dates is broadly consistent with the pattern from the categorised field systems, most notably with

R. Johnston et al.

a gap in the Late Bronze Age/Early Iron Age (c. 1000– 600 BC). The four sites in the Late Bronze Age/Early Iron Age period mainly have single dates. These include a short stretch of L-shaped ditch at Priors Hall (site 101) and Shrubsoles Hill (site 55), where the date is from an enclosure cutting an earlier field ditch system. The date from Dairy Lane (site 91) comes from unspecified mixed fragments of charcoal from various levels within the ditch fill of a system thought to be Middle Bronze Age on the basis of associated pottery, and must be considered dubious, and at Shotton Village (site 74) the only date is from alder charcoal from the fill of a ditch forming part of a T-junction. A small number of uncategorised sites produced early medieval (post-Roman) and medieval to post-medieval dates. The segmented, curvilinear ditches at Newton Bewley (site 66) returned early medieval dates, as did a ditch at Vicarage Lane, Romford (site 71), which appeared to have performed both field boundary and drainage functions. A ditch at Whitelands Farm (site 102) was on the same alignment as Iron Age and Roman field ditches but contained Anglo-­ Saxon pottery consistent with the radiocarbon determination. Four dates from three presumed Iron Age ditches at Rigton Bank (site 15) returned consistent early medieval dates, but were considered to be from intrusive samples. Early medieval dates for activity within stone-built field and enclosure boundaries at Simy Folds (site 96) related to structures overlying the field boundaries and provide a TAQ, though the excavators thought the original boundaries may have been prehistoric, possibly Bronze Age in date. Medieval dates were obtained from a fenced boundary and coppiced brushwood exposed by peat erosion in the Seathwaite Valley (site 46), and from a probable field boundary ditch at the GlaxoSmithCline U–Building site (site 118). Ditches creating a funnelled droveway, probably for bringing sheep or cattle from common ground to ditched enclosures at Lancaster University (site 84) also returned medieval dates, supported by pottery. Stock enclosures and field boundaries of early medieval to medieval date suggested two phases of occupation at Boundary Lane, Whittonstall (site 79), from the eighth to ninth century and eleventh to early twelfth centuries AD.  This was rare evidence for agricultural activity from this period in the North Pennines. Only one ditch was scientifically dated to this period, with a post-medieval date recovered for another boundary phase. Two post-medieval dates were obtained from parallel rock-cut ditches excavated on the Asselby to Pannal pipeline (site 9), and these were interpreted as eighteenth-century enclosure boundaries. Post-medieval dates derived from human bones from two separate ditches at Normanton (site 8) appeared to be intrusive in the contexts, although the similarity of dates suggested that they may have derived from a post-medieval boundary seen crossing the site on historic mapping, which was not visible archaeologically perhaps due to geological conditions.

13  Understanding the Chronologies of England’s Field Systems

13.9 Discussion and Conclusions 13.9.1 Historical Narratives The project recorded 393 scientific dates from 120 sites, with dates covering a variety of different field and boundary types, as well as large scale linear land division, and from all periods ranging from Early Bronze Age to post-medieval. The results confirm that bounded field systems appeared in the English landscape during the first quarter of the second millennium BC and subsequent enclosure was an ongoing, geographically discontinuous process. Large-scale apportionment of land into field systems began in the early centuries of the second millennium BC and became more commonplace after 1700 BC. This revises the long-held chronology of Bronze Age field systems, which placed their inception at around 1600/1500 BC – the Middle Bronze Age (English Heritage 2012; Yates 2007). While individual excavation narratives have noted potential Early Bronze Age origins for field systems, the scientific dating indicates that large-scale land enclosure should be recognised as a phenomenon of the early and middle second millennium BC. A further important insight from the project is that the active apportionment of land and maintenance of field systems was possibly interrupted during c. 1000–600 BC (the Late Bronze Age and earliest Iron Age). This hiatus was observed uniformly across all categories of field systems and throughout the country in the project’s data. The scientific dates from linear boundaries do not fill this gap in the chronology. It is recommended that future research seeks to test this observation that there was a hiatus in land enclosure during c. 1000–600 BC. There is a general low number of scientific dates for Roman field systems across most regions, while post-Roman and medieval fields remain largely inaccessible through scientific dating. Even at sites where there have been strong scientific dating programmes, these have tended to focus on early periods, with later features dated through artefactual material or stratigraphic relationships. Roman, medieval and later landscapes are poorly served by scientific dating, and this is rarely addressed in regional research frameworks. Recent research has suggested that there is a strong case for continuity of field systems from the Roman periods onwards in some areas (Rippon et al. 2015; Oosthuizen 2013), and further scientific dating from these periods could assist in refining these chronologies and understanding these processes.

13.9.2 Methodologies and Data Standards There are a number of biases in the current dataset of scientific dates, many arising from the difficulties in  locating information on projects with scientific dating programmes

197

through HERs and online data repositories. Although more information is now available online, including grey literature reports, it is rare for scientific dating to be recorded in searchable fields. The project has illustrated the requirement for standardised reporting of scientific dates in grey literature reports and publications. This is important, as missing information was frequently found, and this affects the usability of the measurements for appraisal or reassessment of the reliability of the dating. The current lack of regional or national records of scientific dates hinders the potential to undertake wide-scale research or to find comparative dates for excavated sites. A clear strategy for recording and flagging this information, ideally on a national basis, would be an invaluable research and dissemination tool. Geographical biases within the data occur due to the nature of development-led archaeology. Most data arise from regions commonly subject to urban expansion, quarrying and infrastructure projects. There is an attendant under-­ representation of more marginal and upland areas where the majority of fields are recorded through aerial or topographic survey, with dating poorly understood. Features such as the Dartmoor reaves remain poorly dated despite their significance and long history of research. Consequently, we recommend that research funding should be targeted on investigations into field systems in such areas, with provision made for adequate scientific dating programmes. Regional priorities and research agendas influence the scientific dating of field-related features. This is partly period-based, with most regional resource assessments mentioning the desirability of scientific dates for prehistoric fields, but few considering this a priority for post-Roman or later field systems, even where it is recognised that artefactual dating is limited or problematic. The decision to employ scientific dating is influenced by a variety of factors, including funding, the availability of other sources of chronological control, particularly material culture and stratigraphic information. This means that scientific dating is most prevalent where pottery, particularly, is scarce or ceramic chronologies poorly refined. The impact of different sampling and dating strategies can be significant. For coaxial field systems, there is a correlation between the numbers of dates taken for sites and the length of the chronology for the field system. Higher numbers of dates taken tend to correlate with longer chronologies. This implies that on sites where few scientific dates are taken, chronologies are artificially low. There are relatively few sites with multiple (more than 5) dates and more work on this aspect would be needed to understand whether this is the result of a small dataset or a wider phenomenon. The majority of development-led projects understand the appropriate sampling strategies for scientific dating. For smaller projects, or those with limited funds for post-­ excavation work, it appears that few samples are submitted

198

R. Johnston et al.

for scientific dating, and these may be insufficient to provide reliable chronologies. In many areas there is a dearth of scientific dating undertaken for samples relating to Roman and later periods, even where there is limited artefactual material to date features and deposits. There is an opportunity to communicate the importance of dating field systems to the planning archaeologists who can specify the requirements for robust scientific dating strategies and chronological modelling in fieldwork briefs. This will enable the provision for scientific dating to be built into development-led projects at an early stage. The ability to undertake such dating programmes will depend on the presence of suitable contexts and dating material, but greater expectations that this work will be required could lead to more projects incorporating field systems into their scientific dating programmes. Particular attention should be paid to encouraging this dating in regions where field systems are commonly excavated but rarely dated scientifically. Whilst radiocarbon dating has attendant problems of residuality that can make interpretation of results difficult, OSL offers the potential for directly dating sediments within banks and ditches (Arnoldussen, this volume). The utilisation of OSL for dating field systems and linear land division features is currently rare in development-led archaeology. Although results can be variable and age ranges in p­ rehistoric periods are relatively wide, it offers a potential for dating features for which no other dating material is available, and can be used as a comparison with radiocarbon dates from features. It also has the potential to offer more precise dates

for medieval and post-medieval features. This technique should be more widely promoted as a tool for dating field systems and linear land division. In conclusion, the deep chronology and longevity of field systems is an important characteristic of the English landscape. Prehistories and histories of land enclosure are accessible through the material remains of field systems surviving throughout the countryside. There is high potential to research long term and large-scale transformations in landscape organisation through the study and dating of field systems. Designation, research frameworks and fieldwork strategies should prioritise these long chronologies and rebalance our current excessive focus towards periodised perspectives. Scientific dates and chronological modelling offer the data and the tools to understand some of the complexities in formation of England’s field systems. Acknowledgements  Many people provided advice and assistance during this project: Peter Marshall and Pete Wilson of Historic England; the HER, SMR and local authority archaeologists who responded to our complicated data requests; the contracting units who provided access to data, and especially Chris Evans and Sam Smith of Cambridge Archaeological Unit, Peter Dudley and Andrew Young of Cornwall Archaeological Unit, Diccon Hart of ASE, Mary Ruddy and Jim Brown of MOLA, Rachel Newman and Denise Druce of Oxford Archaeology North, and Steve Preston of TVAS; David Yates and David Knight for information and advice; the helpful staff of the Historic England library.

Appendix: Table of Sites

ID 1

Site name Chysauster

County Cornwall

Types (primary listed first) Aggregated

Periods (primary listed first) Bronze Age, Roman

2

Big Moor

Derbyshire

aggregated

Bronze Age

3

North Lincolnshire

Coaxial

Iron Age, Roman

4

Timberland, Scunthorpe Bestwall Quarry

References Walker et al. (1990) and Smith et al. (1996) Barnatt (2001), Heath (2003) and Bayliss et al. (2013) Richardson (2009)

Dorset

Aggregated

Bronze Age, Iron Age

Bayliss et al. (2009a)

5

Gwithian

Cornwall

Aggregated

Bronze Age

6

Moss Carr, Methley

West Yorkshire

Field system

Iron Age, Roman

Nowakowski et al. (2008), Nowakowski (2007), Bayliss et al. (2008) and Roberts (2007) Roberts and Richardson (2002)

7

Doncaster Road, South Elmsall Normanton Industrial Estate AP30 (Asselby to Pannal Pipeline)

West Yorkshire

Field boundary, pit alignment

Iron Age, medieval

Grassam (2008)

West Yorkshire

Field system

Roman, Early medieval

Martin and Harrison (2012)

North Yorkshire

Field boundary

Post-medieval

Gregory et al. (2013)

8 9

(continued)

13  Understanding the Chronologies of England’s Field Systems

ID 10

11

12

13

14

15

16

17

18

19

20

21

22

23 24

25 26 27

Site name Becca Banks, Aberford (Asselby to Pannal pipeline Area 8) Grove Farm, Little Fenton (Asselby to Pannal pipeline site 2) Oldgate Farm, Barkston Ash (Asselby to Pannal pipeline site 18–10) Coldhill Lane, Saxton (Asselby to Pannal pipeline site 18-11B) Rigton Bank, Bardsey (Asselby to Pannal pipeline site 26–2) Rigton Bank, Bardsey (Asselby to Pannal pipeline site 26–3) Stainburn Hill, Kirkby Overblow (Asselby to Pannal pipeline Area 14) North Rigton (Asselby to Pannal pipeline Area 15) Bullerthorpe Lane, Swillington (A1-M1 link road) Swillington Common (A1-M1 link road) Knottingley Road retention pond (A1[M] Darrington to Dishforth) Site M, Castle Hills, Ledsham (A1[M] Darrington to Dishforth) Site WWBP, north of Wetherby (A1[M] Darrington to Dishforth) Wattle Syke, Wetherby Holmfield Interchange, Ferrybridge Glasshoughton Coalfields Link Rd Carr Lodge Farm, Loversall Catesby Business Park, Balby Carr

199 Periods (primary listed first) Bronze Age, Iron Age

References Gregory et al. (2013)

Coaxial

Roman

Gregory et al. (2013)

North Yorkshire

Enclosure

Iron Age

Gregory et al. (2013)

North Yorkshire

stock enclosure

Early medieval

Gregory et al. (2013)

West Yorkshire

Aggregated

Roman, Iron Age

Gregory et al. (2013)

West Yorkshire

Field boundary

Iron Age, Early medieval

Gregory et al. (2013)

North Yorkshire

Field boundary

Iron Age, Roman

Gregory et al. (2013)

North Yorkshire

Aggregated

Iron Age

Gregory et al. (2013)

West Yorkshire

Field boundary

Roman

Roberts et al. (2001)

West Yorkshire

Aggregated

Iron Age, Roman

Roberts et al. (2001)

West Yorkshire

Field boundary

Roman

Brown et al. (2007)

West Yorkshire

Field boundary

Iron Age

Brown et al. (2007)

North Yorkshire

Pit alignment

Iron Age, Roman

Brown et al. (2007)

West Yorkshire

Aggregated

Roman, Iron Age

Martin et al. (2013)

West Yorkshire

Pit alignment, coaxial

Iron Age, Roman

Roberts (2005)

West Yorkshire

Field system

Iron Age, Roman

Moretti (2008)

South Yorkshire

stock enclosure

Iron Age

JSAC (2007)

South Yorkshire

Coaxial

Iron Age, Roman

Jones et al. (2005), ASWYAS (2006) and ASWYAS (2008)

County West Yorkshire

Types (primary listed first) Field boundary, linear earthwork

North Yorkshire

(continued)

200

R. Johnston et al.

County Durham

Types (primary listed first) Aggregated

Periods (primary listed first) Early medieval, Iron Age

References Bayliss et al. (2013) and Gear and Turner (1992)

Durham

Cord rig

Iron Age

Bayliss et al. (2013)

Durham

Coaxial

Medieval, Bronze Age

Bayliss et al. (2013) and Robinson (1993)

Bedfordshire

linear ditch, enclosure

Bronze Age, Iron Age

Cooper and Edmonds (2007)

Bedfordshire

Field system

Bronze Age, Iron Age

Cooper and Edmonds (2007)

Essex

Enclosure

Early medieval

Cooke et al. (2008)

Essex

Field boundary

Bronze Age, Iron Age

Cooke et al. (2008)

Derbyshire

Aggregated

Bronze Age

Isles of Scilly

Field boundary

37

Crab’s Ledge, Tresco North Furzton

Buckinghamshire

stock enclosure

Iron Age, Early medieval Iron Age

38

Samson, East Porth

Isles of Scilly

Field boundary

Bronze Age

39

Stannon, Bodmin Moor Saddlesborough Main Reave site 208, Dartmoor Holne Moor, Dartmoor Shaugh Moor, Cholwich Town, site 203 Shaugh Moor, Wotter Playground site 201 Shaugh Moor, Wotter Common, site 69 Shovel Down, Dartmoor Seathwaite Valley 360–364 Shirley Road, Southampton Ashville Trading Estate, Abingdon Tremough, Penryn

Cornwall

Aggregated

Bronze Age

Devon

Coaxial

Bronze Age

Bayliss et al. (2013) and Barnatt (1994) Bayliss et al. (2013) and Ratcliffe and Straker (1996) Bayliss et al. (2013) and Williams (1988) Bayliss et al. (2013) and Ratcliffe and Straker (1996) Bayliss et al. (2013), Rose (1992a, b) Jordan et al. (1994) and Smith et al. (1981)

Devon

Coaxial

Bronze Age

Devon

Coaxial

Bronze Age

Jordan et al. (1994) and Smith et al. (1981) Jordan et al. (1994)

Devon

Coaxial, lynchet

Bronze Age

Jordan et al. (1994)

Devon

Coaxial

Bronze Age

Jordan et al. (1994) and Smith et al. (1981)

Devon

Coaxial

Bronze Age

Fyfe et al. (2008)

Cumbria Hampshire

Field boundary Field boundary

Medieval Roman, Iron Age

Wild et al. (2001) Russel and Fedorowicz (2013)

Oxfordshire

Aggregated

Iron Age, Roman

Parrington (1978)

Cornwall

Coaxial

Iron Age, Roman

Trenowah, St Austell Croft Road, Spencers Wood, Reading

Cornwall

Coaxial, field boundary

Iron Age, Roman

Gossip and Jones (2007) and Gossip and Jones (2010) Johns (2008)

Berkshire

Field boundary

Iron Age

Taylor and Dawson (2015)

ID 28

29

30

31

32

33

34 35 36

40

41 42

43

44

45 46 47 48 49 50 51

Site name Bowes Moor field system (A66 Stainmore Pass) Coach and Horses Earthwork (A66 Stainmore Pass) Ravock field system, Sites B & D (A66 Stainmore Pass) Hill Lane, Old Warden (Broom Quarry) Ash Covert, Old Warden (Broom Quarry) Area 1A South Gate (SG), Stansted Airport M11 Slip Road, Stansted Airport Eaglestone Flat

(continued)

13  Understanding the Chronologies of England’s Field Systems

ID 52

53 54 55 56 57 58

59 60 61 62 63 64 65 66

67 68 69 70 71

72 73 74 75 76 77 78 79

Site name Charnham Lane, Hungerford, West Berkshire Hitches Lane, Fleet North Bersted, Bognor Regis Shrubsoles Hill, Sheppey Lidsey Landfill, Woodgate Colne Fen, Earith Addenbrooke's Access Road, Site 3, Clay Farm, Trumpington Stencoose Trewellard, North L&’s End Pipeline Penhale Round Harlyn Bay A Harlyn Bay B Elliott Site, Fengate Pegswood Moor Farm Newton Bewley, Hartlepool Monkton Road, Minster in Thanet Edgerley Drain Road, Fengate Barnsdale Bar Quarry, Norton Finningley Quarry Northern Extension Vicarage Lane and Romford Road, London Borough of Newham Lafone Street, Southwark Redscar Wood pit alignment Shotton Village Shotton North-East Huntsman's Quarry, Kemerton Ormesby St Michael Clay Farm, Trumpington Boundary Lane Windfarm, Whittonstall

201

County Berkshire

Types (primary listed first) linear ditch

Periods (primary listed first) Neolithic

References Ford (2014)

Hampshire

Field system

Bronze Age, Roman

Pine (2016)

West Sussex

Field system

Iron Age, Roman

Taylor et al. (2014)

Kent

Field system

Bronze Age

Coles et al. (2003)

West Sussex

Enclosure, field system

Iron Age, Roman

Wallis and Ford (2014)

Cambridgeshire

Coaxial, aggregated

Bronze Age, Iron Age

Evans et al. (2013)

Cambridgeshire

Field system

Bronze Age, Iron Age

Timberlake (2007)

Cornwall

Aggregated

Iron Age, Roman

Jones (2001)

Cornwall

Field boundary

Roman

Lawson-Jones (2013)

Cornwall Cornwall Cornwall Cambridgeshire

Field system Field boundary Field boundary Coaxial, stock enclosure

Roman Bronze Age Bronze Age Bronze Age, Iron Age

Nowakowski and Johns (2015) Jones and Quinnell (2014) Jones and Quinnell (2014) Evans et al. (2009)

Northumberland

Field system, stock enclosure

Iron Age, Roman

Durham

Field system

Early medieval, Roman

Proctor (2002) and Proctor (2009) Platell and Johns (2001)

Kent

Coaxial

Bronze Age

Martin et al. (2012)

Cambridgeshire

Coaxial

Bronze Age

Evans et al. (2009)

South Yorkshire

Coaxial

Iron Age, Roman

ASWYAS (2001)

South Yorkshire

Aggregated

Iron Age, Roman

MAP (2009)

Greater London

Field boundary

Early medieval

Keith-Lucas et al. (2007) and Branch (2003)

Greater London

Bronze Age

Bates and Minkin (1999)

Northumberland

cultivation marks, field boundary Pit alignment

Roman, Early medieval

Northumberland Northumberland Worcestershire

Field boundary stock enclosure Coaxial

Bronze Age Iron Age Bronze Age, Iron Age

Passmore and Waddington (2009) Hodgson et al. (2012) Hodgson et al. (2012) Jackson (2015)

Norfolk

Field system, enclosure

Bronze Age, Iron Age

Cambridgeshire

Coaxial

Bronze Age, Iron Age

Northumberland

Field boundary

Medieval, post-medieval

Gilmour and Mortimer (2012) and Gilmour et al. (2014) Phillips and Mortimer (2012) ASDU (2014)

(continued)

202

ID 80 81 82 83 84

85 86 87 88 89 90 91

92 93 94

R. Johnston et al.

Site name Newark Road sub-site, Fengate Bridge Road, Rainham Roughground Farm, Lechlade Padholme Road sub-site, Fengate Lancaster University Wind Turbine Heathrow Terminal 5 Weir Bank Stud Farm, Bray Vince's Farm, Ardleigh A15 North Borough Bypass, Etton Castle Hill, Feniton Wrotham Quarry, Trottiscliffe Dairy Lane, Nursling, Southampton Slade Farm, Bicester Bar Point, St Mary's

County Cambridgeshire

Types (primary listed first) Coaxial

Periods (primary listed first) Bronze Age, Iron Age

Greater London

stock enclosure

Bronze Age

Gloucestershire

Coaxial

Iron Age, Roman

Meddens (1996) and Meddens and Beasley (1990) Allen et al. (1993)

Cambridgeshire

Coaxial

Bronze Age

Pryor (1980) and Pryor (2001)

Lancashire

Field system

Medieval

Bradley (2013)

Greater London

Coaxial, aggregated

Bronze Age

Healy et al. (2010)

Berkshire

Field system

Bronze Age

Barnes et al. (1995)

Essex

linear ditch, field system

Bronze Age, Iron Age

Brown (1999)

Cambridgeshire

Field system

Neolithic, Bronze Age

French and Pryor (2005)

Devon Kent

Coaxial Coaxial

Bronze Age Iron Age

Fitzpatrick et al. (1999) Malim et al. (2013)

Hampshire

Field system

Bronze Age, Iron Age

Adam et al. (1997)

Oxfordshire

linear ditch

Iron Age

Ellis et al. (2000)

Isles of Scilly

Field system

Iron Age Medieval, post-medieval Bronze Age, Iron Age

Evans (1983) and Jordan et al. (1994) Bayliss et al. (2009b)

References Pryor (1980) and Pryor (2001)

Humberside

Ridge and furrow

95

Sewerby Cottage Farm, Bridlington Cranford Lane

Greater London

Coaxial

96

Simy Folds

Durham

Field system

Early medieval, Bronze Age

97

Phoenix Project, Awe Lynton Way, Sawston Warren Farm, Lockington Quarry Wilby Way, Great Doddington Priors Hall, Kirby Lane, Corby Whitelands Farm, Bicester

Berkshire

stock enclosure

Iron Age, Roman

Booth (2013)

Cambridgeshire

Enclosure

Bronze Age

Weston et al. (2007)

Leicestershire

Aggregated

Iron Age, Roman

Thomas (2013)

Northamptonshire

Aggregated

Iron Age

Thomas and Enright (2003)

Northamptonshire

Field system

Bronze Age

Chapman and Jones (2012)

Oxfordshire

Field system

Early medieval, Iron Age

Martin (2011)

Cambridgeshire

Field system

Iron Age, Roman

Mudd and Upson-Smith (2006)

Somerset Cambridgeshire

Field system Coaxial

Bronze Age Bronze Age

Gent and Reed (2007) Mudd and Pears (2008)

Devon

Field system

Bronze Age

Pearce et al. (2011)

Wiltshire

Field boundary

Bronze Age, Iron Age

Bradley et al. (1994)

98 99 100 101 102

103 104 105 106 107

Alma Road, Peterborough Hillfarrance Tower's Fen, Thorney Old Rydon Lane, Exeter Dunch Hill Midden, Wessex LDP site 081A

Cotton and Elsden (in prep) and Historic England (in prep) Coggins et al. (1983)

(continued)

13  Understanding the Chronologies of England’s Field Systems

ID 108

109

110

111

112

113

114 115 116 117

118 119 120

Site name Weather Hill Linear Ditch, Wessex LDP site 083 The Devil's Ditch, Wessex LDP site 090 Windmill Hill Linear Ditch, Wessex LDP site 091 Brigmerston Down Linear Ditch, Wessex LDP site 092 Haxton Down Linear Ditch/Pit Alignment, Wessex LDP site 099 Sidbury Double Linear Ditch, Wessex LDP sites 100 & 101 Hazel Road, Bognor Regis Needingworth Quarry, Over Plantation Camp, Ingram Wether Hill cross-ridge dyke, Ingram GlaxoSmithKline U Building, Ware Whitemoor Haye Quarry, Alrewas Stanwick, Raunds

203

County Wiltshire

Types (primary listed first) Linear earthwork

Periods (primary listed first) Bronze Age

References Bradley et al. (1994)

Wiltshire

Linear earthwork

Iron Age, Bronze Age

Bradley et al. (1994)

Wiltshire

Linear earthwork

Iron Age

Bradley et al. (1994)

Wiltshire

Linear earthwork

Bronze Age

Bradley et al. (1994)

Wiltshire

Linear earthwork, pit alignment

Iron Age, Bronze Age

Bradley et al. (1994)

Wiltshire

Linear earthwork

Iron Age

Bradley et al. (1994)

West Sussex

Field system

Iron Age

Cambridgeshire

Field system

Bronze Age

Bedwin and Pitts 1978) and Jordan et al. (1994) Evans et al. (2016)

Northumberland

Cultivation terrace

Iron Age

ASUD (1997)

Northumberland

Linear earthwork

Iron Age, Roman

ASUD (1997)

Hertfordshire

Field boundary

Medieval

Kaye (2009)

Staffordshire

Pit alignment

Iron Age

Northamptonshire

Field boundary

Bronze Age

Coates (1999) and Knight and Howard (2004) Harding and Healy (2007)

References Adam, N.  J., Seager Smith, R., & Smith, R.  J. C. (1997). An early Romano-British settlement and prehistoric field boundaries at Dairy Lane, Nursling, Southampton. Proceedings of the Hampshire Field Club and Archaeological Society, 52, 1–55. Allen, T.  G., Darvill, T.  C., Green, L.  S., & Jones, M.  U. (1993). Excavations at Roughground farm, Lechlade, Gloucestershire: A prehistoric and Roman landscape. Oxford: Oxford Archaeology Unit. ASDU. (2014). Boundary lane wind farm near Whittonstall, Northumberland and County Durham: Post-excavation full analysis (Archaeological Services Durham University report 3487). Durham: Archaeological Services Durham University. (Unpublished report). ASUD. (1997). The Ingram and Upper Breamish Valley Landscape Project: Interim report. Durham: Archaeological Services, University of Durham. Unpublished report.

ASWYAS. (2001). Barnsdale Bar Quarry, Norton, South Yorkshire: Archaeological investigations (ASWYAS Report 932). Wakefield: ASWYAS. (Unpublished report). ASWYAS. (2006). First Point, Balby Carr (Zone D1), Doncaster, South Yorkshire (ASWYAS Report 1556). Wakefield: ASWYAS. (Unpublished report). ASWYAS. (2008). Balby Carr: Zone D2, Phase 1, Doncaster, South Yorkshire (ASWYAS Report 1769). Wakefield: ASWYAS. (Unpublished report). Balaam, N.  D., Smith, K., & Wainwright, G.  J. (1982). The Shaugh Moor Project: Fourth report – Environment, context and conclusion. Proceedings of the Prehistoric Society, 48, 203–278. Barnatt, J. (1994). Excavations of a Bronze Age unenclosed cemetery, cairns and field boundaries at Eaglestone Flat, Curbar, Derbyshire, 1984, 1989–1990. Proceedings of the Prehistoric Society, 60, 287–370. Barnatt, J. (2001). Trial excavations of prehistoric field boundaries on Big Moor, Baslow, Derbyshire 1983. Transactions of the Hunter Archaeological Society, 21, 63.

204 Barnes, I., Boismier, W. A., Cleal, R. M. J., Fitzpatrick, A. P., & Roberts, M. R. (1995). Early settlement in Berkshire: Mesolithic to Roman occupation in the Thames and Kennet Valleys (Wessex Archaeology Report 6). Salisbury: Trust for Wessex Archaeology. Bates, J., & Minkin, J. (1999). Lafone street, Southwark – Prehistoric farming and a medieval bridge. London Archaeologist, 8(12), 325–330. Bayliss, A., & Marshall, P. (in prep). Using radiocarbon dating and chronological modelling in archaeology. London: Historic England. Bayliss, A., Cook, G., Bronk Ramsey, C., van der Plicht, J., & McCormac, G. (2008). Radiocarbon dates from samples funded by English Heritage under the aggregates levy sustainability fund 2004–7. Swindon: English Heritage. Bayliss, A., Bronk Ramsey, C., Cook, G., Ladle, L., Meadows, J., van de Plicht, J., Scaife, R., & Woodward, A. (2009a). Radiocarbon dating. In L.  Ladle & A.  Woodward (Eds.), Excavations at Bestwall Quarry, Wareham, 1992–2005. Volume 1: The prehistoric landscape (pp.  126–153). Dorset Natural History and Archaeology Society: Dorchester. Bayliss, A., Fenton-Thomas, C., & Bronk Ramsey, C. (2009b). Radiocarbon dating. In C. Fenton-Thomas (Ed.), A place by the sea: Excavations at Sewerby Cottage Farm. Bridlington/York: On-Site Archaeology. Bayliss, A., Bronk Ramsey, C., Cook, G., McCormac, G., Otlet, R., & Walker, J. (2013). Radiocarbon dates from samples funded by English Heritage between 1988 and 1993. Swindon: English Heritage. Bedwin, O., & Pitts, M.  W. (1978). The excavation of an Iron Age settlement at North Bersted, Bognor Regis, West Sussex 1975–76. Sussex Archaeological Collections, 116, 293–346. Booth, P. (2013). A multi-period site at the Phoenix Project, Awe, Burghfield, Berkshire. Berkshire Archaeological Journal, 81, 53–72. Bradley, J. (2013). Lancaster University Wind Turbine, Lancaster, Lancashire. Archaeological investigation report. Lancaster: Oxford Archaeology North. (Unpublished report). Bradley, R., Entwistle, R., & Raymond, F. (1994). Prehistoric land divisions on Salisbury Plain: The work of the Wessex linear ditches project. Swindon: English Heritage. Bradley, R., Haselgrove, C., Vander Linden, M., & Webley, L. (2015). The later prehistory of Northwest Europe: The evidence of development-­led fieldwork. Oxford: Oxford University Press. Branch, N.  P. (2003). Radiocarbon dating of charred plant material from Romford Road, Stratford, London Borough of Newham (RDV02). London: ArchaeoScape. (Unpublished report). Bronk Ramsey, C. (2009). Bayesian analysis of radiocarbon dates. Radiocarbon, 51(1), 337–360. Brown, N. R. (1999). The archaeology of Ardleigh, Essex: Excavations 1955–1980 (East Anglian Archaeology 90). Chelmsford: Essex County Council. Brown, F., Howard-Davis, C., Brennand, M., Boyle, A., Evans, T., O'Connor, S., Spence, A., Heawood, R., & Lupton, A. (2007). The archaeology of the A1(M) Darrington to Dishforth DFBO road scheme. Oxford Archaeology North: Lancaster. Chadwick, A. (2008). Fields for discourse. Landscape and materialities of being in South and West Yorkshire and Nottinghamshire during the Iron Age and Romano-British Periods. A study of people and place. PhD thesis, University of Wales, Newport. Chapman, A., & Jones, C. (2012). An Early Bronze Age henge and a Middle Bronze Age ditch system at Priors Hall, Kirby Lane, Corby. Northamptonshire Archaeology, 37, 37–67. Coates, G. (1999). Excavations at Whitemoor Haye Quarry, Alrewas, Staffordshire 1997–1998. Birmingham: BUFAU. (Unpublished report).

R. Johnston et al. Coggins, D., Fairless, K.  J., & Batey, C.  E. (1983). Simy folds: An early medieval settlement site in Upper Teesdale, County Durham. Medieval Archaeology, 27, 1–26. Coles, S., Hammond, S., Pine, J., Preston, S., & Taylor, A. (2003). Bronze Age, Roman and Saxon sites on Shrubsoles Hill, Sheppey, and at Wises Lane, Borden. Thames Valley Archaeological Services: Kent. Cooke, N., Fraser, B., & Philpotts, C. (2008). From hunter gatherers to huntsmen: A history of the Stansted landscape. Oxford/Salisbury: Framework Archaeology. Cooper, N.  J. (Ed.). (2006). The archaeology of the east midlands: An archaeological resource assessment and research agenda. Leicester: Leicester Archaeology Monograph 13. Cooper, A., & Edmonds, M. (2007). Past and present: Excavations at Broom, Bedfordshire, 1996–2005. Cambridge: Cambridge Archaeological Unit. Cooper, A., & Green, C. (2016). ‘The good, the bad and the ugly’? History and relationships in archaeological data: The case of the English landscape and identities project. Journal of Archaeological Method and Theory, 23(1), 271–304. Cotton, J., & Elsden, N. (in prep). West London landscapes: Archaeological excavations on the Thames Terraces in the London Borough of Hillingdon: The prehistoric and Roman evidence. London: Museum of London Archaeology. Crawford, O. G. S. (1923). Air survey and archaeology. Geographical Journal, 61(5), 342–366. Duller, G.  A. T. (2008). Luminescence dating: Guidelines on using luminescence dating in archaeology. Swindon: English Heritage. Ellis, P., Hughes, G., & Jones, L. (2000). An Iron Age boundary and settlement features at Slade Farm, Bicester, Oxfordshire: A report on excavations, 1996. Oxoniensia, 65, 211–265. English, J. (2013). Pattern and progress: Field systems of the second and early first Millennia BC in Southern Britain (BAR British Series 587). Oxford: Archaeopress. Evans, J. G. (1983). Excavations at Bar point, St. Mary’s, Isles of Scilly, 1979–80. Cornish Studies, 11, 7–32. Evans, C., Beadsmoore, E., Brudenell, M., & Lucas, G. (2009). Fengate revisited: Further fen-edge excavations, Bronze Age fieldsystems and settlement and the Wyman Abbott/Leeds archives. Cambridge: Cambridge Archaeological Unit. Evans, C., Brudenell, M., Patten, R., & Regan, R. (2013). Process and history: Prehistoric Communities at Colne Fen, Earith. Cambridge: Cambridge Archaeological Unit. Evans, C., Tabor, J., & Vander Linden, M. (2016). Twice-Crossed River: Prehistoric and Palaeoenvironmental investigations at Barleycroft Farm/Over, Cambridgeshire. Cambridge: McDonald Institute for Archaeological Research. FISH. (2020). Archaeological objects thesaurus. http://www.heritage-­ standards.org.uk/fish-­vocabularies/. Accessed 09/02/20. Fitzpatrick, A. P. (2007). Late Bronze Age and Iron Age. In C. Webster (Ed.), The archaeology of South West England: South West Archaeological Research Framework. Resource assessment and research agenda (pp. 117–144). Taunton: Somerset County Council. Fitzpatrick, A. P., Butterworth, C. A., & Grove, J. (1999). Prehistoric sites in East Devon: The A30 Honiton to Exeter improvement DFBO scheme, 1996–9 (Vol. 1). Salisbury: Wessex Archaeology. Fleming, A. (1978). The prehistoric landscape of Dartmoor: South Dartmoor. Proceedings of the Prehistoric Society, 44, 97–123. Fleming, A. (2008). The Dartmoor reaves: Investigating prehistoric land divisions. Oxford: Windgather Press. Ford, S. (2014). Late Neolithic and Early Saxon occupation at Charnham Lane, Hungerford, West Berkshire. Archaeological excavation. Reading: Thames Valley Archaeological Services. (Unpublished report).

13  Understanding the Chronologies of England’s Field Systems French, C., & Pryor, F. (2005). Archaeology and environment of the Etton landscape (East Anglian Archaeology 109). Peterborough: Fenland Archaeological Trust. Fyfe, R.  M., Brück, J., Johnston, R., Lewis, H., Roland, T.  P., & Wickstead, H. (2008). Historical context and chronology of Bronze Age land enclosure on Dartmoor, UK. Journal of Archaeological Science, 35, 2250–2261. Gear, A., & Turner, J. (1992). Changing patterns of landuse during the Late Holocene on Stainmore, Northern Pennines. Durham: Archaeological Services, University of Durham. (Unpublished report). Gent, T., & Reed, S. (2007). A Bronze Age field and figure from Hillfarrance. Proceedings of the Somerset Archaeological and Natural History Society, 150, 1–36. Gilmour, N., & Mortimer, R. (2012). Early prehistoric settlement, field systems and enclosure at Ormesby St Michael, Norfolk. Post-­ excavation assessment and updated project design. Cambridge: Oxford Archaeology East. (Unpublished report). Gilmour, N., Horlock, S., Mortimer, R., & Tremlett, S. (2014). Middle Bronze Age enclosures in the Norfolk Broads: A case study at Ormesby St Michael, England. Proceedings of the Prehistoric Society, 80, 141–157. Gossip, J., & Jones, A. M. (2007). Archaeological investigations of a later prehistoric and a Romano-British landscape at Tremough, Penryn, Cornwall (BAR British Series 443). Oxford: Archaeopress. Gossip, J., & Jones, A. M. (2010). Excavations at Tremough, Penryn, Cornwall, 2000–6. Cornish Archaeology, 48(49), 1–66. Grassam, A. (2008). Land at the junction of Doncaster Road and Field Lane, South Elmsall, West Yorkshire: Archaeological assessment (ASWYAS Report 1867). Wakefield: ASWYAS. (Unpublished report). Gregory, R.  A., Daniel, P., & Brown, F. (2013). Early landscapes of West and North Yorkshire: Archaeological investigations along the Asselby to Pannal Natural Gas Pipeline, 2007–8. Oxford Archaeology North: Lancaster. Hall, D. (2012). The open fields of England. Oxford: Oxford University Press. Harding, J., & Healy, F. (2007). The Raunds area project: A Neolithic and Bronze Age landscape in Northamptonshire. London: English Heritage. Healy, F., Beavan Athfield, N., Cook, G., Debenham, N., Hogg, A., Prior, C., Bronk Ramsey, C., Rhodes, E.  J., & Schwenninger, J.-L. (2010). Absolute dating. In Framework Archaeology (Ed.), Landscape evolution in the Middle Thames Valley: Heathrow terminal 5 excavations. Volume 2. Oxford & Salisbury: Framework Archaeology, Section 20. Heath, A.-M. (2003). Prehistoric settlement and agriculture on the Eastern Moors of the Peak District. PhD thesis, University of Sheffield. Heritage, E. (2012). Designation scheduling selection guide: Agriculture. Swindon: English Heritage. Herring, P., Johnson, N., Jones, A.  M., Nowakowski, J., Sharpe, A., & Young, A. (2016). Archaeology and landscape at the Land’s End, Cornwall: The West Penwith surveys 1980–2010. Cornwall Archaeological Unit: Truro. Historic England. (in prep). Radiocarbon dates from samples funded by English Heritage and dated 1998–2003. Swindon: Historic England. Hodgson, N., McKelvey, J., & Muncaster, W. (2012). The Iron Age on the Northumberland coastal plain: Excavations in advance of development 2002–2010. South Shields: Tyne and Wear Archives and Museums. Jackson, R. (2015). Huntsman's quarry, Kemerton: A Late Bronze Age settlement and landscape in Worcestershire. Oxford: Oxbow.

205 Johns, C. (2008). The excavation of a multi-period archaeological landscape, Trenowah, St Austell, Cornwall, 1997. Cornish Archaeology, 47, 1–48. Johnson, N., & Rose, P. (1994). Bodmin moor – An archaeological survey. Volume 1: The human landscape to c.1800. London: RCHME. Johnston, R., May, R., & McOmish, D. (2020). Scientific dates for England’s field systems. Figshare Dataset. https://doi.org/10.15131/ shef.data.11971599.v2. Jones, A. M. (2001). The excavation of a multi-period site at Stencoose, Cornwall. Cornish Archaeology, 39(40), 45–94. Jones, A. M., & Quinnell, H. (2014). Lines of archaeological investigation along the North Cornish Coast (BAR British Series 594). Oxford: Archaeopress. Jones, L., Gale, R., Greig, J., Smith, D., & Tetlow, E. (2005). Archaeological excavation of a brickwork plan field system at Catesby Business Park, Balby Carr, Doncaster, South Yorkshire, 2002. Birmingham: Birmingham Archaeology. (Unpublished report). Jordan, D., Haddon-Reece, D., & Bayliss, A. (1994). Radiocarbon dates from samples funded by English Heritage and dated before 1981. London: English Heritage. JSAC. (2007). Archaeological field evaluation report, Carr Lodge Farm, Loversall, Doncaster (JSAC report 650/01/07). Newark: JSAC. (Unpublished report). Kaye, D. (2009). Archaeological strip, map and sample excavation: U building, GlaxoSmithKline, Park Road, Ware, Herts. Milton Keynes: Archaeological Services and Consultancy. (Unpublished Report). Keith-Lucas, F., Vaughan-Williams, A., & Moore, P. (2007). Saxon and medieval Newham. London Archaeologist, 11(9), 244–246. Knight, D., & Howard, A.  J. (2004). Trent Valley landscapes: The archaeology of 5000 years of change. King’s Lynn: Heritage Marketing and Publications. Lawson-Jones, A. (2013). Discoveries along the North Land’s End pipeline. Cornish Archaeology, 52, 197–222. Løvschal, M. (2014). Lines in the landscape, boundaries of the mind – The emergence of landscape and settlement boundaries in north-­ western Europe in the first millennium BC. PhD thesis, Aarhus University. Malim, T., Hayes, L., & Wellicome, T. (2013). Wrotham Quarry investigations 2009–2013. SLR Consulting Ltd. (Unpublished report). MAP. (2009). Finningley Quarry Northern extension phases 20 and 21. Assessment report. Malton: MAP. (Unpublished report). Martin, J. (2011). Prehistoric, Romano-British and Anglo-Saxon activity at Whitelands Farm, Bicester. Oxoniensia, 76, 173–240. Martin, L., & Harrison, S. (2012). Normanton industrial estate extension, West Yorkshire. Archaeological excavations (ASWYAS Report 2302). Wakefield: ASWYAS. (Unpublished report). Martin, J., Schuster, J., & Barclay, A. J. (2012). Evidence of an Early Bronze Age field system and spelt wheat growing, together with an Anglo-Saxon sunken-featured building, at Monkton Road, Minster in Thanet. Archaeologia Cantiana, 132, 43–52. Martin, L., Richardson, J., & Roberts, I. (2013). Iron Age and Romano-­ British settlements at Wattle Syke: Archaeological investigations during the A1 Bramham to Wetherby Upgrading Scheme. Wakefield: ASWYAS. McOmish, D. (2011). Introduction to heritage assets: Field systems. Swindon: English Heritage. McOmish, D., Field, D., & Brown, G. (2002). The field archaeology of the Salisbury Plain Training Area. Swindon: English Heritage. Meddens, F.  M. (1996). Sites from the Thames estuary wetlands. England, and their Bronze Age use, Antiquity, 70, 325–334. Meddens, F. M., & Beasley, M. (1990). Wetland use in Rainham. Essex London Archaeologist, 6(9), 242–248. Moretti, D. (2008). Glasshoughton Coalfields Link Road, West Yorkshire: Archaeological evaluation and excavation (ASWYAS Report 1766). Wakefield: ASWYAS. (Unpublished report).

206 Mudd, A., & Pears, B. (2008). Bronze Age field system at Tower’s Fen, Thorney, Peterborough. Excavations at ‘Thorney Borrow Pit’ 2004– 2005 (BAR British Series 471). Oxford: Archaeopress. Mudd, A., & Upson-Smith, T. (2006). Middle Iron Age and Late Iron Age/Early Roman enclosures at the former sports ground, Alma Road, Peterborough. Northamptonshire Archaeology, 34, 19–32. Newman, P. (2011). The field archaeology of Dartmoor. Swindon: English Heritage. Nowakowski, J. A. (2007). Excavations of a Bronze Age Landscape and a post Roman Industrial Settlement 1953–61, Gwithian, Cornwall: Assessment of key datasets (2005–2006) (HES Report 2007R107). Truro: Cornwall Archaeological Unit. (Unpublished report). Nowakowski, J. A. (2009). Living in the sands – Bronze Age Gwithian, Cornwall, revisited. In M.  J. Allen, N.  Sharples, & T.  O’Connor (Eds.), Land and people: Papers in memory of John G Evans (pp. 115–125). Oxford: The Prehistoric Society & Oxbow Books. Nowakowski, J.  A., & Johns, C. (2015). Bypassing Indian Queens  – Archaeological excavations 1992–1994 – Investigating prehistoric and Romano-British settlements and landscapes in Cornwall. Truro: Cornwall Archaeological Unit & Highways Agency. Nowakowski, J.  A., Quinnell, H., Sturgess, J., Thomas, A.  C., & Thorpe, C. (2008). Return to Gwithian: Shifting the sands of time. Cornish Archaeology, 46. Oosthuizen, S. (2003). The roots of the common fields: Linking prehistoric and medieval field systems in West Cambridgeshire. Landscapes, 4(1), 40–64. Oosthuizen, S. (2013). Tradition and transformation in Anglo-Saxon England: archaeology, common rights and landscape. London: Bloomsbury. Parrington, M. (1978). The excavation of an Iron Age Settlement, BA ring-ditches and Roman features at Ashville Trading Estate, Abingdon (Oxfordshire) 1974–1976. Oxford: Oxford Archaeology Unit. Passmore, D. G., & Waddington, C. (2009). Managing archaeological landscapes in Northumberland. Oxford: Oxbow. Pearce, P., Steinmetzer, M., & Quinnell, H. (2011). An Early Neolithic pit alignment, Grooved Ware and Bronze Age field boundaries at the former Royal Navy Stores Depot, Old Rydon Lane, Exeter. Proceedings of the Devon Archaeological Society, 69, 23–51. Phillips, T., & Mortimer, R. (2012). Clay Farm, Trumpington, Cambridgeshire. Post-excavation assessment and updated project design. Cambridge: Oxford Archaeology East. (Unpublished report). Pine, J. (2016). A Middle Bronze Age pit circle and field system, and Roman settlement at Hitches Lane, Fleet, Hampshire (Thames Valley Archaeological Services Occasional Paper). Reading: Thames Valley Archaeological Services. Platell, A., & Johns, R. (2001). Archaeological investigation at Newton Bewley, Hartlepool, May 1998–January 1999. Hartlepool: Tees Archaeology. (Unpublished report). Proctor, J. (2002). An archaeological excavation at Pegswood Moor Farm, Morpeth, Northumberland. Post-excavation assessment report. London: Pre-Construct Archaeology. (Unpublished report). Proctor, J. (2009). Pegswood Moor, Morpeth: A later Iron Age and Romano-British farmstead settlement. London: Pre-Construct Archaeology. (Unpublished report). Pryor, F. M. M. (1980). Excavation at Fengate, Peterborough, England. The Third Report. Royal Ontario Museum: Toronto. Pryor, F. M. M. (Ed.). (2001). The Flag Fen Basin: Archaeology and environment of a Fenland Landscape. Swindon: English Heritage. Ratcliffe, J., & Straker, V. (1996). The early environment of Scilly: Palaeoenvironmental assessment of Cliff-face and Intertidal deposits, 1989–1993. Truro: Cornwall Archaeological Unit. Reimer, P. J., Bard, E., Bayliss, A., Beck, J. W., Blackwell, P. G., Bronk Ramsey, C., Grootes, P.  M., Guilderson, T.  P., Haflidason, H., Hajdas, I., HattŽ, C., Heaton, T. J., Hoffmann, D. L., Hogg, A. G.,

R. Johnston et al. Hughen, K. A., Kaiser, K. F., Kromer, B., Manning, S. W., Niu, M., Reimer, R. W., Richards, D. A., Scott, E. M., Southon, J. R., Staff, R. A., Turney, C. S. M., & van der Plicht, J. (2013). IntCal13 and Marine13 Radiocarbon Age calibration curves 0–50,000 Years cal BP. Radiocarbon, 55(4). Richardson, J. (2009). Iron Age to Roman Settlement at Timberland, Scunthorpe. Wakefield: ASWYAS. Rippon, S., Smart, C., & Pears, B. (2015). The fields of Britannia: Continuity and change in the Late Roman and Early Medieval landscape. Oxford: Oxford University Press. Roberts, I. (Ed.). (2005). Ferrybridge Henge: The ritual landscape. Archaeological investigations at the Site of the Holmfield Interchange of the A1 Motorway. Wakefield: ASWYAS. Roberts, H. M. (2007). Gwithian, Cornwall. Optically stimulated luminescence dating of sands from a Bronze Age Archaeological Site: Scientific dating report (English Heritage Research Department Report Series No. 103–2007). Swindon: English Heritage. Roberts, I., & Richardson, J. (2002). Iron Age and Romano-British settlement enclosures at Moss Carr, Methley, West Yorkshire. Wakefield: ASWYAS. Roberts, I., Burgess, A., & Berg, D. (2001). A new link to the past: The archaeological landscape of the M1–A1 Link Road. Wakefield: ASWYAS. Robinson, P. (1993). Archaeology on the Stainmore Bypass: The A66 project. Durham: Durham County Council. Rose, P. (1992a). Archaeological investigations of a water pipeline adjacent to King Arthur’s Hall and Stannon, Bodmin Moor, Cornwall. Truro: Cornwall Archaeological Unit. (Unpublished report). Rose, P. (1992b). Stannon and King Arthur’s Hall, Bodmin. Cornish Archaeology, 31, 166–167. Russel, J. I., & Fedorowicz, A. (2013). Archaeological watching brief at 360–364 Shirley Road, Southampton (Southampton Archaeology Unit Report 1123). Southampton: Southampton Archaeology Unit. (Unpublished report). Smith, K., Coppen, J., Wainwright, G.  J., & Beckett, S. (1981). The Shaugh Moor Project: Third report – Settlement and environmental investigations. Proceedings of the Prehistoric Society, 47, 205–273. Smith, G., Macphail, R.  I., Mays, S.  A., Nowakowski, J., Rose, P., Scaife, R.  G., Sharpe, A., Tomalin, D.  J., & Williams, D.  F. (1996). Archaeology and Environment of a Bronze Age Cairn and Prehistoric and Romano-British field System at Chysauster, Gulval, near Penzance, Cornwall. Proceedings of the Prehistoric Society, 62, 167–219. Taylor, A., & Dawson, T. (2015). Croft Road, Spencers Wood, Reading, Berkshire: Archaeological excavation (Draft Publication Report). Reading: Thames Valley Archaeological Services. (Unpublished report). Taylor, A., Weale, A., & Ford, S. (2014). Bronze Age, Iron Age and Roman landscapes of the coastal plain and a Late Iron Age warrior burial at North Bersted, Bognor Regis, West Sussex: Excavations 2007–2010. Reading: Thames Valley Archaeological Services. Thomas, J. (2013). Excavations within a developing Iron Age and Roman agricultural landscape at Warren Farm, Lockington. Transactions of the Leicestershire Archaeological and Historical Society, 87, 85–135. Thomas, A., & Enright, D. (2003). Excavation of an Iron Age settlement at Wilby Way, Great Doddington. Northamptonshire Archaeology, 31, 15–70. Timberlake, S. (2007). The Addenbrooke’s Access Road, Clay Farm, Trumpington, Cambridge. The 2007 investigations: Site 3. Cambridge: Cambridge Archaeological Unit. (Unpublished report). Turner, S. (2006). Historic Landscape Characterisation: A landscape archaeology for research, management and planning. Landscape Research, 31(4), 385–398. Walker, A. J., Williams, N., & Otlet, R. L. (1990). Harwell radiocarbon measurements VIII. Radiocarbon, 32(2), 165–196.

13  Understanding the Chronologies of England’s Field Systems Wallis, S., & Ford, S. (2014). Bronze Age and Middle Iron Age occupation and Roman fields at Lidsey Landfill, Woodgate, West Sussex. Reading: Thames Valley Archaeological Services. Weston, P., Newton, A.  A. S., & Nicholson, K. (2007). A Late Bronze Age enclosure at Lynton Way, Sawston, Cambridgeshire. Proceeding of the Cambridge Antiquarian Society, 96, 7–20. Wigley, A. (2007). Pitted histories: Early first millennium BC pit alignments in the central Welsh Marches. In C. Haselgrove & R. E. Pope (Eds.), The earlier Iron Age in Britain and the Near Continent (pp. 119–134). Oxford: Oxbow.

207 Wild, C., Wells, C., Anderson, D., Boardman, J., & Parker, A. (2001). Evidence for medieval clearance in the Seathwaite Valley, Cumbria. Transactions of the Cumberland and Westmorland Antiquarian and Archaeological Society (3rd Series), 1, 53–68. Williams, R.  J. (1988). North Furzton Iron Age settlement. South Midlands Archaeology, 18, 41–45. Yates, D. T. (2007). Land, power and prestige: Bronze Age field systems in southern England. Oxford: Oxbow.

The Prehistoric Agrarian Management Pattern in the Volcanic Landscape of the Campanian Plain, Southern Italy

14

Fabio Saccoccio

Abstract

This chapter discusses patterns of prehistoric agrarian management in the Campanian Plain, southern Italy, in particular along the course of the rivers Clanis/Regi Lagni and Sebeto and in the area of Naples. The region was affected by volcanic activity mainly deriving from the Mt Vesuvius and the Campi Flegrei caldera which preserved and sealed the archaeological deposits. Dating volcanic debris provides a robust chronological framework for the identification of agrarian systems dating to between the 4th and the early 2nd millennia BC in the area considered. These chronologies suggest agricultural use of areas over centuries, even millennia, despite the impact of repeated eruptions. The socio-economic implications of reconstructed patterns are discussed in comparison with wellstudied European contexts. Keywords

Campanian Plain · Southern Italy · Prehistory · Agrarian management

14.1 Introduction The prehistory of the Campanian Plain, southern Italy, was the subject of much discussion in the 1980s and 1990s after Albore Livadie and D’Amore (1981) published a paper on some Early Bronze Age (hereafter: EBA) pottery vessels found intact under the pumice layer of the Plinian Pomici di Avellino eruption of Mt Vesuvius (hereafter PdA; 1950–1820 cal BC; Table 14.1, c) at Palma Campania. This became the type-site for the EBA cultural assemblage of the area considered. The PdA volF. Saccoccio (*) Department of Classics and Archaeology, University of Nottingham, Nottingham, UK e-mail: [email protected]

canic debris was established as the most important EBA chronological marker for the study area because it is dispersed over most of the Campanian Plain, it is an easily recognisable light grey layer of ash and pumice, and because of the exceptional preservation of the evidence which it sealed. Subsequent studies since Albore Livadie and D'Amore (1981), in collaboration between geologists, archaeologists and vulcanologists, have demonstrated that eruptions of not only Mt Vesuvius but also of the Campi Flegrei caldera played their roles in sealing and preserving important evidence related to prehistoric human presence through repeated deposits of eruption debris (Albore Livadie 1986; 1999a; Albore Livadie and Widemann 1990; Guzzo and Peroni 1998). In the last decade or so, papers concerning the Campanian Plain published in international journals mostly focused on EBA sites sealed by the PdA eruption debris (e.g. Di Vito et  al. 2009, 2019; Laforgia et al. 2009; Matarazzo et al. 2010; Saccoccio et al. 2013; Albore Livadie et al. 2019; Saccoccio 2021). Only recently have researchers started to consider the long-term prehistory of the Campanian Plain from a multidisciplinary perspective (e.g. Vogel et al. 2016; Vingiani et al. 2017). In 2019, Vanzetti et al. (2019) began looking at long-term anthropogenic patterns in the Campanian Plain based on the evidence at the site of Gricignano d’Aversa US Navy base, Caserta province, which was built between the mid-1990s and the mid-2000s in an area with rich archaeological evidence (Marzocchella 1998; Fugazzola Delpino et al. 2003, 2007; Saccoccio et al. 2013). Vanzetti et al. (2019) claimed that long-lasting agrarian management was in place at the site between the Late Copper Age (c. 2700 cal BC) and the EBA (PdA event; 1950–1820 cal BC; Table 14.1, c). The area was occupied over many centuries, c. 700–800 years, with at least three different field systems detected through changes in orientation. There were seven settlement areas which relocated over time, with limited spatial shift, possibly in correlation with eruptive events, with at least 14 building phases lasting c. 50 years each (Fig.  14.2). Gricignano d’Aversa US Navy base is one of the few sites excavated in

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 S. Arnoldussen et al. (eds.), Europe’s Early Fieldscapes, Themes in Contemporary Archaeology, https://doi.org/10.1007/978-3-030-71652-3_14

209

210

F. Saccoccio

Table 14.1  OxCal 4.3 (Bronk Ramsey 2017; Reimer et al. 2013) calibrated radiocarbon dates for the eruptions and the agrarian exploitation evidence quoted in the paper; Italic indicates dates obtained using the OxCal 4.3 ‘Combine’ command and TPQ indicates terminus post quem. Eruptions and human exploitation Sample ID evidence a AP2 charcoal

calibrated age BC (at 95.4%) 1620-1430

b

AP1

charcoal

1930-1690

c

PdA

sheep bone

1950-1820

d

Solfatara

charcoal

e

Fossa Lupara

charcoal

f

Astroni 3

charcoal

g

Santa Maria delle Grazie paleosol above AMS debris AMS

charcoal

TPQ 2470-2060 TPQ 2460-2140 TPQ 2460-2140 TPQ 2840-2230 2870-2350

h i

j

k

PaleoAstroni 2 (= Phlegraean tephra B) PaleoAstroni 1

charcoal carbonised wood

2860-2480

charcoal

TPQ 3020-2600

paleosol

de Vita et al. 1999: tab. 4 Nava et al. 2007: 107

Iovine et al. 2017: fig. 8 Di Vito et al. 1999: tab. 3 Orsi et al. 1996: Tab. 1 Sampaolo 2005: 677

charcoal

m

settlement between Agnano 3 and PaleoAstroni 2 Agnano 3

TPQ 2880-2500 3640-3110

-

3650-3380

n

Agnano 1

charcoal

o

Piano Liguori

charcoal

TPQ 3700-3520 4040-3010

p

charcoal

4040-3660

q

agrarian evidence above Pomici di Mercato Pomici di Mercato

charcoal

6680-6490

r

San Martino

charcoal

7460-7080

s

Fondi di Baia

paleosol

t

Pisani 3

paleosol

TPQ 7740-7490 TPQ 9150-8740

l

References Rolandi et al. 1998: tab. 5 Rolandi et al. 1998: tab. 5 Passariello et al. 2009: tab. 4 Albore Livadie et al. 2019: 206 Isaia et al. 2009: 3/6 Di Vito et al. 1999: tab. 3 Di Vito et al. 1999: tab. 3 Isaia et al. 2009: 3/6 Nava et al. 2007: 110 de Vita et al. 1999: tabs 3 and 4 de Vita et al. 1999: tab. 4

Zanchetta et al. 2011: tab. 3 Di Vito et al. 1999: tab. 3 Di Vito et al. 1999: tab. 3 Di Vito et al. 1999: tab. 3

the study area where a thorough reconstruction of the prehistoric occupation has been made on the basis of the settlement and agrarian evidence found allowing the reconstruction of multiple management patterns over time.

There are two main aims of this paper. On the one hand, it aims to provide, for the first time, an overview of the prehistoric agrarian evidence known to date from a selected area of the Campanian Plain between the course of the rivers Clanis/Regi Lagni and Sebeto and in the territory of Naples. Previous publications are mainly limited to preliminary site reports. This enables me to ­reconstruct the human use of the landscape over a geographical extent and time span not matched to date (Fig. 14.1). I will try to understand to what extent continuities in settlement patterns, as discussed and reconstructed by Vanzetti et  al. (2019) for the Gricignano d’Aversa US Navy base, might apply more widely across the study area in order to determine if a diffused and sustainable agrarian management pattern was in place in the Campanian Plain before the Late Copper Age. Moreover, following Marzocchella (2002), in the final part of this chapter a comparison with prehistoric north European agrarian contexts will be used to interpret underlying socio-economic implications related to land tenure. This will be possible only for the positive agrarian evidence sealed by the EBA PdA eruption debris and reconstructed to a significant extent only at Gricignano d’Aversa US Navy base, where the field system is subdivided in elongated agrarian plots of similar size (Saccoccio et al. 2013; Saccoccio 2021).

14.2 C  ampanian Plain: A Geo-­ Archaeological Overview The Campanian Plain, Campania region (southern Italy), is delimited towards the north-west by the natural barrier of the Mt Massico and the quiescent Roccamonfina volcano. The Apennine chain borders the Campanian Plain to the north-east, while the Mts Lattari enclose it to the south-east (Fig. 14.1). The Regi Lagni is the outcome of a seventeenth century land reclamation programme promoted by the Spanish rulers of Naples and, nowadays, it is the main water drainage system of the Campanian Plain. Its course mostly follows the ancient course of the river Clanis, a river mentioned in the classical sources (listed in Giordano 1834: note 105), and its valley is easily identified on maps 183–184 of the Carta Geologica d’Italia (Italian Geological Map; Fig. 14.1, dark grey). As a result of the presence of at least five volcanoes (Roccamonfina, Campi Flegrei caldera, Procida, Ischia and Mt Vesuvius; see Fig. 14.1), the Campanian Plain has been shaped by volcanic debris overlying Quaternary deposits since at least c. 40 kyr cal BC (De Vivo et al. 2001: 59; Di Renzo et al. 2007: 754–755). Volcanic debris was possibly subject to argillation which led to the development of clay minerals by weathering as attested by the Plinian volcano Mt Pinatubo in the Philippines (Sasaki et al. 2003), but this phenomenon has been only little explored in the case-study area (Delibrias et al. 1979). Eruptive quiescence periods led to the development of two prehistoric paleosols, A and B, which are very important

14  The Prehistoric Agrarian Management Pattern in the Volcanic Landscape of the Campanian Plain, Southern Italy

211

Fig. 14.1  Campanian Plain, Campania region (southern Italy), showing the main eruptive centres, the Clanis and Sebeto Holocene valleys and the sites discussed. DTM data from Farr et al. 2007

in archaeological terms as chronological markers (Di Vito et al. 1999). Paleosol A developed during the first phase of quiescence of the Campi Flegrei caldera, between 9150– 8740 cal BC and 7740–7490 cal BC (Table 14.1, t, s), which corresponds to the Mesolithic in central-southern Italy.

Paleosol B developed during the second phase of quiescence of the Campi Flegrei caldera which dates between 7460– 7080 cal BC and 3700–3520 cal BC (Table 14.1, r, n) and corresponds to a period dating between the Late Mesolithic and the Final Neolithic/Early Copper Age in the area consid-

212

F. Saccoccio

Fig. 14.2  Gricignano d’Aversa US Navy base and Cambrannone: settlement and agrarian pattern over time. Grey dots mark the settled area, grey lines mark the field system while the dotted line marks the cart-track; PS stands for Paleosol (after Vanzetti et al. 2019: figs 2 and 9)

14  The Prehistoric Agrarian Management Pattern in the Volcanic Landscape of the Campanian Plain, Southern Italy

213

Fig. 14.3 (a) Gricignano-Fusarello TAV site 5: Final Neolithic apsidal hut; (b) Gricignano-Fusarello TAV site 6: cross-cutting plough-marks bracketed between PaleoAstroni 2 and AMS debris (Marzocchella 1998: figs 8 and 14)

ered. Di Vito et al. (1999: 244) proposed that Paleosol B covered just the Campi Flegrei caldera but the evidence from Gricignano d’Aversa US Navy base suggests a wider distribution, possibly most of the Campanian Plain (Saccoccio et al. 2013: Fig. 3). The third period of activity of the Campi Flegrei caldera dates between the 4th and 3rd millennia BC, marked by the Agnano 1 and Fossa Lupara eruptions (Di Vito et al. 1999: 238; Table 14.1, n, e). Within this phase it is worth mentioning the Agnano Monte Spina eruptive event (hereafter: AMS), dated to 2860–2480 cal BC, described by de Vita et al. (1999: 300) as the highest-­magnitude eruption to have occurred in the last 5000 years deriving from the Campi Flegrei caldera, which had a significant environmental impact over an area of about 1000 km2 (Table 14.1, i). From the point of view of the human presence in the area considered, the second most important eruption after the AMS event is the Vesuvian Plinian eruption of the PdA, dated to 1950–1820 cal BC (Table  14.1, c). At Gricignano d’Aversa US Navy base, the debris linked to this eruption seems to close a long period of human occupation characterised by the presence of paleosols alternating with minor eruption debris possibly deriving from the Astroni volcano (Saccoccio et al. 2013; Vanzetti et al. 2019; Fig. 14.2). The PdA eruption may have affected the climate, probably causing a drop in temperature (Zanchetta et al. 2011). Despite its characterisation as a devastating event, Albore Livadie et al. (2019: 208) recently suggested that even proximal areas to Mt Vesuvius were resettled “no more than a few decades after the Avellino Pumice event” (see also Di Vito et  al. 2019). Moreover, Albore Livadie et al. (2019: 217) argued that there was a cultural change in the Campanian Plain after two Middle Bronze Age Sub-Plinian eruptions, labelled as AP1 and AP2 (Table 14.1, b, a), which critically affected the local population a short time after the PdA event. After the AP2 event, there was a new archaeological culture, the Proto-­Apennine, which it is suggested may reflect the presence of new groups in the area (Albore Livadie et al. 2019: 218).

14.3 P  rehistoric Settlement and Agrarian Evidence in the Campanian Plain The following sections provide an overview of the published prehistoric evidence found in the Campanian Plain between the territory of Gricignano d’Aversa and Naples following the course of the rivers Clanis/Regi Lagni and Sebeto (see Fig.  14.1). Most publications are preliminary reports which record scattered pieces of information and generally do not specify the extent or the orientation of the agrarian evidence found. In the worst cases, only an approximate chronology and the location of the evidence is provided. When it is specified, eruption debris and material culture will be used as chronological markers. It is important to note that most of the agrarian evidence consists of negative plough-marks generated by the tip of the plough/ard cutting into the soil which, depending on the depth reached, can rearrange the content of the affected layers through homogenisation (Vanzetti et al. 2019: 151 and fig. 6). Ploughing is a destructive process, so that often negative plough-marks do not correspond to any overlying positive ridge-and-furrow evidence, which has been homogenised by subsequent ploughing. For this reason, the stratigraphic position of negative plough-marks has to be seen as terminus post quem for the related agrarian exploitation of the study area.

14.3.1 The Territory of Gricignano d’Aversa, Caserta Province The oldest evidence found in the territory of Gricignano d’Aversa (hereafter: Gricignano) is Late Neolithic Serra D’Alto cultural aspect pottery at Collegamento di Gricignano lotto 3 (Fig. 14.1 and Table 14.2, number 1). Although it was found not in situ under the AMS debris (Table 14.1, i), it could be dated to the late 5th/early 4th millennia BC or slightly earlier (Fig. 14.1 and Table 14.2, number 1). Final Neolithic pottery found at Gricignano-Fusarello TAV sites 5 and 6 (Fig. 14.1 and Table 14.2, numbers 2-3) is dated to the

14.3.1

14.3.2 14.3.2 14.3.2

14.3.2

14.3.2

14.3.2 14.3.2

14.3.2

14.3.2

14.3.2

6

7a 7b 8

9

10

11 12

13

14

15

TAV 5-lotto 12/13, site 11

TAV 5-lotto 12 site 10

TAV 5-lotto 17 site 8

TAV 5-lotto 1/18 site 6 TAV 5-lotto 1/18 site 7

TAV 4-lotto 10 site 5

TAV 4-lotto 10 site 4

Gricignano-Centro Sportivo Comunale TAV 4-lotto 1/13 site 1 TAV 4-lotto 1/13 site 2 TAV 4-lotto 7 site 3

Afragola

Afragola

Afragola

Afragola Afragola

Caivano

Caivano

Caivano Caivano Caivano

Gricignano d’Aversa

Settlement with Laterza, Bell Beaker and Capo Graziano cultural aspects pottery between the AMS and Astroni 7(?) debris Subsequently, the site is used for agriculture Stockyards and/or fences and agrarian traces dated to 4040-3660 cal BC above the Pomici di Mercato debris Serra d’Alto and Diana cultural aspects pottery above the agrarian layer Agrarian traces recorded until Agnano 3 debris

Cemetery with Gaudo cultural aspect pottery Ritual pit with animal bones and Laterza cultural aspect pottery Stockyards and/or fences with Serra d’Alto and Diana cultural aspects pottery on top of the Tufo Giallo Napoletano debris Two ritual pits with Gaudo cultural aspect pottery between the Paleosol B and AMS debris Settlement abandoned before the PdA eruption with Laterza cultural aspect, Late Copper and Early Bronze Age pottery Settlement abandoned before the PdA eruption located c. 130 m south of the above Settlement with Final Neolithic and Early Copper Age pottery between under and above the Agnano 3 debris Settlement abandoned before the PdA eruption Settlement between Agnano 3 and PaleoAstroni 2 debris After PaleoAstroni 2 the area was cultivated until the AMS eruption Settlement destroyed by the PdA debris

EBA tumulus

Table 14.2  List of the published prehistoric evidence discussed in this paper ID Section Site Comune Evidence 1 14.3.1 Collegamento di Gricignano Gricignano d’Aversa Serra d’Alto cultural aspect pottery not in situ under the AMS debris lotto 3 Agrarian exploitation under the AMS debris 2 14.3.1 Gricignano- Fusarello TAV Gricignano d’Aversa Apsidal hut with Final Neolithic pottery site 5 3 14.3.1 Gricignano-Fusarello TAV site Gricignano d’Aversa Final Neolithic, Early and Late Copper Age pottery 6 Negative cross-cutting plough-marks between the PaleoAstroni 2 and AMS debris 4 14.3.1 Gricignano-Cambrannone Gricignano d’Aversa Settlement sealed by the PdA debris 5 14.3.1 Gricignano US Navy base Gricignano d’Aversa Negative cross-cutting plough-marks cutting through the ash of the PaleoAstroni 1 eruption Settlement, funerary and agrarian evidence between the AMS and PdA debris

Nava et al. (2007: 106) and Sampaolo (2005: 676−677)

Nava et al. (2007: 114) Nava et al. (2007: 107), Laforgia and Boenzi (2011: 253) Nava et al. (2007: 114), Di Vito et al. (2009) and Laforgia et al. (2009) Nava et al. (2007: 109−112)

Nava et al. (2007: 107)

Nava et al. (2007: 112)

Nava et al. (2007: 109) Nava et al. (2007: 113) Nava et al. (2007: 106−109), Laforgia and Boenzi (2011: 250−252)

De Caro (2003: 584) Marzocchella (1998: 113−130), Fugazzola Delpino et al. (2003, 2007), Saccoccio et al. (2013), Vanzetti et al. (2019) and Saccoccio (2021) Marzocchella (1998: 125 and fig. 31)

Marzocchella (1998: 107 and figs 2.5, 8, 9) Marzocchella (1998: 107, 110, 113 and figs. 2.6, 7, 14, 15)

References De Caro (2003: 582)

214 F. Saccoccio

14.3.2

14.3.2

14.3.3 14.3.3 14.3.3

14.3.3 14.3.3

14.3.3 14.3.3

14.3.3

14.3.3

14.3.3 14.3.3

14.3.3

14.3.3

16

17

18 19 20

21 22

23 24

25

26

27 28

29

30

Fuorigrotta-CNR

Naples-Stazione San Pasquale

Naples-Castel Nuovo Naples-Piazza Santa Maria degli Angeli

Naples-Via Verdi

Naples-Piazza Montecalvario

Naples-Vico della Neve Naples-via Toledo/via Diaz

Naples-Via Donnaregina Naples-Via Settembrini Naples-Sant’Andrea delle Dame Naples-Stazione Duomo Naples-Vico Pallonetto

Casalnuovo di Napoli-piers 125-127, 129-130 and 132 Ponticelli-viadotto Botteghelle

Naples

Naples

Naples Naples

Naples

Naples

Naples Naples

Naples Naples

Naples Naples Naples

Naples

Casalnuovodi Napoli

Agrarian and settlement traces between under PaleoAstroni 1/2 and PdA debris Late and Final Neolithic settlement above the Pomici di Mercato debris Agrarian exploitation until AMS debris with Serra d’Alto and Diana cultural aspects pottery not in situ Agrarian exploitation between under Agnano 3 and AMS debris Agrarian layer overlain by debris from the AMS eruption Agrarian exploitation between under Agnano 3 and PaleoAstroni 2 debris EBA pottery sherds in the layer sealed by the PdA debris Postholes with Serra d’Alto cultural aspect pottery on top of the Pigna San Nicola debris Agrarian exploitation until the paleosol on top of the PaleoAstroni 2 debris Two graves with Gaudo cultural aspect material Late Neolithic postholes on top of the Pigna San Nicola debris Agrarian exploitation between above Paleosol B and the paleosol on top of the PaloAstroni 2 debris Late Neolithic postholes on top of the Pigna San Nicola debris Final Neolithic/Early Copper Age cross-cutting plough-marks above Paleosol B Agrarian exploitation under Agnano 3 debris, and then between PaleoAstroni 2 and AMS debris Pottery sherds between AMS and PdA debris Settlement with Serra d’Alto cultural aspect pottery on top of Pigna San Nicola debris The settlement was then affected by hoe-marks filled by the Piano Liguori debris Agrarian exploitation with Diana cultural aspect and early Copper Age pottery continues until the paleosol on top of the PaleoAstroni 2 eruption Final Neolithic/Early Copper Age and Gaudo cultural aspect pottery under the AMS debris Agrarian exploitation between AMS and Solfatara debris Giampaola et al. (2019: 214, 236−237, 247) Sampaolo (2012: 1328)

Giampaola et al. (2019: 213−214, note 21) Giampaola et al. (2019: 214, 247) De Caro (2001: 884), Giampaola and Boenzi (2013: 39−40) and Giampaola et al. (2019: 210−213)

Guzzo (2010: 1008) and Giampaola et al. (2019: 211−212)

Marzocchella (1985) Giampaola and Boenzi (2013: 39) and Giampaola et al. (2019: 211−213)

Giampaola et al. (2019: 213) Sampaolo (2005: 695) Guzzo (2010: 1008) and Giampaola et al. (2019: 213−214, note 21) Giampaola et al. (2019: 247) Giampaola et al. (2019: 212−213)

De Caro (2002: 647−648) and Giampaola and Stanislao (2007)

De Caro (2001: 880)

14  The Prehistoric Agrarian Management Pattern in the Volcanic Landscape of the Campanian Plain, Southern Italy 215

216

early/mid 4th millennium BC as the emergence of the Copper Age in Campania is dated through the radiocarbon dates from the site of Casalbore to 3630-3520 cal BC (Talamo et al. 2011: 40 and fig. 1). At Gricignano-Fusarello TAV site 5 Final Neolithic pottery sherds were found associated with an apsidal hut (Table 14.2, number 2; Fig. 14.3a). However, the oldest agrarian evidence known so far from the study area is much later and was recorded at Gricignano US Navy base where negative plough-marks were found cutting through the ash of the early/mid 3rd millennium BC PaleoAstroni 1 eruption reaching the underlying Paleosol B (Fig. 14.1 and Table 14.2, number 5; Table 14.1, k). Negative cross-cutting plough-marks were also found between the early/mid 3rd millennium BC PaleoAstroni 2 and the mid 3rd millennium BC AMS debris at Gricignano-Fusarello TAV site 6 (Fig. 14.1 and Table 14.2, number 3; Table 14.1, j, i; Fig. 14.3b), and under the AMS tephra at Collegamento di Gricignano lotto 3 (Fig. 14.1 and Table 14.2, number 1; Table 14.1, i). Continuous occupation is recorded at Gricignano US Navy base and at Gricignano-Cambrannone between the mid 3rd millennium BC AMS and the early 2nd millennium BC PdA eruption debris, c. 700–800 years, with the presence of at least three different field system orientations and at least seven settlements, the latter showing a maximum extension of c. 5 ha per phase (Fig. 14.1 and Table 14.2, numbers 4–5; Table 14.1, i, c). They relocated over time but the spatial shift was never great since all are found within a radius of c. 1 km (Fig.  14.1 and Table  14.2, numbers 4–5). Vanzetti et  al. (2019: fig. 9) were able to associate the three different field systems to four of the seven settled areas (Fig. 14.2). In total, there are c. 14 building phases, each lasting c. 50 years, while field systems seem more persistent over time, spanning different settlement phases (Fig.  14.2, PS1 before the PdA eruption and sealed by the PdA eruption; Vanzetti et al. 2019: 159). Cart-ruts were also found as part of the agrarian infrastructure recorded at Gricignano US Navy base, generally running parallel to the field system. Cart-tracks were also possibly used as communication routes between the different prehistoric settlements in the Campanian Plain (Nava et al. 2007: 114) but, unfortunately, on the basis of the scanty data in our possession, it is not possible to clearly detect the end points of these cart-tracks, as only short sections have been explored. At Gricignano US Navy base, Laterza cultural aspect funerary evidence was found overlying settlement traces between the AMS and the so-­called Flegrea 1/Astroni 3(?) debris, dated to the mid/late 3rd millennium BC, with graves generally following a similar orientation to the underlying huts (Fugazzola Delpino et  al. 2003, 2007; see Fig.  14.2). This pattern also suggests that human groups returned

F. Saccoccio

quickly to the study area while hut timbers were still visible, which allowed them to follow the alignment of the huts. A nearly intact 90 ha field-system was found at Gricignano US Navy base sealed by the PdA debris. SW-NE oriented coaxial elongated agrarian features were identified, with a maximum ascertained length of c. 700  m (Saccoccio et  al. 2013: 86–87; Saccoccio 2021). Furrows, with a 0.35 m inter-­ distance, were bordered by shallow gullies used for water drainage (Saccoccio 2021). Agrarian plots were also hypothesised on the basis of the presence of low hardened-soil banks, regularly distributed about every 56–57 m. A single cart-track was found, heading NE towards the contemporary settled area at Cambrannone but nothing can be said about its destination to the SW (Fig. 14.2). An EBA funerary tumulus was found c. 2 km west of the Gricignano US Navy base at Gricignano-Centro Sportivo Comunale (Municipal Sport Centre; Table 14.2, number 6).

14.3.2 The Prehistoric Evidence Between the Rivers Clanis/Regi Lagni and Sebeto Construction work for the Rome-Naples High-Velocity Train line (Italian TAV, Treno Alta Velocità) in the territory between Caivano and Ponticelli in Naples province led to the discovery of significant prehistoric settlement and agrarian evidence. In the literature, the TAV line in the case-study area is subdivided into two sections: TAV 4, passing through the territory of Caivano, and TAV 5, passing through the territory of Afragola (Nava et al. 2007: fig. 2). In the following section, sites will be labelled following Nava et al. (2007) and discussed chronologically. This means that geographically the narrative will jump from north to south and vice versa between the middle course of the river Clanis/Regi Lagni and the course of the river Sebeto. The oldest material in this area is Serra d’Alto and Diana cultural aspects pottery, the latter aspect dated to the Final Neolithic. Giampaola et al. (2019: 212) suggest that pottery related to these two cultural aspects is usually found associated in the Campanian Plain. This pattern, however, could be in part explained through the agrarian destructive process which tends to homogenise affected layers. Serra d’Alto and Diana cultural aspects pottery was found at TAV 4-lotto 7 site 3, TAV 5-lotto 12/13 site 11, and Ponticelli-viadotto Botteghelle (Fig. 14.1 and Table 14.2, numbers 8, 15, 17). At this last site, a Late and Final Neolithic pluristratified settlement was found on top of the mid 7th millennium BC Pomici di Mercato debris (Table 14.1, q). The settlement was affected by subsequent agrarian exploitation which lasted until the mid 3rd millennium BC AMS debris; Serra d'Alto and Diana cultural aspects pottery was found not in situ under the last agrarian layer (Table 14.1, i; Table 14.2, number 17).

14  The Prehistoric Agrarian Management Pattern in the Volcanic Landscape of the Campanian Plain, Southern Italy

4th millennium BC agrarian evidence is known in the study area at TAV 5-lotto 12/13 site 11 (Fig.  14.1 and Table 14.2, number 15). At this site, negative cross-cutting plough-marks and post-holes, interpreted by the researchers as pertaining to stockyards and/or fences, were found on a paleosol above the Pomici di Mercato eruption debris (Fig.  14.1 and Table  14.2, number 15; Table  14.1, q). The agrarian evidence was dated to 4040–3660 cal BC (Table 14.1, p) thanks to a charcoal sample, and cultivation seems to have continued until the top of the Paleosol B sealed by the mid 4th millennium BC Agnano 3 eruption debris (Table 14.1, m; Sampaolo 2005: 676). 4th millennium BC settlement evidence was found at TAV 4-lotto 10 site 5 and at TAV 5-lotto 1/18 site 7 (Fig. 14.1 and Table 14.2, numbers 10 and 12). At TAV 4-lotto 10 site 5 settlement traces were found both under and above the mid 4th millennium BC Agnano 3 eruption debris associated with Final Neolithic and Early Copper Age pottery sherds (Fig. 14.1 and Table  14.2, number 10; Table  14.1, m). At TAV 5-lotto 1/18 site 7 a settlement was set up after the Agnano 3 event; it is dated to 3640–3110 cal BC based on charcoal found in a post-hole and was abandoned before the PaleoAstroni 2 event, dated to 3020–2600 cal BC, after which the area was cultivated until the AMS tephra (Table 14.1, m, l, j, i; Table 14.2, number 12). At Casalnuovo di Napoli-piers 125-127, 129-130 and 132, agrarian traces were found between the early/mid 3rd millennium BC PaleoAstroni 1/2 and the early 2nd millennium BC PdA eruption debris (Fig.  14.1 and Table 14.2, number 16; Table 14.1, k, j, c). There is Gaudo cultural aspect pottery evidence at a cemetery at TAV 4-lotto 1/13 site 1 and in two ritual pits at TAV 4-lotto 7 site 3 (Fig. 14.1 and Table 14.2, numbers 7a and 8). Gaudo cultural aspect pottery is dated between the early and the mid 3rd millennium BC in the case-study area since it is bracketed between the Agnano 3 and AMS events (Talamo et al. 2011: fig. 1; Table 14.1, m, i). Then, there is Laterza cultural aspect pottery together with ash and animal bones in a pit found at TAV 4-lotto 1/13 site 2 which was interpreted as a ritual pit (Fig.  14.1 and Table 14.2, number 7b). In the case-study area, Laterza cultural aspect pottery is dated to the mid/late 3rd millennium BC as it is bracketed between the AMS and the so-called Flegrea 1/Astroni 3(?) eruption debris (Talamo et al. 2011: fig. 1; Tab. 1, i). There was also Laterza cultural aspect pottery at a settlement at TAV 5-lotto 12 site 10, stratigraphically positioned on top of a paleosol above the AMS eruption debris and dated to 2870–2350 cal BC thanks to a charcoal sample from a hearth (Fig.  14.1 and Table  14.2, number 14; Table  14.1, h). The settlement seems to have been active until “the last Astroni eruption” debris, possibly Astroni 7, dated after 2460–2140 cal BC (Table 14.1, f) but before PdA

217

(1950–1820 cal BC; Table  14.1, c), also suggested by the co-existence of Bell Beaker and Capo Graziano cultural aspects pottery. After the site was abandoned, the area was used for agriculture (Nava et al. 2007: 112). The settlement discovered at TAV 4-lotto 10 site 4 might also be dated to a similar period on the basis of the material culture found and its stratigraphic position (Fig. 14.1 and Table 14.2, number 9). Another two settlements, abandoned before the PdA eruption, were found at TAV 5-lotto 1/18 site 6 and at TAV 4-lotto 10, the latter located c. 130 m south of the above TAV 4-lotto 10 site 4 (Fig. 14.1 and Table 14.2, numbers 11 and 9). In the area considered, there are only two settlements that were demonstrably occupied at the moment of the PdA event: Gricignano-Cambrannone and Afragola TAV 5-lotto 17 site 8 (Fig. 14.1 and Table 14.2, numbers 4 and 13). At Afragola, excavation identified the north and south perimeter of an EBA site destroyed by the PdA eruption enclosing at least 24 huts (Di Vito et al. 2019: 239). Beyond this boundary an extensively wooded zone was identified, dominated by oak (Laforgia et al. 2009: 103).

14.3.3 Naples Underground Lines and Fuorigrotta Archaeological fieldwork between the territory of Naples and Fuorigrotta has brought to light prehistoric agrarian evidence dating to between the early 4th millennium BC and the 3rd millennium BC. Nevertheless, Late Neolithic postholes were found at Naples-­Vico Pallonetto, Naples-via Toledo/via Diaz, Naples-Piazza Montecalvario and Naples-Piazza Santa Maria degli Angeli on top of the humified layer related to the 8th millennium BC Pigna San Nicola eruption debris (Table 14.2, numbers 22, 24-25 and 28). Giampaola et  al. (2019: 212) suggest that in this phase the area considered was occupied only for seasonal exploitation. At Naples-Piazza Santa Maria degli Angeli, the settlement found on top of the Pigna San Nicola debris was associated with Serra d'Alto cultural aspect pottery; the settlement was then affected by negative hoe-marks filled by ash deriving from the Ischia eruption of Piano Liguori dated to the 4th millennium BC (Table  14.1, o; Fig.  14.1 and Table  14.2, number 28). Agrarian exploitation, characterised by negative cross-cutting plough-marks associated with Diana cultural aspect and Early Copper Age pottery, is documented on top of the Paleosol B (Fig. 14.4a) and continues until the paleosol on top of the PaleoAstroni 2 eruption debris, dated after 3020–2600 cal. BC (Table 14.1, j; Table 14.2, number 28). At Naples-Piazza Montecalvario Final Neolithic/Early Copper Age negative cross-cutting plough-marks were recorded on top of the Paleosol B (Fig. 14.1 and Table 14.2, number 25). Occupation dated between the Final Neolithic/

218

F. Saccoccio

Fig. 14.4  Negative cross-cutting plough-marks discovered on top of the Paleosol B at Naples: (a) Piazza Santa Maria degli Angeli; (b) Via Diaz (Giampaola and Boenzi 2013: figs 4 and 5)

Early Copper Age and the Gaudo cultural aspect phase, the latter sealed by the AMS debris, is attested by pottery sherds at Naples-Stazione San Pasquale (Table 14.1, i; Table 14.2, number 29). Final Neolithic/Early Copper Age negative cross-cutting plough-marks on top of the Paleosol B were also found at Naples-via Toledo/via Diaz (Fig. 14.1 and Table 14.2, number 24; Fig.  14.4b). Agrarian exploitation at Naples-Via Toledo/Via Diaz and at Naples-Vico Pallonetto seems to have lasted until the paleosol formed on top of the PaleoAstroni 2 eruption debris (Table 14.2, numbers 24 and 22). At Naples-Via Donnaregina agrarian exploitation was recorded between under Agnano 3 (Table 14.1, m) and the AMS tephra (Table 14.1, i; Table 14.2, number 18). Negative cross-­ cutting plough-marks were found between under Agnano 3 and PaleoAstroni 2 volcanic debris (Table 14.1, m, j) at Naples-­Sant’Andrea delle Dame (Table 14.2, number 20); at Naples-via Verdi agrarian exploitation seems recorded under the Agnano 3 debris, and then between the PaleoAstroni 2 and AMS events (Table 14.1, m, j, i; Fig. 14.1 and Table 14.2, number 26). Fieldwork in advance of the construction of the underground railway line at Via Settembrini (Fig.  14.1 and Table 14.2, number 19) uncovered an agrarian layer located overlain by debris from the AMS eruption (2860–2480 cal BC; Table 14.1, i). Two graves with Gaudo cultural aspect material found at Naples-Vico della Neve may be dated to a similar, or slightly older, chronological phase (c. 2900-2600 cal BC; Talamo et al. 2011: fig. 1; Fig. 14.1 and Table 14.2, number 23). At Fuorigrotta (Fig. 14.1 and Table 14.2, number 30), c. 8 km south-west of the evidence just discussed, two trenches for the construction of the new CNR (National Research Centre) building reached a depth of c. 6 m bringing to light AMS eruption debris (2860–2480 cal BC; Table  14.1, i).

Above the eruptive layer, two different agrarian traces were found: in one trench orthogonal-cross cutting plough-marks, in the other hoe-marks (Sampaolo 2012: 1328). Orthogonal cross-cutting plough-marks were also found stratigraphically located between the PaleoAstroni 3 (dated after 2840– 2230 cal BC; Table  14.1, g) and the Solfatara (after 2470–2060 cal BC; Table 14.1, d) eruption debris (Sampaolo 2012: 1328). Two pottery sherds, stratigraphically located on top of a paleosol between the AMS and PdA debris, were found in a core at Naples-Castel Nuovo (Table 14.2, number 27). Moreover, at Naples-Stazione Duomo EBA pottery was found in the layer sealed by the PdA debris (Table  14.2, number 21); at Naples-­Stazione San Pasquale EBA pottery vessels seem to be recorded only after the PdA event (Giampaola et al. 2019: 236−237).

14.4 Discussion In the previous section, an overview of the published prehistoric evidence recorded along the course of the rivers Clanis/ Regi Lagni and Sebeto and from the area of Naples was presented with the aim of reconstructing the settlement pattern over time (see Table 14.2). In this section, this evidence will be discussed in terms of land exploitation, showing how selected areas were possibly inhabited and exploited over centuries, even millennia, and suggesting the presence of a series of agrarian management patterns over time, although the considered area was repeatedly affected by eruptive events. To date, the oldest known material documenting human presence in the study area is Late Neolithic Serra d’Alto and Final Neolithic Diana cultural aspects pottery (Table 14.2; Fig. 14.5, phase 1). Figure 14.5 provides a diachronic summary of the evidence discussed using eruption debris as key chronological

14  The Prehistoric Agrarian Management Pattern in the Volcanic Landscape of the Campanian Plain, Southern Italy

219

Fig. 14.5  Settlement and exploitation patterns between the late 5th/early 4th millennium BC and the early 2nd millennium BC along the rivers Clanis/Regi Lagni and Sebeto and the territory of Naples, Campanian Plain. DTM data from Farr et al. 2007

brackets. Different symbols indicate the nature of the evidence found at each site and a site exploitation territory is proposed. This territory, indicated by a 2.1 km radius circle, was calculated by Saccoccio et al. (2013: 89–90) on the basis of the distances between those settlements active at the moment of the PdA eruption debris (1950–1820 cal BC; Table 14.1, c) and destroyed by it, the so-called “destruction horizon” (Saccoccio et al. 2013: 89). All the sites considered were contemporary as they were destroyed at the same time. This territory size suggests the presence of a dense settlement pattern in place in the Campanian Plain in the EBA and finds a good parallel in the 2.5 km radius ‘catchment area’ calculated by Flannery (1976: 109) as needed to satisfy the

basic agricultural requirements of the Mesoamerican site of San Josè, dated 1150–850 BC. In Fig. 14.5 each circle, defining a site exploitation territory of 2.1 km radius, was associated to a site, or groups of sites, according to the sites’ spatial distribution and to topography, in that river valleys and highlands were considered as natural boundaries. In the territory of Gricignano the circle fits well north with the Holocene valley of the river Clanis and east with the Holocene valley related to a tributary of the river Clanis (Fig. 14.5, numbers 1-6). In total 30 sites were considered, with the limitation that published information is mostly preliminary in character. The occupation and exploitation of at least 10 different areas

220

across part, or the whole period, between the late 5th/early 4th millennia BC and the early 2nd millennium BC is documented. Occupation is generally marked by one site per phase which I believe is, at least in part, linked to the limited fieldwork to date (see Fig. 14.5). This is shown by the evidence from Naples, a better explored area that documents different sites distributed across the hypothesised exploitation territory, and so suggests that a large area was possibly devoted to agriculture (Fig. 14.5, sites 17–30). Eruptive events possibly led settlement areas to shift a short distance in  location over time, as at Gricignano US Navy base, attesting the return of people to previously occupied areas (see Fig. 14.2). Interestingly, evidence also suggests that the same site could change in use over time: settlement areas becoming field-systems and vice versa (see, for example, Fig. 14.5, numbers 12 and 14). Pollen analysis suggests an open landscape dominated by NAP pollen (70 to 90%), of which cereals account for 10–15% (Albore Livadie 1999b; Vivent and Albore Livadie 2001; Saccoccio et al. 2013: 90 and fig. 11). Stock-rearing was also practised, documented both by ox imprints left on the ash of the PdA debris at Palma Campania and by skeletons of sheep buried by the same eruption at Nola-Croce del Papa (Albore Livadie et al. 1998: 65; 2019: 206). On the basis of the evidence at Gricignano US Navy base and Cambrannone (Fig. 14.5, numbers 4–5), Vanzetti et al. (2019) hypothesised a long-lasting agrarian management pattern between the Late Copper Age (AMS eruption, 2860–2480 cal BC; Table 14.1, i) and the EBA (PdA eruption, 1950–1820 cal BC; Table 14.1, c). This was possibly detected thanks to the huge area investigated, 90 ha over a decade (Saccoccio et al. 2013: 83; Saccoccio 2021). Remarkably, despite small scale shifts in the location of the settled area over time, the agrarian area exploited remains more or less the same throughout the period considered. Looking at the evidence at a regional scale, it is possible to back-date the presence of an agrarian management pattern at Gricignano at least to the 4th millennium BC. This hypothesis relies on three main arguments. First, all the Final Neolithic/Early Copper Age evidence found in the considered area falls into the hypothetical 2.1 km radius site exploitation territory (Fig.  14.5, phases 1–2); second, because ploughing is attested elsewhere in the study area from the 4th millennium BC (Fig.  14.5; phases 1-2), the absence of such evidence at Gricignano itself might simply be explained by the limited area explored to date; and third, the relocation of settlement and agrarian evidence within the defined site exploitation territory at Gricignano suggests continuous occupation throughout the time span considered, linked to successive agrarian management patterns in which the fertility of the soil was periodically enhanced by eruptive debris.

F. Saccoccio

On the basis of Fig. 14.5, I believe it is possible to extend the presence of a similar agrarian management pattern to the whole study area in prehistoric times, with agrarian exploitation of the same territory for centuries, even millennia (see, for example, Fig.  14.5, number 16). Interestingly, funerary evidence seems to locate at the edge of the suggested exploitation territories, possibly marking their possession by the local communities (see Fig.  14.5, numbers 6 and 23). The limited fieldwork to date and preservation might explain the absence of continuous evidence for the exploitation of some territories displayed in Fig. 14.5. Moreover, especially for the Late/Final Neolithic period (Fig. 14.5, phase 1), the lack of evidence might be related to the depth reached during the excavations, which is very often not stated in published reports. This, together with the previous assumptions, might explain the absence of evidence in the areas of sites 7−15 and 17 in some phases. On the other hand, the pattern shown by the aforementioned areas might also be easily explained as abandonment for a certain amount of time or by the lack of preservation of the archaeological record affected by subsequent occupation. The settlement pattern also allows us to estimate the effect of the AMS debris on the local landscape which was described as devastating by de Vita et  al. (1999: 300). Marzocchella (1998: 108, 113) suggested that AMS debris caused a rise in the water table and of the water flow rate of the river Clanis which may have led to floods. As it is possible to see in Fig. 14.5, in phases 3 and 4, the settled and exploited area at Gricignano shifts in  location from the northern outskirts of the site exploitation territory to the southern one, and is no longer characterised by a close proximity to the Clanis riverbed. However, Fig.  14.5 (phases 3–4) shows that the AMS debris did not affect other areas in a similar, destructive, way; the only exception could be the area of Naples where documented human presence after this event is very scarce (see Fig. 14.5, phase 4, number 27). At the same time, it is not possible to be sure if people moved to Fuorigrotta, 8 km away (Fig. 14.5; number 30), which has evidence of occupation only after the AMS event. The excavation at this site, in fact, does not provide evidence for the phases underlying the AMS debris which were not reached during the excavation. In 2019, Albore Livadie et al. (2019: 217) suggested that this agrarian management pattern lasted at least until the AP2 eruption, dated to 1620–1430 cal BC (Table 14.1, a), according to the evidence found in the Campanian Plain following the PdA event (for example Boscoreale and Boscotrecase in Stefani et al. 2001). After this phase, in fact, settlements seem to relocate to hilltop areas, possibly in search of greater security, abandoning the Campanian Plain. New material culture evidence appears: the so-called ProtoApennine cultural aspect (Albore Livadie et  al. 2019: 217–218).

14  The Prehistoric Agrarian Management Pattern in the Volcanic Landscape of the Campanian Plain, Southern Italy

14.5 Conclusions Comprehensive publication of the archaeological research in advance of construction of the Rome-Naples TAV line is still lacking but is a priority. Site descriptions in this paper, in fact, are mainly based on preliminary reports which are, to date, the only literature available. It is, therefore, difficult to reconstruct the tenure pattern and function of the fields as they are mainly recorded as negative plough-marks explored over limited areas. The only positive agrarian evidence which is wellenough explored and published to date that can contribute to understanding this issue is the 90 ha EBA field system sealed by the PdA debris at Gricignano US Navy base. On the basis of this evidence, Saccoccio et al. (2013: 91; see also Saccoccio 2021) suggested the presence of a tribal economy (sensu Polanyi 1944) with single agrarian plots “temporarily attributed to members of the community… correspond[ing] to lineage groups, acting in a corporate way”. Marzocchella (2002) proposed using foreign case-studies as comparisons for the EBA Gricignano US Navy base evidence. Caulfield (1978: 138) suggested “a sizeable organized community working towards a single objective” to explain the Neolithic agrarian evidence at Behy/Glenura, County Mayo (Ireland). This interpretation was based mainly on two lines of evidence: the regular plan of the agrarian features and a single documented instance of forest clearance by fire; Caulfield (1978: 141) interpreted the long strip divisions of the fields as individual farms. Fleming (1988: 64; 122) examined a range of interpretative options from “distribution and perhaps periodic reallocation among various families or other social groups” to perpetuity of land possession or a mixed system combining private and community ownership linked to the presence of stratified societies in order to explain the Bronze Age coaxial field system on Dartmoor, England. However, on the basis of the dispersed settlement pattern with huts aggregated in smaller groups, Fleming (1988: 66) proposed cogent parallels with twentieth century Welsh rural social environments, known as trefgordd. These were characterised by considerable local political strength, and a deep sense of community was sparked by the absence of a central social centre enabling direct relationships between the social groups involved. On the basis of the same evidence, Johnston (2005) gave importance to the connection between burial mounds, huts and field systems in the form of dominant alignments maintained through time. This pattern enabled a direct link to be established between human groups’ ancestry and their claims on land and resources, with the formation and maintenance of social groupings through identity and continuity of ownership. From a geographical point of view, this paper defines 10 different areas mostly occupied at least from the 4th millennium BC linked to the 30 analysed sites by associating to a

221

site, or groups of sites, a 2.1 km radius site exploitation territory. Even if occupation is mainly marked by one site per phase (Fig. 14.5), an observation which may be due to limited fieldwork to date, the evidence suggests that prehistoric farmers returned to occupy previous areas even after devastating volcanic events. At Gricignano US Navy base, reoccupation is marked by graves characterised mostly by poor grave goods made on the site of previously settled areas (Albore Livadie and Marzocchella 1999; Fugazzola Delpino et  al. 2003; 2007). This pattern may lead to two considerations. On the one hand, continuous occupation might suggest that farmers wanted to establish/stress a relationship with the past, with the people that previously lived in and exploited the same landscape. It is not possible to speculate whether newcomers were direct descendants of former farmers, but the need to establish a tie with the past enhances the hypothesis of continuity (Albore Livadie et al. 2019). On the other hand, a series of agrarian management patterns, most probably linked to the fertility of the soil enhanced by eruptive debris over time, allowed continuous settlement of the plain over centuries, even millennia. In most cases, this agrarian management pattern should be seen as developing at least from the 4th millennium BC, but possibly even from the late 5th millennium BC albeit only one radiocarbon date seems to support this statement so far (Table 14.1, p). The only area definitely occupied for the whole period considered is that of Gricignano which, however, only documents evidence for the agrarian exploitation of the local landscape from the 3rd millennium BC (Table 14.2, numbers 1-6; Fig. 14.5). The lack of earlier agrarian evidence was explained as being due to limited fieldwork to date. Limited fieldwork to date is also a possible explanation for the lack of continuous occupation through time, either settled or agrarian exploitation, for eight out of the other nine defined areas. Only in one case is abandonment suggested. This is in the area of Naples where very limited archaeological evidence seems to be recorded after the AMS event, possibly linked to its devastating effects on the local landscape so as suggested by de Vita et al. (1999: 300). However, it is not possible to propose with certainty if at least part of the population moved to Fuorigrotta on the basis of the data in our possession. To conclude. This paper discusses prehistoric management patterns in the Campanian Plain, southern Italy, focusing on the area between the course of the rivers Clanis/Regi Lagni and Sebeto and the territory of Naples. Although the available data are still scanty, it is possible to highlight the presence of at least 10 territories exploited for centuries, even millennia, where communities returned even after devastating eruptive events, re-organising and relocating their settlement and field systems. This pattern allows us to suggest the presence of a series of management patterns, dated at least between the 4th and the early 2nd millennia BC,

222

probably to be linked to soil fertility enhanced by eruptive debris over time. Agriculture, in fact, played a major part in the economy of the prehistoric Campanian Plain as suggested by the widespread presence of agrarian evidence and supported by palaeoenvironmental reconstructions. Acknowledgements  I would like to express my sincerest thanks and appreciation to the editors of the volume for the interest shown in building awareness and developing discussion on prehistoric landscape agrarian exploitation. I would also like to thank Amodio Marzocchella for granting permission to use the images in Fig.  14.3, Alessandro Sabbatini for his help given in procuring bibliography and Mark Pearce, Selena Alota and the two referees for their useful comments on the paper. DTM data services were provided by the OpenTopography Facility with support from the National Science Foundation under NSF Award Numbers 1557484, 1557319, and 1557330.

References Albore Livadie, C. & D'Amore, L. (1981). Palma Campania (Napoli) – Resti di abitato dell’età del bronzo antico. Notizie Scavi di Antichità, 34, 59–101. Albore Livadie, C. (Ed.). (1986). Tremblements de terre, eruptions volcaniques et vie des hommes dans la Campanie antique. Naples: Centre Jean Bérard – Institut Français de Naples. Albore Livadie, C. (Ed.). (1999a). L’eruzione delle Pomici di Avellino e la facies di Palma Campania (Bronzo Antico) (Atti del Seminario Internazionale). Bari: Edipuglia. Albore Livadie, C. (1999b). Territorio e insediamenti nell’agro Nolano durante il Bronzo antico (facies di Palma Campania): Nota preliminare. In C. A. Livadie (Ed.), L’Eruzione Vesuviana delle “Pomici di Avellino” e la facies di Palma Campania (Bronzo antico) (pp. 203– 246). Bari: Edipuglia. Albore Livadie, C., & Marzocchella, A. (1999). Riflessioni sulla tipologia funeraria in Campania tra Bronzo antico e Bronzo medio. In A.  Gravina (Ed.), Ipogei della Daunia: Culti e riti funerari nella media età del Bronzo (19° Convegno Nazionale sulla Preistoria – Protostoria – Storia della Daunia) (pp.  117–134). San Severo: Archeoclub d’Italia. Albore Livadie, C., & Widemann, F. (Eds.) (1990). Volcanologie et Archéologie (Pact 25). Ravello: Centre Universitaire pour le Biens Culturels. Albore Livadie, C., Mastrolorenzo, G., & Vecchio, G. (1998). Eruzioni pliniane del Somma-Vesuvio e siti archeologici dell’area nolana. In P. G. Guzzo & R. Peroni (Eds.), Archeologia e Vulcanologia in Campania. Atti del Convegno (pp. 39–86). Naples: Arte Tipografica. Albore Livadie, C., Pearce, M., Delle Donne, M., & Pizzano, N. (2019). The effects of the Avellino Pumice eruption on the population of the Early Bronze Age Campanian plain (Southern Italy). Quaternary International, 499, 205–220. Bronk Ramsey, C. (2017). Methods for summarizing radiocarbon datasets. Radiocarbon, 59(2), 1809–1833. Caulfield, S. (1978). Neolithic fields: The Irish evidence. In H.  C. Bowen & P.  J. Fowler (Eds.), Early land allotment in the British Isles (British archaeological reports, 48) (pp.  137–144). Oxford: BAR. De Caro, S. (2001). L’attività archeologica della Soprintendenza di Napoli e Caserta nel 2000. In Problemi della Chora coloniale dall’occidente al Mar Nero. Atti del 40° convegno di studi sulla Magna Grecia (pp.  865–905). Taranto: Istituto per la Storia e l’Archeologia della Magna Grecia.

F. Saccoccio De Caro, S. (2002). L’attività archeologica della Soprintendenza di Napoli e Caserta nel 2001. In A. Stazio & S. Ceccoli (Eds.), Taranto e il Mediterraneo. Atti del 41° convegno di studi sulla Magna Grecia (pp. 636–675). Taranto: Istituto per la Storia e l’Archeologia della Magna Grecia. De Caro, S. (2003). L’attività archeologica della Soprintendenza di Napoli e Caserta nel 2002. In A.  Stazio & S.  Ceccoli (Eds.), Ambiente e Paesaggio nella Magna Grecia. Atti del 42° convegno di studi sulla Magna Grecia (pp. 569–621). Taranto: Istituto per la Storia e l’Archeologia della Magna Grecia. de Vita, S., Orsi, G., Civetta, L., Carandente, A., D’Antonio, M., Deino, A., di Cesare, T., Di Vito, M. A., Fischer, R. V., Isaia, R., Marotta, E., Necco, A., Ort, M., Pappalardo, L., Piochi, M., & Southon, J. (1999). The Agnano-Monte Spina eruption (4100 years BP) in the restless Campi Flegrei caldera (Italy). Journal of Volcanology and Geothermal Research, 91, 269–301. De Vivo, B., Rolandi, G., Gans, P.  B., Clavert, A., Bohrson, W.  A., Spera, F. J., & Belkin, H. E. (2001). New constraints on the pyroclastic eruptive history of the Campanian volcanic Plain (Italy). Mineralogy and Petrology, 73, 47–65. Delibrias, G., Di Paola, G. M., Rosi, M., & Santacroce, R. (1979). La storia eruttiva del complesso vulcanico Somma Vesuvio ricostruita dalle successioni piroclastiche del Monte Somma. Rendiconti della Società Italiana di Mineralogia e Petrologia, 35(1), 411–438. Di Renzo, V., Di Vito, M.  A., Arienzo, I., Caradente, A., Civetta, L., D’Antonio, M., Giordano, F., Orsi, G., & Tonarini, S. (2007). Magmatic history of Somma-Vesuvius on the basis of new geochemical and isotopic data from a deep borehole (Camaldoli della Torre). Journal of Petrology, 48(4), 753–784. Di Vito, M. A., Isaia, R., Orsi, G., Southon, J., De Vita, S., D’Antonio, M., Pappalardo, L., & Piochi, M. (1999). Volcanism and deformation since 12,000 years at the Campi Flegrei caldera (Italy). Journal of Volcanology and Geothermal Research, 91, 269–301. Di Vito, M., Zanella, E., Gurioli, L., Lanza, R., Sulpizio, R., Bishop, J., Tema, E., Boenzi, G., & Laforgia, E. (2009). The Afragola settlement near Vesuvius, Italy: The destruction and abandonment of a Bronze Age village revealed by archaeology, volcanology and rockmagnetism. Earth and Planetary Science Letters, 277, 408–421. Di Vito, M., Talamo, P., de Vita, S., Rucco, L., Zanchetta, G., & Cesarano, M. (2019). Dynamics and effects of the Vesuvius Pomici di Avellino Plinian eruption and related phenomena on the Bronze Age landscape of Campania region (Southern Italy). Quaternary International, 499, 231–244. Farr, T. G., Rosen, P. A., Caro, E., Crippen, R., Duren, R., Hensley, S., Kobrick, M., Paller, M., Rodriguez, E., Roth, L., Seal, D., Shaffer, S., Shimada, J., Umland, J., Werner, M., Oskin, M., Burbank, D., & Alsdorf, D. (2007). The Shuttle Radar Topography Mission. Review of Geophysics, 45, RG2004. Accessed 29 June 2019. Available at https://doi.org/10.1029/2005RG000183. Flannery, K.  V. (1976). Empirical determination of site catchment in Oaxaca and Tehuacán. In K.  V. Flannery (Ed.), The early Mesoamerican Village (pp.  103–117). Orlando/San Diego/New York/London/Toronto/Montreal/Sydney/Tokyo: Academic. Fleming, A. (1988). The Dartmoor reaves: Investigating prehistoric land divisions. London: B.T. Batsford Ltd. Fugazzola Delpino, M.  A., Salerno, A., Tagliacozzo, A., & Tiné, V. (2003). Una comunità della facies di Laterza nella Pianura Campana: l’area Forum di Gricignano US Navy (CE). In Atti della 35 Riunione Scientifica IIPP, Le comunità della Protostoria italiana: studi e ricerche sul Neolitico e le età dei metalli (pp. 199–214). Florence: Istituto Italiano di Preistoria e Protostoria. Fugazzola Delpino, M.  A., Salerno, A., Tagliacozzo, A., Tiné, V., & Vanzetti, A. (2007). Villaggi e necropoli dell’area “Centro Commerciale” di Gricignano d’Aversa/US Navy (Caserta). In Atti della 40 Riunione Scientifica IIPP, Strategie di insediamento fra

14  The Prehistoric Agrarian Management Pattern in the Volcanic Landscape of the Campanian Plain, Southern Italy Lazio e Campania in età preistorica e protostorica (pp. 521–537). Florence: Istituto Italiano di Preistoria e Protostoria. Giampaola, D., & Boenzi, G. (2013). Interazione tra attività vulcanica e vita dell’uomo: evidenze archeologiche nell’area urbana di Napoli. Miscellanea INGV, 18, 38–44. Giampaola, D., & Stanislao, I. (2007). Loc. Botteghelle (Napoli). Un insediamento del Neolitico avanzato. In Atti della 40 Riunione Scientifica IIPP, Strategie di insediamento fra Lazio e Campania in età preistorica e protostorica (pp. 606–610). Florence: Istituto Italiano di Preistoria e Protostoria. Giampaola, D., Bartoli, C., & Boenzi, G. (2019). Napoli. Territorio e occupazione in età pre e protostorica. AION, 25(2018), 207–254. Giordano, A. (1834). Memorie istoriche di Fratta Maggiore. Naples: Stamperia Reale. Guzzo, P.  G. (2010). Attività della Soprintendenza per i Beni Archeologici di Napoli e Pompei. In Cuma. Atti del 48° convegno di studi sulla Magna Grecia (pp. 1007–1035). Taranto: Istituto per la Storia e l’Archeologia della Magna Grecia. Guzzo, P. G., & Peroni, R. (Eds.) (1998). Archeologia e vulcanologia in Campania. Atti del Convegno. Naples: Arte Tipografica. Iovine, R. S., Fedele, L., Mazzeo, F. C., Arienzo, I., Cavallo, A., Wörner, G., Orsi, G., Civetta, L., & D’Antonio, M. (2017). Timescales of magmatic processes prior to the ∼4.7 ka Agnano-Monte Spina eruption (Campi Flegrei caldera, Southern Italy) based on diffusion chronometry from sanidine phenocrysts. Bulletin of Volcanology, 79. Accessed 29 June 2019. Available at https://doi.org/10.1007/ s00445-­017-­1101-­4. Isaia, R., Marianelli, P., & Sbrana, A. (2009). Caldera unrest prior to intense volcanism in Campi Flegrei (Italy) at 4.0 ka B.P.: Implications for caldera dynamics and future eruptive scenarios. Geophysical Research Letters, 36, L21303. Accessed 29 June 2019. Available at https://doi.org/10.1029/20.09GL040513. Johnston, R. (2005). Pattern without a plan: Rethinking the Bronze Age coaxial field systems on Dartmoor, South-West England. Oxford Journal of Archaeology, 24(1), 1–21. Laforgia, E., & Boenzi, G. (2011). Nuovi dati sull’Eneolitico della piana campana dagli scavi A.V. in provincia di Napoli. In Atti della 43 Riunione Scientifica IIPP, L’età del Rame in Italia (pp.  249– 255). Florence: Istituto Italiano di Preistoria e Protostoria. Laforgia, E., Boenzi, G., Amato, L., Di Vito, M.  A., Fattore, L., Stanzione, M., & Viglio, F. (2009). The Vesuvian “Pomici di Avellino” eruption and Early Bronze Age settlement in the middle Clanis valley. Méditerranée, 112, 101–107. Marzocchella, A. (1985). L’eneolitico a Napoli. In G.  Macchiaroli (Ed.), Napoli antica (pp. 27–29). Naples: Gaetano Macchiaroli. Marzocchella, A. (1998). Tutela archeologica e Preistoria nella Pianura Campana. In P.  G. Guzzo & R.  Peroni (Eds.), Archeologia e Vulcanologia in Campania (pp. 97–133). Naples: Arte Tipografica. Marzocchella, A. (2002). Le più antiche testimonianze di utilizzazione del territorio: “sistemi agrari” pre- e protostorici. In Il Mondo dell’Archeologia (Vol. I, pp. 905–907). Rome: Treccani. Matarazzo, T., Berna, F., & Goldberg, P. (2010). Occupation surfaces sealed by the Avellino eruption of Vesuvius at the Early Bronze Age village of Afragola in southern Italy: A micromorphological analysis. Geoarchaeology, 25(4), 437–466. Nava, M.  L., Giampaola, D., Laforgia, E., & Boenzi, G. (2007). Tra il Clanis e il Sebeto: nuovi dati sull’occupazione della piana campana tra il Neolitico e l’l’età del Bronzo. In Atti della 40 Riunione Scientifica IIPP, Strategie di insediamento fra Lazio e Campania in età preistorica e protostorica (pp.  101–126). Florence: Istituto Italiano di Preistoria e Protostoria. Orsi, G., Piochi, M., Campajola, L., D’Onofrio, A., Gialanella, L., & Terrasi, F. (1996). 14C geochronological constraints for the volcanic history of the island of Ischia (Italy) over the last 5000 years. Journal of Volcanology and Geothermal Research, 71, 249–257.

223

Passariello, I., Albore Livadie, C., Talamo, P., Lubritto, C., D’Onofrio, A., & Terrasi, F. (2009). 14C chronology of Avellino pumice eruption and timing of human reoccupation of the devastated region. Radiocarbon, 51(2), 803–816. Polanyi, K. (1944). The great transformation. New York: Rinehart. Reimer, P. J., Bard, E., Bayliss, A., Beck, J. W., Blackwell, P. G., Bronk Ramsey, C., Buck, C. E., Cheng, H., Edwards, R. L., Friedrich, M., Grootes, P. M., Guilderson, T. P., Haflidason, H., Hajdas, I., Hatté, C., Heaton, T.  J., Hoffmann, D.  L., Hogg, A.  G., Hughen, K.  A., Kaiser, K. F., Kromer, B., Manning, S. W., Niu, M., Reimer, R. W., Richards, D. A., Scott, E. M., Southon, J. R., Staff, R. A., Turney, C. S. M., & van der Plicht, J. (2013). IntCal13 and Marine13 radiocarbon age calibration curves 0-50,000 years cal BP. Radiocarbon, 55(4), 1869–1887. Rolandi, G., Petrosino, P., & Mc Gehin, J. (1998). The interplinian activity at Somma-Vesuvius in the last 3550 years. Journal of Volcanology and Geothermal Research, 82, 19–52. Saccoccio, F. (2021). Crop field management and social structure at Gricignano d’Aversa (Campanian Plain, Southern Italy) in the Early Bronze Age. Environmental Archaeology: The Journal of Human Palaeoecology, 26(3), 267–280. Saccoccio, F., Marzocchella, A., & Vanzetti, A. (2013). The field system of Gricignano d’Aversa (Southern Italy) and the agrarian impact in the Piana Campana, ca. 3900 cal BP. Quaternary International, 303, 82–92. Sampaolo, V. (2005). L’attività archeologica della Soprintendenza di Napoli e Caserta nel 2004. In Tramonto della Magna Grecia. Atti del 44° convegno di studi sulla Magna Grecia (pp.  663– 705). Taranto: Istituto per la Storia e l’Archeologia della Magna Grecia. Sampaolo, V. (2012). L’attività archeologica della Soprintendenza di Napoli e Pompei. In Alle origini della Magna Grecia. Mobilità, migrazioni, fondazioni. Atti del 50° convegno di studi sulla Magna Grecia (pp.  1309–1353). Taranto: Istituto per la Storia e l’Archeologia della Magna Grecia. Sasaki, R., Yoshida, M., Ohtsu, Y., Miyahira, M., Ohta, H., Watanabe, M., & Suzuki, S. (2003). Soil formation of new lahar materials derived from Mt. Pinatubo. Soil Science and Plant Nutrition, 49(4), 575–582. Stefani, G., Fergola, L., Albore Livadie, C., & Di Maio, G. (2001). Le Somma – Vésuve et al région p­ ompéienne durant la Préhistoire: premiers résultats géorachéologiques à Boscoreale et à Boscotrecase. In E. Juvigné & J. P. Raynal (Eds.), Tephras – Cronologie et archéologie. Actes du Symposium international Téphrochronologie et co-­ existence hommes-volcans, Brives-Charensac 24-29 août 1998, Les dossiers de l’Archéo-logis (Vol. 1, pp.  211–228). Laussonne: Clermond Ferrand. Talamo, P., Passariello, I., Lubritto, C., & Terrasi, F. (2011). Evoluzione culturale in Campania: indagine cronologica sistematica tramite datazioni radiocarboniche. In Atti della 43 Riunione Scientifica IIPP, L’età del Rame in Italia (pp. 39–48). Florence: Istituto Italiano di Preistoria e Protostoria. Vanzetti, A., Marzocchella, A., & Saccoccio, F. (2019). The Campanian agrarian systems of the late Copper-Early Bronze Age (ca. 4.5503.850 BP): A long-lasting agrarian management tradition before the Pomici di Avellino eruption. Quaternary International, 499, 148–160. Vingiani, S., Minieri, L., Albore Livadie, C., Di Vito, M. A., & Terribile, F. (2017). Pedological investigation of an early Bronze Age site in southern Italy. Geoarchaeology. Accessed 29 June 2019. Available at https://doi.org/10.1002/gea.21625. Vivent, D., & Albore Livadie, C. (2001). Contribution de la palynologie à l’étude des comportements humains en contexte volcanique actif: l’exemple de l’Âge du Bronze en Campanie (Italie). In E. Juvigné & J. P. Raynal (Eds.), Tephras – Cronologie et archéologie. Actes

224 du Symposium international Téphrochronologie et co-existence hommes-­ volcans, Brives-Charensac 24–29 août 1998, Les dossiers de l’Archéo-logis (Vol. 1, pp.  245–254). Clermond Ferrand: Laussonne. Vogel, S., Märker, M., Rellini, I., Hoelzmann, P., Wulf, S., Robinson, M., Steinhübel, L., Di Maio, G., Imperatore, C., Kastenmeier, P., Liebmann, L., Esposito, D., & Seiler, F. (2016). From a stratigraphic sequence to a landscape evolution model: Late Pleistocene

F. Saccoccio and Holocene volcanism, soil formation and land use in the shade of Mount Vesuvius (Italy). Quaternary International, 394, 155–179. Zanchetta, G., Sulpizio, R., Roberts, N., Cioni, R., Eastwood, W.  J., Siani, G., Caron, B., Paterne, M., & Santacroce, R. (2011). Tephrostratigraphy, chronology and climatic events of the Mediterranean basin during the Holocene: An overview. The Holocene, 21(1), 33–52.

Index

A Abandoned, 9, 59, 64, 72, 75, 77–79, 91–93, 96, 107, 137, 138, 153, 172, 173, 176, 177, 217 Abandonment, 7, 9, 17, 67, 74, 87, 92, 119, 125, 137, 170, 186, 220, 221 Accelerator mass spectrometry (AMS), 31, 32, 155, 156, 187, 190, 213, 216, 217, 220 date, 31, 80 dating, 3, 79, 190 eruption, 217, 218 event, 213, 217, 220, 221 Achany Glen, 87, 89, 91–93, 95–98 Acidic, 116 Aegean, 102, 112, 143, 161–163 Aerial photographs, 8, 9, 13, 18, 23, 29, 30, 90, 118, 119, 122, 146, 152, 159, 167–169, 172–176, 178, 179, 182 Aerial photography, 8, 122, 168, 172, 174, 176, 178 Aggregated, 3, 5, 63, 188, 189, 194, 198–202, 221 fields, 188, 194–195 field systems, 189, 194, 195 Agrarian evidence, 210, 213–218, 220–222 Agricultural boundaries, 5, 29 calendar, 109, 110, 112 installations, 102, 105, 107, 108 intensification, 163 labour, 101, 109–112 land, 74, 108, 134, 137, 139 layers, 31, 34 possibilities, 54, 60 practices, 3, 36, 81, 82, 95, 101, 137, 139, 145, 176 production, 8, 9, 74, 88, 105, 108, 109 revolution, 51 strategy, 5, 34 systems, 9, 32–34, 88, 97 terraces, 71–74, 77, 78, 101–103, 105, 107, 111 terracing, 71, 80 work, 110, 111 Alstrup Krat, 168, 177, 179 Animal, 2, 21, 32, 33, 51, 54–61, 64–66, 73, 82, 83, 87, 94, 101, 110, 112, 171, 177, 180, 190, 196, 217 handling, 56 husbandry, 15, 54, 55 labour, 109, 161 Annual cycle, 95, 96 Arable, 9, 15, 16, 21, 22, 29, 30, 34, 57, 59, 61, 63, 65, 66, 80, 81, 88, 95, 105, 108, 161, 162, 168, 171, 173, 177

cropping, 16, 54–59, 61, 63, 64, 66 crops, 53, 56, 58, 59, 66, 173 fields, 167–182 production, 59, 61, 63–66 stubbles, 58, 59 Archaeobotanical data, 54, 59, 64, 66 Ard-marks, 5, 29 B Ballinvally, 118, 122–127 Bank, 1–3, 5, 7, 9, 14, 18, 23, 29–36, 38, 48, 50, 57, 61, 89, 91, 94, 119, 120, 122–124, 127, 137, 155, 168, 172, 174, 175, 186, 187, 191, 192, 194, 196, 198, 199, 202, 216 aggradation, 31–34, 36 sections, 31, 33 sediments, 33, 34 Barley, 15, 17, 31, 34, 35, 61, 80, 108 Barrows, 8, 27, 36, 37, 43, 118, 120, 168, 169, 172–177 Belderg Beg, 116 Bernières, 132 Bestwall Quarry, 194, 198 Big Moor, 186, 191, 192, 194, 198 Bille river, 43 Boiotia, 143, 160–163 Borremose, 169–171, 175, 176, 179 Botanical remains, 108 Boundaries, 5, 7–10, 21–23, 29, 38, 43, 44, 46, 48–50, 53, 54, 56, 57, 59–61, 63, 65, 88, 93, 98, 116, 119–127, 136, 172, 174, 177, 179, 180, 185, 186, 188–197, 201, 217, 219 Box terraces, 102, 107, 109–111 Bronze Age, 9, 13, 15, 17, 21, 23, 27, 29, 36, 38, 53, 57–60, 63, 72–76, 79–83, 87–98, 101, 102, 111, 116, 118, 131–139, 145, 156–160, 163, 168–170, 172–182, 193, 196, 198–203, 221 farm, 76, 79 field, 13, 18, 21–23, 28, 72, 143–163, 191, 193, 197 landscape, 22, 59, 66, 88–90, 92, 96, 98, 107 settlement, 20, 29, 72, 75, 89, 91, 92, 95, 96, 98, 103, 107, 108, 169, 173 Buckelgräber, 46, 48 Buildings, 3, 17, 18, 43, 57, 59, 61, 65, 76, 79, 82, 89, 91, 92, 97, 111, 112, 133, 139, 144, 152, 159–161, 171, 172, 174, 178, 191, 192, 196, 203, 209, 216, 218, 222 Burial cairns, 13, 19, 20, 23, 91, 92, 174 Burial mound, 43, 44, 46, 48, 50, 134 Burnt mound, 23, 88–90, 92–93, 95–97

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 S. Arnoldussen et al. (eds.), Europe’s Early Fieldscapes, Themes in Contemporary Archaeology, https://doi.org/10.1007/978-3-030-71652-3

225

226 C Cadbury Castle, 54, 59–66 Cadbury Hill, 61, 64, 65 Cairns, 9, 14, 19, 23, 89, 92, 95, 96, 108, 118, 120, 122, 124, 125 Cairon, 132, 135, 136 Campanian Plain, 9, 209–222 Campi Flegrei, 72 Campi Flegrei caldera, 9, 209–211, 213 Canals, 5, 146–149, 153, 154, 159, 161 Carbonised residue, 190, 192, 194 Carlit Mountain, 72, 75, 77, 79–81, 83 Caserta, 209, 213–216 Castle hills, 195, 199, 202 Céide Fields, 116, 127 Celtic field, 3, 7–9, 27–38, 43–50, 167, 189 agriculture, 34, 38 banks, 30–37 complex, 36, 38 economy, 34–35 embankments, 43, 44, 50 systems, 8, 28, 38, 46, 49, 50 usage, 30, 35 Cerdanya mountains, 76, 77 Cereal, 31, 35, 61, 75, 76, 80, 81, 83, 88, 93, 95, 98, 137, 171, 173, 220 cultivation, 15, 16, 21 processing, 76, 83 Channel, 102, 131, 136–138, 146–149, 153, 158, 160 Chemin Saulnier, 133 Chronology, 3, 7, 8, 10, 21, 23, 27–38, 54, 59, 66, 90–92, 94, 96–98, 103, 120, 132, 133, 145, 155, 156, 161, 185–198, 213 Cladh Hallan, 91, 95, 97 Clanis, 210, 211, 216–220 Clearance cairns, 9, 14, 15, 18, 95, 120, 172 Clearance heaps, 89, 92–95 Climate, 7, 10, 51, 55, 56, 65, 74, 76, 81, 87, 89, 101, 125, 132, 133, 137, 139, 162, 213 Coaxial and coaxial field systems, 5, 8, 61, 116, 132, 188, 189, 195, 197–202, 216, 221 Cohesive, 5, 107, 116, 189, 195 Colluvial deposits, 78 Colne Fen, 195, 201 Communities, 2, 5, 21, 22, 27, 28, 34, 38, 55, 65, 66, 74, 80, 82, 83, 95, 108, 132, 137, 139, 161, 221 Contour terraces, 102, 107, 109 Cord rig, 93, 94, 200 Corinthia, 5, 10, 101, 103 County Durham, 189, 195 Cracked stones, 90, 93 Cranford lane, 195, 202 Cropping, 54, 57, 58, 63, 65, 110 Crops, 3, 14, 17, 21, 23, 34, 38, 50, 55, 57, 58, 61, 64, 66, 75, 76, 80, 82, 101, 108–110, 146, 147, 158, 162, 171, 176 cultivation, 3, 5, 28, 29, 34, 35, 87 fields, 71–84 processing, 147 production, 76, 81, 82 Cultivation, 9, 10, 15, 21, 32, 33, 43, 47, 48, 50, 51, 56, 57, 64, 74, 76, 80–83, 87, 89, 90, 92–94, 96, 101, 102, 107–112, 116, 119, 125, 159, 168, 170–173, 175–180, 182, 186, 191, 192, 201, 203, 217 activities, 76, 82 ridges, 116, 118–120, 122–124 strategy, 171, 173, 176, 180

Index Cursus monuments, 118, 122, 123 Cyclopean walls, 144, 148, 158, 160 D Dating, 1, 3, 9, 17, 22, 23, 27, 30, 31, 33, 35, 36, 48, 50, 60, 61, 65, 71–73, 79, 90, 95, 104, 108, 116, 119, 126, 127, 132, 133, 135, 136, 155, 167, 172, 177–179, 186, 188, 190–193, 196–198, 211, 217 methods, 23, 187, 190–191 sample, 187 sites, 17, 23 strategies, 3, 31, 186, 191–192, 197, 198 Ditch, 1, 2, 5, 7, 29, 32, 36, 56, 57, 59, 61, 71, 90, 98, 120, 123, 124, 132–137, 146–150, 152–155, 159, 186, 187, 190–196, 198, 200–203 fills, 191, 196 profiles, 153–155 sections, 152 Drainage, 56, 57, 72, 74, 102, 121, 143–145, 147, 149, 155, 158–162, 169, 196, 216 Drainage system, 136, 144–149, 156, 157, 160, 161, 210 Draught animals, 108, 111 Drift sand area, 174, 175 Drumsawry, 118, 120, 122, 123, 125–127 DTM data, 211, 219, 222 Dung, 31, 33, 34, 57, 94, 171, 177 Dykes, 17, 93, 123, 144, 148–150, 158–162, 203 E East Anglia, 57, 59, 123 Eastern Mediterranean, 156, 160, 161 Eastern Pyrenees, 9, 82 East Lothian, 90, 97 East Perthshire, 92, 97 Economy, 1, 2, 5, 7–10, 29, 36, 55, 63, 64, 66, 76, 83, 87, 88, 108, 111, 132, 137, 139, 161–163, 221, 222 Elite, 108, 110, 111, 134, 137, 139 Embanked fields, 5, 32 Embanked field systems, 29 Embankment, 44–46, 48–50, 71, 148, 154 Enclosed, 5, 8, 22, 46, 48, 57–59, 63, 88–90, 92, 98, 116, 120, 132, 134, 136–138, 167–182, 185, 189, 194 field system, 8, 9, 131, 167 landscapes, 88, 93, 98, 182 settlements, 132–134 Enclosure, 5, 13, 14, 17–21, 48, 57, 58, 60, 63, 64, 75, 76, 82, 83, 88, 90, 91, 93–95, 97, 98, 105, 108, 116, 118, 120, 122, 124, 125, 133, 134, 137, 176, 185–188, 190, 191, 193, 194, 196, 197, 199–202 Erosion, 1, 49, 56, 72, 78, 82, 83, 91, 92, 101, 153, 162, 171, 177, 186, 191, 193, 196 Eruption, 9, 73, 209, 210, 213, 216, 217, 220 Evolution, 8, 66, 74, 83, 91, 98, 132, 139 Exploitation, 55, 64, 72, 73, 76, 83, 93, 137, 168, 175, 177, 218–220 Extensive, 2, 5, 8, 9, 13, 15, 18–20, 32, 35, 54–56, 58, 59, 61, 64–66, 87, 89–94, 96, 98, 103, 108–110, 116, 119, 124, 131, 132, 144, 157, 159, 162, 170, 173, 175–177, 185, 192 cultivation, 54, 108, 109, 161, 180 field systems, 9, 19, 116 grazing, 15, 59, 61, 65

Index F Fallow, 3, 34, 35, 38, 83, 93, 96, 98, 177 cycles, 5, 34, 35 fields, 177 Farm, 8, 58, 59, 61, 65, 71, 82, 83, 90, 92, 93, 96, 110, 111, 120, 123, 126, 131–139, 148, 167, 177, 179, 180, 191, 195, 196, 199, 201, 202, 221 work, 108, 110–111 workers, 110 Farming, 9, 13, 15, 28, 32, 48, 56–58, 65–67, 72, 81, 83, 95–98, 101–112, 116, 126, 137, 139, 146, 161, 162 Faunal, 9, 54, 59, 61, 64, 65 assemblages, 53–55, 60, 63, 64, 66, 67 remains, 54, 60, 65 Fences, 5, 7, 29, 32, 34, 37, 56–59, 94, 95, 176, 180, 181, 191, 195 Field, 2, 5, 7–10, 13–23, 27–38, 49, 53, 54, 56–59, 61, 64–67, 71–73, 75–79, 82, 83, 87, 88, 90–95, 97, 98, 101–112, 115, 116, 119, 120, 122, 124, 125, 127, 131, 135, 147, 148, 153, 157–160, 162, 167–169, 171, 175–180, 185–189, 191–197, 221 banks, 8, 32, 167, 173–175, 177–181 boundaries, 5, 46, 59, 115, 116, 119, 120, 122, 124–127, 146, 172, 175, 179, 180, 186, 187, 189–191, 193, 195, 196, 198–203 marks, 146–149, 152 plots, 3, 28, 29, 31, 32, 34, 35, 48, 173, 175, 180 preparation, 109–111 survey, 9, 13, 14, 17, 23 systems, 1–10, 13–23, 28–31, 35, 54, 57–61, 64–66, 71–73, 79–81, 87, 88, 90, 94, 98, 101, 103, 104, 116, 119, 120, 122, 124–127, 131–139, 143–163, 167–182, 185–203, 209, 210, 212, 216, 220, 221 walls, 13, 18, 22, 23, 116, 193, 194 Flax, 34, 35, 61, 64 Flögeln, 3, 29, 31, 35 Flood, 58, 101, 144, 158, 160, 220 Foddering, 55, 58, 59, 64 Forest, 2, 29, 43, 44, 47, 49, 50, 73–76, 78, 79, 89, 168–170, 177, 221 Fortress, 143, 146, 148, 162 Funerary contexts, 133–134, 136 Funnel, 29, 36, 57, 59, 61, 64 Furrow fields, 46, 48 G Gallow Hill, 18, 19, 22 Garden, 27, 28, 102, 110–112, 127 plants, 109, 110 terrace, 110, 112 Gates, 56, 57, 148, 159, 200 Gender, 110 Gendered agricultural spaces, 10 Geophysical survey, 60, 122, 124, 145 Grassland, 34, 54–58, 75–80, 82, 83, 167, 169–171, 173, 177 Grave, 29, 36, 46, 135, 137, 157, 178, 216, 218, 221 Grazing, 2, 7, 9, 15, 21, 22, 27, 34, 35, 55–59, 61, 63–66, 75, 76, 82, 94–98, 112, 170–173, 175–177, 180, 182, 191 indicators, 93 land, 167–182 landscape, 170, 171, 182 management, 56, 57, 64 Gundersted heath, 174, 175, 180 H Hamburg, 8, 43–51 Hasenbekshorst, 46, 48, 50

227 Heath, 7, 8, 50, 167, 170–176, 180, 186, 191, 192, 198 Heathrow Terminal 5, 190, 191, 202 Helladic, 144, 145, 157, 159–162 Herkenbosch, 28 Hijken, 36, 37, 180, 181 Hillfort, 9, 54, 59, 60, 63–66 Hilversum, 28 Himmerland, 8, 167–182 Holne Moor, 195, 200 Holstein, 50 Homestead enclosures, 13, 17–19, 21–23 Homestead enclosures mapped, 18 Hordeum/triticum type, 34 Hordeum vulgare, 75, 80 House, 9, 13–19, 21, 23, 36–38, 51, 58, 64, 71, 73–76, 95, 97, 105, 116, 127, 172, 177–182, 192 Household refuse, 9, 31, 34 House structures, 36, 173, 177–178 Hut-circles, 89–98 Hydraulic engineering, 145, 160–163 Hydraulic technologies, 160 I In extenso, 22, 96 Infield, 15, 22 Intensification, 10, 16, 21, 54, 56, 88, 98, 137, 161, 162 Intensive, 2, 13, 54, 56, 64, 66, 76, 83, 87, 94, 98, 103, 107–109, 126, 136, 157, 180, 182, 191 Iron Age, 8, 29–32, 37, 38, 57, 59, 60, 63–66, 72, 89, 92, 93, 97, 98, 139, 167–169, 172, 174, 176–178, 193–203 Irrigation, 74, 75, 79, 109–111, 146, 147, 149, 155, 156, 158, 159, 161, 162 K Kalamianos, 101–112 Kin group, 107, 108 Kolisbos, 31 L Lake Navnsø, 170 Land division, 53–67, 119, 185, 193, 195–198 Land enclosure, 7, 185, 186, 190, 197, 198 Land holding, 53, 59, 65, 105, 119, 120, 125, 126 Land management, 3, 5, 131, 132, 137, 139 Land occupation, 135, 137, 138 Land-use, 2, 3, 7, 8, 18, 21–23, 30, 53, 59, 63, 64, 66, 82–84, 92, 95, 97, 98, 105, 116, 119, 126–127, 167–169, 171, 173, 175–177, 180 Layout, 2, 14, 17, 19, 21, 30, 38, 51, 61, 64–66, 105, 180, 188 Light Detection and Ranging (LIDAR), 2, 8, 30, 44, 103, 115–127 data, 2, 29, 43, 72, 118–120, 122, 123, 125, 126 surveys, 8, 117–119, 122 Liver fluke, 56 Livestock, 2, 5, 9, 14, 17, 21, 23, 29, 53–67, 72, 76, 82 husbandry, 54, 56, 58, 60, 66, 67 management, 56, 58, 59 production, 54, 55, 57, 67 Lockington Quarry, 194, 202 Long-term, 1, 3, 5, 7, 9, 10, 28–29, 38, 66, 71, 74, 81, 83, 84, 88, 95, 96, 116, 137, 169, 171, 179, 180, 191, 198, 209 Loughcrew, 5, 8, 115–127 Lowland, 2, 9, 28, 35, 59, 83, 89–93, 95, 96, 98, 105, 186, 194

228 Luminescence age, 72, 155, 156, 187, 190, 192 Lundby heath, 175 Lynchets, 2, 5, 9, 13, 14, 29, 36, 57, 90, 93, 94, 98, 120, 167, 168, 173–175, 177, 179, 180, 187, 189, 200 M Magnetic anomalies, 145, 146, 148, 152 Magnetic susceptibility, 73, 145, 152, 153 Magnetometry, 9, 145–148, 153, 154, 157 data, 148–149, 153, 154 results, 145, 147, 150, 152 Manuring, 3, 5, 7, 9, 34, 35, 38, 51, 56, 65, 74, 75, 82, 83, 94, 95, 108–110, 157, 168, 177, 180, 186 Mapping, 2, 3, 8–10, 13, 17, 18, 20–23, 30, 105, 119, 120, 168, 176 Marína, 147–149 Medieval, 31, 34, 43, 48, 58–60, 76, 81, 83, 89, 91, 92, 95, 118, 119, 122, 125, 126, 157, 174, 186, 188, 192, 193, 195–203 Micromorphological analysis, 21, 33 Middle Ages, 76, 79–82, 134, 195 Migration Period, 43 Mounds, 46, 93, 118, 173, 175 Mountain, 58, 71–84, 126, 144 Mount Fteliá, 146, 147 Mount Lebanon, 74, 82 Mount Vesuvius, 9 N Naples, 210, 213, 216–221 Neolithic, 8, 9, 13–23, 27, 28, 51, 59, 73, 75, 76, 79, 80, 83, 88, 92, 96, 115, 116, 118, 123, 127, 157, 162, 172–174, 201, 202, 220, 221 Neolithic/Bronze Age, 17, 21–23, 81, 83, 174 Nucleated settlement, 167, 182 Nutrients, 3, 35, 56–58, 95, 96, 110, 171, 173, 180 O Old Drift plateau, 43, 44, 48 Optically stimulated luminescence (OSL), 3, 9, 31, 104, 155, 186, 187, 190, 198 date, 3, 31, 32, 73, 155, 187, 190, 192 dating, 10, 73, 192 samples, 31, 32 Orchard crops, 109, 110 Ordnance Survey (OS), 87, 119, 125, 127 Organic matter, 73, 79, 82 Organisation, 1, 2, 5, 7, 9, 10, 54, 55, 60, 62, 65, 135, 182, 185, 198 Outfield, 15, 22, 58, 59, 177 P Palaeosol, 71–75, 77–82, 103 Palatial, 108, 110–112, 157 centre, 102 loans, 111, 112 Palma Campania, 220 Palynology, 3, 15, 16, 33, 75, 88, 93 Parliamentary field, 5, 189 Passage tombs, 116, 118 Pastoral, 2, 5, 7, 9, 16, 54, 57, 59, 61, 66, 76, 78, 82, 83, 97 activities, 73, 75, 76, 82 huts, 72, 76, 79, 80 sites, 76

Index Pastures, 8, 15, 22, 28, 51, 55, 56, 58, 67, 73–77, 79, 82, 88, 93–97, 116, 118, 161, 168, 180 Peat, 13, 15, 17, 95, 137, 169–171, 177, 180, 190, 191, 194–196 Peat deposits, 171, 177 Pedoarchaeology, 71–84 Pedosedimentary deposits, 72, 73, 77, 78 Pedosedimentary stratifications, 74 Permanent settlements, 76, 83, 88 Pinus uncinata, 75, 78 Pit, 89, 90, 92, 93, 95, 178, 179, 181, 190, 191, 195, 196, 217 Pit alignments, 189, 193, 195–196, 198, 199, 201, 203 Plant macrofossils, 9, 190 Plough, 3, 13, 48, 51, 102, 109, 147, 148, 171, 175, 176, 213, 218 Ploughing, 30, 33, 34, 48, 49, 56, 57, 63, 72, 90, 98, 109, 133, 146, 147, 168, 188, 213, 220 Polder, 9, 143–145, 147–149, 157–162 Polder dykes, 149, 150, 157, 160 Pollen, 15, 33–35, 72, 76, 80, 93, 137, 167, 169–171, 176, 177, 179, 220 data, 8, 168, 176, 182 diagrams, 96, 168–171, 173, 176, 177, 180, 182 Post-Roman, 59, 65, 186, 195–197 Pottery sherds, 80, 190–192 Pre-Roman Iron Age, 8, 50, 169–171, 173, 175–182 Pyrenees, 3, 5, 73, 74, 81 R Raatakkers, 27, 29, 32 Radiocarbon, 17, 31, 73, 92, 96, 98, 156, 161, 186, 187, 190–192, 194, 196 dates, 3, 15–17, 32, 75, 79–81, 87, 89, 90, 93, 97, 155, 156, 179, 187, 190–192, 194, 195, 198, 210 dating, 15, 17, 23, 31, 72, 186, 187, 190, 198 Radiometric dates, 157, 159, 190 Regi Lagni, 210, 216 Revetted canal, 146–149, 152–154, 157–159 Ribwort plantain, 15, 170 Ridge, 43, 46, 48, 76, 88, 105, 108, 119, 120, 172–175, 202 Ring forts, 138, 139 Ritual landscape, 118, 126, 127 River Clanis, 210, 216, 219, 220 Roman, 51, 59, 65, 80, 105, 160, 186, 190, 192, 193, 195–203 era, 27 field systems, 186, 192, 195, 197 Iron Age, 48, 173, 177–180 jug, 31 period, 1, 30, 71, 193–197 times, 72, 79, 80 Routes, 3, 125, 133, 137, 216 Rye, 35, 51, 64, 76, 80, 171 S Sachsenwald, 8, 43, 44, 48, 50 Sandy deposits, 169, 175 Sandy soils, 8, 29, 168, 169, 180 Satellite data, 146–148, 159 Scale, 1–10, 14, 15, 18, 19, 21, 23, 29, 36–38, 50, 54, 59, 61, 65, 66, 72, 73, 83, 98, 101, 104, 109, 120, 139, 149, 160, 171, 185, 186, 197, 220 field systems, 23 land allotment, 2 land division, 185, 193 landscape change, 10

Index

229

Seasonal, 21, 55, 56, 58, 59, 61, 64–66, 82, 83, 95, 102, 111, 112, 144, 147, 148, 154, 158, 161 Sebeto, 210, 211, 216–219 Secale cereale, 35, 75, 80 Secondary products, 54, 64 2nd millennium BC, 1, 3, 8, 54, 59, 61, 83, 87, 89, 90, 92, 95, 97, 216, 219, 221 Sedentary, 3, 8, 51, 95, 97 Sediment, 9, 30, 31, 72, 74, 92, 102–104, 145, 149, 152, 153, 155, 171, 177, 186, 190–192, 198 deposition, 171 traps, 92, 94 Seeds, 10, 33, 34, 55, 56, 61, 80, 94, 95, 97, 108, 109, 145, 159, 171, 190, 194 Settlement, 3, 5, 7–10, 13–15, 21–23, 27, 29, 33, 35, 36, 49, 55, 58–61, 63, 65, 66, 71–73, 78–80, 82–84, 87–98, 101–112, 116, 118, 119, 131–134, 137–139, 149, 158, 160–162, 167–169, 172–182, 191, 192, 194, 210, 212, 216, 217, 219–221 areas, 209, 220 history, 168, 171–175 pattern, 83, 161, 167–182, 218–221 sites, 14, 17, 29, 64, 116, 147, 157, 159, 173 traces, 152, 158, 169, 172–175, 177–179, 216 Severn Estuary, 87, 88, 95 Shaugh Moor, 57, 186, 189, 193, 195, 200 Sheep, 53–67, 73, 76, 95, 98, 170, 180, 196, 220 Shetland, 5, 9, 13–23, 90–94 Skånsø, 170 Skørbæk Heath, 173, 175 Social, 2, 3, 5, 7, 21, 38, 53, 54, 58, 65, 66, 81–84, 98, 101, 109, 131–133, 137, 145, 162, 163, 189, 221 elite, 136, 139 implications, 10 organisation, 7, 66, 134 status, 132, 134, 139 Soil, 3, 5, 9, 10, 15, 17, 18, 21, 29, 30, 32–35, 43, 48, 49, 51, 54–57, 59–61, 64–66, 71–83, 89, 90, 92, 94, 98, 101, 102, 104, 108, 110, 116, 120, 125, 132, 145, 146, 152, 155, 160, 162, 167–173, 175, 177, 179, 180, 186, 191, 195, 213, 220–222 horizons, 72, 73, 152, 191 improvement, 50, 82, 171, 173, 180 nutrients, 7, 171, 173 profiles, 72, 73, 75, 78, 152, 153 science, 9, 73–75, 81 types, 29, 30, 55, 64, 82 Someren, 30, 31, 34, 35 Somerset, 9, 53–67, 202 South Cadbury environs, 54, 59, 62, 63, 67 South Cadbury valley, 61, 64 South Houllan, 20, 23 Standing stones, 23, 91, 118, 123, 125 Stiri, 101–112 Stock, 55–59, 61, 64, 71, 73, 88, 193, 196, 199–202 handling features, 57–59, 64 proof, 57, 61 Store Binderup, 168, 169, 175–176, 180 Store Vildmose bog, 168, 171 Surveyed field system, 168, 174, 177

Tenure, 2, 7, 22, 38, 53, 55, 95, 125, 221 Tenurial, 5, 23, 131 Terrace, 1, 2, 5, 30, 51, 71–83, 101–112, 160, 162, 189, 203 construction, 102, 105, 111–112 farming, 108, 109 fields, 9, 10, 72, 73, 75, 77–82, 109, 112, 194 walls, 71, 72, 77–79, 103–108, 111 Territorial organization, 81–84 Territory, 53, 55, 73, 75, 82, 83, 96, 97, 161, 213, 216, 217, 219–221 Texts, 9, 60, 67, 82, 88, 103, 108, 110–112, 145, 159 Threshing floor, 103, 107, 108, 110 Traction animals, 101, 102, 108–110 Tradition, 7, 27, 31, 115, 159–161, 177 Trees, 96, 102, 105, 109, 147, 169, 170, 173, 176, 177, 180

T Tatihou, 132–137

Z Zeijen, 31, 33–36

U Upland, 9, 58, 87, 89, 90, 93, 95–97, 103, 186, 188, 191, 194, 197 Upstanding features, 116, 122, 124, 126, 127 Use-rights, 7, 38 V Vaassen, 29, 31, 35, 36 Vesuvius, 209, 210 Viking Age, 168, 173, 179, 180 Villages, 17, 20, 43, 75, 81–83, 101, 103, 138, 139, 148, 168, 173, 175, 176, 178–180, 188, 196, 201 Visual inspection, 120, 122, 124 Volcanic, 72, 73, 90, 209–222 Volcanic activity, 72 Vrystiká Sinkhole, 147–149, 159 W Wall, 15, 17–20, 22, 45, 55, 56, 59, 71–74, 77–80, 82, 83, 91, 94, 95, 97, 102–107, 111, 122, 124, 143, 153, 159–161, 172, 191 Water, 3, 7, 55–57, 61, 64, 65, 75, 90, 93, 102, 109–111, 143, 144, 146–149, 153, 156, 158, 159, 161, 162, 210, 216, 220 Watering hole, 190, 191 Waterlogged wood, 190, 191 Wekerom, 31, 33–37 Wells, 3, 7–10, 17, 18, 27, 33–36, 50, 51, 55–57, 59, 61, 64–66, 73, 74, 83, 88, 95, 96, 101–104, 107–110, 116, 118, 120, 123, 125, 131, 134, 144, 145, 152, 153, 157, 160–162, 171–173, 177–180, 187, 188, 193, 195, 197, 219 Westeinde, 31, 33–36 Western Ballinvally, 120 West Mainland, 13–18, 20, 21, 23, 91 West Yorkshire, 191, 192, 195, 196, 198, 199 Wetlands, 9, 33–35, 58, 60, 64, 91, 93, 143, 161, 169, 174, 175 Wheat, 31, 34, 35, 61, 63, 80, 108, 144 Wild resources, 17, 55 Wind erosion, 171, 177 Winter grazing, 55, 171, 180 Woodland, 8, 15, 16, 30, 57, 58, 75, 93, 169–171, 176 Written sources, 76, 173, 174