Investigating Animal Burials: Ritual, mundane and beyond 9781407308128, 9781407322070

Table of contents :
Front Cover
Title Page
Copyright
Preface and Acknowledgments
Table of Contents
List of Figures
List of Tables
Chapter 1. Introduction to a Deposit Type
Chapter 2. Creation of the Faunal Record
Chapter 3. Neolithic and Bronze Age
Chapter 4. Iron Age Southern England
Chapter 5. Iron Age Yorkshire
Chapter 6. Romano-British Southern England
Chapter 7. Romano-British Yorkshire
Chapter 8. Medieval Southern England
Chapter 9. Medieval Yorkshire
Chapter 10. Patterns and Trends
Chapter 11. Assigned Meaning
Chapter 12. Assigning Meaning; Animal Biographies
Chapter 13. Conclusions
Bibliography
Appendix 1: Sites with ABGs from Southern England
Appendix 2: Sites with ABGs from Yorkshire
Appendix 3: Sites from southern England with no ABGs reported
Appendix 4: Sites from Yorkshire with no ABGs reported

Citation preview

l na tio ne di nli ad l o ith ria W ate m

BAR 535 2011  MORRIS  INVESTIGATING ANIMAL BURIALS

Investigating Animal Burials Ritual, mundane and beyond

James Morris

BAR British Series 535 9 781407 308128

B A R

2011

Investigating Animal Burials Ritual, mundane and beyond

James Morris

BAR British Series 535 2011

ISBN 9781407308128 paperback ISBN 9781407322070 e-format DOI https://doi.org/10.30861/9781407308128 A catalogue record for this book is available from the British Library

BAR

PUBLISHING

Preface and Acknowledgments In recent years, zooarchaeology has started to move beyond the purely economic towards social interpretations. In particular, these ‘social’ interpretations have often concentrated upon complete or partial animal burials rather than the disarticulated and fragmented faunal remains more commonly recovered from archaeological sites. This book presents a study of animal burials, referred to as associated bone groups, from the Neolithic to late Medieval periods of southern England and Yorkshire. Not only does it present data on over 2000 deposits, it also discusses their interpretation, arguing that most are based on generalised period-based assumptions. It is proposed that a biographical approach would allow the investigation of the specific above-ground actions behind the deposits creation, moving away from generalisations towards individual interpretations. The study shows the value of not only utilising specialist data, but integrating such knowledge with other archaeological evidence and theoretical approaches. This book represents the publication of the author’s PhD thesis. It has been edited, tweaked and in places expanded, however I have resisted the temptation to rewrite the whole volume. This book therefore maintains the structure of the original thesis, with some of its data heavy chapters. For this I remain unapologetic as I feel it is important to, in effect, show your workings. The final chapters of this book may be concerned with theoretical interpretive arguments, but these are built upon and informed by the more classically ‘processual’ data analysis conducted in the period specific sections. It is hoped that through this book the reader will appreciate not only the interesting and variable nature of these faunal remains, but the contribution they can make to the study of past societies. For the aims of zooarchaeology and archaeology are the same, ‘to shed light on humanity’s past’, the zoo- part only designates the material we use to do this. Finally, this volume could be seen as part of an increasing trend towards social zooarchaeology. Although steeped in processual methodological tradition this should not and, I hope this volume shows, does not limit our engagement with archaeological theory. Book Structure This book is divided into three main sections. The first chapter outlines the previous work regarding animal burials, the nature of this study and its methodologies. In Chapter 2 the underlying taphonomic processes in the creation of animal burials are discussed. In the second section the dataset is examined for each period and region. It starts with Chapter 3, in which the Neolithic and Bronze Age data for both southern England and Yorkshire are examined. The regional datasets are discussed in separate chapters for the subsequent periods

due to their larger size. Hence Chapter 4 deals with the Iron Age data for southern England and Chapter 5 the Iron Age data from Yorkshire. This pattern continues until the last chapter in this section, Chapter 9, which concerns the Medieval dataset from Yorkshire. At the beginning of each chapter the general background to the period is laid out. However, it is not the aim of this publication to discuss the general nature of each period in depth. The structuring of the discussions does differ between the chapters, dependant upon the quantity and quality of the data available. The third section concerns interpretation. Chapter 10 discusses any biasing effects within the datasets, and the overall inter-period trends encountered. Chapter 11 deals with the interpretations given to and their associated problems. Chapter 12 suggests a new way of investigating and interpreting these deposits and shows that such a method can bring us closer to human actions and meanings. The final chapter summarises the results, offers recommendations and directs us towards future research in this area. The appendix at the back of this book consists of four tables. The first and second tables provide a summary of the sites with associated bone groups present from southern England and Yorkshire respectively. The third and forth tables contain lists of sites with faunal remains but no associated bone groups reported as present. The rest of the data utilised for this project, including butchery, taphonomy, detailed body area information, pathology and associations can be found on the accompanying CD. This contains all the data collected during the project in a Microsoft Access database file and a series of Microsoft Excel files (please see note on pg ii). Sections in Chapters 1,10, 11 and 12 reproduce or contain text that is similar to that subsequently published elsewhere (Morris 2008; 2010a; 2010b; in press-a; in press-b; Morris and Jervis 2011; Serjeantson and Morris 2011) but has been included here to maintain the study’s coherence. As the book maintains a ‘PhD’ structure, readers with a general interest in animal burials are directed to chapters 1, 10, 11 and 12 first. The more detailed period specific chapters are suitable for those requiring detailed knowledge of deposits within a specific timeframe. Acknowledgments It would not have been possible to complete a project of this nature without the help and support of colleagues, peers and family. First and foremost my eternal gratitude must go to my supervisors, Mark Maltby and Ellen Hambleton. They have at all times shared their Yoda like wisdom and Doc Brown-esque encouragement.

Investigating Animal Burials; Ritual, Mundane and Beyond I am most appreciative to my examiners Michael Fulford (Reading University) and Miles Russell (Bournemouth University) for making my viva an enjoyable process and also for the advice, help and support they after shown afterwards.

I will be forever grateful for the help and support of Justine Biddle, especially her efforts in the ‘battle of the H’s’ and subsequent help producing this book. Finally, I must acknowledge my family, for the moral and financial support without which this project would never have been completed. The fact they do not understand what I do makes their support even more generous.

Many people have offered comments on specific aspects or themes covered, often at conference sessions, and I must acknowledge in particular; Adrian Chadwick, Alex Pluskowski, Alice Choyke, Andy Tullett, Ben Jervis, Claire Randall, Dale Serjeantson, David Orton, Howard Williams, Pam Crabtree, J D Hill, Jacqui Mulville, Jessica Grimm, Jody Joy, Julie Hamilton, Kate Waddington, Kate Welham, Krish Seetah, Mark Dover, Martin Pitts, Mick Monk, Mike Lally, Naomi Sykes, Paul Buckland, Rachel Pope, Richard Madgwick, Sheila Hamilton-Dyer, Sue Stallibrass, Terry O’Connor, Tim Darvill and Umberto Albarella.

Please note that the CD referred to on page i has now been replaced with a download available at www.barpublishing.com/additional-downloads.html

Polydora Baker, Andy Hammon and Fay Worley from English Heritage provided wonderful assistance with accessing and making sense of the Ancient Monument Laboratory reports. Penny Dale, and Janet Coles of Bournemouth University library and Robin Nove (Yorkshire Archaeological Society) provided much assistance with finding particular reports. For general help and understanding with all the little things, I must acknowledge the help and support of the archaeology community at Bournemouth University, in particular Bronwen Russell, Clare Randall, Ehren Milner, Linda Osborne, Louise Pearson and Paul Cheetham. I must also make a special mention of my PhD colleague Niels Brouwers. A number of the figures in this book have been reproduced with the kind permission of individuals and societies, my grateful thanks to; Antiquity for the reproduction of head and hoof burial illustrations (Figure 1.2). Barry Cunliffe and the Council of British Archaeology allowed photos and illustrations from the Danebury and Danebury Environs Projects to be used, (Figure 1.2 and 12.4). The Hampshire Field Club and Archaeology Society for the Oakridge Well illustration (Figure 2.4). The East Riding Archaeology Society for the plan of Whitegrounds entrance grave (Figure 3.1). The Society of Antiquities for the Fussell’s Lodge Long Barrow plate (Figure 3.4). The Prehistoric Society for the plan of Hemp Knoll (Figure 3.6). Dorset Natural History and Archaeological Society for the reproduction of plates from Crab Farm (Figure 3.8) and Greyhound Yard (Figure 6.33). English Heritage for the reproduction of the plans from Gussage All Saints (Figure 4.8) and the Garton Station illustrations (Figure 5.3 and 12.6). Martin Millett and the Yorkshire Archaeology Society for the reproduction of the Hasholme Logboat plan (Figure 5.2) and the site plan of Shiptonthorpe (Figure 7.6).

ii

Contents Preface and Acknowledgments Contents List of Figures List of Tables

i iii xiii xix

Part 1: Introducing Associated Bone Groups 1. 1.1. 1.2.

1.3. 1.4. 1.5.

2. 2.1. 2.2. 2.3. 2.4. 2.5.

2.6. 2.7. 2.8.

Introduction to a Deposit Type Introduction ABGs and zooarchaeology 1.2.1. Zoologicoarchaeologists 1.2.2. Palaeoeconomists 1.2.3. Bucking the trend; Neolithic animal burials 1.2.4. Bucking the trend; Romano-Celtic religion 1.2.5. Structured deposition 1.2.6. Iron Age influence; ‘special animal deposits’ 1.2.7. Iron Age influence; sceptical zooarchaeologists 1.2.8. Iron Age influence; a menagerie of deposits 1.2.9. Social zooarchaeologists Research aims Defining ABGs Achieving the aims; methodology 1.5.1. Data sources 1.5.2. Data recording and categories 1.5.3. Data analysis

1 1 1 2 2 3 4 5 7 9 10 11 12 13 13 13 15

Creation of the Faunal Record Introduction Death Death to ground; skeletonisation Death to ground; disarticulation Death to ground; butchery 2.5.1. Primary butchery 2.5.2. Secondary butchery 2.5.3. Tertiary and further butchery Ground to trowel Trowel to report Summary

16 16 17 17 20 21 21 22 22 24 24

Part 2: Period Specific Results 3. 3.1. 3.2. 3.3. 3.4. 3.5. 3.6. 3.7.

3.8.

Neolithic and Bronze Age Introduction Neolithic and Bronze Age ABGs from Yorkshire Neolithic and Bronze Age species proportions Neolithic site types Neolithic ABG composition Neolithic - Bronze Age transition Bronze Age ABGs 3.7.1. Round barrows 3.7.2. Wilsford shaft 3.7.3. Bronze Age settlements, enclosures and middens Summary

25 25 26 28 28 31 32 32 35 37 39

4. 4.1. 4.2.

Iron Age southern England Introduction Species proportions

40 40

iii

Investigating Animal Burials; Ritual, Mundane and Beyond 4.2.1. Wild species 4.2.2. Domestic animals Nature of the assemblage and influential sites Site types and context ABG composition Butchery ABG associations 4.7.1. Multiple ABGs 4.7.2. Associated deposit groups Summary

40 42 43 45 48 53 55 55 55 57

Iron Age Yorkshire Introduction Species proportions and context Hasholme logboat Funerary sites Settlements 5.5.1. Ferrybridge and Parlington Hollins 5.5.2. Dalton Parlours 5.5.3. Garton and Wetwang Slack Summary

59 59 61 61 62 64 64 64 65

Romano-British southern England Introduction Species proportions 6.2.1. Wild species 6.2.2. Domestic species 6.2.3. Greyhound Yard and the middle Romano-British assemblage Nature of the assemblage Owslebury 6.4.1. Owslebury ABG species proportions 6.4.2. Owslebury ABG placement 6.4.3. Owslebury ABG ageing data 6.4.4. Owslebury ABG composition Oakridge well Greyhound Yard, Dorchester 6.6.1. Greyhound Yard ABG species proportions 6.6.2. Greyhound Yard ageing data 6.6.3. Greyhound Yard ABG composition ABG site type differences 6.7.1. Town and rural settlements 6.7.2. Military and villa sites 6.7.3. Funerary sites Composition and butchery data ABG associations Summary

66 66 66 67 69 69 71 71 73 75 75 78 81 82 83 85 86 86 90 91 92 94 97

7.4. 7.5. 7.6. 7.7. 7.8.

Romano-British Yorkshire Introduction Species proportions and context Shiptonthorpe 7.3.1. Shiptonthorpe species proportion and assemblage composition 7.3.2. Shiptonthorpe ABG placement Rudston Roman villa Trentholme Drive, York Bainesse Farm, Catterick Overall ABG composition Summary

99 99 101 102 103 105 106 106 107 108

8. 8.1.

Medieval southern England Introduction

109

4.3. 4.4. 4.5. 4.6. 4.7. 4.8 5. 5.1. 5.2. 5.3. 5.4. 5.5.

5.6. 6. 6.1. 6.2.

6.3. 6.4.

6.5. 6.6.

6.7.

6.8. 6.9. 6.10. 7. 7.1. 7.2. 7.3.

iv

Contents 8.2.

8.3. 8.4. 8.5. 8.6. 8.7. 8.8. 8.8. 9. 9.1. 9.2. 9.3. 9.4. 9.5.

9.6.

Species proportions 8.2.1. Birds 8.2.2. Wild mammals 8.2.3. Domestic mammals Nature of the assemblage Site type Feature type ABG Composition Butchery Associated ABGs Summary

109 109 110 111 111 112 114 115 116 117 118

Medieval Yorkshire Introduction Species proportions and site type Anglo-Scandinavian High Medieval Late Medieval 9.5.1. Manorial 9.5.2. Monastic 9.5.3. Towns Summary

120 120 121 122 122 122 123 123 124

Part 3: Trends, Interpretations and Conclusions 10. 10.1. 10.2.

Patterns and Trends Introduction Are the patterns real? 10.2.1. Sample size and ABGs 10.2.2. Taphonomy 10.2.3. Archaeologists How common are ABGs? Changing species 10.4.1. Southern England 10.4.2. Yorkshire Composition; not all ABGs are the same 10.5.1. Complete ABGs 10.5.2. Partial ABGs Butchery Animal health; pathology Depositional context Human remains Summary

125 125 125 125 127 128 129 130 132 133 134 136 139 144 145 147 148

11.5. 11.6. 11.7. 11.8.

Assigned Meaning Introduction A pantheon of interpretations Structured deposition Domestic ‘food’ mammals 11.4.1. Complete ABGs; unfit for consumption, or sacrifices? 11.4.2. Partial ABGs; butchery waste or ritual rubbish? Domestic dogs and cats; population control or sacred animals Domestic birds; offerings and waste Wild species; accidents, natural or special deposits? Meta-Level categories and the problem with ‘ritual’

149 149 151 153 153 155 159 162 163 165

12. 12.1. 12.2. 12.3.

Assigning Meaning; Animal Biographies A way forward Building ABG biographies The taphonomic model

167 167 168

10.3. 10.4. 10.5. 10.6. 10.7. 10.8. 10.9. 10.10. 11. 11.1. 11.2. 11.3. 11.4.

v

Investigating Animal Burials; Ritual, Mundane and Beyond 12.4. 12.5. 12.6.

12.7. 13. 13.1. 13.2. 13.3. 13.4. 13.5.

13.6.

How humans create ABGs Assigned meaning Looking again at ABGs 12.6.1. Windmill Hill outer ditch section V 12.6.2. Suddern farm pit 197/7 12.6.3 Kirkburn burial K5 12.6.4. Rudston Roman villa pit 47 12.6.5 Coppergate, York 12.6.6. Winnall Down pit 6596 revisited, again! What are ABGs?

169 172 172 172 174 176 177 178 178 180

Conclusion Associated Bone Groups; their nature Associated Bone Groups; their meaning Developing methodologies Recording recommendations Future directions 13.5.1. Availability of data 13.5.2. Regionality 13.5.3. Associated deposit groups 13.5.4. Intra-site study 13.5.5. Metrical analysis 13.5.6. Post-depositional taphonomic action 13.5.7. Animal behaviour 13.5.8. Moments of transition Concluding remarks

181 181 182 182 183 183 184 184 184 184 185 185 185 185

Bibliography

186

Appendix Appendix 1 Sites with ABGs from southern England Appendix 2 Sites with ABGs from Yorkshire Appendix 3 Sites from southern England with no ABGs reported Appendix 4 Sites form Yorkshire with no ABGs reported

221 226 228 236

Please note that the CD referred to throughout the text has now been replaced with a download available at www.barpublishing.com/additional-downloads.html

vi

List of Figures 1.1 1.2 1.3 1.4 1.5

Postcard produced 1937 during the Maiden Castle excavations Illustrations of head and hoof burials Partial articulated cattle skeleton in association with chalk blocks, Danebury Herarchy of animals indicated by their presence in culinary garbage Graph showing the percentage of archaeological prehistoric publications regarding ‘ritual’

2 3 6 8 10

2.1 2.2 2.3 2.4

Statistical scheme of disarticulation for an ideal skeleton of domestic cow Diagram illustrating the elements expected to be articulated after natural disarticulation Dismemberment practices of different ethnic groups Section of the top 4 meters of the Oakridge well

19 19 23 24

3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8

Plan of the burials present in the Whitegrounds entrance grave Percentages of cattle, sheep/goat and pig found on selected Neolithic and Bronze Age sites Percentages of body parts present, from 16 partial cattle from Windmill Hill Cattle ABGs from within Fussell’s Lodge long barrow Plan of the pond barrow at Down Farm Plan of the central grave at Hemp Knoll The Wilsford Shaft indicating approximate locations of ABGs Cow ABG pregnant with calf from Crab

26 28 30 31 33 34 36 37

4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 4.10 4.11 4.12 4.13

Percentages of the most common ABG species ABG species percentages from Balksbury Camp Species percentages for the largest middle Iron Age ABG samples from southern England ABG species percentages from Danebury Percentage of southern England ABGs per species deposited on hillforts and non-hillfort sites Correspondence analysis showing relationships between species and feature types Correspondence analysis showing relationships between species and feature types, excluding pits Plan of cow and calf ABGs from Gussage All Saints Percentage of complete and partial Iron Age sheep/goat ABGs by Age Percentage of complete ABGs per species for the early, middle and late Iron Age Percentage of complete and partial Iron Age pig ABGs by age Percentage of complete and partial Iron Age dog ABGs by age Percentages of body areas represented from partial cattle, horse, sheep/goat, pig and dog ABGs

41 44 44 44 45 47 47 49 50 51 51 52 52

5.1 5.2 5.3

Tripolar diagram showing the percentage of cattle, sheep/goat and pig for non-funerary sites Plan of the Hasholme Logboat showing overall bone distribution Garton Station grave-group, GS6

60 62 63

6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 6.10 6.11 6.12 6.13 6.14

Percentages of the most common ABG animals in the Romano-British periods Percentages of the most common ABG animals, excluding dogs, in the Romano-British periods Percentage of ABGs recorded from middle Romano-British context, excluding Grayhound Yard Percentages of the main ABG species from Owselbury for the Romano-British sub-periods Percentages of the main species NISP for the overall faunal assemblage from Owslebury Percentage of each species per feature type from Owslebury. Percentage of each species per feature type for the overall faunal assemblage from Owslebury Attrition of species population, for individuals that became ABGs, at Owslebury Attrition of species population, using data from the whole Owslebury faunal assemblage Attrition of the ABG and non-ABG pig population from Owslebury Percentage representation of body areas for partial dog ABGs from Romano-British Owslebury Percentage representation of body areas for partial cattle ABGs from Romano-British Owslebury Percentage representation of body areas for partial horse ABGs from Romano-British Owslebury Percentage representation of body areas for partial sheep/goat ABGs from Romano-British Owslebury Percentage representation of body areas for partial sheep/goat ABGs from Iron Age Owslebury Section of Oakridge well with relative position of ABGs indicated Percentages of the main species from the late Romano-British Oakridge Well deposits Percentage NISP for the main species from the total faunal assemblage Oakridge Well Attrition of species population, for individuals that became ABGs, at Oakridge Well Percentages of the main species from Greyhound Yard for the Romano-British sub-periods

68 68 68 72 72 72 73 74 74 74 76 76 77

6.15 6.16 6.17 6.18 6.19 6.20

77 77 79 80 80 80 82

Investigating Animal Burials; Ritual, Mundane and Beyond

6.21 6.22 6.23 6.24 6.25 6.26 6.27 6.28 6.29 6.30 6.31 6.32 6.33

Percentages of the main species overall faunal assemblage NISPs from Greyhound Yard Attrition of species population, for individuals that became ABGs, at Greyhound Yard Percentage representation of body areas for partial dog ABGs from Greyhound Yard Percentage representation of body areas for partial sheep/goat ABGs from Greyhound Yard Percentage representation of body areas for partial cattle ABGs from Greyhound Yard Percentage of species represented in the ABG assemblages from rural settlements and towns Proportion of ABGs per species from early Romano-British rural settlements Proportion of ABGs per species from late Romano-British rural settlements Proportion of ABGs per species from late Romano-British towns ABG species proportions for late Iron Age and early Romano-British rural settlements ABG species proportions from Porchester Castle Body area percentage for Romano-British partial ABGs Associated dog ABGs, allegedly tied together at the throat from Greyhound Yard

82 83 84 84 85 87 88 88 88 89 90 92 94

7.1 7.2 7.3

100 101

7.5 7.6 7.7

Percentages of the most common animals to constitute ABGs, Romano-British Yorkshire Triplot of NISP counts for cattle, sheep/goat and pig for Romano-British sites from Yorkshire Percentages of the most common domestic mammal ABGs from middle and late Romano-British rural settlements, Yorkshire Percentages of the most common domestic mammals ABGs from late Romano-British town and villa sites, Yorkshire Percentages of ABGs per species from Shiptonthorpe Site plan for phase 4 (middle Romano-British) Shiptonthorpe Percentage of ABGs per species from Bainesse Farm, Catterick

8.1 8.2 8.3

Percentages of the most common animals to constitute Medieval ABGs, southern England Species percentages of ABGs from rural and urban Anglo-Saxon sites, southern England Body area percentages for Medieval partial ABGs, southern England

112 114 117

10.1 10.2 10.3 10.4 10.5 10.6 10.7 10.8 10.9 10.10 10.11 10.12 10.13

Total NISP plotted against number of ABGs present for each site with available data Total NISP plotted against number of ABGs present for sites with less than 60 ABGs Percentage of sites per region and period with ABGs present Total percentage NISP for the most common species per period from southern England sites Total percentages of ABGs from each period for southern England Total percentage NISP for the most common species per period from Yorkshire sites Total percentage of ABGs from each period for Yorkshire Mortality profiles of complete ABGs per species Mortality profiles of partial ABG per species Percentage of complete ABGs present for the main species per period Triplot of the body area proportions for partial dog ABGs per period Triplot of the body area proportions for partial cattle ABGs per period Triplot of the body area proportions for partial sheep/goat ABGs per period

126 126 129 131 131 133 133 135 135 135 137 137 138

11.1 11.2 11.3

Percentages of different interpretation categories by period Percentages of different interpretations given to ABGs from Iron Age sites per decade Percentages of different interpretations given to ABGs from Romano-British sites per decade

150 151 151

12.1 12.2 12.3 12.4 12.5 12.6 12.7

General model concerning the creation of ABGs Taphonomic model of the creation of ABGs from primary butchery Taphonomic model of the creation of ABGs from secondary butchery Plan of layer 7 within pit 197 at Suddern Farm Proportion of cattle ABG and non-ABG elements from Suddern Farm pit 197 Plan of burial K5 from Kirkburn Model of the changes and process behind the creation of the partial cat ABG from pit fill 6473

169 170 171 174 176 177 178

7.4

viii

101 102 103 104 107

List of Tables 1.1

Age stages for the main domestic mammals

15

2.1

Rank order of joint disarticulation in cattle, sheep/goat and donkey

18

3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8

Summary of the number of sites with ABGs for the Neolithic and Bronze Age Number and percentage of ABGs per species for the Neolithic and Bronze Age period Number of ABGs per species recorded on different Neolithic site types Total number of Neolithic to Bronze Age site types with no ABGs present Formation of partial ABGs by species and site type for the Neolithic period Number of ABGs per species recorded from Late Neolithic-early Bronze Age site types Number of ABGs per species recorded from different Bronze Age site types Formation of partial ABGs by species and site type for the Bronze Age

25 26 29 29 29 32 32 37

4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 4.10 4.11

Number of ABGs per species for the southern England Iron Age Number of ABGs found on southern England Iron Age sites The seven largest Iron Age ABG assemblages Number of complete, partial or unknown ABGs per species Number of ABGs with butchery marks present Number of butchery marks recorded per species and body Number of instances of inclusion in multi-ABG deposits per species Number of ABGs in multi-ABG deposits Number of species represented in multi-ABG deposits Multi-ADG deposits from southern England Iron Age sites Number of species associated with other material deposits

41 43 43 49 54 54 54 56 56 56 56

5.1 5.2 5.3 5.4 5.5

Number of ABGs per species for the Yorkshire Iron Age Number of ABGs recorded from different site types per species Number of ABGs found on different sites in Yorkshire Composition of partial pig ABGs from square barrow sites Sites with faunal assemblages but no ABGs present

60 60 61 63 64

6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 6.10

67 70 70 70 70 76 81 85 90

6.12 6.13 6.14 6.15 6.16 6.17 6.18

Number of ABGs per species for the Romano-British period The five largest Romano-British ABG assemblages Number of sites with ABGs present Total number of ABGs recorded for each site type Number of sites recorded with no ABGs present Composition of Romano-British ABGs from Owslebury Composition of Romano-British ABGs from Oakridge Well Composition of Romano-British ABGs from Greyhound Yard Number and proportion of ABGs per species from villa sites Number of complete and partial ABGs per species from late Iron Age and early Romano-British funerary contexts Number of complete and partial ABGs per species from middle and late Romano-British funerary contexts Number and composition of ABGs with butchery marks present Number of butchery marks recorded per species and body area Type of butchery marks observed on ABGs in the Romano-British period Number of multi-ABG deposits per period/site type and feature Species inclusion in multi-ABG deposits per site type for the Romano-British periods Number of ABGs within multi-ABG deposits Number of different species present multi-ABG deposits

7.1 7.2 7.3 7.4 7.5 7.6 7.7

Number of ABGs per species for the Romano-British period, Yorkshire Number of ABGs found on different types of Romano-British site in Yorkshire Number of complete and partial ABGs per period and species from Shiptonthorpe Number of ABGs per species and period from Rudston Roman villa Numbers of complete or partial ABGs per species from Romano-British Yorkshire Body area information per species, for partial Romano-British ABGs from Yorkshire Butchery marks recorded on Romano-British Yorkshire ABGs

100 102 103 105 107 107 108

6.11

91 91 93 93 93 95 96 97 97

Investigating Animal Burials; Ritual, Mundane and Beyond

8.1 8.2 8.3 8.4 8.5 8.6 8.7 8.8 8.9

Number of ABGs per species from southern England Medieval sites Number of medieval sites from southern England with ABGs Total numbers of Medieval site types from southern England recorded by period Number of ABGs per species and site type for the Anglo-Saxon period, southern England Number of ABGs per species and site type for the high and late Medieval periods Number of ABGs per species and feature type for the Medieval period, southern England Composition of Medieval ABGs, southern England Butchery information from Medieval ABGs, southern England ABGs per species deposited in multiple ABG deposits from Medieval sites, southern England

110 113 113 113 114 116 116 118 118

9.1 9.2 9.3

Number of ABGs from Medieval sub-periods, Yorkshire Number of Medieval site types with ABGs from Yorkshire Number of ABGs per species recorded from late Medieval site types, Yorkshire

121 121 123

10.1 10.2 10.3 10.4 10.5 10.6 10.7 10.8 10.9 10.10 10.11 10.12 10.13 10.14 10.15

Number of ABGs reported per period for each decade when data were collected The percentage of sites published by decade for each time period Percentages of complete ABGs reported in each decade Number of ABGs recorded per region and period Percentage of complete, partial and unknown ABGs for the total assemblage Species body area proportions for non-complete ABGs for all periods and regions Description of butchery marks recorded from cattle ABGs Description of butchery marks recorded from horse ABGs Description of butchery marks recorded from sheep/goat ABGs Description of butchery marks recorded from pig ABGs Description of butchery marks recorded from dog ABGs Summary of the pathologies recorded on ABGs per species and period Summary of the number of ABGs per feature type per period Number of ABGs per species recovered from grave contexts Number of ABGs per species recorded in association with human remains from feature other than graves

128 128 128 129 134 136 140 141 142 143 143 145 146 147

11.1 11.2 11.3 11.4 11.5 11.6 11.7 11.8 11.9 11.10 11.11 11.12 11.13

Summary of how ABGs have been interpreted by reporting authors by period Number of pits and ABGs from Iron Age and Romano-British sites Numbers of ABGs per period interpreted as animals that died of disease Numbers of ABGs per period interpreted as natural deaths Ritual interpretations of complete ABGs of cattle, sheep/goat and pig, per period Total number of domestic mammal ABGs per period interpreted as ‘waste’ Total number of domestic mammal partial ABGs per period interpreted as ritual Number of dog and cat ABGs per period interpreted as the result of culling Number of dog and cat ABGs per period interpreted as natural mortality Total number of dog and ABGs per period interpreted as ritual depositions Total number of domestic birds per period and their interpretation Number of wild mammals per interpretation and period Number of wild birds per interpretation and period

150 152 153 153 154 156 158 160 160 160 163 163 164

12.1

Summary information of ABGs from Suddern Farm Pit 197/7

175

x

147

Chapter 1. Introduction to a Deposit Type 1.1

Introduction

In the last couple of decades zooarchaeology has undergone important growth and development. It has begun to expand from what can be perceived as purely processual explanations of sociocultural change, which were often dependant upon economic and ecological factors, and has started engaging more with the dialogue of post-processual paradigms. Most archaeologists are now aware that the study of faunal remains can offer much more than answering the classic question of ‘so what did they eat?’ The anthropological work of LéviStrauss (1962, 127-8) with the much used quote, ‘natural species are chosen, not because they are ‘good to eat’ but because they are ‘good to think’, has now become accepted within archaeology. However, Levi-Strauss’s work was concerned with totemism and, although offering a useful new perspective, underemphasised the economic factors in the relationship between animals and humans. Recently Gilhus (2006, 4) has pointed out that animals are also ‘good to feel’, in that they give emotional value and impetus to anything they are linked with. However, this may be drawing too much on our modern views of ethical issues concerning the treatment of animals. Douglas (1990, 33) points out how we think about animals relating to one another on the basis of our own relationships. Therefore human social categories are extended in to the animal world. The challenge is to utilise such postulations when analysing a faunal dataset. The majority of faunal assemblages are fragmented and biased by taphonomic factors. Therefore the interpretation of faunal remains by what could be called ‘social zooarchaeology’, has concentrated on specific types of deposit. Faunal material recovered from archaeological sites is normally found in a state of disarticulation and fragmentation. Occasionally remains of an individual animal are found still in articulation, or close association. These types of deposits have long been noted in the archaeological record and have been subject to a number of descriptions, often heavily loaded with interpretation. Examples include ‘animal burials’ (Wheeler 1943, 115), ‘butchery waste’ (Maltby 1985a), ‘culled deposit’ (Maltby 1981a), ‘fall victim’ (Maltby 1993), ‘feasting waste’(Armour-Chelu 1991), ‘sacrificial offerings’ (Ross 1968) and ‘special animal deposit’ (Grant 1984a, 533; Wait 1985, 122). One of the most influential pieces of work on the subject was Grant’s (1984a) study on the faunal material from the Iron Age hillfort of Danebury, Hampshire. A large number of articulated animal skeletons were encountered during the excavation. Grant (1984a) labelled these as ‘special animal deposits’ and argued they resulted from a distinct type of ritual activity. Grant’s work has been discussed and built upon by a number of authors, and has become the dominant interpretation for such deposits, with the majority of ‘special animal deposits’ from

prehistoric sites now being interpreted as forming part of a society’s ritual framework. This form of interpretation has drawn from and influenced the rise of post-processual theoretical debate within archaeology, which, contrary to Hawkes’ (1954) ladder of inference, considers that understanding a society’s ritual/religious nature is as feasible as understanding its economic nature. Indeed, they are closely interlinked. Following on from Grant (1984a; 1991) was Hill’s (1995) work on the nature of possible ‘special’ deposits within Iron Age pits from sites in Wessex. Hill (1995, 27), in order to be more objective in his analysis of these deposits, utilised the term Articulated or Associated Animal Bone Group (ABG). This countered the problem of using Grant’s term of ‘special animal deposit’. It removed the inherent assumption that the deposit is of a ‘special’ or ‘ritual nature’. Throughout the term ABG has been utilised for the same reason. 1.2

ABGs and zooarchaeology

This section considers previous influential studies and discusses how ABGs were examined and interpreted, against the background of zooarchaeology’s development as a discipline. 1.2.1

Zoologicoarchaeologists

The discipline of zooarchaeology has a long and varied history, drawing information and skills from many different fields. Zooarchaeology grew out of zoologists’ interests in the history of, and changes in, animal populations. The first reference to it was by John Lubbock, Lord Avebury (Lubbock 1865, 169), who used the term zoologicoarchaeologist when referring to the work of Rütimeyer (1862) on the animal bone assemblages from Neolithic lakeside dwellings in Switzerland. Although zoologists such as Dawkins and Jackson (Dawkins and Jackson 1917; Jackson 1925; 1948a) did build on Rütimeyer’s work, animal bones from excavations were often ignored. When they were examined, the results were not always published, such as the work of D .M .S Watson on the faunal material from the causewayed enclosure at Windmill Hill, Wiltshire (Grigson 1999). Often these reports were concerned solely with species identification and metrical analysis. Work on the Glastonbury Lake village assemblage from Bulleid and Gray’s excavations has indicated that the retrieval strategies of the time concentrated on the retention of whole bones (J Morris 2000). The value of analysing a fragmentary faunal assemblage had not yet been realised and efforts of the founding zooarchaeologists concentrated upon biometrics, for which whole bones are required.

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 1.1 Postcard produced 1937 during the Maiden Castle excavations With the emergence of the ‘Palaeoeconomy’ School at Cambridge under the leadership of Eric Higgs, zooarchaeology began to develop at a greater pace in the 1970’s heyday of processual archaeology (Milner and Fuller 1999). Concentrating on the areas highlighted earlier by Chaplin (1965), mainly concerned with the procurement and consumption of food, as well as the growing field of taphonomic studies, the wider archaeological community began to realise the value of zooarchaeology in the study of subsistence economies.

During these early years in the development of zooarchaeology, ABG deposits were described but rarely discussed. Pitt-Rivers (1888, 198) was one of the first archaeologists to note such deposit types. As zooarchaeology developed, through the work of individuals such as J. W. Jackson, more ABGs were commented upon. Interpretations were commonly what could be considered functional. Articulated limbs were interpreted as joints of meat gone bad (Jackson 1943) or waste (J. W Jackson 1948a). Complete skeletons of domestic animals such as sheep were interpreted as fall victims (Jackson 1925).

The majority of work involving ABGs at the time reflects this mindset. For example, a complete ABG of a red deer and twelve associated foxes were interpreted as the result of pit falls (Jones 1976b; 1977). Other deposits were commonly interpreted as the result of butchery waste (Griffith 1976; Harcourt 1979b), or not commented upon at all (Grant 1975; King 1970). During the 1970’s there appears to have been no ‘ritual’ interpretations offered for ABGs by zooarchaeologists (see 11.2). This corresponds to the general trend that archaeologists of the 1960’s and 1970’s were reluctant to investigate the role of ritual and religion (Renfrew 1994). The ‘palaeoeconomic’ school at Cambridge took a hard line in stating that ‘the soul leaves no skeleton’ (Higgs and Jarman 1975). ‘Ritual’ was viewed by many as unimportant in the past and not worth studying.

However ‘ritual’ interpretations were occasionally suggested, the majority of the time for complete skeletons (Wheeler 1943, 98) (Figure 1.1), complete skull deposits (Fox and Wolseley 1928), or ABGs in association with human remains (Collins 1953). 1.2.2

Palaeoeconomists

Despite the examples mentioned above, studies of animal bone were still a rare occurrence prior to 1970. Chaplin (1965, 204), writing in ‘Antiquity’, called for animal remains to be subject to the same studies as other archaeological material; ‘domestic animal remains of no less stratigraphical significance than pottery vessels are surely deserving of equal study along with the more familiar artefacts of the archaeologist such as the brooches, rim sherds and tool types.’

1.2.3 Bucking the trend; Neolithic animal burials Although the above approach could be considered the majority view within British zooarchaeology at the time, interpretations by Piggott (1962a), Alcock (1966; 1970) and Ross (1967; 1968; Ross and Feachem 1976) of Neolithic, Iron Age and Romano-British assemblages were drawing conclusions that would influence future interpretations of ABGs.

Chaplin pointed out the important contribution animal remains can make to the understanding of dietary and economic factors. However, possible contributions to understanding social factors were not discussed.

2

Chapter 1. Introduction to a Deposit Type In association with human remains:

Piggott’s (1962a) synopsis of Neolithic ‘head and hoof’ burials discussed deposits of the skull and the articulated feet of cattle and horses (Figure 1.2). These deposits were being discovered with increasing frequency in the south of Russia, often in association with human remains. This association lead Piggott (1962a) to conclude that the deposits were of a ritual nature, and should be expected in the rest of Europe. However, at that time few British examples were known, although others have been discovered since (Robertson-Mackay 1980).

Sociological – the dead are brought into prominence, compared with those without an animal gift; Spiritual – the animal may be a guide or sacrifice; Emotional – the animal may be a favoured pet, or gift by mourners. Not in association with human remains: Foundation deposit – for the spiritual/divine blessing of a new construction Animal cults – burials of animals deemed to be ‘special’, feared or worshipped (this aspect has also been interpreted as emotional (Pollex 1999) but (Behrens 1964) does not refer to it as such).

Piggott (1962a) and subsequent authors were also influenced by the work of the German archaeologists Lidia Gabalówna (1958) and Hermann Behrens (1964). Gabalówna (1958) discussing Neolithic cattle burials, suggested the important point that humans and animals can be subjected to the same ritual activity. Behrens’ (1964) investigation into ‘Animal skeleton finds of the Neolithic and Early Metallic Age’ covered the staggeringly wide geographical areas of Europe, Africa and Asia. Behrens discovered 459 cases of complete animal skeletons from 268 sites, although the term ‘complete’ is never defined. Over 50% of the cases consist of dog skeletons, including unusual deposits such as a dog skeleton in association with a human child’s cranium and cow’s skull.

Interestingly, complete cattle skeletons were not often associated with human remains, whereas the majority of dog skeletons were. It was suggested that this may indicate that dogs were being utilised as a guide in the afterlife, and/or economically it was favourable to sacrifice a dog rather than a cow, which has a greater economic value. The above work on Neolithic/Bronze Age animal burials did influence the interpretation of such finds within Britain. Complete dogs in associated with human remains started being interpreted as ‘ritual’ (Bailey 1967; Bunting et al. 1968; Grinsell 1959). However, such interpretations were the exception rather than the rule. It is also noteworthy that in the theoretical climate in which it was written, it is surprising that Behrens’ (1964) rarely suggests a ‘functional’ interpretation throughout the text. However his work does predate the main period of processual archaeology.

In correspondence with Piggott’s (1962) findings, the majority of Behrens’ (1964) cases were either in direct association with, or close to, human burials. After dog, the second most common species was cattle, which made up around 30% of the cases. As well as gathering an extraordinary amount of data, Behrens (1964, 81-82), inspired by the earlier work of Gabalówna (1958), suggested a number of interpretations which can be summarised as;

1.2.4 Bucking the trend; Romano-Celtic religion Influenced by archaeological thinking on mainland Europe, animals were beginning to be considered as part of a society’s socio-religious make-up within Neolithic studies. During the later part of the 1960’s and early 1970’s animals started to be viewed in a new light by authors examining ‘Celtic’ religion. Anne Ross (1967) drew on literature (Irish and Roman), iconography and archaeological examples from mainland Europe in her investigation into ‘Pagan Celtic Britain’. Similar work regarding the role of dogs in RomanoGaulish religion had been carried out by Frank Jenkins (1957), but Ross greatly expanded on the study to look at all animals. In chapters on ‘sacred and magic birds’ and ‘divine animals’, she summarised the iconographic and literature evidence for ‘Celtic’ beliefs regarding animals. Her work indicated that certain animals were viewed as having connections with specific deities, with roles covering aspects of health, fertility, death and the underworld. There is, though, little correspondence made between ABG deposits and the information provided by the iconographic and literature studies. Ross (1967, 2425) does draw the connection between votive deposits in

Figure 1.2 Illustrations of head and hoof burials: top left Sorte Muld hide-burial Bornholm; bottom left burial with ox-hide South Russia; right burial with ox-hides, South Russian Catacomb Grave culture (Piggott 1962a, Figure 2). 3

Investigating Animal Burials; Ritual, Mundane and Beyond pools and lakes and those placed in wells, shafts and pits having a comparable significance.

2.

In further work, Ross (1968) undertook a survey of deposits from pits, shafts and wells from Iron Age and Roman Britain. She suggested that a full interpretation of ABGs (in this case complete articulated skeletons) and other ‘strange’ deposits in Iron Age pits, shafts and wells was hard to achieve. However, she advocates that when such deposits are examined in the light of ‘Celtic’ practice evidenced from iconography and literature then;

3.

Cisterns and wells were on occasion filled. 3.1. Naturally and slowly. 3.2. When the water had become impure or had dried up. 3.3. On abandonment by the Roman forces to remove material that enemies could use. 3.4. On the return of the Roman forces to clean the site up.

Unfortunately no conclusion is reached as to which reasons would have resulted in the filling of specific pits. One important lasting contribution is the suggestion that features may have been excavated for what could be deemed ‘functional’ terms, but later used for ‘ritual’ deposition (for a recent discussion of the Newstead pits see, Clarke in press).

‘The human skulls, the dog remains, the heads of ravens, the smooth stones, the smashed pottery, the bronze vessels at the bottom of votive wells, the venerated weapons, the equipment of the divine smith, the votive hazelnuts and acorns, the sacred trees, the full equipment for the otherworld feast, the animal sacrificed for prognostication and other ritual purposes, all these can be found regularly in Celtic religion’ (Ross 1968, 275).

Although a ‘ritual’ purpose for the deposition of ABGs was being advocated by Ross (1967; 1968; Ross and Feachem 1976), very few ABGs were being considered in this light at the time. One of the few exceptions were the complete and partial skeletons which were being discovered at the hillfort of South Cadbury, Somerset (Alcock 1970). These deposits were interpreted as sacrifices because of their proximity to a putative Iron Age shrine. However, significantly other ABG deposits from the site were being interpreted as waste, as was the convention at the time.

Ross only included those pits, shafts and wells which seemed to have a clear ritual importance. However, there does not appear to be any defined criteria for what makes a feature, or the deposits associated with it, ‘ritual’. Influenced by Ross (1967; 1968) and Jenkins (1957), Jocelyn Toynbee (1973) produced a study investigating animals in Roman life and art. Like Ross’s (1967) study, the majority of the evidence is drawn from literature and art sources with little attempt to integrate the results with archaeological remains. However, Toynbee (1973) does indicate that many animals had a ritual role in Roman society.

One important point to make here is the lack of involvement in interpretation by zooarchaeologists. The above examples of ‘ritual’ interpretations are offered by conventional archaeologists or historians. At the time zooarchaeology was still in its infancy, and through the influence of the ‘paleoeconomic school’ was more concerned with themes advocated by processual theory. Even Anne Grant who later would become an important advocate for a ‘ritual’ interpretation of ABGs, was interpreting the ABG deposits from Portchester Castle, Hampshire, along conventional ‘functional’ lines (Grant 1975). The work of individuals such as Ross (1967) was also to some extent dismissed by the archaeologists working in the period. For example, in Cunliffe’s (1974) first edition of Iron Age Communities in Britain, ABGs and the ideas of Ross (1967) are not mentioned.

Petres (1972) also alludes to the ritual role dogs and horses had in Celtic and Roman society. Investigating the ABGs from Pákozd, Hungary, Petres links the archaeological and literature evidence to suggest that two groups of ABGs exist; those which were sacrificed and those which were consumed at sacrificial feasts. This is one of the first times ABGs were suggested to be the remains from feasts, a theory that later became popular. However, Petres does not expand upon the point, as the majority of the article is concerned with human sacrifice. Ross (Ross and Feachem 1976) expanded upon her earlier work by investigating pits from the Roman fort at Newstead in the aptly named ‘Ritual Rubbish –The Newstead Pits’. Ross and Feachem argue that it is the nature of the deposits within the pits which classify the features as having a ritual nature. Such deposits often included cattle, dog and horse skulls, animal skeletons, human remains, complete pots, ornaments, metalwork, and stone work.

1.2.5

Structured deposition

The interpretation of the Neolithic henge enclosure of Durrington Walls, Wiltshire, offers a useful example of how archaeological theory has developed, which in turn has affected the interpretation of animal bones and ABGs in particular. When the site was first reported upon by Wainwright and Longworth (1971), the finds (indicating possible feasting) and the large external bank (from where people could be spectators to what was taking place in the interior) suggested that Durrington Walls was

A number of reasons for the filling of pits were suggested; 1.

Cisterns or wells originally made for storage or water supply were on occasion used and filled ritually.

Shafts or pits were on occasion designed and used only for ritual purposes, and filled ritually. 4

Chapter 1. Introduction to a Deposit Type deposits can be described as ritual deposits. For example, during the recent foot and mouth crisis in the United Kingdom a large number of trenches where excavated and slaughtered cows were placed within these pits. If future archaeologists were to excavate these features they would encounter a large number of articulated cattle skeletons, and the bones could be seen as a ‘structured deposit’. The interpretation of the deposit would be dependant upon the sources available to the archaeologist. Such a deposit could be viewed as ‘structured’, having a specific purpose, but is not of a ‘ritual’ nature. As Hill (1995, 96) notes, various ethnographic and archaeological studies have shown that garbage, settlement space, preparation, cooking and consumption of food can be seen to be structured through deep-rooted cultural norms.

probably a ceremonial centre. Soon afterwards Wainwright (1979) investigated the Iron Age settlement enclosure at Gussage All Saints, Dorset, which had very similar finds to Durrington Walls (Bradley 2005, 12-14). These similarities led Wainwright (1975) to reinterpret Durrington Walls as a major settlement. Later investigations of the finds assemblage from Durrington Walls by Richards and Thomas (1984) led to a further re-evaluation. They noted that certain items had been carefully placed at the foot of individual posts in the timber circles present on the site. They also noted that there were non-random patterns in the distributions of pottery and animal bones. They used the term ‘structured deposit’ to describe this treatment of the artefacts and argued that the deposits could be shown, through a highly structured mode of deposition, to be ritual because the performance of ritual involves formalised, repetitive actions, which may be detected archaeologically.

1.2.6

Richards and Thomas did recognise that structured deposition is not a litmus test that defines ritual and suggested that structured deposition relies on showing that: x

the archaeological deposits are recurrently patterned in terms of associations and disassociations between finds and their spatial distribution;

x

the patterns are not due to natural taphonomic processes, but are the product of cultural transformations.

Iron Age influence; ‘special animal deposits’

Influenced by the work taking place on ‘structured deposits’ in the Neolithic, Grant (1984a) interpreted the ABGs from Danebury hillfort in a similar way. A number of ‘special animal deposits’ were noted either through their association with other bones or by the manner of their deposition (Grant 1984a, 533). Three types of ‘special animal deposit’ where distinguished; x x x

This led Wainwright (Wainwright 1989, 50) to return to his original interpretation that Durrington Walls was a ceremonial centre, and also greatly influenced the interpretation of Iron Age sites. However, recent work has indicated that the finds assemblage studied by Richards and Thomas (1984) may have been biased by the way the finds were stored (Albarella and Serjeantson 2002).

Animal burials – fully or partially articulated skeletons (Figure 1.3). Skulls and horse mandibles – complete or near complete. Articulated limbs – complete limbs or portions of limbs, with upper limb bones considered as one group and lower limb bones (below the carpals or tarsals) as another group.

The animal burials were considered to consist of the articulated remains of complete or partially complete single animals, with no evidence of butchery upon them. However, later work has indicated that some ABGs of all types do have butchery marks (Knight 2001). Grant (1984a) also separated the animal burial data by age when it was possible. Animals with a Mandible Wear Stage (MWS) (see Grant 1982) of five or less were excluded from the analysis. These animals were very young, having lived for a month at most, and were considered to be natural deaths. A large proportion of the animal burials with a MWS score available fell into this group. In addition, Grant (1984a) comments that neonatal animal burials are less likely to be spotted by the excavation team, which means that more neonatal animal burials could have been present at Danebury. Grant also makes the assumption that animal burials with a MWS of five or less died of natural causes and are therefore ‘functional’ in nature, whereas those with a MWS of six or above were deposited in a ‘ritual’ manner.

The term ‘structured deposition’ has been utilised by a number of authors since, especially when studying Neolithic assemblages (Bradley 1990, 4; Pryor 1988). The term has become universal shorthand for ‘ritual’. At the ICAZ (International Council for Archaeozoology) conference in 2006, a number of papers were presented utilising the term structured deposition when discussing ‘ritual’ ABG deposits. However Hill (1995, 95-100) has cautioned that just because a deposit is structured does not mean it is ‘ritual’. In fact: ‘Structured deposition can be seen as a more secure way of showing that deposits contain well preserved material, whatever their origin’ (Hill 1995, 96). In essence, the majority of classes of ritual deposits can be described as structured deposits. But not all structured 5

Investigating Animal Burials; Ritual, Mundane and Beyond The third criterion is the one Grant (1984a) draws on the most. If the ‘special animal deposits’ represent sacrifices made for ritual/religious reasons, then it is argued that it would be better to sacrifice animals of less economic importance. As the remains of sheep were the most common species represented in the ‘normal’ faunal assemblage at Danebury, it was assumed to have been the most important species economically. It would therefore be better to sacrifice horses or dogs, which were not as important as sheep in the economy. This, Grant argues, is why horse and dog are more common in ‘special animal deposits’ than in the ‘normal’ faunal assemblage. The same reasoning is also used to explain the limb deposits. It is cheaper to make a ‘ritual’ deposition of a portion of an animal, rather than a whole animal. Other explanations for the presence of horse and dog remains were also suggested. They may be ‘special’ animals, being neither fully wild nor domestic, and therefore a mighty sacrifice to make. However, the majority of Grant’s arguments are economic in nature, which possibly reflects the undercurrent of economic/environmentally determinist explanations associated with animal bones at that time. In this, Grant was being influenced by the work on the Romano-British shrine at Uley, Gloucestershire, where 80% of the animal remains consist of goat. It had been suggested that goats were being sacrificed because of their limited economic value (Ellison 1980), which is also suggested by Roman literature sources (Toynbee 1973).

Figure 1.3 Partial articulated cattle skeleton in association with chalk blocks, Danebury (Cunliffe 1984, Figure 9.32). ‘Special’ skull deposits were distinguished from ’nonspecial’ skull remains by their completeness and lack of butchery marks upon them. Horse mandible deposits, which consisted of undamaged, still articulated, left and right halves were also considered to be ‘special deposits’. Limb bone groups that were recorded as articulated by the excavators were assigned as ‘special animal deposits’. However, those which were only noted to be part of the same animal during the post-excavation stage were not included. Also, the limb bone deposits were split into two groups; x

Those consisting of the upper limb bones.

x

Those of the lower limb bones, the carpals, tarsals, metapodials and phalanges.

Wait’s (1985) investigation into the nature of Iron Age ritual and religion followed on from Grant’s (1984a) work. He identified two main deposit types; human remains and ‘special animal deposits’. Wait’s study covered the south of England and drew similar conclusions that the animals involved were not those which were of the most economic importance to the communities. Wait (1985, 141-151) suggests five ways ‘special animal deposits’ can be identified;

Grant separated them as she considered the lower limb bones to be possible butchery waste, although they are not excluded from the ‘special animal deposits’.

1. They consist of animals or their parts, which are not exploited in a normal manner.

Three main criteria are considered by Grant (1984a) to indicate why these deposits should be considered ‘special’;

2. The proportion of ‘special animal deposits’ does not correlate with the ‘normal’ proportion of species on the site. 3. There is a consistency in the body parts chosen for partial ‘special animal deposits’.

1. The deliberate positioning of two or more animals together. Grant (1984a) argues that the possibility of two natural deaths of different species and ages of animal occurring is unlikely.

4. There is evidence of care in the placement of the remains and association with other ’special’ deposits.

2. The associations and position of the deposits. Some were in association with sling stones and chalk blocks, others were placed on the bottom of empty pits.

5. ‘Special animal deposits’ only occur in pits, not ditches.

3. The species represented as ‘special animal deposits’ did not reflect the relative proportion of species present on the site. Horse and dog ‘special animal deposits’ were much more common than would be expected from counts of individual bones.

Wait (1985, 138) went further than previous work in arguing that ‘special animal deposits’ occurred in specific places spatially. He suggested that ‘special animal deposits’ were deposited in the interior of the occupation 6

Chapter 1. Introduction to a Deposit Type sacrifice as their remains from ‘normal’ deposits are not found amongst the ‘special’ deposits (Figure 1.4).

sites close to either houses or paddocks. Also, the deposits only occurred within pits, which, drawing on Ross’s (1968) work, Wait (1985, 141) suggested were disused storage pits and not therefore constructed for a ‘ritual’ purpose. However the claim that they do not occur in ditches has since been disproved (for example Bullock and Allen 1997; Maltby 1987c)

Miranda Green (now Aldhouse-Green) (1992) conducted a similar study utilising artefactual and written evidence, although unlike Méniel (1992) the emphasis is on the ancient literature evidence provided by Roman and later Irish authors and not the faunal remains. Green (1992), greatly influenced by Ross’s (1967) early work, shows that many animal species had connections with ‘Celtic’ deities, hence their utilisation as sacrifices. She also makes an important point, later expanded upon by Hill (1995), that;

Such theories concerning ABGs had greater support on mainland Europe compared to Britain (see below). This is emphasized by the holding of a conference on animal sacrifice in 1988 at Compiègne, France (Méniel 1989). The only paper concerning ABGs from Britain was by Grant (1989a), who expanded on her work at Danebury by showing that it might be possible to identify ritually deposited animal remains in Britain from the Neolithic to the Romano-British period, but says little concerning the significance of the deposits. Grant (1989a, 79) also comments;

‘It is quite impossible to separate the profane and spirit worlds, or the ritual from the secular aspects of society’ (Green 1992, 4). 1.2.7

‘In fact it has until very recently been rather unfashionable to talk about ritual for fear of being labelled as part of the lunatic fringe of archaeology.’

With the exception of Grant and Méniel the majority of the previous work discussed above was conducted by non-zooarchaeologists. But the zooarchaeological community as a whole was starting to take note. The animal remains section in the 1991 Danebury publication (Cunliffe and Poole 1991) is very revealing. The section on ‘ritual’ deposits is limited to half a page where Grant (1991, 482) comments;

Excavations on continental Iron Age sites such as the sanctuary at Gournay sur Aronde (Brunaux et al. 1985) had been producing large numbers of ABGs and influenced their interpretation in France and elsewhere in mainland Europe. Two different types of ABG deposits were identified at Gournay (Brunaux 1988, 120); x

Cattle and horse complete/partial ABGs with little or no butchery evidence deposited in the outer ditch of the enclosure.

x

Sheep and pig partial ABGs with numerous butchery marks.

Iron Age influence; sceptical zooarchaeologists

‘There is also, it must be added, an undercurrent of scepticism about these animal deposits, and some have argued, privately and publicly, though not necessarily in the press, that they represent nothing more than natural deaths of animals that died in circumstances that render them unfit for human consumption.’ Grant goes on to note that as the ‘special animal deposits’ were possibly linked with the deposits of other materials, the publication of the 1979 to 1988 findings was delayed so the discussion could be integrated with other finds groups. This discussion has still not been published. It appears that in contrast to the continent, a ‘ritual’ explanation for ABGs was not readily accepted by zooarchaeologists in Britain. Hill (1995; 1996) suggested that there is a divide that appears to eliminate the ability of any archaeological evidence to illuminate the real world as soon as it is labelled ritual.

The cattle and horse ABGs found in the outer ditch of the enclosure were from old individuals, whereas the sheep and pig remains came from young animals. Utilising ancient Roman literature, it was suggested that the remains present in the outer ditch were sacrifices for the deities, who can feast on tough flesh. The inhabitants sacrificed young sheep and pigs to be feasted upon by themselves (Brunaux 1988, 120). Therefore the ABGs at the site represent two different ritual activities. Méniel (1992) added further to the work at Gournay and investigated the evidence for animal sacrifices taking place in Iron Age Gaul, utilising evidence from faunal remains, and iconographic and ancient literature. Méniel (1989) suggested that ABGs were formed by a number of different practices; as sacrifices for deities, ritual feasting and food offerings for the dead. Pointing out that ‘normal’ butchery waste is also found amongst the ‘special’ deposits, Méniel (1992, 141-143) argued that horses may be viewed as the most important animal for

‘It is perhaps because they feel any bone labelled ‘ritual’ cannot be used to reconstruct diet, herd management and other practical matters of the economy’ (Hill 1996, 23). However Grant’s (1984a) work had drawn attention to these deposits and they were now being regularly

7

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 1.4 Hierarchy of animals indicated by their presence in culinary garbage (non-ABGS) from habitation sites, burial sites (necropole) and sanctuaries (altered from Méniel 1992, 142) therefore represent evidence for population control (Maltby 1985a; 1987c;1987d; 1988; 1993). Although the reason for dumping neonatal fatalities in pits, rather than above ground was not explained (Hill 1995, 28), it could in fact result from better preservation within pit deposits.

commented upon by zooarchaeologists working on Iron Age material. But the majority of zooarchaeologists were sceptical about describing such deposits as ‘ritual’. The majority of the work on faunal material from the south of England in the 1980’s was carried out by the English Heritage Wessex Faunal Remains Unit, based at the time at the University of Southampton. The majority of the Iron Age sites with ABGs present were reported on by Mark Maltby. He noted that the ABGs from Winnall Down (Maltby 1985a) could, as Grant (1984a) had put forward, be split into different types; 1.

Burials of complete or partially complete carcasses

2.

The burial of foetal or neonatal carcasses

3.

Small groups of articulated bones (e.g. limbs)

The small groups of articulated bone discovered from sites dating to all periods, including the Iron Age were invariably interpreted as butchery waste by staff of the Faunal Remains Unit (for example see Coy 1984b; for example see Maltby 1981a; Maltby 1985a;1985c; 1987a; 1993). The view of most zooarchaeologists, in the 1970’s and 80’s to ABGs can be summarised by Maltby’s (1987c) comment; ‘..that the large proportion of articulated bones were not of any particular significance that cannot be explained by the events normally associated with pastoral farming’.

However unlike Grant (1984a), Maltby’s (1985a) explanations were not ‘ritually’ based. As Hill (1995, 28) points out, interpretation of the first group at Winnall Down is limited. However, members of the Faunal Remains Unit and other zooarchaeologists at that time, interpreted similar groups in reports from other sites. These were variously viewed as the remains of diseased animals being buried (Buckland-Wright 1987; Maltby 1987c), natural deaths (Bourdillon 1990a; BucklandWright 1987; 1990; Maltby 1987c; 1988) or animals (specifically canids) that were subject to population/pest control (Maltby 1987c; 1988).

However this does not mean that ‘ritual’ possibilities were overlooked. A large number of the reports did not come down in favour of one explanation and gave a balanced view. Pits that contained a large number of small groups of associated bone were considered to represent a large butchery event (Maltby 1985a). However, it was conceded that such a butchery event may have been associated with ritual feasting (Armour-Chelu 1991; Maltby 1985a). However, feasting was considered to be a very different form of ‘ritual’ activity compared with sacrificing animals as Grant (1984a) had suggested.

The large numbers of young animals found were considered to be natural deaths or, in the case of dogs, deliberate population control. Groups of neonatal puppies were discovered at Winnall Down as well as at other Iron Age and Romano-British sites. It appears unlikely that large number of puppies, possibly from several litters, would all die at the same time. Their presence may

One of the main critics of a ritual interpretation for ABGs has been Bob Wilson (1992; 1996; 1999). Wilson saw the majority of ABGs to be a product of butchery practices. They were ‘special’ because of their unusually good preservation, but this did not mean they were the result of ‘ritual’. Wilson (1992) argued that ABGs were created on Medieval sites by normal processing activities,

8

Chapter 1. Introduction to a Deposit Type E. Deposits of the pelvic girdle.

not by ‘ritual’ activity (Wilson and Edwards 1993). If the majority of faunal material is secondary refuse, not reflecting the original position of its deposition (i.e. in middens), then what makes ABGs special is that they have survived this process or have been deposited straight into open features as Maltby (1985a) suggested. Wilson (1985; 1992) has shown that deeper features contained better preserved assemblages. Therefore we can expect more ABGs to be found at the bottom of pits. Wilson (1992) also pointed out that the taphonomic history of ABGs had been generally overlooked and to some extent this is still the case. 1.2.8

F.

Deposits of individual bones

The categories bear a striking resemblance to those that define the ‘special animal deposits’. However, Grant (1984a) and Wait (1995) do not go as far as to suggest that it meant a similarity in ‘ritual’ treatment (Fitzpatrick 1991). The interpretation of human remains within Iron Age pits has also developed with Bradley (1984, 159; 1990, 164) and Cunliffe (1991a; 1991b, 518; 1992) both arguing that the deposition of human remains is a rite possibly linked with fertility. Brunaux (1988, 121, 126-127) proposes that in Gaul, in addition to animals, other material types were also deposited as offerings, such as plants, weapons and humans. Cunliffe (1992) suggests that there might have been a ‘pit belief system’, in which human, animal, wood, textiles and food products such as cheese and drinks were all deposited in a ‘ritual’ manner. As the pits were probably originally utilised for the storage of grain (see Bowen and Wood 1968; Reynolds 1974), then after their disuse items were ‘ritually’ placed in the pits to ensure future fertility and/or as a offering to the deities/spirits of the underworld for their protection of the grain.

Iron Age influence; a menagerie of deposits

One of the central points Ross (1967; 1968), Ross and Feachem (1976), Grant (1984a), Wait (1985), Brunaux (1988) and Green (1992) make is that the ABGs are ‘special’ deposits because of their association with the deposition of human remains in the Iron Age. Human remains have long been recognised on Iron Age sites and have been discussed by such archaeological pioneers as Pitt-Rivers (1888, 60-61), who considered the complete corpses in pits to be the normal burial practice, with the pits utilised due to their convenience. Another aspect of the disposal of human remains that was often commented upon was the haphazard nature of the deposits showing ‘casual treatment of dead bodies’ (Cunliffe 1974, 316). Such discussion of Iron Age human burial lead Hodson (1964) to famously claim that there was no coherent burial tradition in the British early and middle Iron Age, and that a clear burial tradition was absent until the late Iron Age when Aylesford-Swarling type cremation rites were introduced.

Hill (1995; 1996) utilised the idea of different material types being subject to similar ‘ritual’ deposition in his seminal investigation into the nature of ‘ritual and rubbish’ in the Iron Age. Following Grant’s (1984a) work at Danebury, Hill’s (1995) investigation and conclusions can be seen as one of the most influential works on the subject. As noted above, he criticised the use of such terms as ‘special animal deposits’ instead opting for Associated Bone Groups (ABGs), which is a term this author has also adopted. Hill also took on board the discussions of zooarchaeologists, Maltby and Wilson in particular, regarding ABGs (see above).

Such an attitude to human remains from the Iron Age remained dominant until Whimster’s (1981) reexamination of Iron Age mortuary practices and his proposal of a pit burial tradition. The emphasis was upon complete human remains being placed in disused pits, but this is not the only burial rite evident. A large number of human remains discovered on Iron Age sites did not come from complete burials. For example the excavations during 1969-78 at Danebury produced 24 deposits of complete human remains, but also 85 deposits of isolated human bone (Walker 1984, M31:A3-A5). The human remains from Danebury were classified into a number of different deposit types (Walker 1984, 442);

Hill’s (1995) innovative approach involved the examination of the detailed excavation records of a number of sites within the Iron Age ‘Wessex’ region. He considered not only ABGs, but human remains, pottery, plant remains and small finds, producing a large analysis of the material on feature, ‘feature-thirds’ and where possible context levels. From this research Hill reached a number of very influential conclusions. Hill (1995, 97) argued that it is not possible to create a detailed ‘check list’ of ritual and the attempt at a universal criterion to identify ritual archaeology (as argued by Levy 1982; as argued by Renfrew 1985; Richards and Thomas 1984) would prove unsuccessful. Hill (1995, 97-98) suggested two approaches;

A. Inhumations consisting of complete articulated individuals. B. Individual deposits of incomplete partial skeletons.

1.

C. Charnel pits, with the deposition and mixing of a number of individuals. D. Deposits of skulls and frontal bones. 9

A universal litmus test for ritual is impossible. We should instead concentrate on the interpretation of the specific material.

Investigating Animal Burials; Ritual, Mundane and Beyond 2.

We should outline clearly what each of us means when using the word ‘ritual’, by giving examples of the type of practices in different societies we recognise as rituals.

Hill (1995, 98-100) further argued that the main problem with interpreting ABGs had been their association with domestic occupation sites. However, this was only causing a problem because of the archaeologically imposed sacred and profane dichotomy. Instead we should; ‘see ritual as a particular form of practice distinct from everyday practices which typify the ordinary, the commonplace, the routine, but which are still practices…ritual does not belong in a box separate from other activities’. (Hill 1995, 98).

Figure 1.5 Graph showing the percentage of archaeological prehistoric publications regarding ‘ritual’ for publications found during a keyword search of the Bodleian Library, Oxford (altered from Bradley 2005, Figure 1.12)

Therefore rituals will use the same practices as the mundane, so ABGs would have been butchered in the same manner, and using the same tools as ‘normal’ butchery.

fracture patterns etc whereas the tendency of nonspecialists was to see these matters as trivial (Milner and Fuller 1999), did not help the situation. But during the 1990’s zooarchaeology started to become involved in the post-processual dialogue taking place at the time, possibly due to the influence of people such as Hill. As mentioned above, the majority of the zooarchaeological community regarded ABGs as a result of butchery waste or other ‘functional’ activity. However, that started to change, especially for those examining Iron Age material. Zooarchaeologists started adopting a ‘ritual’ explanation for ABGs, which today is the norm (for example Grant 2000; for example Grigson 1999; Hamilton-Dyer 1999b; 2002a; Maltby 2002b; Powell et al. 2005; Sykes 2003). This occurs against a background where there was a general increase in the study of ritual, due to changes in archaeological theory and the influence from anthropology (Bradley 2005, 31-32) (Figure 1.5).

Hill’s (1995) work also showed that there were statistically significant relationships between the presence and absence of a whole range of different artefacts. ABGs could be shown to be placed in features in a certain order compared to other material such as human bone, pottery, metalwork etc. But such deposits were not made regularly. If Winnall Down was typical of middle Iron Age non-hillfort settlements, then such deposits were probably only made once every ten to twenty years or more (see11.3). The possible irregularity of such deposits and the strong similarity between ABGs and human remains led Hill (1995:100) to suggest that ABGs are ritual in nature; ‘those archaeologists who accept that the treatment of human remains is ritual must extend this interpretation to animal remains, pottery or small finds treated in similar ways.’

Wilson (1999) also appears to accept the general premise that ABGs are created by a form of ‘ritual’ behaviour, although he argued that the explanation was not as simple as seeing them as sacrificial offerings as suggested by Grant (1984a), Green (1992) and Wait (1985). Drawing on the different types of ABGs, positioning and butchery evidence Wilson (1999, 301) stated;

However, Hill did note that the human remains could on the other hand be rubbish, a point often lost by authors referring to his work. He also did not see the ‘ritual’ production of ABGs in the same way as Grant (1984a), who saw them as sacrifices. Hill (1995, 101) saw the creation and deposition of ABGs in the context of a gathering involving a large number of people, possibly feasting. 1.2.9

‘Either these skeletal remains were waste from butchery and normal consumption of meat or from feasting, or they were buried as a sacrifice of bones, which might normally be broken for marrow. The evidence may also mean that rituals of various kinds are represented, e.g the ABGs in the grain storage pits inside Danebury hillfort may have been deposited for somewhat different reasons to the ABGs in the pits outside and near the entrance.’

Social zooarchaeologists

After Hill’s (1995) study the argument that ABGs and other artefacts have a ritual nature appears to have become widely accepted. At the time zooarchaeology had become separated from the wider field of archaeology, which was neither necessary or helpful (O'Connor 1998). The impression that zooarchaeologists focussed on methodological issues such as taphonomy, bone densities,

By the time of Wilson’s (1999) discussion the ‘ritual’ nature of ABGs from prehistoric sites appears to have been accepted and was starting to influence the interpretation of ABGs from other time periods. Fulford 10

Chapter 1. Introduction to a Deposit Type 1.3

(2001) urged archaeologists to take a closer look at ABG and other ‘special’ deposits found in Romano-British pits. He pointed out the high proportion of dog ABGs discovered on Romano-British settlement sites and argued that such deposits might represent a continuation of Iron Age ‘ritual’ practices and not population control as suggest by earlier zooarchaeologists (Maltby 1987c; 1988). King (2005) has also recently claimed that ‘ritual’ animal deposits can be found on a number of RomanoBritish temple sites.

Research aims

The above demonstrate some of the different concepts and theories which have been put forward concerning ABGs. It is worth pointing out that zooarchaeologists have to a great extent been too passive with regards to studies involving ABGs. This may not always be the fault of the zooarchaeologist. For example the ‘special deposits’ from the Danebury Environs investigations were not discussed and in some cases possibly not examined by the zooarchaeologists (e.g. Poole 2000b). From the above examples only Grant (1984a; 1989a; 1991), Maltby (1985a), Méniel (1992) and Wilson (1992; 1999) are zooarchaeological specialists. The majority of literature is written by non-osteologists. Although zooarchaeology will always be linked with questions of an economic nature, it has started to become more integrated with current archaeological theory.

Influenced by Fulford’s (2001) argument, Woodward and Woodward (2004) revaluated the ABGs found within a number of Romano-British shafts at the Greyhound Yard site in Dorchester. The majority of the ABGs consisted of complete or partial dog skeletons and had originally been interpreted as the result of natural deaths or culling of infant puppies to control the population (Maltby 1993; Woodward 1993). However, using new theories and further information regarding the deposits, they were now interpreted as forming ‘structured deposits’ relating to the foundation of the Romano-British town (Woodward and Woodward 2004). Interestingly, the original faunal report is not mentioned and some of the information utilised, such as the evidence that two of the complete dog ABGs were tied together at the neck was not available to the zooarchaeologists when the faunal material was originally analysed (pers comm. M, Maltby).

Zooarchaeologists may have been criticised for concentrating too much on methodological issues, but these areas are of great importance not only to the faunal analysis, but more importantly, to the interpretation of archaeological deposits. Matters such as taphonomy remain a central part of zooarchaeology regardless of the theoretical rhetoric being employed (MacDonald 1991). However, very little taphonomic analysis has been carried out on ABG assemblages, possibly in some cases due to the lack of zooarchaeological involvement (Wilson 1992). Hill (1995, 24-25) does briefly discuss taphonomy and is one of the few authors to discuss the physical processes behind the creation of ABGs.

Although the majority of the literature regarding ABGs is concerned with the prehistoric and Roman periods, ABGs dating to the early Medieval period of Britain are starting to the reconsidered. ABGs from this time period are normally associated with human burials such as horse and dog burials (Evison 1994, 29; Hills 1999; Lucy 2000, 9094). However, a recent survey by Hamerow (2006) proposed 42 possible ‘special deposits’ from AngloSaxon settlements, of which 24 are ABGs (the rest consist of human remains), although the published table (Hamerow 2006, table 1) suggests there were more. The results indicate there may have been an association between ABGs, human remains and boundaries in the Anglo-Saxon period. Like previous authors Hamerow (2006) argued that the correlation between the treatment of human remains and ABGs implied a ‘ritual’ nature to the deposition of ABGs. Interestingly, the term ‘special deposits’ was still utilised despite knowledge of Hill’s (1995) justifiable criticism.

The previous literature shows that although they are faunal deposits, very little attention has been given to the actual zooarchaeological information obtainable from ABGs. Research aim 1 To record, but also to look beyond the species deposited as ABGs, and investigate their osteological nature. For instance, which elements are present? What was the age at death of the animal? How healthy was the animal? Are taphonomic indicators such as gnawing or butchery marks present? The majority of the previous studies concentrate more on the deposits associated with human remains, and the supporting evidence from iconography and ancient literature. The correlation between human and animal bones is an area utilised by many authors to explain the ‘ritual’ nature of the human bone deposits, and few authors have taken on board Hill’s (1995, 100) caution, that the human bones may be rubbish. The utilisation of ancient literature sources by authors such as Green (1992) has also contributed to some animals being viewed as ‘most special’, such as horses, dogs and rooks/crows, when they are not necessarily the most common of ABG deposits. Also the majority of texts are concerned with ABGs deposited within pit features.

At the time of writing zooarchaeology has certainly come a long way since its conception. Once conceived as a purely economic discipline, it has now become entwined in current theoretical debates, especially where ABGs are concerned. Zooarchaeology can now be seen as investigating the life history of an animal, comprising different steps executed at different levels and scales such as human-animal relationships, food acquisition, preparation, cooking, eating and disposal (Marciniak 2005, 2). We are now moving towards a social zooarchaeology.

11

Investigating Animal Burials; Ritual, Mundane and Beyond southern England (classified for this study as Dorset, Hampshire and Wiltshire) and Yorkshire. To carry out such a study, it is also important to investigate by noting negative results, how common ABGs are. This will enable the comparison of the ABG assemblages from two regions that are comparable in size but which have contrasting cultural and environmental characteristics.

Research aim 2 To investigate the contextual associations and the features in which ABGs are deposited. For example, are they mainly deposited within pits? Are human remains deposited in association with them? What other material types are in association? There are certainly variations in the interpretation of articulated animal deposits from different time periods. A large number of deposits found dating to the prehistoric and Romano-British periods have been interpreted as forming part of a society’s ritual framework. In comparison, the majority of deposits from sites dating after the Romano-British period are interpreted as being the result of socio-economic factors, such as butchery processing. The majority of projects investigating the character of faunal remains such as Hambleton (1999; 2007) and King (1978; 1984; 1999b) are normally limited to one time period. Also, the majority of literature examining the nature of ABGs such as Hill’s (1995) work has concentrated on deposits from the Iron Age.

Research into the previous literature regarding ABGs has highlighted that a number of different interpretations have been offered to explain the presence of ABGs in the archaeological record. As noted above, the majority of interpretations can be divided into ‘functional’ or ‘ritual’ categories. The most recent literature on ABGs generally argues that these represent ‘ritual’ deposits. However, many of these explanations appear to be generalised and are not based on the nature of individual ABG deposits. Research aim 5 To investigate the interpretations offered for ABGs and the reasons why they have been used. To utilise the data collected regarding the nature of ABGs and to assess whether our current explanations are valid and, if deemed necessary, develop new ways of interpreting these deposits.

Research aim 3 To move on from the Iron Age basis for ABGs and investigate deposits from other time periods. Although ABGs are present from Palaeolithic sites (for example Bratlund 1996; Chazan and Horwitz 2006; Kooyman 2006; Nadel et al. 2004) they appear to be formed either at kill sites or through natural processes, and they are not present in the areas investigated in this project. One of the critiques of a ‘ritual’ explanation of ABGs by Wilson (1992) is that they are present in the Medieval period. Therefore this study will investigate the nature of ABGs from the Neolithic to late Medieval period.

1.4

Defining ABGs

It is necessary at this point to define an ABG for the purpose of this study. A number of previous studies such as Grant (1984a; 1991), and Hill (1995) have included deposits of single bones in their examination of ABGs. This is because they were examining ‘special animal deposits’ within Iron Age features, which were defined by Grant (1984a, 533) as consisting of three types of deposits; animal burials, skulls (plus horse mandibles) and articulated legs. The animal burials and articulated legs consisted of more than one element, still in articulation when excavated. However, a number of single bones were also included in the classification of special animal deposits. Deposits of complete or nearly complete skulls were included even if there was no sign of them being deposited in articulation with other elements, although sometimes the mandibles were still attached. Complete mandibles, again deposited as a single unarticulated element, have also been classified as special animal deposits.

The majority of previous projects investigating faunal remains have been limited to one archaeological period, but these projects often covered the majority of the British Isles. Due to time restraints it would not be possible to collect data from all of the time periods for the whole of Britain. Although a project of such extent would be invaluable to the zooarchaeological and larger archaeological community, the amount of data would be extraordinary and is beyond the scope of this project. Therefore this project will be restricted by geographical area. ‘Wessex’ is an important region to include in the project as a number of studies which have influenced interpretation of these deposits on a national level, primarily Hill (1995), utilised data from this region. However, it is important to also include ABGs from other regions to test if the ‘Wessex’-based assumptions regarding these deposits are valid.

This project is concerned with deposits of ABGs. This type of deposit can be defined as constituting three types of animal remains; 1.

Research aim 4 To investigate regionality in the ABG assemblage by the collection of data from 12

Animal remains which have been deposited with some portion of the flesh or connective tissue still attached, which has caused them to remain in articulation.

Chapter 1. Introduction to a Deposit Type 2.

Animal remains, which had been deposited in articulation but became disarticulated through the taphonomic processes which are then consequently recognized and identified as constituting a single animal by the zooarchaeologist.

3.

Animal remains which constitute disarticulated remains when deposited, but are deposited in association, and subsequently identified as being from the same animal by the zooarchaeologist.

as information regarding the ABGs recovered from the 1979-1988 excavations has not been published (Grant 1991). Therefore the Danebury data included in this study comes only from the first faunal report (Grant 1984a) covering the 1969-1978 excavations. Hill (1995, 32) also encountered these problems with the Danebury material. An important aspect in achieving the above aims is also the recording of negative results. Therefore, sites with a faunal assemblage but no ABGs present were also recorded. This enables us to examine if ABGs are more common in one period or region, and whether there are differences in the prevalence between site types.

Therefore, single bone deposits are not included in the analysis within this project. This does not mean that skull deposits are discounted, but they will only be included if they are in association with other elements. 1.5

To collect such a large amount of data, a systemic approach was used where key publications were searched for faunal reports. Each faunal and associated site report was then examined for the presence of ABGs.

Achieving the aims; methodology 1.5.2

The aim of this project is to move beyond previous studies undertaken on this subject, in particular the work of Hill (1995). To achieve this we need to approach the topic with a wider-ranging dataset. The review of the previous literature has shown that investigations into ABGs have either been site specific, such as the work of Grant (1984a), or have concentrated on a single time period using a small number of sites (Fulford 2001; K Smith 2006; Wait 1985). For example, Hill (1995, 35-36) used data from eight sites. The use of a small number of sites for this study would be invalid, because two of its main aims are to investigate broad chronological and regional variations. Therefore an extensive systematic approach is more appropriate. However, some limits to the data collection were required to enable the project to progress at the required rate. 1.5.1

Data recording and categories

The large amount of data produced has required the use of data management techniques. Therefore a relational database was constructed using Microsoft Access and this was used to store and analyse the data. When constructing a database one must be aware of the nature of the data and the queries you wish to ask. One of the problems with using such a data management tool is that the data need to be placed into defined categories. It was not the aim of this study to place the ABG data into categories redefined by this author and therefore changing its nature. To avoid this, the database was constructed using the published reports with the most detailed amount of ABG information. Therefore the database is over-constructed and the majority of sites did not provide the detailed level of information it is capable of recording.

Data sources The data recorded from each site, can be placed into one of three levels of data. The first level concerns general information regarding the site. The second level recorded general information regarding the ABG including the feature it was recovered from, species, composition and age. The third level recorded specific information regarding the ABG, including body areas, taphonomic indicators such as butchery and gnawing, and associations with other ABGs or material culture.

As this project will be developing wide ranging conclusions based on secondary sources, only publicly available ABG data will be collected. This also offers a chance to review the current state of publications regarding zooarchaeological data. Therefore data for this study will be drawn from published site records and the English Heritage Ancient Monument Laboratory (AML) faunal reports, which are publicly available. After reviewing a selection of pre-1940s publications for ABG data and finding the data held within them very limited, it was deemed prudent to only collect data published after 1940. Therefore the collected dataset represents the publicly available data on ABGs recovered from excavations of Neolithic to late Medieval sites published post-1940 within the study areas.

As the data for this project come from a search of excavation reports, the nature of a ‘site’ needs to be considered and defined. Multiple excavations have taken place on a number of locales within the study regions. The majority are from towns, such as Dorchester, Winchester and York, but there are also large prehistoric sites like Maiden Castle. Unfortunately, the information from different excavations within the same locales is rarely integrated. Therefore, to maintain consistency, each excavation has been treated as a separate site. For example the Greyhound Yard (Woodward et al. 1993) and Colliton Park (Aitken and Aitken 1982) excavations within Dorchester are recorded as separate sites. When

Such an approach means that a much clearer picture of the nature of ABGs can be formed as all sites with ABGs will be recorded, not just those with large assemblages. However, it does mean the project is reliant upon ABG information being published. This does affect, for example, the data it is possible to collect from Danebury, 13

Investigating Animal Burials; Ritual, Mundane and Beyond deposition, or missed by the excavation team. In order to synthesise the data and make them comparable, definitions of complete and partial ABGs were required. Therefore for the purpose of this study complete ABGs are defined as deposits for which the skull, mandibles, all long bones and pelvis are present and all other body areas are represented. Any ABG not fulfilling these criteria was recorded as a partial ABG, unless no data was present, in which case it was recorded as unknown.

possible, the results from the same locales will be discussed together. Another aspect that needs to be consistent is the recording of chronologies. This consists of two types, the overall chronology for the use of the site, and the specific date for each ABG. Very rarely are specific dates given for ABGs, and the majority of them are assigned a chronological range based on the phasing of the site. The range is recorded for each ABG as an earliest and latest date using the below chronology defined by Hunter and Ralston (1999). Neolithic

- Early Neolithic Middle Neolithic Late Neolithic

Bronze Age

- Early Bronze Age - 2600-1800 BC Middle Bronze Age - 1800-1300 BC Late Bronze Age - 1300-700 BC

Iron Age

- Early Iron Age Middle Iron Age Late Iron Age

RomanoBritish

- Early RomanoBritish Middle RomanoBritish Late RomanoBritish Early Medieval - Early Anglo-Saxon Middle AngloSaxon Late Anglo-Saxon Later Medieval - High Medieval Late Medieval

The second category is used to record the basic body areas which were present for partial ABGs. Often, within publications, an ABG would be described as consisting of ‘the majority of the body and a front limb’. Therefore, a number of basic ABG categories were created to enable recording of ABGs with limited detail. They are based upon body areas and consist of;

- 4000-3500 BC - 3500-3200 BC - 3200-2600 BC

x x x x

- 700-400 BC - 400-100 BC - 100 BC-AD 50

Head (skull and mandible) Axis (the main body including the vertebrae, ribs and pelvis) Leg (appendicular elements) Mixed (all of the above body areas present in a partial ABG)

A partial ABG can consist of combinations of the above such as ‘axis+leg’ or, if elements from the head, axial and appendicular body areas are all present in a partial ABG, it would be recorded as ‘mixed’.

- AD 50-150 - AD 150-300

The third category of element recording consisted of detailed body areas. Some publications gave further element data. Therefore for this study a series of body area categories was created and their presence or absence was noted. These body areas are;

- AD 300-450 - AD 450-600 - AD 600-850

x x x x x

- AD 850-1050 - AD 1050-1300 - AD 1300-1550

When possible, specific dates are utilised. However, often the date range is the only form of dating available. When this is the case it is important that the use of such a range is consistent throughout the study. Therefore the latest date will always be utilised. In effect such a designation is arbitrary, the earliest date could also be used, but using the latest date does give a terminus ante quem for the ABG. However, it will be noted in the period discussions when the use of the latest date may be underestimating the chronological age of a deposit.

x x x x x x

From the initial data search it became apparent that detailed element information is very rarely given for ABGs. Therefore three key categories were created.

x

Firstly the completeness of the ABG was recorded as complete, partial or unknown. However, completeness is a very subjective term, as many authors describe ABGs as complete skeletons, despite there being a few missing elements, which had either been disturbed post-

Head (consisting of the skull and mandible) Cervical Vertebrae Thoracic Vertebrae Lumber Vertebrae Vertebrae (used when the specific vertebra present were not described) Ribs Pelvis Upper front limb (consisting of the scapula, humerus, radius and ulna) Lower front limb (costing of the metacarpals, carpals and phalanges) Upper back limb (consisting of the femur, tibia and fibula) Lower back limb (consisting of the metatarsals, tarsals and phalanges) Foot (used when just the phalanges are present and the limbs they originate from are unknown)

If known, the side of the body the limb was from was also recorded. Although not ideal, this study is limited by the dataset available and such categories allow us to investigate the general composition of the ABGs recorded. 14

Chapter 1. Introduction to a Deposit Type Another category that required definition was the age of the ABG. Within a normal faunal assemblage the mortality profiles will be investigated by utilising either toothwear and/or epiphysial fusion data. These are then used to analyses the population as a whole. However, the raw data, such as the Grant mandible wear stage, are usually not provided for individual ABGs. Many authors only give the general age of ABGs, such as ‘juvenile’, ‘young adult’ etc, or an age range, such as ‘2 to 3 years old’. Therefore, the age of the ABG needs to be recorded in these general terms. To do this, the age stages of the main domestic mammals have been defined using the same terms as the current English Heritage regional reviews for cattle, S/G, pig and horse (Hambleton 2007) (Table 1.1). The age structure for dogs follows the same conventions as Maltby (1993, 327) used for the Greyhound Yard assemblage.

1.5.3

Data analysis

The analytical approach to the ABG data collected is both descriptive and quantitative. To achieve the aims set out for this study, a number of analytical processes were followed for each defined period and region. Firstly, the number of sites with and without ABGs present was examined. This was also compared to the site type data, to look for any associations. The number of ABGs per site was also calculated to allow biases in the sample size to be taken into account for the next level of analysis. Then, the basic number of ABGs per species was calculated for each sub-period, site type and feature type. These data were then compared to the species composition of the non-ABG faunal assemblage. When only a few assemblages are present, or the overall assemblage is dominated by a number of large samples from a small number of sites, these sites are discussed in detail.

At present the age terms zooarchaeologists utilise, such as juvenile, are descriptions that everyone uses, but with no standard definition for each species. Also, we must be aware that such classifications are based purely on skeletal development and tooth degradation. Such classifications would not have been used by the inhabitants of the time periods investigated in this study. However, they offer us a means of investigating the mortality profiles of individual species.

After investigating species representation, the composition of the ABGs was examined. This utilises the levels of data recorded firstly to compare and contrast complete and partial ABGs by species, sub-period, site type and feature type, and then to investigate the body areas present for partial ABGs. When possible, ageing and butchery data are incorporated into this analysis. Finally, any contextual associations between different ABGs and other material types are quantified, and compared by sub-period, site and feature

To investigate the interpretation history of these deposits it was also necessary to record the explanations given by the original authors. This was carried out for each individual ABG as many authors would offer different explanations based upon the species deposited or the associated feature. No predefined categories were created for the recording of these interpretations. Instead, the categories were built up during the data recording process as new interpretations were encountered. However, it soon became apparent that only a few interpretations are repeatedly used by zooarchaeologists, probably due to the influential publications discussed above. This process also ensured that different interpretations were not ‘lumped together’ because of this study’s recording categories. Although during the analysis such a process did prove to be valuable, it was based upon the underlying paradigm associated with each interpretation.

The quantification methods utilised in this analysis may be viewed as simplistic as they rarely move beyond percentages, but such an approach is regulated by the nature of the study and the quality of the data. The results are shown using bar and line graphs, tables and triplots. If appropriate, further statistical methods such as correspondence analysis have been used (see Cool and Baxter 2002; Shennan 1997). This project has resulted in the collection of a large and complex dataset. However, there is no statistical ‘magic bullet’ that will provide all the answers.

Table 1.1 Age stages for the main domestic mammals based on Hambleton (2007) and Silver (1963) Age category Neonate Juvenile Subadult Young adult Adult Old adult

Cattle 0-1 month 1-18 months 18-40 months >40 months 4-8 years >8 Years

S/G 0-2 months 2-12 months 1-2 years 2-4 years 4-6 years >6 years

Pig a few weeks c.1-14 months 15-26 months 27-36 months 3-6 Years > 6 years

15

Horse 0-2 months c.2-18 months c.18-42 months c. 3.5-5 years c.5-8 years c.>8 years

Dog 0-2 months c.2-18 months c.18-42 months c. 3.5-5 years c.5-8 years c.>8 years

Chapter 2. Creation of the Faunal Record 2.1

Introduction

As mentioned in the previous chapter animal remains are one of the most common classes of find retrieved from archaeological sites. They have the potential to aid the archaeologist when investigating all aspects of society from technological choice to spiritual beliefs. That is not to infer that all investigations are easy. There are, as with all aspects of archaeology, limiting factors. As discussed earlier one of the weaknesses of the previous work on ABGs is the lack of taphonomic analysis of the material. Taphonomy is the study of the transition of organics from the biosphere into the lithosphere (Lyman 1994, 1). The term was formally defined by I. A. Efremov (1940). However, William Buckland (1823) with his experiments on the gnawing caused by hyenas set some of the foundations of the study of taphonomy (Boylan 1997). Within archaeology the study of taphonomy is also extended to include the retrieval and analysis of archaeological finds (see below). Before we can discuss the results from the analysis of animal remains, it is important to understand the underlying processes that affect all archaeological finds and faunal remains in particular. We must understand the limitations of our data before we use it to infer behaviour or beliefs in past societies. This chapter looks at how faunal remains come to be deposited on archaeological sites and the processes which affect them during their formation into an archaeological faunal assemblage, effectively their taphonomic history. 2.2

Death

To gain an understanding into the nature of ABG deposits, we must first consider how animal bones enter the archaeological record. Animal bones undergo a number of influencing factors that can all be considered under the general title of taphonomy. It was once considered that the ‘animal bone assemblage is derived in a straightforward manner from an animal population herded in the immediate area’ (Mountenay 1981). However, this has been clearly shown not to be the case. A number of factors affect the ‘death’ assemblage and we cannot draw direct comparisons between the ‘dead’ and ‘living’ assemblages. These taphonomic factors can be split into a number of sub-groups which affect the faunal record as it is transformed through a number of stages subsequent to the death of an animal. Indeed, the first stage is the transformation from living to dead. During this stage the faunal record can be influenced by environmental and human agents. The

environmental factors can be seen as being outside of human control. This is not to say that this transformation is wholly subject to environmental determinism, but it is an important factor. The animals living around a settlement are subject to environmental influences, such as climate, geology, vegetation, food supply and availability of water sources. All of these form the habitat and as such influence what type of faunal community is able to live in the area. Although many domesticates utilised by past societies have a wide and variable habitat range, the habitat will still affect their relative abundance. For example cattle require a ready supply of drinking water, which sheep do not, since on good pasture sheep require very little water (Wright and Ashton 1978). Sheep are therefore much more suitable to higher hilltop areas, whereas cattle are better suited to lower lying valley areas. Grant’s (1984b) comparison of Wessex and Upper Thames Valley Iron Age assemblages showed there is a link between the height of the settlement and the faunal assemblage recovered from it. Sites situated over 76m Ordnance Datum (OD) tended to have a higher proportion of sheep; whereas those below 76m (OD) had a higher proportion of cattle. Hambleton (1999, 51), rightly points out that although topography may be a factor, the areas compared could also have been used by different human communities, and the differences in species abundance could be due to social choice. She tested Grant’s (1984b) theory on a wider range of sites and showed that although further work is needed, there is a difference in species proportions for sites above and below 76 OD. Socio-economic choices will also affect the relative abundance of species within the assemblage. Although an animal may be available, it does not mean that it will be killed and consumed by the local population. A good example is the faunal assemblage from the possible Jewish enclave in Buda, Hungary, which shows that although pork was readily available, due to religious beliefs it was not consumed (Daróczi-Szabó 2004). The influence of social factors upon the faunal record can also be seen when a chronological change occurs. For example, there is a general change in the species proportions between Iron Age and Romano-British sites, with cattle replacing sheep as the dominant animal (King 1978; 1999b). Sites such as Droitwich (Locker 1992b) which was continually occupied in both periods, shows a change in the proportion of sheep, cattle and pig, as well as the introduction of new species and breeds. Changes may also be dependant on the site type. Rural settlements do not show changes to the degree that urban settlements appear to do in the transition from the Iron Age to Romano-British periods. What is important about these changes is that the local environment has not been transformed but the socio-economic factors have. This is also shown by variability between different regions.

Chapter 2. Creation of the Faunal Record become disarticulated as the soft tissue is removed through bacterial and insect action (Hill 1979a;1979b).

This is the great strength of faunal records. The environment may be an influencing factor, but so are the past societies which inhabit it. The faunal data can provide insight into that society’s mindset. Therefore the species comprising ABGs on a site are influenced by a number of factors such as the availability of a suitable habitat, social choice in what animals were kept or imported and also social choice in what animals to utilise. 2.3

Hill (1979b) described a technique for the determination of the disarticulation sequence of a mammal. This was then applied by Hill and Behrensmeyer (1985), using data collected by Behrensmeyer and Dechant Boaz (1980), to a number of species including cattle, sheep/goat and donkey (Table 2.1). The data have been modified from the original Hill and Behrensmeyer (1985) data using Lyman’s (1994, 145) method of placing the skeletal joints in a rank order of disarticulation. This enables us to know the possible order of disarticulation and apply it to the archaeological material. To aid this, it is possible to split the disarticulation sequence up into a number of groups of elements based on the amount of time which it may have taken for disarticulation to occur. Hill and Behrensmeyer (1985) were able to calculate the order that disarticulation stages occurred, using bone weathering stages (Behrensmeyer 1978), which were developed on a sample of 35 cattle carcasses with a known time of death.

Death to ground; skeletonisation

Therefore the animals which make up the faunal assemblage are present due to a number of factors, environmental and social. However there are other influences upon the faunal record, which take place between the time the animal died to when the remains were deposited. If the animal died naturally and was not subject to the influence of the local human population, we would expect skeletonisation to take place where the animal had died, or where it was transported to by predators and scavengers. The process of skeletonisation involves a number of factors. Consumption of the soft tissue by scavengers can take place. The majority of the archaeological literature on this subject is concerned with Palaeolithic archaeology (e.g Binford 1978; Brain 1981; Marean and Spencer 1991), but scavenging normally by canids (dogs) has been noted in most faunal reports relating to the time periods covered in this project. A number of different types of insect will also act as agents of soft tissue removal (Micozzi 1991, 44). Environmental factors such as temperature and moisture will influence the extent to which this occurs (Lyman 1994, 142). Finally micro-organisms including fungi from outside and inside the carcass are a major contributing agent to the decomposition of soft flesh (Micozzi 1991, 37).

There is much less available information for the disarticulation of sheep/goat and domestic donkey, compared with cattle. Therefore it was only possible to split the cattle disarticulation order into stages. This was done using the weathering stages and time differences indicated between the disarticulation of different joints, based on an ideal cattle model (Figure 2.1). The stages do not correspond to a scaled time period as there are a number of possible problems with using weathering data to calculate the length of time passed since death. What weathering data does do is measure the time of exposure and not the time since death as the bones will be exposed to weathering agents at differing rates as the flesh is removed. Todd’s (1983; 1987) experiment using 20 pairs of left and right domestic cattle femora from individuals with the same time of death showed that the weathering stages of Behrensmeyer (1978) can be variable with only eight pairs displaying the same weathering stage, and all six weathering stages were present even though the time of death was the same.

All the factors noted above will also affect animals that have died under the control of humans. The main difference is that human activity upon the carcass will normally occur before these factors can take place. Skeletonisation of animals destined for human consumption is normally attributed to butchery practices. 2.4

Lyman and Fox (1989) suggest a wave model with the bones of a carcass weathering progressively but each at a slightly different rate, and therefore each has a slightly different exposure history. However, the general model of progressive weathering of bones, with those which disarticulate first displaying longer exposures, appears to hold true, with some variability only to be expected (Lyman 1994, 146). We can therefore use the disarticulation stages proposed, with the caution that they do not relate to a scaled chronology. Stage 0 would consist of a completely articulated skeleton. After a cattle skeleton has reached stage 1 we would expect the head and distal phalanges of the limbs to have become disarticulated. The limbs would also have become disarticulated from the rest of the axial body. (Figure 2.2).

Death to ground; disarticulation

As well as human agency a number of other taphonomic factors can result in disarticulation. If an animal dies and is not subject to human influences, then we can expect disarticulation to take place due to the action of scavengers, insects, bacteria, weathering and diagenesis. As scavengers (such as dogs) consume the soft tissue around the bone, they will also pull apart the skeleton and in some cases fragment the bone as well, resulting in disarticulation (Haynes 1980). A carcass which is not subject to human or animal scavenger activity will also

17

Investigating Animal Burials; Ritual, Mundane and Beyond Table 2.1 Rank order of joint disarticulation in cattle, sheep/goat and donkey. Numbers denote the order of disarticulation 1=first joint to disarticulate. (after Hill and Behrensmeyer 1985; Lyman 1994, 145) Body area Head Upper front leg

Lower front leg

Body

Upper Back leg

Lower Back leg

Joint Cranium - Mandible Cranium - Atlas Forelimb - body Scapula - Humerus Humerus - Radius, ulna Radius - Ulna Radius - Carpals Carpals - Metacarpals Metacarpals – 1st phalanx First phalanx – 2nd phalanx (fore) Second phalanx – 3rd phalanx (fore) Atlas -Axis Axis – 3rd cervical vertebra Cervical - Cervical vertebra 7th cervical – 1st thoracic vertebra Thoracic - Thoracic vertebra Thoracic - Rib 12th Thoracic – 1st Lumbar Lumbar - Lumbar vertebra Fifth Lumbar - Sacrum Sacrum - First caudal Caudal - Caudal Sacrum - Innominate Innominate - Femur Femur - Tibia Tibia - Tarsal Tarsal - Metatarsals Metatarsals – 1st phalanx 1st phalanx – 2nd phalanx (hind) 2nd phalanx – 3rd phalanx (hind)

Joint type S-Hinge S-Hinge None S-Hinge S-Hinge (F-syndesmosis) S-plane S-plane

Cattle 2 6 1 5 16 27 14 18

Sheep / Goat

Donkey

4.25 16

10

S-Hinge

12

S-Hinge

7

S-Hinge S-Pivot

3 13

C-Secondary

24

4.25

C-Secondary

25

9

C-Secondary

22.5

4.25

C-Secondary S- condyloid

26 21

14 8

5 4

C-Secondary

28

15

7.33

29 30 20 9 22.5 8 15 15 19

13 10.33 4.25 3 10.33

6

10.33 0.5 0.5

7.33 1.33 1.33

C-Secondary C-Secondary C-Secondary C-Secondary S-plane S-ball and socket S-condyloid S-plane S-plane S-Hinge

11

S-Hinge

10

S-Hinge

4

18

7.33

1.33

Chapter 2. Creation of the Faunal Record

Figure 2.1 Statistical scheme of disarticulation for an ideal skeleton of domestic cow (adapted from Hill and Behrensmeyer 1985, Figure 2)

Figure 2.2 Diagram illustrating the elements expected to be articulated after natural disarticulation stage 1-3. (images adapted from Barone 1976)

19

Investigating Animal Burials; Ritual, Mundane and Beyond burials in places sheltered from transportation forces.’

Stage 2 consists of the disarticulation of the remaining elements of both the hindlimbs and forelimbs, with the radius and ulna remaining articulated. The ribs also become disarticulated during this stage, as do the caudal vertebrae. The vertebral column, sacrum and pelvis remain in articulation.

When studying articulated material from an archaeological context, we must also add to Schäefer’s (1972, 10) description the human agent. It is humans that determine who is buried and when but also how and where (Henderson 1987). If articulated bone was left above ground in a midden for a period of time, it may well have been disarticulated by gnawing although this does depend upon scavengers being present on sites and having access to the material. We can expect that an ABG left above ground in, for example, a midden on a site with scavengers present, to be at the least disarticulated and possibly totally destroyed. Therefore for ABGs created by possible butchery practices to survive in articulation, they would need to be deposited in the ground before major scavenger activity takes place.

Stage 3 sees the disarticulation of the radius and ulna, and the disarticulation of the majority of the vertebral column, except for the lumbar vertebrae and sacrum. These become disarticulated during Stage 4, which results in the final disarticulation of the whole skeleton. The majority of the skeleton should be disarticulated by the time Behrensmeyer’s (1978) stage 4 weathering had been reached (6-15 years), with the majority of the disarticulation over by about five years after death (Hill and Behrensmeyer 1985). One problem noted with Hill and Behrensmeyer’s (1985) work on disarticulation is that they did not identify and distinguish between the activity of decomposition bacteria and insects from the larger scale activity of scavengers, in their case hyenas (Micozzi 1991, 50).

2.5

Death to ground; butchery

The investigation of butchery methods used on animal carcasses involves the examination of marks left on the bones. These tool marks, either made by a flint blade, metal knife, metal cleaver or other tools can be recorded by the zooarchaeologist. Butchery is still part of the taphonomic process, and has been defined by a number of authors (Binford 1978, 386; Lyman 1987; Rixson 1988) . The general consensus is that the main aim of butchery is to reduce an animal carcass to a number of primary products, which can result in skeletonisation. Although the majority of archaeological faunal material is assumed to have undergone some form of butchery, most bones show no identifiable butchery marks. For example a detailed study of the faunal material from Dudley Castle (c.AD1100-1750) identified butchery on 1185 (26%) of the cattle assemblage from the site (Thomas 2005, 38). This, when compared with other sites, is a relatively high proportion of butchery marks. It is possible for a skilled butcher to completely process a carcass without leaving a butchery mark, as the aim of the butcher is not to cut bone, but meat. Whether a butchery mark is left can be dependant upon a number of factors including the skill of the individual, the tools available, the morphology of the animal (Binford 1981, 91), which cuts and standards of meat are desired (Maltby 2007; Seetah 2006) and how long butchery occurs after death as this affects the toughness of the meat (Potter 2005).

However, the dataset this project is dealing with would have undoubtedly been subject to some form of human activity. It is probable that the majority of archaeological faunal material has undergone some form of butchery. Therefore when scavenger activity takes place it will be on bone, which has had the majority of the flesh removed, allowing mainly access to bone and connective tissue. Brain (1967; 1981) in his study of goat bones deposited around a South African village noted the attrition of the bones due to the damage caused by dog gnawing. Binford (1981, 51-77) furthered the work, by investigating the effects of gnawing on faunal assemblages from Alaskan Nunamiut camps, showing that a number of diagnostic characteristics can be left on bones by animal gnawing. Both these and other studies have shown that gnawing by large canines can result in the fragmentation and destruction of particular elements especially the ends of limb bones, probably because these retain bone grease (Blumenschine 1988; Marean and Spencer 1991) Much of the material being examined in this study would have still been articulated when deposited. Therefore, if the material was naturally formed, it would not have progressed as far as the end of disarticulation Stage 4 (for cattle) (Figure 2.2). In the natural environment articulated skeletons are rare, often the result of natural traps of specific sediment activity (Lyman 1994, 135-136). Schäefer (1972, 10) noted that:

Butchery also has a social dimension. The basic requirement of butchery is to remove the primary products, such as meat and marrow from a carcass for human consumption. The technology and associated human societies may have changed over time, but not this basic requirement. However, because humans are not just functional animals, butchery can also be seen to change due to social factors, ‘the Nunamuit butchered animals one way and the Navajo do it another’ (Binford 1978, 47).

‘..usually enough time has passed for the skeletal parts to have lost their original connections or for external agents to have separated them. Hence there are only two ways in which complete skeletons of vertebrates can be preserved: either quick burial in areas of rapid sedimentation or

20

Chapter 2. Creation of the Faunal Record lower foot bones (phalanges and metapodials) and the skull. These bones may be removed with the skin, as the lower feet and skull can break away from the carcass with the skin by applying a small amount of pressure. Such practices are assumed for the Neolithic head and hoof burials (Piggott 1962a; Robertson-Mackay 1980). Alternatively, if the skin was a priority the animal may be carefully skinned completely resulting in cuts around the phalanges, mandible and skull, with these elements remaining attached to the carcass (Binford 1981, 103, 107).

Although butchery can differ between societies, there are a number of major stages that are normally progressed through when a carcass is fully butchered. In a modern setting all these stages are carried out, but it would be inadvisable to assume that the same always occurred in the past. Lyman (1987) has suggested that there are three main stages of butchery, kill-butchery followed by possible transport of carcass portions, secondarybutchery with redistribution within the site of consumption and the final butchery-consumption stage. Although based on Palaeolithic archaeology, as are a large proportion of butchery studies, Lyman does raise the important point that the different stages of butchery will not necessarily all take place at the same time, or in the same place. This is evident in the Romano-British period with cattle carcass elements being spread across towns for specialist processing (Dobney 2001; Grant 1989b; Maltby 1989d; 2007). Rixson (1988) highlighted five stages of butchery; 1.

Primary butchery – including slaughter, skinning and evisceration.

2.

Secondary butchery – including main dismemberment of the joints and/or main muscle sites.

3.

Tertiary butchery – processing of the main elements into smaller portions suitable for cooking.

4.

Utilisation for marrow – involving breaking of bones to extract the marrow.

5.

Once skinned, we could expect the next step to be the evisceration of the carcass, with some of the internal organs being utilised dependant upon social tastes and requirements. Once this has been completed, the carcass may then undergo secondary butchery. The skin may be transported to another area for processing, possibly taking the head and foot bones with it. 2.5.2

Once the animal has been skinned and gutted the carcass can be utilised for its meat. This is of course assuming the animal requires skinning. Some smaller animals may not require such processing and could be cooked whole or with the skin still attached. For example, a hog roast will require just evisceration and the carcass can then be cooked with the skin still attached. Assuming the animal has underdone primary butchery the next step would be to process the carcass. This initially may involve dismemberment of the carcass into more manageable portions that may then be transported elsewhere. For example, on the Medieval site of Launceston Castle (Albarella and Davis 1996) the main elements present for red deer and fallow deer are the haunches (back leg). This may indicate that the deer were butchered at the kill site with the haunches being transported back to the castle and the remaining parts of the carcass redistributed amongst the lower status huntsmen, foresters or parkers (Sykes 2005). Alternatively the resulting meat units from carcass dismemberment are distributed across a settlement, possibly in line with social values such as rank (Grant 2002).

Bone working – the transformation of the bone to artefacts.

This project is mainly concerned with the primary to tertiary stages as the later processes will be very unlikely to result in the creation of an ABG. 2.5.1

Secondary butchery

Primary butchery

The first activity to take place would normally be the skinning of the animal. The skin is not an edible product, but can be used for clothing etc, and is therefore of value. Also to get to the meat, the skin requires removal. Therefore an animal will be skinned before being further processed. In modern butchery, butchers start at the head and work their way down the underside of the carcass. Similar ethnographic practices have also been noted (Jones 1980) There are few places where skinning can result in leaving a butchery mark, the two most likely places being the head and feet (Binford 1981, 107; Shipman 1986), although Binford’s (1981) assumptions have recently been questioned (Abe et al. 2002). From a number of different societies cuts have been noted on the lower limb bones, mandibles and skulls, (Binford 1981, 107; Crabtree 1989b; Dobney et al. 1996; Maltby 1985e; Wilson 1978). These are often interpreted as cut marks from skinning the animal. This has consequences for the

Binford (1981, 91-126), investigating dismembering strategies of the Nunamiut Eskimo, drew on evidence from a number of other ethnographic studies (Binford 1978; Gifford 1977; Yellen 1977a;1977b). Binford (1981, 91-92) ascertained that several dismemberment practices are common across a number of groups (Figure 2.3);

21

x

Disarticulation of the head from the neck, and neck from the vertebral column

x

Separation of the front and back legs from the axial skeleton

Investigating Animal Burials; Ritual, Mundane and Beyond x

the radius/ulna, with each bone and corresponding meat, representing a ‘cut’. Also vertebrae and ribs may be dismembered into groups. It was noted by Binford (1981, 91-92) that all groups generally butcher the spine and ribs into smaller units, some keeping the vertebrae and ribs still attached. Unfortunately the vertebrae and ribs are bones which are often neglected in butchery studies.

Separation of the lower feet from the legs in most cases

The studies Binford drew upon are mainly of huntergatherer communities, and therefore some of his findings may not be applicable to some of the periods in this study. However, the majority of butchery recorded from the ethnographic record he used, were created by knives, which appear to be the main tool utilised in the late prehistoric period covered in this study. Tools such as the cleaver and saw do not appear to have been greatly used until the Romano-British period (Maltby 2007). Butchery in the Neolithic and possibly in some later prehistoric periods may sometimes have been carried out using flint implements (Humphrey 2003), although more work is needed on this subject. However, the butchery capability of flint tools and metal knifes are very similar (Jones 1980).

If such further dismemberment was taking place we could expect the possible production of ABGs, for example articulated vertebrae and ribs. However this is dependant upon what happens to the meat units and the cooking processes used. If further processing for marrow, or for material to make bone tools takes place, we could expect the complete fragmentation of the meat unit and therefore no ABG will be produced. It will also depend upon the society utilising the carcass, as butchery practices can be seen to differ between societies and periods. These differences can in turn affect the faunal assemblage.

The butchery of hunter-gatherer groups is also influenced by the need to transport the carcass to their base camp. This would probably not have been necessary for many of the domestic animals utilised in the periods covered by this study, which may have been slaughtered close to the settlement, or brought to it. Some hunting did take place, possibly even of domestic animals, as indicated by the flint arrowheads embedded in domestic pig bones from the late Neolithic site of Durrington Walls (Albarella and Serjeantson 2002).

2.6

Faunal material can enter the archaeological record due to human activity via two main ways; burial and deposition. Although very similar burial and deposition are two different events which can affect faunal remains in different ways. For this project burial is defined as the deliberate placement of material (in this case faunal) in a defined feature, with the deliberate covering of the material with sediments.

It is possible that some aspects of secondary butchery could have been skipped. Once skinned the meat may have been filleted straight from the carcass, possibly resulting in the deposition of a complete skeleton. However to access all the meat and to obtain marrow it would have been necessary to dismember the carcass. Also the butchery marks recorded from archaeological sites in the prehistoric period are similar to those Binford (1981) recorded and seem to indicate dismemberment of the carcass into portions at sites of natural disarticulation (Wilson 1978, 137). 2.5.3

Ground to trowel

Deposition can be defined using Lyman’s (1994, 406) description of; ‘dynamic placement either on a land surface or in an existing sedimentary unit.’ Therefore faunal material which is deposited can be placed in a pit, but not covered and may be subject to taphonomic effects such as weathering and trampling. However faunal material which is buried will be covered over soon after been placed in the ground. Burial in effect creates another archaeological context, as does deposition, but bones which are deposited can become sedimentary clasts. That is they become a part of the sedimentary context (O'Connor 2003, 207) , which most faunal remains from archaeological sites are. The faunal material comes from a context, and the bones are spread throughout the sedimentary material which constitutes the context.

Tertiary and further butchery

Tertiary butchery involves further dismemberment of the carcass into what might be termed cuts. These are smaller portions of carcass suitable for cooking. Further processing into cuts may not be necessary if the meat is filleted off the bone, and is also dependant on the size of the animal. Our modern view of meat is one of certain ‘cuts’ normally no longer attached to their corresponding bone. Today’s standard cuts of meat have developed from those used during the Medieval period, as tastes, fashion and equipment have changed.

Bones which are deposited may therefore suffer from fragmentation and erosion due to weathering and trampling. They may also be still accessible to scavengers and suffer damage from gnawing. This is demonstrated by the poorer preservation of faunal material in the top third of pits from Balksbury (Maltby 1985c; 2001).

Any butchery marks left on the bone are dependant upon the nature of the cut. It may just involve the dismemberment of two bones, such as the humerus from

22

Chapter 2. Creation of the Faunal Record

Figure 2.3 Dismemberment practices of different ethnic groups (altered from Binford 1981, Figure 4.01) crushing (Lyman 1994, 423-432). If an ABG did suffer from fragmentation or crushing due to overburden, then we could expect some of the fragments to survive as conjoining pieces. But as Lyman and O’Brien (1987) note, fragmented bones tend to be analytically absent, because as the fragment size decreases so do the bones’ diagnostic anatomical marks. If the ABG becomes fragmented, bioturbulence, the movement of soil due to animal and plant action, may also become a factor, with smaller fragments more likely to be moved out of place by bioturbulence such as earth worm activity (Stein 1983).

Therefore if an ABG is deposited and not buried, disarticulation may still occur due to the same factors as discussed above. If taphonomic effects are recorded upon the ABG it could indicate that it was deposited and not buried. However, it is at present not possible to identify whether the damage to the bones occurred before or after final deposition. The ABG may have been placed above ground in a midden before being deposited in a pit. But using Lyman’s (1994, 406) definition of deposition, placement above ground in a midden before being placed in a pit, are all part of the same depositional process. The remains may eventually be subject to secondary deposition within a pit or other feature, thus becoming sedimentary clasts within a feature’s fill.

Soil movement such as slumpage, caused by the rotting of organic matter placed within a feature may also result in the disarticulation of an ABG. The Romano-British well at Oakridge, Hampshire is a good example of how overburden and slumpage can affect a faunal assemblage. The well measures just over 26 metres deep and shows good evidence of slumpage in the top section (Figure 2.4). The faunal remains from the lower contexts were highly fragmented probably due to the effects of overburden (Maltby 1988; 1993). At Oakridge it was possible that the final stage of large scale deliberate dumping may have originally filled the feature, but due to slumpage the material may have sunk by as much as 50 feet (Maltby 1993; Oliver 1992), almost half the depth of the feature. Maltby (1993, 55) was very clear on what effect this would have on the faunal assemblage;

Once buried in the ground faunal material is subject to what could be defined as diagenesis (Lyman 1994, 417). The chemical makeup of the sediments the bones are within can greatly affect the preservation, with the sediment pH being a major factor (Gordan and Buikstra 1981). The ideal pH for the preservation of bone is 7.8 to 7.9 (Reitz and Wing 1999, 117). Therefore in certain geologies faunal remains will not be well preserved. It has long been noted on British archaeological sites that certain regions, such as the chalk downlands have better bone preservation than other regions with a more claybased geology, due to the pH levels of the soil. Therefore some regions will produce a ‘better’ faunal ABG record and the local geology will need to be taken into account when looking at their spatial distribution (see 3.2).

‘It is important to recognise that the ultimate resting place of the bones was largely determined by the substantial post-depositional slumping of the fills.’

Another factor that can affect faunal remains is sediment overburden. The weight of the soil on top of the faunal remains can resulting in deformation, fragmentation and 23

Investigating Animal Burials; Ritual, Mundane and Beyond A large number of ABGs were present within the well, but the majority of them had become separated due to the slumpage, causing intermingling of elements and making it impossible to assign all the elements to an individual (Maltby 1988; 1993). Therefore a feature may contain a number of ABGs but due to the post-burial/deposition taphonomic effects the ABG may become disarticulated and therefore possibly not spotted by the archaeologist. This has consequences for this project, as there are many occasions where such disturbed ABGs were only noted by the zooarchaeologist during the post-excavation stage. 2.7

place. Also during excavations features may only be half sectioned, thus limiting the possibility of ABGs being discovered from a pit, and if an ABG is discovered in the pit, associations may not be spotted. One major limiting factor for the study of ABGs is data. Once the faunal remains, including ABGs, have been recovered they will often, if sample size and finances allow, be sent to the zooarchaeologist. The zooarchaeologist can be seen as one of the final taphonomic factors. This includes their ability to identify bones, their methodologies and any constraints which may be placed upon them, including the amount of space they have available to publish the data.

Trowel to report

The final taphonomic stage the faunal material has to undergo is the archaeological process. Once an archaeological site is discovered or undergoes required excavation, a number of limiting factors are placed upon the site by the archaeologists. The skill of the excavators to spot and recover ABGs is essential. Also the treatment of the faunal material after excavation can affect the data, for example severe cleaning of the bone could destroy butchery and pathological information. Often only a small percentage of the site is excavated, meaning an incomplete picture may be gathered. However it would be inadvisably negative to see this as a major limiting factor, as all sites are subject to some form of archaeological sampling. It is however important to be aware of the issue when investigating the spatial distribution of deposits.

2.8

Summary

Faunal remains are not as simple as it was once thought. Bones present in archaeological assemblages do not equal the proportion of animals eaten. The faunal material undergoes a number of different taphonomic processes before the zooarchaeologists get their hands on them. The processes briefly discussed above have a limiting effect upon the faunal assemblage, meaning that the faunal assemblage is but a small proportion of the living assemblage. However we should not view this in a negative light. There may well be issues with our data, but we also have the means to identify when an assemblage has been biased. Also, some of those taphonomic factors are created by the past cultures which we are studying . Therefore the study of taphonomy is also the study of past human activity upon an assemblage, such as butchery and deposition discussed above. The human agents and their choices affect the faunal data and are, for the majority of ABGs found upon settlements, the primary factor in their creation. This point is discussed further in relation to the interpretation of ABGs.

One area of major research especially in the 1970’s and 1980’s concerned bone recovery. Payne (1972; 1975), showed that smaller fragments of bone can be missed if the spoil is not sieved. However as this project is concerned with articulated remains, sieving is not a major limiting factor when investigating the nature of ABGs. Their discovery is much more dependant upon being collected by hand and therefore being spotted in the first

Figure 2.4 Section of the top 4 meters of the Oakridge well (Oliver 1992, Figure 7).

24

Chapter 3. Neolithic and Bronze Age 3.1

Introduction

In this section the data regarding ABGs from Neolithic (4000-2500BC) and Bronze Age (2500-700BC) sites are examined. The two periods are discussed together due to the nature of the archaeological record. Although within this report and within archaeology in general, time has been divided up into a series of convenient bundles, chronological boundaries are often fuzzy. Generally, the start of the Bronze Age is taken as 2500 BC, when we see the introduction of the ‘Beaker Culture’ to the British Isles. With the ‘Beaker package’ the use of metals appeared alongside new pottery forms and burial styles (Parker-Pearson 1993, 84-85). However, such changes were not instantaneous, cultural change may have occurred slowly, with some aspects of the ‘Beaker package’ were present during the Neolithic, such as the development of round barrows (Kinnes 1979; Woodward 2000, 37). The last two decades of archaeological work has shown that there was no clear break in cultural or economic life. Change did occur, but the process was a long term one (Barber 2003, 11). Although these periods have the greatest time spans, they have produced a small number of sites compared with later periods. Only 27% of the sites recorded for the project come from these two periods. In total, 140 sites were recorded, with most sites 88% (123) being situated in southern England (Table 3.1). The majority of the sites 72% (102) had no records of ABGs. Relatively more Neolithic sites (34%) have ABGs present compared to the Bronze Age (20%). The majority of Neolithic sites recorded come from Wiltshire. This is probably because the importance of Stonehenge and Avebury has prompted a number of research projects in their environs. There is a much more even geographical distribution of Bronze Age sites recorded, although the majority of these consist of round barrows with no associated ABGs. The relative number of ABGs found on Neolithic and Bronze Age sites is even smaller in comparison with other time periods. A total of 115 ABGs were recorded, representing only 5.5% of those discussed in this project. Or, to put it into perspective, the total number of Neolithic and Bronze Age ABGs is less than the number of ABGs retrieved from a single well at the RomanoBritish site of Oakridge (Maltby 1993). This is perhaps a Table 3.1 Summary of the number of sites with ABGs for the Neolithic and Bronze Age. The bracketed figures indicate the figures from the Yorkshire region Not Total Total. Period Present Present Sites ABGs Neolithic 14 (1) 28 (8) 42 (9) 54 (1) Bronze Age 24 72 (7) 96 (7) 61

surprising result, considering how influential past interpretations of Neolithic deposits have been (see 1.2.3). However, this must be viewed within the context of the size of the faunal assemblages. Windmill Hill in Wiltshire has produced one of the largest faunal assemblages from the Neolithic, with 1,676 identified fragments. Most Neolithic assemblages are much smaller. In later periods the Windmill Hill assemblage would be considered to be very small. For example, the RomanoBritish assemblage from Greyhound Yard, Dorchester amounts to 18,138 identified domestic mammal fragments (Maltby 1993). In addition, we must consider that the Neolithic and Bronze Age periods have much fewer settlement assemblages than in later periods. If we split the data between the two research areas, 5.8% of the southern England information comes from these two periods, compared to only 0.5% of the Yorkshire data. This corresponds to the small amount of faunal data from the Neolithic and Bronze Age recorded from Yorkshire. Only one Neolithic site and none of the Bronze Age sites from Yorkshire had ABGs present. 3.2

Neolithic and Bronze Age Yorkshire

Whitegrounds, Barrow 1 (Riggott and Williams 1984), was the only site, discovered during this study, with an ABG deposit present. The site consists of an oval cairn, with a walled entrance passage, closed by a stone infilling. The associated human remains are some of the earliest dated from East Yorkshire, 5040±100 BP (40403640 cal BC) and 5260±200BP (4510-3640 cal BC) (Brewster 1984, 17). The site is part of a series of ‘nonmegalithic long barrow’ structures recognised to have been constructed before long barrows (Kinnes 1992; Manby et al. 2003a). A Bronze Age round barrow was later constructed over the tomb. The ABG consists of a partial fox skeleton (positively identified by the dentition), found on the floor of the cairn under a flint nodule and the legs of a decapitated juvenile human skeleton (Figure 3.1). The juvenile skeleton is thought to be the first burial placed in the entrance grave. The excavation report suggests the fox may have been a pet, due to its old age. It was argued that the human burials would have been more disturbed, if the tomb had been used as a foxes’ den (Riggott and Williams 1984). The paucity of ABGs from the Neolithic and Bronze Age of Yorkshire, compared with southern England, may simply reflect the small amount of faunal material recorded from the area. Stallibrass (1995b), in her review of animal remains from northern England, bemoaned the lack of data from both periods. She pointed out that, although large quantities of animal bones may have been present at many sites, they were not collected or curated, as many sites were investigated in the nineteenth and early twentieth centuries.

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 3.1 Plan of the burials present in the Whitegrounds entrance grave. The arrow indicates the area of the fox ABG (Riggott and Williams 1984, Figure 19). Table 3.2 Number and percentage of ABGs per species for the Neolithic and Bronze Age periods. The Sheep/Goat, (S/G) counts include all ABGs identified to either species Early Middle Late Early Middle Late Bronze Bronze Bronze Species Neolithic Neolithic Neolithic Age Age Age Total Cattle 19 (58%) 5 (31%) 4 (80%) 4 28%) 5 (20%) 13(60%) 50 S/G 4 (12%) 5 (35%) 20 (80%) 3 (13%) 32 Pig 3 (9%) 6 (38%) 1 (4%) 10 Horse 1 (8%) 1 (4%) 2 Dog 6 (18%) 1 (20%) 3 (21%) 4 (19%) 14 Roe deer 5 (31%) 5 Wild Cat 1 (3%) 1 Sea Eagle 1 (8%) 1 Total 33 16 5 14 25 22 115 (Maltby 1990c) and a partial wild cat from Windmill Hill (Pollard 1999). The only wild fauna from the Bronze Age are the remains of the white-tailed eagle from the Coneybury Henge, Wiltshire (Maltby 1990e; Richards 1990, 123-58). The eagle was deposited in the primary fill of the southern ditch. Other bones within the ditch were dated to 3100-2450 BC, placing the deposition at the late Neolithic-early Bronze Age boundary. As is the convention for this study the latest date is utilised. No other Neolithic or Bronze Age sites produced wild mammal or bird ABGs. The Coneybury Anomaly produced several bones of a trout/salmon, which may well have been from one fish, but there was insufficient evidence to determine whether they formed a closely associated group (Maltby 1990c). Some other sites, such as the late Neolithic enclosure at Thomas Hardye School, Dorset (Smith 2000) have a number of antler deposits, which have been interpreted as ‘special’ offerings, but such deposits do not fall within the scope of this project.

In addition, the underlying geology of many sites is not conducive to good bone survival. The faunal evidence from Yorkshire is biased towards the calcareous conditions found in the Magnesian limestone belt, Craven and Hambleton Hills and some of the lowland peat and silts. Most of the lowlands and Pennines sites have unfavourable conditions and produce very small faunal assemblages (Gaunt and Buckland 2003; Manby et al. 2003a). As only one ABG for these periods comes from the Yorkshire region the following analysis concentrates on deposits from southern England. 3.3

Neolithic and Bronze Age species proportions

Domestic animals dominate both the Neolithic and Bronze Age ABG assemblages, making up 89% and 98% respectively. The Neolithic wild faunal remains consist of five roe deer from the Coneybury Anomaly pit, Wiltshire

The overall Neolithic ABG assemblage is generally dominated by cattle. However, this changes in the Bronze

26

Chapter 3. Neolithic and Bronze Age dog deposits were recorded from Neolithic sites, five of which came from Windmill Hill. The inclusion of ABGs in the NISP counts for the Early Neolithic Windmill Hill faunal assemblage mean that dog elements account for 10.9% (241) of the identified remains (combining the assemblages from the 1925-29, 1957-58 and 1998 excavations). Unfortunately there is not enough published information to calculate the NISP figures excluding ABGs. However, all the dog ABGs appear to have been discovered during the 1925-29 excavations and were included with the rest of the available faunal material in Jope’s (1965) analysis of the 1957-1958 material. Therefore, if we compare the dog NISP counts for the earlier excavations with the 1988 results we can get a picture of the proportion of disarticulated dog remains. Of the 241 fragments, 226 come from the earlier excavations. Only 15 were discovered during the 1988 excavations, representing only 1.4% of that assemblage. If the 1988 assemblage is typical of the low percentage of disarticulated dog remains represented at Windmill Hill, it strongly suggests that the majority of the dog remains from the earlier excavations are from ABGs and their inclusion has inflated their NISP percentage. More importantly, it also shows that dog ABG deposits from this site are much more common than would be expected from the disarticulated faunal assemblage. This is a pattern seen for dogs in all of the time periods examined in this study.

Age, with a large increase in the number of sheep/goat (Table 3.2). This change mirrors what is seen in the ‘normal’ (non ABG) faunal assemblage, where sheep/goat are fairly poorly represented on the majority of Neolithic sites, with cattle the most common species in sites from the early and middle Neolithic (Figure 3.2). In the late Neolithic there is a rise in the exploitation of pig (Grigson 1982), as demonstrated from sites such as Durrington Walls. During the early Bronze Age there is a decrease in the relative number of cattle and pig bones recovered from archaeological sites, and sheep/goat (particularly sheep) appears to become the most commonly exploited domestic animal. However, this is a simplistic overview and there are many inter-site differences. For example, the late Bronze Age faunal assemblage from Potterne, Wiltshire (Locker 2000) contains an unusually high proportion of pig remains. Two important factors also need to be considered when discussing the non-ABG species proportions. Firstly, as explained above, the quantification method, be it NISP or MNI can affect the result (for example see O'Connor 2000, 54-67; Pilgram and Marshall 1995; N. P Winder 1991). Even in cases where sheep/goat or pig produced the highest number of individual bone fragments, cattle, due to their large size, may still have provided the most meat. In addition, the activities taking place on the site need to be taken into account. For example, the faunal remains deposited at funerary monuments such as long and round barrows may not be representative of ‘domestic’ waste deposited at an occupation site.

The proportion of dog ABGs is also higher in the Bronze Age sample than would be expected from the disarticulated faunal assemblages. For example, disarticulated remains of dog account for only 1% of the identified assemblage from the late Bronze Age enclosed settlement at Dean Bottom, Wiltshire (Maltby 1992). Dogs also make up a similarly small proportion of the disarticulated faunal assemblage from a number of other sites including Bishop Cannings Down (Maltby 1992), Potterne (Locker 2000), Milton Lilbourne Barrow 4 (Grigson 1986) and Avebury Barrow G70 (Christie 1964).

Cattle during the earlier Neolithic and sheep/goat during the Bronze Age are the most common species overall from both ABG and non-ABG assemblages. There are sub-period differences, with cattle ABGs the most common in the later Bronze Age, which does not correspond with the overall faunal assemblage. However, this may be due to the influence of Poundbury, Dorset, which has eight cattle deposits dating to this sub-period. Species proportions also differ between the site types. The remains of pig and dog best represent this point. For the Neolithic assemblages, pig ABGs make up 15% of the total sample from the south of England. This is roughly in-line with the proportion of pig remains seen in disarticulated faunal assemblages from the period. However, no pig ABGs were recorded on the late Neolithic to early Bronze Age sites such as Durrington Walls (see below). In the Bronze Age the proportion of pig remains in both the disarticulated and ABG assemblages falls. In fact only one pig deposit has been recorded, discovered at the late Bronze Age settlement at Bell Street, Romsey, Hampshire (Coy 1993).

The species proportions of the ABG assemblage indicates that different choices/processes were taking place regarding which species were deposited in this form compared with the disarticulated faunal assemblage. Dog deposits are more common than would be expected. Conversely, the proportion of pig is less than in the disarticulated faunal assemblage. Also, the proportions of Bronze Age cattle and sheep/goat differ from those found in the disarticulated assemblage, with the percentages of sheep/goat ABGs being slightly lower and cattle ABGs higher than expected.

Dog are more common than would be expected if the ABG assemblage mirrored the disarticulated one. Seven

27

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 3.2 Percentages of cattle, sheep/goat and pig found on selected Neolithic and Bronze Age sites in southern England. Windmill Hill (Grigson 1999) and Maiden Castle Causeway (Armour-Chelu 1991) = early Neolithic; Durrington Walls (Stone et al. 1954) = late Neolithic; Middle Farm (Bullock and Allen 1997) = early to middle Bronze Age; Dean Bottom (Maltby 1985b) and Potterne (Locker 2000) = late Bronze Age. NISP sample sizes are in brackets 3.4

Neolithic site types

Funerary monuments, in this case long barrows, also produced a number of ABG deposits. Six cattle are present from three sites: one from Alington Avenue long barrow (Maltby 2002b), two from Maiden Castle bank barrow (Jackson 1943), and three from Fussell’s Lodge (Grigson 1966). Only cattle ABGs have been recorded at long barrows. The remains of cattle also dominate the disarticulated assemblage from these sites.

Species proportions also vary between different types of Neolithic and Bronze Age site. This section discusses the differences and similarities on different types of sites dating to these periods. The two most common sites with ABGs present are causewayed enclosures and pit complexes. Causewayed enclosures are a relatively well known site type (Oswald et al. 2001). ‘Pit complexes’ is a generic term that refers to sites where the majority of the features recorded are pits. They may in some cases such as the Coneybury Anomaly (Richards 1990), consist of one isolated pit, or like Rowden Pasture (W4) a number of pits (Woodward 1991).

Nearly all the Neolithic ABG deposits come from causewayed enclosures, pit complexes or funerary monuments. Absence data shows a different pattern. The most striking difference is that all causewayed enclosures recorded in this study have produced ABGs. However, they are absent from 47% of the funerary monuments and 42% pit complexes from southern England (Table 3.4). The other sites with faunal remains present but no ABGs, are the Stonehenge Lesser Cursus (Maltby and Richards 1990) and a number of isolated ditch complexes around Durrington Walls (Hamilton-Dyer 2004a) and Poxwell, Dorset (Jones 1986).

The majority of the ABGs dating to the Neolithic come from causewayed enclosures (Table 3.3). This is mainly due to the dominance of the Windmill Hill assemblage. Not only did the excavations at Windmill Hill produce one of the largest disarticulated faunal assemblages from this period, but it also produced one of the largest ABG assemblages. Of the 54 deposits from this period (not including ABGs dating to the late Neolithic-early Bronze Age), 26 (48%) come from Windmill Hill. The relative frequency of species from pit complexes is very different to those found on causewayed enclosures. Pig and roe deer are the most common ABGs from these sites. However, the data are dominated by two sites. Five of the pig deposits are from pits excavated at the Old Sarum Spur site (Powell et al. 2005) and the five roe deer are from the Coneybury Anomaly (Maltby 1990c). Finally at Silbury Hill, Wiltshire, a partial dog, consisting of articulated vertebrae was discovered in the build-up layers of the mound (Gardner 1997), (Recorded as site type ‘Other’).

3.5

Neolithic composition

Of the 54 ABGs recorded, only 6 consist of complete skeletons, all, perhaps significantly, from causewayed enclosures. One of the complete deposits is a dog skeleton found during excavations at Maiden Castle (Jackson 1943). A complete juvenile sheep were recovered from Whitesheet Hill (Maltby 2004a). Although the majority of the foot bones are missing, Maltby suggests their absence may be due to sieving not being utilised during recovery. The remaining complete

28

Chapter 3. Neolithic and Bronze Age Table 3.3 Number of ABGs per species recorded on different Neolithic site types. The number of sites is indicated in brackets. Causewayed Funerary Pit complexes Species enclosures (5) monuments (3) (4) Other (1) Total Cattle 22 6 28 S/G 3 1 4 Pig 3 6 9 Dog 6 1 7 Wild Cat 1 1 Roe Deer 5 5 Total 36 6 12 1 54

Table 3.4 Total number of Neolithic to Bronze Age site types with no ABGs present for southern England and Yorkshire. The number of Yorkshire sites is indicated in brackets Late Neolithic Site Type Neolithic Early Bronze Age Bronze Age Cursus 1 1 Ditch complex 1 1 3 Enclosure 4 (1) 2 Henge 4 (1) Funerary monuments 12 (2) 1 Other 1 Pit complex 14 (5) Ring ditch 3 Round barrow 3 43 (6) Rural Settlement 6 Total 28 15 57

Table 3.5 Formation of partial ABGs by species and site type for the Neolithic period (information not available for all sites) Site type Species Axial + head Axial Axial + leg Leg Causeway enclosure Wild cat 1 Cattle 7 1 14 Dog 2 1 Pig 1 1 Sheep 1 Funerary monuments Cattle 1 1 4 Other Dog 1 Pit complex Pig 1 1 1 3 Roe deer 2 3 S/G 1 Total 3 16 3 27 vertebrae from a cow found in Inner ditch XVI. The third a pigs pelvis and back legs found in pit 3020 at the Old Sarum Spur site.

ABGs come from Windmill Hill, and consist of the skeletons of two dogs, a calf, a pig and a goat. Table 3.5 shows the main body areas that constitute partial ABGs from the Neolithic period. The majority of them consist of elements from the limbs or the axial skeleton only. There are only three deposits which contain elements from both the axial and appendicular skeleton. Two of these come from Windmill Hill. One consists of the back legs and associated tail vertebrae of a wild cat, although the pelvis appears to have been missing (Grigson 1999). The second is comprised of the back legs, pelvis and

The rest of the partial ABGs consist of, limb elements, vertebrae/ribs, or skulls with vertebrae. Deposits consisting of the skull associated with vertebrae are the rarest. As already discussed, isolated skull deposits are not included in this study, although they are present on a number of Neolithic sites (Russell 2002, 33). For example, five incomplete cattle skulls were recovered 29

Investigating Animal Burials; Ritual, Mundane and Beyond from other sites are too small to be able to compare ABGs to this level of detail.

from Maiden Castle (Jackson 1943). With the exception of dogs, limb deposits are the most common body area represented for each period. The majority of the dog partial ABGs consist of axial elements. This may be because axial elements are in closer association within the body, as only limited movement within their joints is required. Therefore appendicular elements may become disassociated more easily, assuming that compete dog skeletons are being deposited (see 2.4). However, as appendicular elements are common for other species, it seems unlikely this is the case. It would seem that different body parts of dogs were being deposited compared with other domestic mammals. However, we should be cautious with such conclusions due to the small sample size.

The composition of the deposits from funerary monuments, in this case long barrows, is worth further consideration. As discussed above, the majority of funerary monuments do not have ABGs present. When present, they appear to be mainly recovered from the ditches around the monument and are mainly composed of axial skeleton elements. For example, within the ditch at Fussell’s Lodge a cattle ABG was deposited, consisting of vertebrae and ribs. However, this deposit may not have been recovered in articulation. The report describes how some of the vertibra were in the incorrect anatomical positions (Grigson 1966). The two other ABGs from Fussell’s Lodge are very different. They were discovered within the long barrow chamber, possibly in association with human remains.

Windmill Hill provides us with the largest sample for further investigating which elements make up partial ABG deposits. Figure 3.3 shows the combined body area percentages for the 16 partial cattle from Windmill Hill. In this case it was not possible to separate out the different vertebrae types, as the required information is not available from the Keiller excavations. Although the figure has all the vertebrae highlighted, in most cases only a small number of vertebrae were recovered. For example, an ABG from ditch D, context 413, consisted of just three thoracic vertebrae. The results show that the majority of ABGs contained elements either from the vertebral column or the lower limbs, with the front lower limb more common than the back. The majority of partial cattle ABGs from Windmill Hill do not consist of meatbearing upper limb bones. Unfortunately the sample sizes

Both consist of lower limb elements (Figure 3.4). Early antiquarians such as John Thurnam (1868, 182) noted that the bones of cattle from within long barrows were frequently those of the skull and feet. He suggested that the carcasses had been eaten and the head and hooves held together by the skin were then placed in the barrow. Such deposits became known as ‘head and hooves’ burials (Ashbee 1966; 1970, 75; Piggott 1962a). Ashbee (1966) suggested that the ABGs from within Fussell’s Lodge represented such a deposit because of a close by cattle skull. However, there is no evidence to confirm that the skull came from the same animal as the foot bones. However, the presence of a caudal vertebra and the

Figure 3.3 Percentage of body parts present, from 16 partial cattle ABGs from Windmill Hill

30

Chapter 3. Neolithic and Bronze Age sesamoids (displayed on either side of the metapodials in Figure 3.4) would suggest that the skin was still present for at least one of the Fussell’s Lodge deposits. Such ‘head and hooves’ burials appear to be rare within the study area as the examples from Fussell’s Lodge are the only ones. However, the tradition may also have carried on into the Bronze Age at sites such as Bishop Cannings (Robertson-Mackay 1980). Unfortunately no butchery data are available for the Fussell’s Lodge deposits. In fact, none of the Neolithic ABGs have butchery data recorded. It is unknown whether this is due to a genuine absence of butchery or just lack of recording/reporting. 3.6

Neolithic - Bronze Age transition

Only a small number of ABG deposits date to the late Neolithic or early Bronze Age period. As discussed above, the boundary between the Neolithic and Bronze Age is not a clear one, with sites such as Coneybury Henge and the possible enclosure ditches at Alington Avenue, Dorchester, dating to this period. The majority of the sites recorded as dating to the late Neolithic- early Bronze Age are round barrows.

Figure 3.4 Cattle ABGs from within Fussell’s Lodge long barrow (Ashbee 1970, Plate XXXIIb).

Eleven ABG deposits were recorded from this period, making up just 0.5% of the overall assemblage recorded for this project. Cattle and sheep are the most common (Table 3.6), with the majority of these being recovered from round barrows. Only one is recorded from the ditches at Alington Avenue. It consists of a partial dog, represented by 12 elements from the head and a front limb (Maltby 2002b). A small number come from other monument types. The partial skeleton of a cow, represented by 8 elements was found in a pit, associated with a possible late Neolithic enclosure at Thomas Hardye School, Dorset (Smith 2000). At the Marden enclosure, Wiltshire, a partial sheep, consisting of a front limb, was also discovered (Harcourt 1969b; 1971c).

that their past ranges included inland areas. They are opportunistic feeders and have been recorded a feeding on small mammals, carrion, and acting as kleptoparasite (stealing food from other animals). They have also been recorded travelling and nesting far inland away from water sources (Wille and Kampp 1983). The bird is one of the largest found in Europe with a wing span of over two metres. Currently this is the only ABG recorded for this species from the study area. However, other members of the raptor family are found as ABG deposits, including buzzards from the Iron Age and Romano-British periods (see 4.2.1 and 6.2.1), as well as hawks and falcons from the late Anglo-Saxon and Medieval periods (see 8.2.1). Also Neolithic white-tailed eagles have been found, famously, in the ‘Tomb of the Eagles’ Isbister (Hedges 1983; 1984).

A partial male dog was recovered from the southern part of the ditch at Coneybury Henge. This consisted of 52 elements, from the axial and appendicular skeleton. Also found within the henge ditch was a partial white-tailed eagle (Haliaeetus albicilla), with 13 elements from the thoracic region and left wing (Maltby 1990e). Both deposits probably date to the late Neolithic as animal remains from the primary fill were radiocarbon dated to 3100-2450BC, although Beaker pottery was present in the upper fills of the ditch (Darvill 2007, 104). Interestingly, the shaft of the sea eagle’s humerus was distorted, with exostosis present, possibly indicating the presence of a healed break. This raises the interesting hypothesis that to survive such a break the sea eagle may have been feed and cared for?

Another interesting aspect is that Coneybury is the only henge recorded with ABGs present, especially considering that the deposition of animal remains on these sites is argued to be of a ‘ritual’ nature (ParkerPearson et al. 2006; Pollard 1995; Richards and Thomas 1984; Thomas 1999, 80-83). However, current excavations at the eastern entrance to Durrington Walls are beginning to recover ABGs in the form of articulated pig vertebrae and foot bones (Parker-Pearson et al. 2006; Parker-Pearson et al. 2007, 10; Parker-Pearson et al. 2005, 27, 37 and 66). As well as henges, three other enclosures were recorded with faunal remains present, but no ABGs. The Sanctuary at Overton Hill (Rouse 2001), the palisade enclosure at West Kennet (Edwards and Horne 1997) and the possible timber circle enclosure at Greyhound Yard, Dorchester (Maltby 1993). Although

Today white-tailed eagles are mainly found in coastal regions, but their distribution has changed and it is likely

31

Investigating Animal Burials; Ritual, Mundane and Beyond The majority of Bronze Age deposits come from sites dating to the later part of this period. Enclosures, middens (particularly Potterne) and settlements make up 66% of the sites with ABGs present (Table 3.6). The earlier Bronze Age deposits come from very different site types, namely pit complexes, round barrows and shafts. This corresponds with the general divide seen in the archaeological record. From the beginning of the middle Bronze Age, c.1500 BC, there is an increase in the visibility of the places where people lived and farmed (Barber 2003, 12), and this is also seen in the site type data.

there are no ABGs present at Greyhound Yard, Maltby (1993) suggests that pig limb bones discovered within the packing layers of the post-pits may have been deliberately deposited for possible ceremonial/ritual reasons. The lack of ABG deposits from henge enclosures could be due to several factors. They are the only site type with publications predominately from earlier decades; therefore deposits may have been missed or not reported. However, all the other site types have publications of similar dates and do have ABGs reported. Recent excavations at Durrington Walls show that such deposits are present but it is perhaps significant that the majority are partial ABGs consisting of only a small number of elements, which are easier to miss, compared to complete skeletons. However, only five complete skeletons were recorded from Neolithic sites published between 1940 and 1969, which represents 25% of the ABGs reported during this time frame. Therefore partial ABGs on excavations were being recovered and reported at the time, suggesting that although recent ABGs have been found, the disparity of ABG deposits on henge sites may be a real phenomenon.

Some sites such as the pit complex at Old Sarum Spur, Wiltshire (Powell et al. 2005), have a continuation of use, with ABG deposits present in Neolithic (see above) and middle Bronze Age pits. Two partial cattle are present on the site in contexts 3344 and 3346. The report suggests the ABGs would have been deposited complete, but become disarticulated and fragmented due to modern ploughing. The other Bronze Age pit complex ABG comes from Easton Lane, Winchester (Fasham et al. 1989). This consists of 15 elements from the front limb of a sheep/goat from Pit 6053.

Most of the domestic mammal ABGs have been recorded from round barrows. These continued to be constructed and utilised well into the middle Bronze Age (Woodward 2000, 43). Therefore, the late Neolithic-early Bronze Age barrows are discussed in more detail below. 3.7

3.7.1

Round barrows

Although associated with the Bronze Age, Kinnes (1979) has shown that round barrows were present in the early Neolithic, developing from crematoria. However, round barrows, as recognised by the classic descriptions of Grinsell (1959), did not become prevalent until the late Neolithic and early Bronze Age. Round barrows make up 37% (46) of the sites recorded from the Neolithic and

Bronze Age

Table 3.6 Number of ABGs per species recorded from Late Neolithic-early Bronze Age site types. Other Enclosure Species Ditch complex (1) Henge (1) (2) Round barrow (3) Total Cattle 1 2 3 S/G 1 3 4 Pig 0 Horse 1 1 Dog 1 1 2 Sea Eagle 1 1 Total 1 2 2 6 11 Table 3.7 Number of ABGs per species recorded from different Bronze Age site types. * includes one pregnant ABG, mother and foetus counted separately Enclosure Midden Pit complex Round Settlement Species Shaft (1) Total (3) (1) (2) barrow (3) (6) 3* 1 2 1 16 22 Cattle 4* 1 4 15 24 S/G 1 1 Pig 2 2 Horse 1 1 2 1 5 Dog 8 2 3 3 22 17 55 Total 32

Chapter 3. Neolithic and Bronze Age

Figure 3.5 Plan of the pond barrow at Down Farm, the later discovered ABGs have been added in their approximate positions (hashed features) (adapted from Green 1982, Figure 1) left-hand sides with their legs flexed (Legge 1991b). The sheep were also deposited within shallow pits, at the margins of the bowl of the barrow. One was a complete skeleton of an old ewe. It was placed in the pit on its belly, with the hind legs pulled forward to either side of the body and the head twisted to the left. The excavator reports that it was only just big enough to fit into the pit. The second sheep deposit is a poorly preserved partial ABG, consisting of the head and front limbs. Legge (1991b, 75) suggests that it would have consisted of a compete animal when deposited, but was disassociated due to the poor preservation, and was not recognised in the field.

Bronze Age periods in the southern England study region. The majority of the barrows date to the Bronze Age, and make up 60% (40) of the sites recorded from southern England for this period. However, only four barrows dating to the late Neolithic-early Bronze age and two barrows dating to the Bronze Age have recorded finds of ABGs. Therefore it would appear that ABGs are a rare phenomenon on barrow sites, which is reflected in only 9 (13%) being recovered (Table 3.6 and 3.7). The majority of the ABGs from round barrows date to the late Neolithic-early Bronze Age, with nearly half recorded from the Down Farm pond barrow, Dorset (2 cattle and 2 sheep/goat). The two cattle deposits are complete adult skeletons that were placed within oval pits adjacent to the barrow (Figure 3.5), both placed on their

The positioning of the ABGs around the barrow appears to be unique. The pits containing the cattle are positioned 33

Investigating Animal Burials; Ritual, Mundane and Beyond It is suggested that the animal’s hide may still have been attached to the elements, and the deposit represents a hide bundle possibly placed as a grave good for the deceased. If this is the case, the bones may have been of no importance, and the hide could have been of more significance. It may also explain why no other ‘head and hooves’ burials from this period have been discovered so far, as other burials may only contain the hide and not the attached elements. Given the often ‘ritual’ interpretation of ABGs it is interesting that the majority of ABGs recorded for the Neolithic and Bronze Age are not from sites where clear ritual activity, in the form of human burial, took place, despite the large number of burial mounds known.

diametrically opposite each other, one on the south-east side, and the other on the north-west side of the barrow. The pits containing the sheep were positioned in a similar manner, one on the north side and the other opposite it on the south side. It appears that all the pits and presumably the depositions were created at the same time as the barrow. At the late Neolithic-early Bronze Age Winterbourne Stoke barrow 44, a sheep/goat ABG was also discovered within an oval pit, closely associated with the barrow. It consisted of the hind limbs and the majority of the axial skeleton. The scapulae were also present, both with butchery marks associated with the disarticulation of the front limbs (Green and Rollo-Smith 1984). A late Neolithic-early Bronze Age horse ABG was recovered from the ring ditch at Flagstones barrow, near Dorchester (Bullock and Allen 1997). Unlike the other deposits discussed above, this consisted of only a small portion of the animal, with four elements from the left front limb present. This is the earliest occurrence of a horse ABG recorded by this study. Unfortunately it is unknown if this is from a domesticated or wild individual.

However, other animal remains from contemporary sites outside the study area do point to some form of ritual activity involving the deposition of animals. The round barrows at Irthlingborough, Northamptonshire (Davis and Payne 1993) and Gayhurst, Buckinghamshire (Chapman et al. 2004), both contain very large deposits of disarticulated cattle remains. The Irthlingborough barrow contained 184 cattle skulls which were deposited within the make-up of the barrow. Detailed analysis of the faunal remains showed that a limited number of other elements were deposited on the site and that some of the skulls may have been exposed for a period of time. This suggests that the skulls were curated for a period of time before being brought to the site for deposition, rather than deriving from slaughtering and feasting in the vicinity of the site.

Three ABGs come from round barrows constructed during the Bronze Age (Table 3.7). However, the species recorded are different to those recovered from the late Neolithic- early Bronze Age barrows, which were mainly from cattle or sheep/goat. Two of the later barrows had dog ABGs recovered from them. One of these deposits comes from North Down Barn, barrow 23, Dorset (Grinsell 1959, 142). Limited information is given, but the ABG appears to consist of a complete dog skeleton, which was deposited in a pit of middle-to-late-Bronze Age date, dug into the barrow mound. The re-use of round barrows appears to have been relatively common, with many barrows showing remodelling of the mound with further burials (human, often cremations) inserted into them (for example Christie 1967; Donaldson 1977; Grinsell 1959; Woodward 2000). This, however, is the only example of the secondary deposition of an ABG within a barrow. Other ABGs appear to have been associated with the primary construction phases, although we cannot discount that the Down Farm pond barrow ABGs may be later. Although barrows may have fulfilled a number of important roles within the Bronze Age, for example as land/resource markers, they are primarily linked with human burial. Therefore it is perhaps surprising that from the study area the only ABG directly associated with human remains is the often quoted ‘head and hooves burial’ from Bishop Cannings barrow 81, Hemp Knoll, Wiltshire (Robertson-Mackay 1980) (Figure 3.6). The deposit consists of the four feet, skull and mandibles almost certainly from the same cow (Grigson 1980). The ABG is separated from the grave, unlike some of the closer associated grave goods, such as the beaker pot. The burial was placed in a possible wooden coffin, and the ABG was deposited within the backfill around the coffin.

Figure 3.6 Plan of the central grave at Hemp Knoll (Robertson-Mackay 1980, Figure 7).

34

Chapter 3. Neolithic and Bronze Age eruption. The other was possibly foetal, as the third and fourth metapodials were unfused, leading to the suggestion it may have been aborted and the neonatal animal may have been a stillbirth. All the other ABGs from the shaft consist of partial remains of different individual animals. These are either composed of appendicular (7) or vertebral remains (6) (Table 3.8). All but two of the appendicular ABGs consist of elements from the lower parts of the limbs.

Gayhurst is equally unusual, with the remains from at least 300 cattle being deposited in the ditch of the barrow, consisting mainly of skulls and upper limb elements. It is suggested that the cattle were slaughtered and butchered away from the barrow, then left to rot in a protected area to prevent scavenging, after which certain body parts were selected for deposition on the barrow (Chapman et al. 2004). However, another possible explanation could be that a feast took place at the barrow, for which people transported meat with them on the bone. Such possible transportation is indicated by the finds from the Hasholme Logboat, Yorkshire (Stallibrass 1987). Although it would appear that animal remains may have been used in some ritual fashion at round barrows, there are a limited number of examples. It would appear that the deposition of ABGs and large quantities of possible feasting waste was rare in both the study areas, although antiquarian investigations of hundreds of barrows in Wessex and Yorkshire may have destroyed such evidence, if it ever existed. 3.7.2

Interestingly, the ABGs have possibly been deposited in groups. Figure 3.7 shows the relative position of the ABG deposits within the shaft. As discussed above, the complete lamb ABGs were deposited close together. The two partial ABGs discovered above them consisted of cervical vertebrae from juvenile animals, and their proximity suggests they could have been deposited around the same time as each other. There is then a gap and the next ABG deposits consist of six partial sheep/goats. Five of these were discovered in context 1961:89, at 89ft 6in (26.9m). The deposits consisted of one upper-back limb from a juvenile animal, one upper fore limb, one lower fore limb and two sets of vertebrae remains, one from an adult the other from a juvenile. Slightly above in context 1961:88, at 89ft (26.7m) a lower hind limb ABG was also present. It is unknown if these ABGs were discovered articulated, but their associations were noted by the zooarchaeologist. Also deposited in these contexts were the disarticulated remains of at least six foetal/neonatal sheep/goat. It is likely that they may have originally formed ABGs, which became disarticulated due to post-depositional movement. However it remains uncertain whether the remains are from the same individuals.

Wilsford shaft

The Wilsford shaft was originally thought to be a pond barrow, which may have been created by the spoil from the shaft excavations. The shaft is cut into bedrock and is about 30m deep (Ashbee et al. 1989). The site produced the largest number of ABGs from any one site dating to this period (Table 3.7). The purpose and date of construction are still matters of debate. The earliest date comes from a wooden container at the base of the shaft, possibly used for removing the spoil during excavation. This was radiocarbon dated to 3650-3100 BC (Ashbee et al. 1989, 68-69). The other radiocarbon dates obtained from the material deposited indicate that filling of the feature began in the middle Bronze Age and that the shaft was completely filled by the beginning of the Iron Age (Figure 3.7) (Darvill 2007, 181). It would therefore appear that the shaft may have been constructed in the late Neolithic and a small amount of deposition allowed to occur in the very bottom, including the wooden artefacts. Then towards the middle Bronze Age more substantial deposition occurred, when the original use of the shaft as a well (Bell 1989) or a ritual feature (Ashbee 1989), ended. The majority of the ABGs recovered from the shaft belong to sheep/goat. Two sheep ABGs were encountered towards the bottom of the shaft. However, both were discovered in two separate groups and it was only during post-excavation analysis that Caroline Grigson, the zooarchaeologist, realised that the elements were from the same individuals. She concluded that the ABGs were likely to have been deposited as complete animals, but had become disarticulated due to sediment movement caused by the decomposition of organic matter and fluctuation in the water table. Wool preserved in the waterlogged deposits also suggests that the individuals were deposited as complete carcasses (Grigson 1989, 106). The two complete sheep were from very young animals, one was neonatal on the basis of the tooth

The next ABG deposits were discovered much further up the shaft in contexts, 1960:G37, G38, G39, G40, at the depth of 44ft (13.2m). Three ABGs were discovered at this depth consisting of lower hind limbs from juvenile animals. Sheep/goat skull fragments were also present in this context, but no other sheep/goat bones were found. Grigson (1989, 112) suggested that these ABGs could represent fleeces being deposited in the shaft, or were a ritual deposit similar to ‘head and hooves burials’. The remains could also be from skinning waste, and may not have been deposited with the fleece attached. The next deposit consists of two sheep/goat ABGs from contexts 1960:G3 and G8, at a depth of around 20ft (6m). Both ABGs consist of vertebrae, and may have come from the same animal. Two horse ABGs were also discovered within the weathering cone. The remains of a hind limb foot, consisting of three elements were found in context 1960:121, at 8ft-8ft 6in (2.4-2.5m). Another back limb was recovered from context 1960:125, at 9ft (2.7m). We cannot discount that these elements came from the same animal. Also deposited in these contexts, was a possible dog ABG (Grigson 1989, 114), but the evidence only consists of two metatarsals (the only dog remains discovered from the shaft).

35

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 3.7 The Wilsford Shaft indicating approximate locations of ABGs. * complete sheep ABGs from foetal/neonatal individuals (adapted from Ashbee et al. 1989, Figure 7; Darvill 2007, Figure 20) animals. The presence of neonatal and juvenile ABGs, aged to between 13 and 24 months, could indicate that their deposition occurred during the summer period. Grigson (1989, 121) argued that the presence of these immature animals points to pastoral activity taking place close to the site.

Radiocarbon dating indicates that the majority of the ABGs were deposited around the middle Bronze Age, except for the four near the top of the shaft which are late Bronze Age –early Iron Age in date. The dates for the majority of the shaft fills appear to be around 1530 to 1250 BC. Another notable feature is the age of the sheep/goat ABGs, with the majority coming from young 36

Chapter 3. Neolithic and Bronze Age Table 3.8 Formation of partial ABGs by species and site type for the Bronze Age (information not available from all sites) Site type Species Mixed Head+ Axis Axis Axis + leg Leg Enclosure Cattle 1 Enclosure Dog 1 Midden Cattle 1 Midden Dog 1 Pit complex S/G 1 Rural Settlement Cattle 4 9 2 Rural Settlement Dog 1 Rural Settlement Pig 1 Rural Settlement S/G 1 2 Shaft Horse 2 Shaft S/G 6 7 Total 2 4 17 2 16

Figure 3.8 Cow ABG pregnant with calf (indicated by the arrows) from Crab Farm (Papworth 1992, Plate 4). 3.7.3

Bronze Age settlements, enclosures and middens

seems likely that a number of the ABGs deposits may be from the same animal(s). The report states that the ABGs represent the remains of at least six cattle (BucklandWright 1987).

The majority of the sites where we can be certain pastoral farming took place date to the later part of the Bronze Age period. Cattle are the most common ABGs from settlement sites largely due to their frequency in the boundary ditch at Poundbury, Dorset (Green 1987). These consist of 12 partial cattle ABGs. These ABGs are all from the axial part of the skeleton and either are comprised of the skull and associated vertebrae (4 cases) or just vertebrae (8). None of the ABGs therefore have limb bones present. Unfortunately the information available from the site is limited as the identification of which vertebrae were present is not available. Some of the ABGs consist of only a few elements, and it therefore

Two of the ABGs, one consisting of 12 vertebrae and the other 7, show possible signs of butchery. In both deposits the vertebrae are lacking the transverse or spinous process. The absence of blade marks and the appearance of the fractured bone surface led Buckland-Wright (1987, 4, A6) to suggest that the processes were removed with the flesh attached. The ABGs also show few signs of exposure and no gnawing. Therefore they may have been deposited and covered soon after the butchery took place. The deposits occur at the same time the settlement appears to have been abandoned at the beginning of the 37

Investigating Animal Burials; Ritual, Mundane and Beyond This deposit makes up only 2% of the ABG assemblage from the Bronze Age. Although non-ABG pig remains become less common in this period, the paucity of pig ABGs is surprising, especially considering that many of the unusual faunal deposits from this period are thought to be associated with feasting, and it has been suggested that pigs are utilised for such activities especially in the late Neolithic (Parker-Pearson 2003). Perhaps other species such as cattle start to be utilised in this way, as suggested by some barrow deposits.

late Bronze Age. It was suggested that the abandonment of the settlement was accompanied by the slaughter of the cattle and partial backfilling of the boundary ditch (Green 1987, 31), or the cattle were slaughtered in association with a raid (Buckland-Wright 1987, 4, A5). The other settlement sites with cattle ABGs present, Middle Farm, Dorset (Bullock and Allen 1997), Shearplace Hill, Dorset (King 1962) and Dean Bottom, Wiltshire (Maltby 1985b; 1992) have either two (Middle Farm) or only one cow ABG present. Compared to Poundbury the composition is different, with no skulls recorded as present. Only one of the deposits, from Middle Farm, consists of just vertebrae (3 cervical), the rest of the ABGs from settlement sites consist of limb bones (Table 3.8).

ABGs are also recorded from a number of enclosures dating to the middle or late Bronze Age. Only a small percentage come from enclosure sites, compared to the open settlements discussed above. The majority of the ABGs recorded from this site type were found on the Crab Farm enclosure, Dorset (Papworth 1992). Two cattle and four sheep/goat ABGs were recorded from the site. Unusually, all consist of complete skeletons, and two, one cow and one sheep, were pregnant. The two cattle ABGs consist of the mother and calf (Figure 3.8). The cow was dated to 2930±70 BP, placing it in the late Bronze Age, and was found on the base of the enclosure ditch. Tooth eruption data suggests the mother was around 30 months old, so this would probably have been her first pregnancy. The tooth eruption of the calf suggests it was close to term. The only butchery mark present is a knife mark under the diastema of the left mandible of the cow.

Dean Bottom also produced a complete cow ABG. This consists of 99 elements from a neonatal calf, deposited in a shallow pit, sealed by sarsen blocks, associated with a terrace floor, possibly of a structure (Gingell 1992, 27). Also present on this site was a mixed dog ABG, deposited in a shallow scoop. This ABG consisted of 74 elements from the head, axis and fore limbs, but the hind limbs were not present. No butchery marks were observed (Maltby 1985b), and it is therefore unknown whether the hind limbs were removed on purpose. However, the presence of the ABG in a small feature with no other ‘rubbish’ could indicate it was purposely deposited and not included as part of general waste. Dean Bottom and Shearplace Hill, Dorset (King 1962) are the only Bronze Age settlement sites to have ABGs of more than one species present.

The pregnant sheep ABG, dated to around the same period, 2990±80 BP, was discovered within a shallow pit. The adult sheep was found to be pregnant with twin lambs. The lambs were close to birth on the evidence of their size and the development of the deciduous 4th premolar. The other ABG from Crab Farm consisted of a complete adult sheep, although the skull is missing due to truncation by a later pit. The report argues that if the pregnant animals had died giving birth, then it is expected they would have been butchered and eaten. Therefore the remains are interpreted as votive offerings. If this was the case, offerings of this nature were very rare, as these are the only deposits from this period that belong to pregnant animals. Also, as this chapter has shown, complete ABGs are uncommon. Of the 55 ABGs recorded from this period, only 12 consist of complete animals, six of which are from Crab Farm. The other finds from enclosures consist of a partial cow from South Lodge Camp, Dorset (Legge 1991b) and a partial dog from the enclosure near Badbury Rings, Dorset (Maltby 1986c; Vatcher and Vatcher 1965). It is suggested that the cattle ABG from South Lodge Camp, may have been a complete skeleton, but has been disturbed by modern ploughing. Both deposits are from the enclosure ditches and when present, Bronze Age ABGs appear to be commonly deposited on settlement boundaries.

A partial cow ABG consisting of limb bones was recorded from Shearplace Hill, as were eight elements of a partial sheep/goat ABG recovered from a post-hole. The report does not state which parts of the body are represented by the ABG (therefore these data are not included in Table 3.8). The remainder of the sheep/goat ABGs from settlements were all recovered from Bishop Cannings Down, Wiltshire (Gingell 1992). The ABGs all come from layer 138, but it is unknown which features this layer was associated with. The excavation was of a limited area containing two round houses, and therefore the context may have been associated with them. The deposits are all partial ABGs, one comprised of vertebrae and ribs, the other two consisting of appendicular elements from the upper and lower front limb. The ABGs were discovered in three discrete groups, but we cannot discount the possibility that they are all from the same animal. The only pig ABG dating to this period comes from the late Bronze Age settlement at Bell Street, Romsey, Hampshire (Coy 1993). The faunal assemblage from this site is very small, with only 30 identified fragments dating to this period, 23 of which come from the pig. The deposit was recovered from layer 320, and consists of a mixed ABG with all areas of the axial skeleton represented, along with some of the front and back limbs.

One of the later sites recorded from the Bronze Age is Potterne, Wiltshire (Lawson 2000). Compared to the other sites discussed Potterne is unique, consisting of a midden which covers over 3.5 hectares and is comprised 38

Chapter 3. Neolithic and Bronze Age The majority of the Bronze Age ABGs have been recorded from settlement sites and the Wilsford Shaft. The two largest Bronze Age assemblages are very different in nature. Wilsford Shaft consists of just one feature, into which 17 ABGs were deposited, all sheep/goat apart from the two partial horses near the top of the feature. The age of the sheep/goat ABGs would suggest that the deposits are being made at a specific time of year. Also, the ABGs deposited at the very bottom of the feature are complete, which could be an indication that slumpage has affected the composition of the ABGs further up the shaft. In comparison, the deposits from the later Bronze Age Poundbury site are mainly from cattle. All 12 of the ABGs were found within the site’s ditch and have been interpreted by the report’s authors as the result of either a ‘raid’ or the abandonment of the settlement. Unusually, all the cattle ABGs consist of axial elements, whereas the rest of the ABGs from Bronze Age settlements consist of appendicular elements.

of dark, anthropogenic deposits up to 2m thick. The site appears to have been formed over 500 years by material from an undiscovered settlement. The excavators favoured interpretation is that stock, predominantly cattle, were corralled there and the site was used periodically as a place for meetings/exchange. Potterne produced one of the largest faunal assemblages for the period with 130,000 fragments. But only two ABGs were recovered, a partial cow comprised of vertebrae and a partial dog consisting of the upper front limb. Suggestions have been made regarding repetitive ‘ritual’ deposition at Potterne (Waddington 2010). However, if this was the case, the deposition of ABGs does not appear to have been included in the practice. 3.8

Summary

Only a small number of Neolithic and Bronze Age ABGs have been recorded for this study in comparison to later periods. Only one Neolithic ABG has been recorded from the Yorkshire region and none from the Bronze Age. This may be a reflection of bone preservation in the locales of the Yorkshire Neolithic and Bronze Age sites. However, it is also noteworthy that all the Bronze Age sites recorded from Yorkshire with animal bone present are round barrows. The southern England data indicates that ABGs have rarely been recorded from such sites.

In summary, the results from the Neolithic and Bronze Age periods have shown that the nature of the ABGs is variable. There is a lack of such deposits from the Yorkshire study region which would merit further investigation. The southern England assemblage has shown that the composition and presence of ABGs differs between these two periods and also between site types

Therefore virtually all the Neolithic and Bronze Age data comes from southern England. Cattle appear to be the most common ABG species in the Neolithic, which corresponds with their prominence in the non-ABG faunal assemblages. It is worth noting that no late Neolithic pig ABGs have been recorded despite the increase in pigs in the non-ABG faunal assemblages of this time. Cattle remain the most common ABG species in this sub-period. The majority of the recorded assemblages date to the early Neolithic. However, this is due to the Windmill Hill assemblage, which is by far the largest for this period. Although a number of species were deposited as ABGs at Windmill Hill, cattle were the most common. The majority of the cattle deposits from the site consist of either vertebrae or lower limb elements. In addition to Windmill Hill, all other causewayed enclosures with surviving animal remains had ABGs present, a pattern not found on funerary sites and henge monuments. Although previous literature has described ABGs from Neolithic and Bronze Age funerary monuments, they appear to be very rare in the southern England study region. For example, Fussell’s Lodge provides the only Neolithic record from the region of a ‘head and hooves’ deposit. Also, ABGs were only present on six of the 49 round barrows recorded with animal remains present. Of the henge monuments recorded only Coneybury had ABGs present, although recent excavations at Durrington Walls have started to reveal ABGs there. It appears that such deposits are rare phenomena on henge and funerary sites.

39

Chapter 4. Iron Age Southern England 4.1

Introduction

As with the late Neolithic and early Bronze Age, the transition between the late Bronze Age and the early Iron Age is not clear cut, and the distinction is largely artificial. As discussed above, there is a change in the type of archaeological sites found in the middle to late Bronze Age, as settlements become more visible in the archaeological record. This development appears to lay the foundations for the changes that occur in the late Bronze Age and early Iron Age, such as the development of hillforts. Many later Bronze Age settlement areas, such as Poundbury, continue to be utilised in the Iron Age. There is also the development of new site types in the form of hillforts. The construction of these sites began around 800 BC (Cunliffe 2005, 50), at the same time as a reorganisation of the landscape was taking place using linear earthworks often associated with the hillforts. Cunliffe (2005, 589) suggests that these earthworks imply a greater control over land and livestock, which may have become a symbol of status after the collapse of a bronzebased prestige goods economy. Some Wessex hillforts developed during the Iron Age, becoming more complex in terms of defences and entranceways. Some of these middle Iron Age (300-150 BC) ‘developed’ hillforts show signs of being intensively occupied and probably acted as central places for the communities that used them (Cunliffe 2005, 590-591) . By the late Iron Age significant social change is evident in the archaeological record, at a time when there was increased contact between continental Europe and southern Britain. This contact is argued to have stimulated social change in southern Britain (Cunliffe 1988; Haselgrove 1982; 1989). However, such a coreperiphery model has been criticised as it does not fit the archaeological data particularly well (Millett 1990; Willis 1994). Fitzpatrick (2001) and Hill (2007) have argued that long-distance exchange with Gaul and the Roman world was important, but did not induce change. They have argued instead that continental trade is a symptom of a continuation of social change from the middle Iron Age, driven by internal forces (Hill 2007).

areas this is largely due to the discovery of several Iron Age sites during archaeological fieldwork in the wake of large road construction projects, (the M3 motorway and the Dorchester by-pass respectively). The grouping around Danebury is the result of the large research project undertaken to investigate a number of later prehistoric sites in this area (Cunliffe and Poole 1991a), as well as sites discovered during a number of commercial excavations. The paucity of sites with ABGs from Wiltshire compared to the other counties should not be taken as a sign that they were more rarely deposited on sites in the region. The pattern is created because there has been a greater emphasis on studying Iron Age sites in Dorset and Hampshire compared to Wiltshire, where Neolithic and Bronze Age archaeology is more prevalent. More archaeology from commercial development has also taken place in these counties, especially in Hampshire (Darvill and Russell 2002, 61). Finally, the dataset is also dependant on the publication of excavation and faunal data. There are (at the time of writing) a number of unpublished reports concerning Iron Age ABGs from Wiltshire (e.g. Hambleton and Maltby 2004) as well as reports that have been published or discovered after data collection had been completed for this project (e.g. Powell et al. 2006). The sites where no ABGs have been recorded are more evenly distributed between the three counties and there appears to be no significant patterning in their distribution, compared with the sites with ABGs present. 4.2

As in the Neolithic and Bronze Age sample, domestic mammals dominate the ABG assemblage, making up 89% of recorded ABGs. Wild mammals and birds make up 4% and 6% of the assemblage respectively, the remaining 1% consists of domestic fowl. 4.2.1

A large proportion of the ABGs recorded for this study come from the Iron Age. In total 784 ABGs were recorded, the majority (746 – 95%) from sites in southern England, with the remaining 38 from Iron Age sites in Yorkshire (see 5.1). The south England Iron Age assemblage forms 41% of the overall ABG assemblage from the region. In total, 73 Iron Age sites were recorded from southern England, of which 50 (68%), have ABGs recorded as present, a much higher percentage than in the preceding periods. The majority of the sites with ABGs present are recorded from either Dorset or Hampshire, whereas only 10 are from Wiltshire. These include clusters of sites around Dorchester, Basingstoke and Danebury. In the first two

Species proportions

Wild species

The majority of the wild mammal ABGs date to either the middle or late Iron Age (Table 4.1). Foxes, followed by cats and then hare, are the most common wild mammals. However, both the fox and cat assemblages are dominated by those found in associated deposit groups. Of the 15 fox ABGs recorded, 12 come from a single deposit found at Winklebury Camp hillfort, Hampshire (Smith 1977). The deposit comes from a fill towards the base of a late Iron Age pit (context 3920). Alongside the fox ABGs was a complete red deer ABG. Due to slight post-depositional movement it was not possible for the zooarchaeologist to identify each fox element to the individual, but the evidence suggested that the foxes were

Chapter 4. Iron Age Southern England Table 4.1 Number of ABGs per species for the southern England Iron Age. The Sheep/Goat, (S/G) counts include all ABGs identified to either species Species Early Iron Age Middle Iron Age Late Iron Age Total Domestic Cattle 32 42 39 113 Mammal S/G 43 150 68 261 Sheep 14 41 32 87 Goat 1 5 3 9 Pig 16 40 10 66 Horse 24 57 16 97 Dog 30 67 36 133 Wild Fox 2 13 15 Mammal Cat 1 6 7 Pine Martin 1 1 Stoat 1 1 Weasel 2 2 Red Deer 1 1 Hare 3 2 5 Domestic Domestic bird Fowl 2 4 6 Wild bird Raven 12 16 3 31 Rook/Crow 3 3 Buzzard 1 1 Cormorant 1 1 Lark 1 1 Other Snake 1 1 Total 160 385 201 746

Figure 4.1 Percentages of the most common ABG species. Sample size in brackets (Grant 1991). However, as discussed above, the data from these excavation seasons cannot be included in this study (see 1.5.1).

all deposited complete (Jones 1976b; 1977). The red deer ABG was from an adult stag. There is no evidence of butchery on any of the ABGs, but the antlers and ends of the red deer’s limbs do display signs of gnawing. The development of the red deer’s antlers suggests the animal died in the late autumn or winter. The pit also appears to have been rapidly filled after the ABGs were deposited. The original unconvincing explanation for the deposit is that the red deer was a pitfall victim, and that the foxes were also fall victims that gnawed the deer bones once they were in the pit (Jones 1977). Red deer ABGs are rare, although there is mention of a complete neonatal skeleton from the 1979-88 excavations at Danebury

The majority of cat ABGs came from a single unknown late Iron Age feature at Gussage All Saints, Dorset (G. J Wainwright 1979). The feature (likely to be a pit) contained the associated remains of five complete neonatal cat ABGs. Although some have argued that domestic cats were not introduced to Britain before the Roman period (Clutton-Brock 1999, 135), these ABGs may be from domesticated cats. O’Connor (in press) has argued that domestic cats were present as far north as 41

Investigating Animal Burials; Ritual, Mundane and Beyond Orkney in deposits dated between the 1st and early 5th century AD and therefore may have been introduced to Britain earlier than the Romano-British period, unless these represent evidence for the local domestication of native wild cats. The presence of five neonatal cat ABGs deposited together suggests the animals were all from the same litter (Harcourt 1975). If this was the case, it would indicate that the animals were born in close proximity to human settlement, which would support the case that they were domestic. It is interesting to note that the cat ABG from Danebury, Hampshire was also from a kitten (Grant 1984a), whereas the only adult cat ABG comes from Owslebury, Hampshire (Maltby 1987c).

(1984c) as birds killed to prevent their scavenging of carcasses. However, more recent interpretations, such as those made for the raven ABG from a late Iron Age pit at Silchester, Hampshire (Grant 2000), view such deposits as ritual. Authors such as Green (1992, 125-127, 177181), following Ross (1967, 255-285), have placed great emphasis on the ritualised symbolic nature of ravens in Iron Age society. It is, however, interesting to note that, compared to domestic mammals, ravens make up a very small percentage of the ABG assemblage (Figure 4.1). This indicates that, if all ABGs were deposits of a ritual/symbolic nature, the deposition of domestic fauna was either more important than wild, or simply domestic animals were more easy to acquire for such purposes.

There is an increase in the number and variety of wild bird ABG deposits in the Iron Age compared to previous periods. Only one pre-Iron Age wild bird ABG was recorded - the white-tailed sea eagle from Coneybury Henge (Maltby 1990e). Wild bird ABGs were recorded from 10 (16%) of the sites with ABGs present, suggesting their deposition was a fairly rare phenomenon during the Iron Age. Some sites have produced several species of wild bird ABGs. For example, a buzzard, lark and two rook/crow ABGs were recovered from late Iron Age pits at Owslebury (Maltby 1987c). The majority of sites only have one bird ABG recorded as present. The exceptions are discussed below.

4.2.2

Domestic animals

The most common domestic mammal ABGs are those of sheep/goat, dogs and cattle. There are only nine definite goat ABGs, including three from Poundbury (BucklandWright 1987) and four from Owslebury (Maltby 1987c). This low number would correspond with the disarticulated faunal material which suggests that goats were not kept in great numbers on Iron Age settlements (Maltby 1981b). Therefore the majority of the sheep/goat ABGs are probably sheep. Indeed, 86 of the 260 sheep/goat ABGs have been specifically identified as sheep (Table 4.1).

In the early and middle Iron Age assemblages, raven ABGs are the most common amongst all wild species, and, if the multiple ABGs of foxes from Winklebury (see above) are excluded, raven ABGs are also the most common in late Iron Age sites. However, raven ABGs are only present on a small number of sites. The majority come from Danebury. In total, 21 raven ABGs were recorded, with eleven, two and eight of these coming from early, middle and late Iron Age deposits respectively. As discussed above, only the ABGs from the 1969 to 1978 excavations were recorded in this study, due to the lack of available published data from the later excavations. The only published data on the bird ABGs is in Coy’s (1984c) report, in which only limited information is provided regarding individual ABGs. Therefore, some of the ravens from Danebury may not strictly be ABGs as defined for this project. However, even if this is the case, there is still an unusually large number of raven ABGs from Danebury. The majority appear to consist of partial skeletons, with three of the late Iron Age deposits consisting of just wing bones. However, the make-up of other ABGs is unclear.

In all phases of the Iron Age sheep/goat are the most common species recorded as ABGs. However, differences are apparent between the sub-periods. In the early Iron Age, sheep/goat constitute 27%, cattle 20% and dog 18% of the total ABG assemblage. In the middle and later Iron Age samples, sheep/goat ABGs become more dominant. This ties in with the general species proportions seen in faunal assemblages from the South of England, where there is a progressive increase in the percentage of sheep/goat elements during the Iron Age (Hambleton 1999, 59). Hambleton’s findings show that the proportion of sheep/goat to cattle and pigs is highest in the late Iron Age for sites from southern England. The ABG data do differ slightly. Sheep/goat ABGs are at their most common in middle Iron Age deposits, whereas there is a decrease in the proportion of sheep/goat ABGs in the late Iron Age sample and a corresponding increase in the percentages of dog and cattle ABGs. Both pig and horse ABGs are most common in early Iron Age deposits. The percentage of pig ABGs remains the same in the early (10%) and middle (10.4%) Iron Age assemblages but decreases (5%) in the late Iron Age sample. The proportion of horse ABGs shows a similar trend, with those dating to the early Iron Age representing 15% of the ABGs recorded. By the late Iron Age this has dropped to 8%.

Two other sites have multiple deposits of ravens. Balksbury Camp, Hampshire (Maltby 1985c; 1987b; 1995a; 2001) and Boscombe Down West RAF Station (King 1951; Platt 1951) have both produced three raven ABGs in middle Iron Age features. All six consist of partial skeletons. The deposits from Boscombe Down all consisted of parts of the axial skeleton or legs. The excavator (Richardson (1951, 125) suggested that the ABGs represent the remains of making raven stew. The ravens from Danebury were initially interpreted by Coy 42

Chapter 4. Iron Age Southern England Table 4.2 Number of ABGs found on southern England Iron Age sites Site Type 1 to 4 5 to 9 10 to 29 30 to 49 50 to 99 Hillfort 4 2 3 Non-Hillfort 19 9 5 4 1 Total 23 11 8 4 1 Percentage total sites 46.9% 22.4% 16.3% 8.2% 2.0% Percentage No. ABGs 6.4% 10.0% 20.8% 22.9% 7.6%

>100 2 2 4.1% 32.3%

Table 4.3 The seven largest Iron Age ABG assemblages Site Name Site type No. ABGs Percentage of total ABG assemblage Balksbury Camp Hillfort 152 20.3% Danebury Hillfort 123 16.4% Suddern Farm Non-Hillfort 57 7.6% Winnall Down Non-Hillfort 49 6.5% Old Down Farm Non-Hillfort 48 6.4% Owslebury Non-Hillfort 47 6.3% Nettlebank Copse Non-Hillfort 30 4.0% 4.3

Nature of the assemblage and influential sites

from Balksbury are from horse. There are also a small number of cattle, sheep/goat and dog ABGs (Figure 4.2). All ten horse ABGs are partial skeletons and eight of these consist of limb elements only. This corresponds with Balksbury’s disarticulated faunal data where 75% of the horse elements are from the appendicular parts of the skeleton. However, the ABG and disarticulated faunal assemblage species proportions show very different patterns. The majority of identified elements in the ‘normal’ faunal assemblage belong to sheep/goat or cattle, whereas only 10.8% are from horse (Maltby 1995a; 2001).

One of the aspects that must also be considered is the proportion of ABGs per site. During his study Wait (1985, 122) recorded 21 sites with ‘special animal deposits’ present, but used data from just six sites for much of his analysis (see below). The data from all sites recorded in this study have been utilised, but the majority of sites have very few ABGs. Of the 49 sites recorded, 46% only have between one and four ABGs present. However, the ABGs from these sites represent only 6.4% of the total Iron Age ABG assemblage (Table 4.2). A few sites have large assemblages with 67.6% of the ABGs coming from just seven sites.

The species proportions from middle Iron Age contexts at Balksbury are similar to the overall middle Iron Age pattern, although there are slightly lower proportions of cattle and pig ABGs. Of the 385 middle Iron Age ABGs from southern England, 34% (132) are from Balksbury. Therefore we should expect some similarity between the site and overall species proportions. The Danebury assemblage also makes up a large proportion of the middle Iron Age dataset (22%). The remainder of the middle Iron Age data comes from thirteen other sites, some of which (Winnall Down, Suddern Farm, Old Down Farm, Owslebury) each have between 20 and 50 ABGs. Sheep/goat commonly makes up a high proportion of the ABGs from these sites. However, each site is different with cattle, pig and dog ABGs each dominating the assemblages from some sites (Figure 4.3). Thus, although sheep/goat ABGs are always common, the proportion of ABGs from other species fluctuates between sites.

Two hillforts, Balksbury Camp and Danebury, have assemblages of over 100 ABGs. Although they represent only 4% of the sites with ABGs, 36.7% of all the ABGs recorded come from them. Suddern Farm, has an assemblage of 57 ABGs and four non-hillfort sites have assemblages of between 30 and 49 ABGs (Table 4.3). Therefore, although the Iron Age is synonymous with ABGs, the majority of sites from southern England have produced only small ABG assemblages. Balksbury Camp has the largest Iron Age ABG assemblage. Wainwright (1969) carried out initial excavations on the site in the 1960’s and more recently Ellis and Rawlings (2001) excavated a small area of the site in 1995 to 1997. ABGs were recovered from both excavations, but the majority of the ABG data can be found in the Ancient Monument Reports (Maltby 1985c; 1987b) from the substantial 1973 and 1981 excavations (Wainwright and Davies 1995). In total, 152 ABGs have been recorded from the site which represents 20.3% of the Iron Age ABGs, the majority, (132), coming from middle Iron Age contexts. The remaining ABGs are dated to the early Iron Age. This early Iron Age assemblage differs from the overall pattern where sheep/goat and cattle are the most common species (Figure 4.1). Half of the early Iron Age ABGs

Danebury has the second largest Iron Age assemblage, consisting of 123 ABGs. As discussed above, the Danebury data only derives from the 1979-88 excavations. Also, single bone ‘special deposits’ were not included in this study, and so a large number of the 151 skull deposits reported by Grant (1984a) have not been included. 43

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 4.2 ABG species percentages from Balksbury Camp

Figure 4.3 Species percentages for the largest middle Iron Age ABG samples from southern England

Figure 4.4 ABG species percentages from Danebury

44

Chapter 4. Iron Age Southern England

Figure 4.5 Top: percentage of southern England ABGs per species deposited on hillforts. Bottom: percentage of ABGs per species deposited on non-hillfort sites. Sample sizes in brackets given the size of pig litters, this is a distinct possibility (see below).

ABGs were recorded from early and middle Iron Age sub-periods. There is a high proportion of sheep/goat ABGs from Danebury throughout (Figure 4.4). However, cattle and then raven are the most common types of early Iron Age ABGs from Danebury. This differs from the overall trend, but these species are generally more common in the early Iron Age than in later sub-periods. However, this is because the large Danebury assemblage (24% of the early Iron Age sample) has biased the overall results. Excluding the early Iron Age Danebury data, a sheep/goat and dog-dominated pattern becomes more apparent.

The late Iron Age data from southern England show a different pattern. The sample is not dominated by any large assemblages. Winklebury Camp contributes 14% (29) of the ABG dataset, followed by Suddern Farm and Whitcombe, which each contribute 9% (19). The rest of the dataset is made up of small assemblages from 34 other sites. 4.4

The pattern at Danebury changes in the middle Iron Age deposits, which include a larger proportion of pig ABGs. These represent 22% of the assemblage, in which the percentage of sheep/goat ABGs also increases slightly. Therefore the Danebury assemblage differs from the overall sheep/goat-dominated pattern. A large proportion of the middle Iron Age pig ABGs (47%) are from Danebury. 60% of these consist of complete neonatal skeletons. Unfortunately the data are not available to determine whether they were deposited together, but

Site types and context

Investigating patterns in the site-type data for Iron Age ABGs does present some problems. During the Iron Age there is an increase in the types of site encountered in the archaeological record. As well as hillforts, a number of different types of enclosures, boundary ditches and open settlements appear. Settlement sites also change and develop during the Iron Age, with many of the early Iron Age open settlements becoming enclosed in the later parts of the period (Hingley 1996). It must also be noted 45

Investigating Animal Burials; Ritual, Mundane and Beyond study’s results show this is not the case. The majority (690) of ABGs do come from pits, (86%), but this is to be expected as archaeological excavations tend to sample a higher proportion of a site’s interior. For example, the excavations at Suddern Farm (Cunliffe and Poole 2000b) consisted of two trenches, one investigating the ditches around the site, which covered 190 m2. The other, investigating the interior of the settlement, covered 1200m2. In essence we excavate a greater volume of Iron Age pits than we do ditches. Therefore the majority of our data come from pits.

that site type definitions are created by archaeologists and in all likelihood bear no direct relevance to the past, as it seems unlikely that Iron Age people would have viewed their settlement sites in the same way we do. One distinction we can draw is between hillforts and nonhillfort sites. Although the interpretation of hillforts is itself problematic, there are a number of distinct differences between hillforts and other sites, including positioning within the landscape, size of area enclosed and the potential for grain storage (Fitzpatrick 1997). Therefore, for this part of the analysis, sites have been placed into two groups, hillforts and non-hillforts, which is similar to Wait’s (1985, 126) division of hillforts and settlements.

After pits, ditches are the most common feature containing ABGs, with 41 (5%) of the sample recovered from them. The majority of these are boundary ditches, showing that ABGs are not just found in the interior of settlements, contra to Wait (1985). Correspondence analysis (see 1.5.3) was carried out to investigate links between species and feature type. The majority of the species and feature types are clustered together, due to the overwhelming number of ABGs from pits compared to other features (Figure 4.6). The correspondence analysis is based on two separate factors, the species proportions within an individual feature type and the overall proportion of species. For example, pig remains comprise 66% of the ABG assemblage from graves, but overall, only 17% of the total pig ABG assemblage is recovered from graves.

Comparing the main domestic mammals, the pattern of species deposited as ABGs on hillfort and non-hillfort sites in the early Iron Age period is relativity similar, with the exception of the percentage of cattle and sheep/goat. Cattle ABGs appear to be much more common on hillfort and sheep/goat more common on non-hillfort sites (Figure 4.5). In the middle Iron Age sample, there is a drop in the percentages of cattle and horse ABGs found on hillforts. In the middle and late Iron Age samples cattle ABGs are more commonly found on non-hillfort sites. However, many of these late Iron Age cattle ABGs are from Lains Farm (Coy 1991). The proportion of horse ABGs fluctuates, with horse being more common on nonhillfort sites in the middle Iron Age, but more common on hillforts in the late Iron Age. Such fluctuations are probably due to the small size of the horse ABG sample. Pig remains make up only a small part of the ABG assemblage. In the early Iron Age there is little difference between the two site types. However, in the middle and late Iron Age pig ABGs make up a higher proportion of the hillfort assemblage, although the majority of these are from Danebury. Dog ABGs are slightly more common on non-hillfort sites until the late Iron Age.

The analysis does show that certain species are more commonly associated with different features. Although the majority of the species are closely associated with pits, domestic fowl, pig and cat show different associations. Both domestic fowl and pig are closely associated with grave contexts, with 33% and 17% of their ABG assemblages respectively, coming from this feature type. Despite a higher proportion of the domestic fowl ABG assemblage coming from graves, pig constitute 66% of the ABGs from this type of feature. Pig ABGs also make up a similar proportion of the assemblage from gullies. However, the majority of pig ABGs (75.3%) have been recovered from pits, which is why pigs are spatially close to pits and gullies.

Wait (1985, 132) (see 1.2.6) carried out similar analysis and described a number of patterns, most of which are contradicted by the results of this study. The most likely reasons for the disparity between the results of this study and Wait’s (1985) survey are probably related to sample sizes and criteria used for defining ‘special deposits’. Wait used data from just two hillfort sites, Danebury (Grant 1984a) and Winklebury (Jones 1977), and four settlements, Little Somborne (Locker 1980a), Old Down Farm (Maltby 1981a) and, from outside this project’s study area, Ashville (Parrington 1978) and, Twywell (Jackson 1975). In comparison, the results from this study are from 49 separate sites, 11 hillforts and 38 non-hillfort sites. Also, Wait (1985) included single bone deposits in his analysis, which are excluded from this study.

The majority of the five cat ABGs, (71%) come from an unknown context at Gussage All Saints. Cat ABGs are, also recorded from a quarry and pit feature. Cattle sheep/goat and wild birds are also recorded from quarry and unknown features. Dogs are recorded from pits, ditches and middens. Although the vast majority (94%) of the dog ABGs are recorded from pits, dogs make up a large proportion of the ABG assemblage from middens. To investigate any further associations between species and feature type, we need to exclude ABGs recovered from pits, because of its dominating effect on the assemblage. By removing pit ABGs, Figure 4.7 shows that a number of other associations are present. The patterns discussed above are still evident, and the association between dogs and middens becomes clearer.

Wait (1985, 138) noted that only ten of the ‘special animal deposits’ he recorded had not come from pits, but from interior gullies and ditches, summarising that, unlike human remains, ‘special animal deposits’ are always from the interior of a settlement. However, this

46

Chapter 4. Iron Age Southern England

Figure 4.6 Correspondence analysis showing relationships between species and feature types

Figure 4.7 Correspondence analysis showing relationships between species and feature types, excluding pits

47

Investigating Animal Burials; Ritual, Mundane and Beyond raven from Cowdown (Harcourt 1968b). As discussed above, the exact composition of a number of raven ABGs from Danebury is unclear. It is stated that ravens are represented by whole skeletons or partial skeletons, especially wings (Coy 1984c, 527), but later in the report the majority of the ravens are described as partial (Coy 1984c, 530).

Wild mammals and horses are closely associated with ditches. However, the wild mammal sample consists of a single ABG, a fox from Nettlebank Copse (Poole 2000b). Cattle make up a large proportion of the ABGs from layers, quarries and unknown feature assemblages. However, the majority of cattle and horse ABGs are from ditches. Both make up 26% of the ditch ABG assemblage. This fits with the pattern seen in the disarticulated faunal assemblages for the period, where cattle and horse bones tend to be more common on the outskirts of settlements (Maltby 1985a; Wilson 1996, 23). Sheep/goat make up 21% of the ditch ABG assemblage and all ABGs discovered from post-holes are from sheep/goat. 4.5

Only a small amount of body area information is available for wild bird ABGs, but it indicates that elements from the axial skeleton and legs are most common, with only three confirmed deposits from Danebury consisting of just a raven wing (Coy 1984c, 530). Perhaps wings were valuable for their flight feathers and hence removed before deposition, as possibly indicated by butchery marks found on the wing elements from a middle Iron Age partial raven ABG from Danebury (Coy 1984c). Also, preservation factors affect elements differently and wing elements are more vulnerable to non-cultural taphonomic transformations (Higgins 1999).

Composition

The majority of ABGs dating to this period consist of partial skeletons. Of the 756 recorded, 535 (72%) are partial, 157 (21%) are complete and information is not available for the remaining 55 (7%). However, some species appear to have been deposited as complete ABGs more often than others. Apart from hare, nearly all other wild mammal ABGs consist of complete skeletons (Table 4.4). The exceptions are one fox ABG from Old Down Farm, Hampshire (Maltby 1981a) and one cat ABG from Owslebury, Hampshire (Maltby 1987c). The only wild mammal to have more partial than complete ABGs recorded is the hare. Of the five deposits, two complete hare ABGs are recorded from a middle Iron Age feature at Winnall Down, Hampshire (Maltby 1985a) and another was found in a late Iron Age deposit at Suddern Farm, Hampshire (Poole 2000d).

Domestic fowl show a different pattern, with the majority consisting of complete ABGs, including two from Winklebury Camp (Jones 1976b; 1977) and one from Houghton Down (Poole 2000c). However, the sample size is very small. There is a difference in the contexts partial and complete domestic fowl are deposited within. Partial domestic fowl ABGs have been recovered from graves. The presence of complete domestic fowl in pits, might tie in with Caesar’s comments that suggest domestic fowl were not consumed by ‘native’ Britons during the Iron Age (Green 1992, 125). Horse and cattle show a pattern dominated by partial ABGs. Only four complete horse ABGs are recorded from the study area for this period, two from Danebury (Grant 1984a), one from Tolpuddle Ball (Hamilton-Dyer 1999b) and one from Berwick Down (Bird 1968). Other complete horse ABGs are known from outside the study area (see Bendrey et al. 2010), but the majority of horse ABGs consist of partial remains (Table 4.4). The majority of cattle ABGs also consist of partial remains. Of the 99 cattle ABGs, only fourteen are complete skeletons. Most of the complete cattle ABGs are from Three ABGs from Danebury do not have age data recorded. The two complete cattle ABGs from Gussage All Saints consist of a mother and calf. Both must have Danebury, where ten are recorded. Of the remaining four, two come from Gussage All Saints (Harcourt 1975; 1979b), one from Suddern Farm (Poole 2000d) and one from Winklebury Camp (Jones 1976b; 1977). The Suddern Farm and one of the Gussage All Saints deposits are adult cows, the rest where noted, were neonates died during labour as the calf remains were discovered half way through the pelvic cavity (Figure 4.8). In contrast the complete horse ABGs are all from adult individuals. This would correspond with Harcourt’s (1979b) suggestion that horses were not bred during the Iron Age but captured from wild populations and trained.

Red deer, the largest wild herbivore recorded in the assemblage, consists of only one complete ABG, in the unique deposit at Winklebury (Smith 1977). Red deer elements are rare on Iron Age sites and often consist of the remains of antler (Maltby 1981b; 1996). For example, 15 of the 16 red deer elements recovered from Maiden Castle are antler fragments (Armour-Chelu 1991). Therefore red deer may have supplied only a very small proportion of meat to the Iron Age diet. It is interesting that the carnivorous wild mammal ABGs were mainly deposited as complete skeletons. From the point of view of our westernised diet, it would make sense for carnivorous mammal ABGs, not required for food, to be deposited complete, and hares, animals we still eat today, to be deposited more frequently as partial ABGs, possibly due to butchery having taken place. However, we cannot discount that wild mammals such as the pine martin and stoat were eaten. The wild bird ABGs show a different pattern to wild mammals with the majority deposited, or surviving only as partial ABGs. The only exceptions are a complete raven deposited spread-eagled on the bottom of a pit at Winklebury Camp (Jones 1976b; 1977) and a complete 48

Chapter 4. Iron Age Southern England Table 4.4 Number of complete, partial or unknown ABGs per. The Sheep/Goat, (S/G) counts include all ABGs identified to either species Percentage Species Complete Partial Unknown Total partial Domestic Cattle 14 98 1 113 87% Mammal S/G 42 203 16 261 78% Pig 33 27 6 66 41% Horse 4 93 97 96% Dog 32 81 20 133 61% Wild Fox 14 1 15 7% Mammal Cat 6 1 7 14% Pine Martin 1 1 0% Stoat 1 1 0% Weasel 2 2 0% Red Deer 1 1 0% Hare 2 3 5 60% Domestic Domestic bird Fowl 3 2 1 6 33% Wild bird Raven 2 18 11 31 58% Rook/Crow 3 3 100% Buzzard 1 1 100% Cormorant 1 1 100% Lark 1 1 100% Other Snake 1 1 100%

Figure 4.8 Plan of cow and calf ABGs from Gussage All Saints. FL=forelimbs of foetus, FS=foetal skeleton, SK=skull and X=position of hind limbs (G. J Wainwright 1979, Figure 109).

49

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 4.9 Top: Percentage of complete sheep/goat ABGs by age. Bottom: Percentage of partial sheep/goat ABGs by age hillfort settlements by Bullock and Allen (1997) and Hamilton-Dyer (1999b), particularly from Dorset, occurring in small shallow pits often with no other associated finds. Complete sheep/goat ABGs have also been recorded from hillfort and non-hillfort sites in Hampshire and Wiltshire. This trend continues into the early Romano-British period where 30% of the sheep/goat ABGs consist of complete skeletons. There is no significant change in the site types or features in which sheep/goat ABGs were deposited to explain the increase in the proportion of complete ABGs in the late Iron Age. Buckland-Wright (1987) suggests that the animals at Poundbury had succumbed to disease or, in the case of young animals, natural death. However, a number of the sheep/goat ABGs from other sites (Danebury, Suddern Farm, Barton Field, Viables Farm) have been given a ritual interpretation (see 11.4.1).

Like horses and cattle, the majority of sheep/goat ABGs are partial. This appears to be a continuation of the trend seen in the Bronze Age. A small number of complete sheep/goat are recorded from the early and middle Iron Age (Figure 4.10). Only two of the complete ABGs are identified as goat, one dating to the early Iron Age from Danebury, the other from a late Iron Age deposit at Poundbury (Buckland-Wright 1987). As observed for cattle, the majority of the sheep/goat ABGs are from young animals. Of the 29 complete sheep/goat ABGs with ageing data present, only one (from a late Iron Age pit at Suddern Farm) is from an adult animal (Poole 2000d). As Figure 4.9 shows, the majority of complete sheep/goat ABGs of known age are from neonatal or juvenile animals. This differs significantly from the age of sheep/goat deposited as partial ABGs. Although a number of neonatal and juvenile partial ABGs have been recorded, a substantial proportion are from adult animals. For example, 39% of the partial sheep/goat ABGs from the middle Iron Age are from adults (Figure 4.9).

Pig and dog show a very different pattern to the other species discussed, with much higher proportions found as complete ABGs. Complete ABGs represent 58% of the total pig Iron Age assemblage (Table 4.4). Surprisingly, considering the emphasis placed on dogs by Green (1992, 111-113) and Smith (2006), complete pig ABGs are more common than complete dog ABGs. This may partly be the result of underestimating the number of complete dogs in multiple neonatal dog burials.

A different pattern is seen in the late Iron Age, where 38% of the sheep/goat ABGs consist of complete skeletons (Figure 4.10). As with the preceding subperiods, the majority of complete sheep/goat ABGs are from younger animals (Figure 4.9). Complete sheep/goat ABGs have been noted at a number of late Iron Age non-

50

Chapter 4. Iron Age Southern England

Figure 4.10 Percentage of complete ABGs per species for the early, middle and late Iron Age

Figure 4.11 Top: Percentage of complete pig ABGs by age. Bottom: Percentage of partial pig ABGs by age from young individuals (Figure 4.11). Only one complete adult pig ABG was recorded, from an early Iron Age pit at Suddern Farm (Poole 2000d). Similarly to sheep/goat, more of the partial pig ABGs are from adult individuals, although it should be noted that the sample size for partial pig ABGs of known age is small. A noticeable difference between the complete pig and complete cattle and sheep/goat ABGs is that the pig deposits are often found in groups. For example, at Danebury, Houghton Down and Nettlebank Copse, groups of neonatal complete pig

For example, 26 partial dog ABGs were recorded from Balksbury Camp, the majority of a very young age, which Maltby (1995a) interpreted as burials of newborn puppies in pits. Therefore, although the majority of the ABGs are partial, they were thought to be deposited as complete animals, many of whose bones had not survived or been recovered. As in the case of the complete cattle and sheep/goat ABGs, the majority of complete pig ABGs are 51

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 4.12 Top: Percentage of complete dog ABGs by age. Bottom: Percentage of partial dog ABGs by age

Figure 4.13 Percentages of body areas represented from partial cattle, horse, sheep/goat, pig and dog ABGs If dog and pig deposits do represent culling of litters or natural deaths soon after birth, this would fit with the biology of these species. Cattle normally give birth to one calf, and Soay sheep, the breed most closely related to the sheep which would have been present during the Iron Age, often have single births (Clutton-Brock et al. 1992). By contrast wild pigs and dogs have an average litter size of around six (Okkens et al. 1993; Wood and Barrett 1979). Therefore if a whole litter of dogs or pigs was to die soon after birth and be deposited as complete carcasses, we could expect to find multiple associated ABGs. This appears to be case for dogs in this and the following Romano-British period, but multi-pig deposits

ABGs were discovered in the same context, whereas the cattle and sheep/goat ABGs were found isolated. A similar pattern exists with dog ABGs. As discussed above, a number of deposits of partial neonatal dog ABGs have been recorded from sites such as Balksbury Camp, and are often interpreted as the disposal of members of an unwanted litter originally deposited as complete individuals. If these deposits do represent the disposal of groups of complete neonatal puppies, then it would appear that, like pig neonatal ABGs, dog ABGs were also often deposited in groups, whereas cattle and sheep/goat neonatal ABGs are deposited in isolation.

52

Chapter 4. Iron Age Southern England In total, 59 ABGs were recorded with butchery marks present, which represents just 7.7% of the Iron Age ABG assemblage. All of the butchery marks were made by knives.

are much rarer. One of the main differences of the complete dog ABGs compared with other species is the number of deposits derived from adult individuals (Figure 4.12). As already discussed, the majority of cattle, sheep/goat and pig complete ABGs are from neonatal individuals. Although some of the neonate partial dog remains may have been deposited as complete skeletons, there are still a larger number of adults represented compared to any other species. It therefore appears that the remains of adult dogs were treated differently to adults of other species.

Butchery marks have been recorded from only one wild mammal (Table 4.5), a complete fox ABG from Nettlebank Copse. Knife marks are present on the front and back lower limbs (Table 4.6) suggesting that the animal was skinned before the remains were deposited. No other butchery marks were observed, indicating the carcass was not dismembered or filleted. Therefore it was probably deposited soon after being skinned. As the majority of wild mammals are deposited complete, and no butchery marks are reported except for the fox ABG, it seems likely that these animals were not commonly processed for primary products. Only one wild bird, a raven from a middle Iron Age deposit at Danebury, has butchery marks present. They consist of knife marks on the wing bone, possibly resulting from the removal of the flight feathers before deposition. The majority of wild bird ABGs consist of partial skeletons, but no butchery marks indicating dismemberment have been recorded. However, it is not necessary to utilise a knife for the disarticulation of bird remains, as it is possible to dismember a bird skeleton by hand, especially if cooked.

A number of different patterns are also present when the body parts which constitute partial ABGs are investigated. Cattle, horse and sheep/goat all show a similar pattern, with the remains from just the appendicular skeleton being the most common, making up, for example, 70% of all the partial horse ABGs (Figure 4.13). A much smaller proportion of partial cattle, horse and sheep/goat ABGs consist of just elements from the axial skeleton. A very small number of cattle, horse and sheep/goat are mixed ABGs (consisting of elements from all body areas, head, axial and appendicular). The pattern for partial pig and dog ABGs is very different; 35% of partial pig and 30% of partial dog ABGs consist of mixed deposits (Figure 4.13). Partial remains consisting of just elements from the appendicular skeleton are the second most common type of pig and dog ABGs. The high percentage of partial pig and dog ABGs consisting of mixed deposits reflects the deposition of neonatal litters from the species.

Butchery marks have been recorded on some ABGs of all domestic mammal species except pigs. As discussed above, a high proportion of pig ABGs consist of complete skeletons, mainly from neonatal or juvenile animals (Figure 4.11). The lack of butchery marks combined with the number of complete ABG deposits could indicate that these ABGs were deposited without being utilised for primary products. A number of partial pig ABGs are present, however, and it is possible these were subjected to some form of butchery that left no marks. Only a small proportion of the other partial domestic mammal ABGs have recorded butchery marks (Table 4.5). However, the lack of butchery marks does not necessary mean that no processing took place, and the composition of most ABGs suggests that some form of butchery would have taken place. The problem with butchery analysis is that a skilled butcher can process a whole animal and leave no trace of a cut mark.

As discussed above, a number of deposits from these species had to be recorded as partial remains, but were probably deposited as complete animals before postdepositional activity disturbed them, creating partial ABGs. The differences in body areas represented between cattle, horses and sheep/goat, on the one hand, and pigs and dogs, on the other, may be a consequence of this depositional activity. Therefore most of the partial ABGs for cattle, horse and sheep/goat may have been created by human activity prior to or at the time of deposition, as opposed to being created by postdepositional taphonomic events. If this is the case, then the patterning seen for pigs and dogs in Figure 4.13 may provide a means of recognising deposits in which complete carcases have been deposited, but the remains have become disassociated. 4.6

Table 4.6 shows the number of butchery marks from different body areas. In total, 75 occurrences were recorded with some ABGs having marks on bones from more than one area. The axial elements, followed by the lower back limb and the front limbs are the most common areas with butchery marks present. The marks from different areas indicate that the individual skeletons were subjected to a number of butchery processes before being deposited as an ABG. If we combine the front and back limbs, then 33.3% of the marks observed occur on these elements. Many of these marks appear to result from the skinning process and/or the removal of the feet from the upper limbs and normally occur around the carpals/tarsals or metapodials.

Butchery

As the majority of ABGs from the Iron Age consist of partial skeletons, they could have been created by one or more of a number of processes. Dismemberment of the individual animals would have taken place naturally or by human activity, or both. Butchery is one of the main activities likely to have created partial ABGs. To investigate this further, butchery marks were recorded when present.

53

Investigating Animal Burials; Ritual, Mundane and Beyond Table 4.5 Number of ABGs with butchery marks present Early Iron Middle Iron Late Iron ABG type Species Age Age Age Total Complete S/G 2 1 3 Dog 2 1 1 4 Fox 1 1 Partial Cattle 1 12 13 S/G 4 8 5 17 Dog 3 4 1 8 Horse 4 4 3 11 Raven 1 1 Unknown Cattle 1 1 Total 15 19 25 59 Percentage of whole assemblage 9.4% 5.5% 9.7% 7.7%

Percentage of species ABG type 7.0% 11.1% 7.1% 12.3% 7.9% 9.3% 11.8% 5.6% 100.0%

Table 4.6 Number of butchery marks recorded per species and body area Upper Lower Upper Lower fore fore hind hind Period Species Head Axial Pelvis limb limb limb limb Early S/G 4 1 2 1 Iron Horse 1 3 Age Dog 2 1 3 3 Middle Cattle 1 Iron S/G 2 1 2 3 Age Horse 2 1 1 Dog 2 3 1 Raven 1 Late Cattle 2 2 1 1 5 1 2 Iron S/G 3 4 1 1 1 Age Horse 3 1 Dog 1 1 Fox 1 1 Total 6 17 5 12 10 8 15 Total Percentage 8.0% 22.7% 6.7% 16.0% 13.3% 10.7% 20.0%

Unknown

1 1

2 2.7%

Table 4.7 Number of instances of inclusion in multi-ABG deposits per species. The number in brackets indicates the number of multi-ABG deposits Early Iron Age Middle Iron Age Late Iron Age Period (9) (5) (24) Total Cattle 1 8 23 32 S/G 6 9 41 56 Pig 6 1 1 8 Horse 5 8 3 16 Dog 11 2 9 22 Cat 5 5 Fox 12 12 Red Deer 1 1 ABGs total 29 28 95 153 Percentage of ABGs 18% 9% 37% 20%

54

Chapter 4. Iron Age Southern England the highest number of different species. Normally, ABGs from small multiple ABG deposits are from the same species.

The butchery marks observed on the upper limbs normally occur on the extremities of the ABG (for example on the proximal end of the femur or humerus, and the distal end of the radius/ulna or tibia) and result from dismemberment of the ABG elements from the rest of the skeleton. The majority of the axial skeleton butchery marks are observed on vertebrae or ribs and are often interpreted as the result of filleting meat from the elements or dividing the trunk into two roughly equal halves.

Two of the large deposits contain the ABGs from just two species. One from Winklebury Camp, consists of 12 foxes and one red deer (see above). The other is an early Iron Age deposit in pit 226 from New Buildings (Poole 2000d), which contains seven dog ABGs and one horse ABG. The dog ABGs consist of a complete adult and six neonatal puppies. It is unknown whether they are complete. The horse is a partial ABG comprised of six elements from the left lower back leg, with skinning marks present. The horse ABG is also deposited in association with iron ore, which is thought to be metalworking waste. The largest multiple deposit consists of 15 ABGs from the middle Iron Age pit 197, from Suddern Farm (Poole 2000d). This deposit has ABGs of five cattle, five horses, four sheep/goat and one pig (see 12.6.2). This is an unusual deposit as the majority of other ABG deposits from this site are found in isolation. The sediment matrix and the excellent preservation of the ABG elements indicate that the deposit was created over a short period. The majority of the ABGs from this deposit consist of elements from the axial skeleton, whereas it is more common for ABGs in this period to be comprised of appendicular elements. The ‘normal’ faunal assemblage from this feature also contains a large number of elements from the axial skeleton. This could indicate that more ABGs of a similar nature were deposited in the pit and became disarticulated due to post-depositional movement, and/or that the ABG and ‘normal’ faunal assemblage are derived from the same event.

The analysis indicates all major butchery processes (except breakage for marrow extraction) are evident amongst ABGs. It should be borne in mind that the butchery marks discussed are from only a small proportion of the ABG assemblage. However, butchery marks are also only rarely observed in ‘normal’ faunal assemblages. For example, only 3.3% of the 12,000 fragments forming the assemblage from Suddern Farm had butchery marks present (Hamilton 2000), and the Danebury assermblage has a similarly low proportion of butchery observations (Knight 2003, 290) . Although butchery marks are not observed on the majority of the partial ABGs, it is highly likely that some butchery would have taken place during their creation. 4.7

ABG associations

As discussed above, some of the ABGs from the Iron Age are found in close association with other finds such as human remains as well as other ABGs. 4.7.1

As well as producing one of the largest multiple ABG deposits, one of the ABGs in this pit is associated with the deposition of a quern fragment. A partial sheep ABG, consisting of the spine and ribs, referred to as ‘animal deposit R’ in the microfiche (Poole 2000d, 5:C4), was deposited near to the side of the pit, with a quern fragment lying between it and the side of the pit. No other close associations between Iron Age ABGs and quern stones have been recorded. However, the recording of ABGs deposited in association with other material types is reliant on the information being available. The Danebury Environs excavations are the best sites to obtain such information, as details of ABGs and other deposits deemed ‘special’ are available. We must be aware that such records are inherently biased, as they rely on the excavators deciding a find is ‘special’ for it to be included.

Multiple ABGs

Of the 764 ABGs from this period, 153 (20%), are from just 38 multiple ABG deposits. The early and middle Iron Age deposits are each from four sites, with the late Iron Age multiple deposits coming from thirteen sites. Dogs are the most common species to be recovered from early Iron Age multiple ABG deposits in contrast with the overall ABG species proportions (Figure 4.1, Table 4.7). Sheep/goat are the most common species to be included in multiple ABGs for the middle Iron Age, closely followed by cattle and horse. The proportion of sheep/goat in multiple ABG deposits increases in the late Iron Age (Table 4.7). The number of ABGs included in each multiple deposit is variable. In some cases in all sub-periods they consist of just two ABGs (Table 4.8). In contrast, three multiple deposits, one from the middle Iron Age and two from the late Iron Age, contain between ten and fifteen ABGs. The majority (63%) of the multiple ABG deposits are from just one species (Table 4.9). A small number include ABGs from three or more species. There is a positive correlation between the number of ABGs and the number of different species. For example, the deposit from Suddern Farm (see below), which has the highest number of ABGs recorded from a multiple ABG deposit, also has

4.7.2

Associated deposit groups

Including the Suddern Farm pit 197 deposit, 25 associated deposit groups (ADGs) were recorded from the Iron Age from eleven different sites. ABGs from grave contexts have not been included in this discussion as formal burial is a late Iron Age tradition that continues into the Romano-British period. Those deposits are 55

Investigating Animal Burials; Ritual, Mundane and Beyond Table 4.8 Numbers of ABGs in multiple ABG deposits Period 2 3 to 5 6 to 10 10 to 15 Early Iron Age 4 (44%) 4 (44%) 1 (11%) Middle Iron Age 3 (60%) 1 (20%) 1 (20%) Late Iron Age 13 (54%) 6 (25%) 3 (13%) 2 (2%) Table 4.9 Number of species represented in multiple ABG deposits Period 1 species 2 species 3 species 4 species Early Iron Age 5 2 2 Middle Iron Age 3 1 1 Late Iron Age 16 7 1 Total 24 (63%) 9 (24%) 4 (11%) 1 (3%)

Table 4.10 Multi-ADG deposits from southern England Iron Age sites Site Material Species Human Bone Articualated (partial), Human Bone Pig (1), Dog (1), Horse Early Iron Age Suddern Farm Unarticualted, Pottery sherds (1) sheep/goat (1), Pig (2), Suddern Farm Pottery sherds, Daub Dog (1) Pottery sherds, Flint object, Middle Iron Age Houghton Down daub, chalk block Dog (1) Late Iron Age Flagstones enclosure Human bone unarticulated Cattle (1), Horse (1) Flagstones enclosure Pot Complete Cattle (1), Dog (1) Hod Hill Human bone Articulated Sheep/goat (1), Pig (1) Cattle (5), sheep/goat Viables Farm Human bone Articulated (2), horse (2) Period

Table 4.11 Number of species associated with other material deposits. From 25 separate Iron Age ADGs from southern England. HB=Human bone, Art=Articulated, unart=unarticulated, obj=object, Pot C= Complete pot, Pot S= Pottery fragment HB HB Bone Flint Pot Pot Quern Chalk Metal Period Species art unart obj obj C S frag Daub block waste Total Early Iron Age

Middle Iron Age Late Iron Age

Total

S/G Pig Horse Dog Domestic fowl S/G Dog Cattle S/G Pig Horse Dog

1 1 1 1

1 1 2

2 1

3 3 1 3

1 1 1

1

11

5

1 1

1 1

1 5 2 1 2

1 2

1

1 1

1 2

2 15

2 3

1 1

1

5

2

1 1

7

56

1

7 7 3 6

1

4 8 2 1 4 3 50

Chapter 4. Iron Age Southern England or indeed other objects, ABGs are associated with. The majority of the pottery fragments recorded as ADGs are from the base of a pot. Only one ADG, from Flagstones (Bullock and Allen 1997), has been recorded that includes a complete pot. Hill (1995) views the deposition of querns as a rare event that may have had ‘ritual’ connotations due to the longevity of their use and associations with agriculture and transformations. However, only one ADG involving the deposition of a quern fragment is recorded. The rest of the material/objects associated with ADGs could also be considered as ‘normal’ rubbish.

therefore discussed in the southern England RomanoBritish chapter (see 6.7.3). Seven of the 25 ADG deposits are multiple ADGs (Table 4.10). Of the five ADG deposits recorded involving human remains (excluding those deposited in formal graves), four are multiple ADG deposits. Only one deposit from Nettlebank Copse (Poole 2000b) consists of human remains (cranium fragment) and a single ABG (a complete butchered dog). Although sheep/goat ABGs are the most common type during the Iron Age, only three examples are recorded in association with human remains. Of the 17 animal ABGs associated with human material, thirteen have element data available. These indicate that the majority consist of elements from limb bones, with the exception of dog ABGs and the Viables Farm deposit (see below). Dog ABGs show a different pattern, as all those associated with human remains consist of complete skeletons.

4.8

Summary

As with the Neolithic and Bronze Age assemblages from southern England, the majority of ABGs from Iron Age sites in this region are from domestic mammals. However, a larger range and proportion of wild mammals and birds are represented. Ravens are the most common of these, the majority of which are recorded from Danebury. In comparison to the previous periods, the Iron Age assemblage is much larger, contributing 41% of the overall southern England dataset. However, it is dominated by the assemblages from Balksbury Camp, Danebury, Suddern Farm, Winnall Down, Old Down Farm, Owslebury and Nettlebank Copse, which make up 62.8% of the ABG assemblage for this period and region. Sheep/goat are the most common ABG species, which corresponds with their high abundance in the non-ABG faunal assemblage. However, the proportion of sheep/goat ABGs compared with those of non-ABG sheep/goat remains does differ, as do the proportions of the other species, with dog and cattle the second and third most common ABG species respectively. There are some differences between hillforts and non-hillfort sites, with sheep/goat ABGs more common on hillforts. There is also some variation in the types of feature specific species are recovered from. These and other results contradict those proposed by Wait (1985).

The largest deposit of ABGs associated with human remains comes from pit 5 at Viables Farm. The ABGs are all associated with Inhumation 2, an adult female, 25-30 years old (Millett and Russell 1982). Four of the five cattle ABGs recorded are possibly from the same individual; these separate ABGs include the left and right lower back limbs, ribs with the cervical and thoracic vertebrae, and the lumber vertebrae and pelvis, which were found, articulated, south of the left shoulder of the human inhumation. Although these ABGs may be from the same individual, knife marks, possibly due to disarticulating the skeleton are present on the vertebrae and pelvis, indicating that the separate ABGs had been created before deposition. The two horse ABGs consist of elements from the head and axial skeleton, again possibly from the same individual. The two sheep/goat ABGs from this pit are likely to be from sheep and both are complete. One was discovered beneath the chest of Inhumation 2, partially disarticulated due to slumpage (Maltby 1982b). The other is possibly from a ram or castrated male and was found with the head of the inhumation resting on the neck of the ABG. It is interesting that the association of ABGs with human remains from pit contexts appears to be so rare. It has been argued that the similarity in treatment between human remains and ABGs suggests that ABGs must be of a ‘ritual’ nature (see 1.2.8). Russell (2010) has shown that there are differences in the species proportions and taphonomic preservation of the ‘normal’ faunal assemblages between Iron Age pits with and without human material present. She shows there is a higher proportion of pig and cattle remains in features with articulated and disarticulated human remains present, a pattern also seen in the ABG data. However, most ADGs do not involve human remains.

The composition of ABGs also differs by species. The majority of wild mammals recovered consist of complete skeletons. In comparison with the other domestic species a high proportion of the pig and dog ABGs have also been recovered as complete skeletons. With the exception of dog, the majority of the complete domestic species ABGs are from neonatal or juvenile animals. The majority of partial dog ABGs are from neonatal animals, which may have been complete when deposited but have became disarticulated and partially destroyed due to taphonomic and retrieval processes. The partial ABGs of other species show a different pattern including a high proportion of adult animals. Although the majority of ABGs are partial, only 7.7% have recorded butchery marks, most associated with disarticulating the skeleton. However, it would seem likely that the majority of partial ABGs have been created by a butchery process.

Twenty-five ABGs have been recorded in association with other materials, with pottery fragments being the most common (Table 4.11). However, this sets us a problem, as how to interpret fragmented pottery remains, 57

Investigating Animal Burials; Ritual, Mundane and Beyond In contrast to the previous period, a large percentage of ABGs, 20% (152), were recovered from just 32 multiple ABG deposits. Sheep/goat and cattle are the most common species in such deposits. Most multiple ABG deposits consist of just two ABGs, but a few very large deposits, containing as many as 15 associated ABGs are present. The majority of multiple ABGs contain just one species. Only a small number of ABGs have been recorded in association with other material types, with pottery and human remains the most common. Although complicated by a number of factors, including the quality of the dataset, the Iron Age results from southern England have shown that the ABG record is much more variable in its composition than would be expected from the previous literature on the subject.

58

Chapter 5. Iron Age Yorkshire 5.1

Introduction

The ABG data from Yorkshire are limited compared to the data available from southern England. In total, 22 sites were recorded from Yorkshire dating to the Iron Age, of which 11 had ABGs present. The majority of the sites date to the late Iron Age period, with one of the ABG sites, and four of the non-ABG sites dating to the Iron Age-Roman transition period. Only four ABGs were recorded from middle Iron Age contexts. In total, 38 ABGs are recorded from Yorkshire for this time period, representing 24% of the ABG sample for this region. This differs from southern England where a large proportion of the ABGs recorded date to the Iron Age (see 4.1). The large difference between the Yorkshire and southern England ABG assemblages are mirrored in the ‘normal’ faunal assemblages. In 1995 Stallibrass states; ‘we are still almost totally ignorant of the preexisting Iron Age economies and faunal environments for the region [north England]’ (Stallibrass 1995b, 131) Although further faunal assemblages have since been recovered, analysed and published, the depth of the faunal record is still small compared to the south of England. For example, only six sites from Yorkshire had large enough faunal assemblages to be included in Hambleton’s (1999) study of Iron Age animal husbandry. The majority of the sites used in this study are from the Yorkshire Wolds and Permian ridge regions. These are the areas with the best bone preservation conditions due to the underlying cretaceous (virtually all chalk) and Permian rocks (Gaunt and Buckland 2003). However, as discussed previously, Yorkshire was chosen as a region to provide a contrast to the often used ‘Wessex’ region. Although a much smaller number of sites are recorded from Yorkshire, there are some corresponding patterns with the southern England results. A much higher percentage of Iron Age sites have ABGs recorded as present compared to the Bronze Age results. This is probably due to the increase in the number of settlement sites found in the Iron Age. Half of the Iron Age sites from Yorkshire have ABGs recorded as present. Therefore, as with the results from southern England, ABGs appear to be common on sites in this period. This has also been noted by other reviews of this region (e.g Manby et al. 2003b, 271). 5.2

Species proportions and context

All the ABGs recorded from the Iron Age of Yorkshire were domestic mammals, with cattle and pig the most common species. No ABGs were recorded from early Iron Age contexts, which means there is a large gap in the ABG record, with none recorded by this project from the late Neolithic to middle Iron Age. In comparison 20 and

18 ABGs were recorded from both the middle and later Iron Age (Table 5.1). There is a distinct difference in the species proportions for the middle and late Iron Age assemblages. However, this reflects a difference in the type of sites the ABG data comes from. The middle Iron Age data comes from the Hasholme log boat and a number of Arras culture graves. All the cattle remains from the middle Iron Age assemblage are recorded from the Hasholme log boat (see below). All the late Iron Age ABGs are recorded from settlement sites with cattle dominating the assemblages. Although some funerary sites were still in use in the late Iron Age, unfortunately there is no ABG data present for them. Although data are sparse, settlements such as Garton and Wetwang Slacks (a settlement between the slacks of Garton and Wetwang) (Brewster 1980) are shown to be contemporary with later Iron Age Arras culture burials in close proximity to the settlement’s boundary. What the settlement and funerary sites used in this study are showing is the difference between the ‘domestic’ and ‘funerary’ context. Although such concepts are increasingly seen as unrealistic for Iron Age sites in southern England (Fitzpatrick 1997), they may hold true for the Arras culture landscape where domestic and funerary activities, or at least deposition, appear to have been separated (Bevan 1999). Consequently the ABG data show a very different species pattern for those deposited in association with human remains compared to those deposited in ‘domestic’ features. The emphasis on cattle and sheep/goat ABGs from settlement sites ties in with the general trends seen in the faunal assemblage data. If we look at cattle, sheep/goat and pig NISP counts for settlement sites in Yorkshire, we see a variety of patterns (Figure 5.1). Pig remains are not common on any of the sites, which is also true for the ABG data. For the majority of the sites, remains of cattle constitute over 50% of the assemblage (of these three species), which, again, is a trend seen in the ABG data. This differs from the sheep/goat dominated pattern seen both from the ‘normal’ and ABG assemblages in southern England. Only two sites, Garton and Wetwang Slack (68%) and Dalton Parlours (71%) have high percentages of sheep/goat in the faunal assemblage. Interestingly, both only produce two sheep/goat ABGs each, with Garton and Wetwang Slack producing a large number of cattle ABGs (see below). Therefore, although some trends are similar among the ‘normal’ and ABG assemblages, this differs from site to site. Just because one species is the most common in one type of assemblage does not mean it will be in the other. Unlike the southern England assemblage, the majority of Yorkshire sites produced a small ABG assemblage. Only, one site, Garton and Wetwang Slack, produced more than 10, with 12 recorded from the site. The majority of sites,

Investigating Animal Burials; Ritual, Mundane and Beyond Table 5.1 Number of ABGs per species for the Yorkshire Iron Age. The Sheep/Goat, (S/G) counts include all ABGs identified to either species Species Middle Iron Age Late Iron Age Total Cattle 4 10 14 S/G 1 4 5 Pig 15 1 16 Dog 3 3 Total 20 18 38 Table 5.2 Number of ABGs recorded from different site types per species Period Site type Cattle S/G Pig Dog Total Middle Iron Age Funerary 1 15 16 Other 4 4 Late Iron Age Settlement 10 4 1 3 18

Figure 5.1 Tripolar diagram showing the percentage of cattle, sheep/goat and pig for non-funerary sites and including data from Stanwick (Rackham forthcoming) southern England, but the number of ABGs is also small in comparison. Such a trend is also seen from funerary sites. However, this is to be expected as the NISP counts include the ABGs. Also ‘normal domestic refuse’ is not thought to have been deposited on these sites. For example, the excavations at Garton Station produced a small assemblage of 23 animal bones, all of which come from a partial pig ABG associated with a cart burial (Legge 1991a, 142).

72%, produced between one and four ABGs (Table 5.3). Such small ABG assemblages appear in this instance to correspond with the size of the faunal assemblages for the settlement sites. For example Dalton Parlours (4) and Garton and Wetwang Slack (12) produced the highest number of Iron Age ABGs from settlement sites, and also have some of the largest assemblages of identified fragments at 1,011 and 2,916 respectively. Such assemblages are small in comparison with those from

60

Chapter 5. Iron Age Yorkshire Table 5.3 Number of ABGs found on different sites in Yorkshire Period Middle Iron Age Late Iron Age Percentage Sites Percentage ABGs 5.3

Site type Funerary Other Settlement

1 2 2 40% 11%

Hasholme logboat

2 to 4 2 1 1 40% 32%

5 to 10 1 10% 24%

10 to 15 1 10% 32%

beached on a low tide, or could have been sunk in deeper water. The position of the boat within the marine channel and the depth of water at the time (0.9m at its lowest), suggests that the boat was sunk and not lost through a mooring accident (Millett and McGrail 1987, 146). A portion of the upper bow is displaced, but found close to the main part, which is suggested to represent a failed attempt to dismantle the boat or haul it from the estuary’s bed. If the boat had been purposely sunk as part of a ceremonial activity, there would be no reason to try and recover it. It therefore seems that the ABGs represent cargo from the boat, showing that parts of animal carcasses were transported along the water way, either for trade or as a food supply for a journey.

The only middle Iron Age ABG recorded comes from the chance find of a log boat. This was discovered at Hasholme during drainage ditch cutting through peaty deposits, which represented a silted-up river channel (Millett and McGrail 1987). Tree ring analysis and the marine transgression indicated the boat dated to between the mid-3rd and mid-2nd centuries BC. Four partial cattle ABGs were discovered in close association with the log boat. Individual one consisted of 32 elements from all four legs of a young adult, possibly male. Individual two is a sub-adult and is represented by 15 elements from both the left and right back legs. Two further ABGs were found. One consisted of 12 right side ribs found in a close bundle. The other comprised 15 vertebrae (cervical, thoracic and lumber) with the spine split sagitally and only the right half of each vertebra present. It is unknown if the rib and vertebra ABGs are from the same individual, but it seems likely that they were from either individual 1 or 2 (Stallibrass 1987). They had already been butchered prior to being placed within the boat, and therefore represent separate ABGs.

5.4

Funerary sites

As discussed above, a large number of the middle Iron Age ABGs come from funerary sites (Table 5.2), although possibly we should not view these as separate sites, but more separate contexts. The majority of these data come from the Yorkshire Wolds and as shown above, there is a close relationship between settlements and funerary monuments, in the form of square barrows. The same people would have used both areas. The main difference between the ABGs from these two areas is their association with human remains. All of the ABGs discussed in this section are found within human graves. Five separate funerary sites are recorded in this study with ABGs present, four are from Stead’s (1991) excavations around the Yorkshire Wolds area, north and west of Driffield. The other site is Grindale square barrow II, North Yorkshire (Manby 1980). Within the central grave at Grindale was a complete juvenile pig ABG. The site had been disturbed in the 19th century so limited information was available. However, the presence of pig ABGs appears to be a trend with square barrows.

The majority of the elements constituting the ABGs are all found in tight anatomical groups (Figure 5.2), with most discovered around the bow of the boat (the northern part). However, a number of the lower limb elements from individuals one and two were discovered towards the stern (the southern part). The boat was found upside down in the sediment and the excavators believe that the cattle parts were stored towards the bow and remained in place when the boat was sunk. However, as the flesh around the bones decomposed, some of the lower foot elements became disarticulated from the upper leg, and were light enough to be washed further downstream, but became caught in the stern of the boat (Millett and McGrail 1987, 144). Butchery marks are present on the ABGs around areas of articulation. None of the marks indicate that meat was stripped from the bone. This suggests that they represent joints of a carcass with meat still attached. A small number of other animal bones were discovered near the log boat, including a sheep/goat skull, a canine-gnawed cattle femur and a few fragmented long bones. These are interpreted as refuse thrown into the river from the banks (Stallibrass 1987).

Of the 15 pig ABGs recorded from Yorkshire all but one are associated with square barrows. All the other pig ABGs recorded from square barrows are partial. The majority of the partial pig ABGs are composed of elements from the head and leg (Table 5.4). Elements of the axial skeleton appear to be rare; only two ABGs, one from Garton Station and the other from Kirkburn consist of mixed deposits. Pig ABGs consisting of just the axial skeleton do not appear to have been recovered from square barrow sites.

A number of reasons for the sinking of the boat have been put forward, which affect the interpretation of the ABGs. The boat could have been abandoned, could have been

The majority of graves did not have ABGs present. For

61

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 5.2 Plan of the Hasholme Logboat showing overall bone distribution (Millett and McGrail 1987, Figure 30) 1991a). There is also a possible association between sheep/goat and pottery vessels being placed within graves at Rudston (Stead 1991, 177). Such deposits were not included in this study as they do not fullfill the ABG criteria (see 1.4). This though raises the interesting pattern that all but one of the ABGs from these grave groups are from pigs, whereas the majority of the single bone deposits are from sheep/goat. If, as argued by Legge (1991a) and Stead (1991), such deposits represent gravegood food offerings, it shows a difference in the treatment of the offering dependant upon the species it derives from. Also, Legge (1991a) suggests that the butchery seen on the ABG pig skulls infers they were deposited as bones with most of the flesh removed, also some of the ABGs may have been recreated with elements not in their correct orientation. He therefore sees these deposits as symbolic food offerings, consisting of just the bone and connective tissue. Of course the removed meat may have been deposited within the grave as well but would not be traceable. If this were the case, it also shows a difference in the treatment of the bodies and bones of different species.

example, only one pig ABG was recovered from Burton Fleming, from a total of 58 graves and of the nine burials at Garton Station only one, a cart-burial, contains a pig ABG. This consists of the left and right halves of the skull, both left and right forelimbs and seven ribs, all from the same individual. Butchery marks on the right scapula indicate the limbs were dismembered from the carcass before deposition, which would also explain the lack of vertebrae. The ribs and head would also have been dismembered from the rest of the carcass before deposition. The ABG was then placed over the human skeleton (Figure 5.3). However, only the animal remains deemed as ‘grave goods’ were recorded. ‘Normal’ fragments within the graves were not. Therefore we do not know if other parts of the individual were deposited within the grave, but not within proximity of the human remains, or became disturbed due to post-depositional movement. The majority of ABGs associated with square barrows come from Rudston Makeshift. However, this may be due to the overall sample size, as 176 graves were excavated, with 5.1% (9) containing pig ABGs. Again, the majority of the pig ABGs consisted of elements from the head and front limbs. Only one ABG is recorded that is not from a pig, and that is a sheep/goat ABG from Burton Fleming grave BF50. It consists of the upper portion of all four limbs and also the pelvis. Fusion indicates that the individual was aged three to six months at death, and the morphology of the bones suggests it may possibly have been goat. The ABG was positioned over the top of the human skeleton.

5.5

Settlements

The settlement evidence from Yorkshire differs to that of southern England. Like southern England, the Yorkshire landscape, especially in the vale of Pickering, was divided by linear earthworks and pit alignments (Bradley 2007, 244). However, unlike the south of England, hillforts do not dominate the archaeological record for this period, although a small number are present, mainly on the Pennines (Manby et al. 2003b, 121). Many of the later settlements consist of large open sites, where it is unclear if structures are contempory, or how much of the occupation area was used at one time. Hill (1999) compared open settlements in the east of England with

Sheep/goat bones are common within the graves deposited as single bones on some sites, with 38 graves from Rudston and Burton Fleming combined containing such deposits. The majority of these deposits consist of the humerus, mostly from the left hand side (Legge

62

Chapter 5. Iron Age Yorkshire Table 5.4 Composition of partial pig ABGs from square barrow sites with data present. Partial indicates head, axis and leg elements are present Site Name Head + leg Leg Mixed Total Burton Fleming Garton Station Kirkburn Rudston Makeshift Total

1 1 8 10

1 1 1 1

2

1 1 2 9 13

Figure 5.3 Garton Station grave-group, GS6, the ABGs are shaded (Stead 1991, Figure 122). All of the sites with no ABGs reported in the faunal assemblage are settlement sites. Although a number of these sites were excavated and reported before ABGs were greatly recognised in the archaeological record, reports on over half were published after Grant’s (1984a) work on Danebury (Table 5.5). Therefore we can assume the authors were aware of ABGs and would have reported them, if present. It appears that ABGs are not as common on settlement sites from Yorkshire as they are from southern England.

‘wandering settlements’ of the same date in northern Europe, where structures were abandoned and replaced after a limited period of time. Bradley (2007, 260) expands the argument to northeast England, suggesting that such settlements may have served large communities similar to hillforts in southern England. Only Grimthorpe (Jarman et al. 1968) and Stanwick Camp (Wheeler 1954) are classified as hillforts and have published faunal reports present at the time of writing. Although, Stoertz (1997) categorised Grimthorpe as a defended enclosure.

63

Investigating Animal Burials; Ritual, Mundane and Beyond Table 5.5 Sites with faunal assemblages but no ABGs present. Year of report indicates the year the sites report with faunal data was published Site Name County Earliest Period Latest Period Year of report Bursea Grange Catterick Racecource Settlement Driffield RAF Station Grimthorpe Hillfort

East Yorkshire

Early Iron Age

Late Iron Age

1999

North Yorkshire East Yorkshire East Yorkshire

Early Iron Age Early Iron Age Early Iron Age

Late Iron Age Late Iron Age Late Iron Age

2003 1960 1968

Rock Castle Rillington Staple Howe Bursea House South Lawn 'Ladder' settlement

North Yorkshire North Yorkshire North Yorkshire East Yorkshire

Middle Iron Age Middle Iron Age Middle Iron Age Late Iron Age

Late Iron Age Late Iron Age Late Iron Age Late Romano-British

1994 1983 1963 1999

East Yorkshire

Late Iron Age

Late Romano-British

1999

Stanwick camp Topham Farm

North Yorkshire South Yorkshire

Late Iron Age Late Iron Age

Early Romano-British Early Romano-British

1954 2003

5.5.1

Ferrybridge and Parlington Hollins

mentioned in text, but a photograph of the deposit does show a number of other associated bones (Wrathmell and Nicholson 1990, Plate 34).

The number of ABGs found on these settlement sites is normally small. At both Ferrybridge and Parlington Hollins, in west Yorkshire, only one ABG was found. The Parlington Hollins ABG consists of an unknown number of cattle elements from the lower leg, with the hoof and hock still articulated. The ABG is deposited in a pit (pit 2066) associated with cattle bones from the skull, limb and other foot bones. The author suggests that two individuals are represented, and uses the presence of the ABG to suggest that this represents the processing of an entire carcass close by and subsequent rapid deposition within the pit (Richardson 2001). A partial pig ABG consisting of an unknown number of elements was recorded from Ferrybridge, found in isolation within a gully ditch associated with a round house (structure 5). This represents the only pig ABG from Yorkshire not found in association with human remains (see below). Due to its isolation and association with a house structure, the same author interprets this ABG as a deposit of a ‘ritual’ nature (Richardson 2005).

The other dog ABG was also found towards the bottom of a shallow pit (pit 3454) and in a similar position except lying on its right hand side. This one was not as complete as the other dog ABG, as the lower leg bones were missing. There is no mention in the published report of butchery marks being found to indicate skinning, during which the foot bones are often removed, and the elements do not appear to have been found in the pit fill. However, it is unknown if sieving took place during the excavation so the smaller lower limb bones may have been missed. All the ABGs from Dalton Parlours are interpreted as being the result of ‘ritual’ activity (Berg 1990a). The report shows the influence Grant’s (1984a) Danebury report had on the interpretation of these deposits when found in Iron Age contexts, with Berg referring to these ABGs as ‘special animal deposits’. 5.5.3

5.5.2

Garton and Wetwang Slack

Dalton Parlours As already discussed, a large proportion of the known archaeological records from Yorkshire is concentrated around the Yorkshire Wolds, East Yorkshire. The presence of the Hasholme and also the South Carr Farm logboats is possibly associated with an iron industry utilising bog ore in this area (Halkon 2003). However, there is only a small amount of published settlement evidence from the area. Late Iron Age ABGs are recorded from only Garton and Wetwang Slack (Brewster 1980; Noddle 1979). The site consists of a large open settlement, with round houses, four-post structures and possible grain storage pits present.

The only other site from west Yorkshire with ABGs present is Dalton Parlours (Berg 1990a). Four ABGs are recorded from the late Iron Age. Two consist of complete and near complete dog ABGs. One, consisting of 112 elements, was discovered on the bottom of a shallow pit, close to a round house (structure B). The dog was deposited lying on its left hand side with its back curled around the edge of the pit. The front portion of the skeleton was fully articulated, however, the back part had been disturbed. Tooth wear indicated the dog was an old adult when it died. Two sheep/goat ABGs were also recovered from this pit. One consisted of an unknown number of vertebrae found in articulation; the other consisted of nine elements of an unsided leg. Their association with each other and the dog ABG is not

Garton and Wetwang Slack produced the largest number of ABGs from Yorkshire, with 12 recorded. The majority (nine) are cattle ABGs, with two sheep/goat and one dog 64

Chapter 5. Iron Age Yorkshire 5.6

ABG also recorded from the site. Unlike the other sites discussed, all of the ABGs consist of complete animals. Unfortunately it is unknown if any of the ABGs are closely associated with each other as detailed faunal information is limited, but none of the ABGs are from the same contexts. Quern and carved chalk fragments are present within the fill of silo five which also contains a cattle ABG, but the associations between the finds is unknown. A human infant burial is also deposited next to this pit (Brewster 1980, 312).

Summary

In comparison with the southern England Iron Age assemblage the one from Yorkshire is very small. It also makes up a smaller proportion of the total ABG dataset from Yorkshire compared to the southern England Iron Age assemblage. It would therefore appear that although common in southern England ABGs are not as common from Iron Age sites in the Yorkshire region. All the ABGs are from domestic mammals with pig and cattle the most common species. This differs greatly from the pattern seen in the southern England assemblage. There is a significant variation in species between funerary and settlement sites. All but one of the pig ABGs are recorded from Arras culture graves. It is interesting to note that sheep/goat remains are also common from such graves, but not as ABGs. In comparison all the cattle ABGs are recorded from settlement sites dating to the later Iron Age. This does correspond with the non-ABG assemblage recorded from most sites in the region where cattle are also the most common species. In this respect the pattern is similar to that seen from southern England Iron Age settlements where the most common non-ABG species is also the most common ABG.

The majority of the ABGs are recorded from basal pit deposits. Two of the cattle ABGs are from small pits close to or within round house structures, which is a pattern also seen on some Yorkshire early RomanoBritish sites (see 7.3.2 for example). The cattle ABGs are all from juvenile animals and all appear to have been deposited in a similar manner, lying on one side with their front legs tucked up next to or under the rib cage. The two sheep/goat ABGs are also complete, deposited in pits and with the bodies in similar positions to the cattle ABGs. Unlike the cattle ABGs, the sheep/goat deposits are adult individuals. The two dog ABGs also consist of complete adults. One, an adult male dog, was deposited within a shallow pit (pit 14) within house structure two. Like the other ABGs from the site it was positioned on its side with its front legs flexed close to the chest. The author suggests it may have been deposited either for spiritual reasons connected to the house, or for sentimental reasons (Brewster 1980, 613).

Although limited in size the Yorkshire Iron Age dataset has shown that the ‘wessex’ pattern should not be taken as the norm, and reveals the possibility of regionality in the nature and composition of ABGs.

The other dog ABG dates from the late Iron Age to the early Romano-British period and is deposited within a well. The individual was an old adult, and pathology is present on the upper limb bones indicating a possible healed injury. Neonatal dog remains were also recovered from the same context, indicating that a MNI of six puppies were also deposited within the well. It is unknown if these were found as ABGs, but it is possible they were deposited as complete individuals. The Garton and Wetwang Slack ABGs are only described as ‘ritual’ deposits in the main report (Brewster 1980). Although this site is unusual in containing mainly complete ABGs, we could assume that partial ABGs may not have been recorded as both sites were excavated and published before Grant (1984a) highlighted the significance of partial ABGs. The interpretation of these deposits as the result of ‘ritual’ activities, shows the time depth of the concept. Also, at Garton and Wetwang Slack, the association between domestic and funerary spaces have influenced the interpretations. Close to the open settlement are a number of areas of square barrow burials and in some cases the square mounds were built over the round houses. This close proximity led Brewster (1980) to see much of the site as having a ‘ritual’ nature.

65

Chapter 6. Romano-British Southern England 6.1

Introduction

The boundary between the Iron Age and Romano-British periods is normally taken as AD43, when Claudius conquered Britain. However, contact had been occurring between British Iron Age communities and the Roman Empire for a long period before the invasion. This contact has long been argued to instigate a process of ‘Romanization’, with some of the British elite adopting Roman culture for their own advantage driving social change during the late Iron Age (Fitzpatrick 1989; Wells 2001). In contrast, Hill (2007) has recently suggested internal forces may be the stimulation for such changes. Studies by Goodman (1999) and Wells (2001) have shown that after the conquest many communities were not eager to adopt Roman styles and the process of postconquest ‘Romanization’ was a slow one. Therefore, after the conquest we see changes in some aspects of society, but a continuation in others. The main change seen in the southern England faunal record is a shift in emphasis towards cattle husbandry in comparison to the largely sheep/goat dominated Iron Age. The rise in the utilisation of cattle was possibly due initially to the need to feed the Roman army, which may have had a cultural preference for beef (King 1978). Albarella (2007) has argued there is minimal difference in animal husbandry between the late Iron Age and early Romano-British periods, which plays down the effects of Romanization after the conquest. Although we see the introduction of new types of site in the establishment of urban centres, beyond the towns, changes in rural life appear to have been slow. If this was the case we could expect little change in the ABG data in the early Romano-British period. ABGs from the Romano-British period represent one of the largest data sets, with 820 recorded from southern England. This represents 43% of the ABG assemblage from this region. Although the majority of previous studies have focused on ABGs from Iron Age sites, they are also common in the Romano-British period. This has started to be recognised (for example Fulford 2001; Woodward and Woodward 2004) but this study represents the first attempt to synthesise a large proportion of the available data. In total, 77 sites were recorded from southern England for this study, of which 43 (55%) had ABGs present. This is very similar to the Iron Age results from this study area, confirming that ABGs are a common phenomenon not only on Iron Age sites. Deposits were recorded from all three counties used for the south of England, with no one county dominating. There are concentrations of sites around the towns of Dorchester, Winchester and Silchester. This is partly due to separate excavation reports being counted as individual sites and the high levels of archaeological activity in these areas.

6.2

Species proportions

As with the preceding periods domestic mammals dominate, making up 81.8% of the assemblage. Wild mammals and wild bird remains constitute 4% and 8.4% of the assemblage respectively, with domestic fowl making up the rest of the assemblage. 6.2.1

Wild species

Almost all the wild mammals are from late RomanoBritish contexts (Table 6.1), which appears to correspond with a national trend in faunal assemblages noted by King (1991). Polecats are the most common wild mammal, followed by hare. However, all the polecat ABGs are from Oakridge Well (Maltby 1988; 1993). In fact, the majority of the wild mammal remains, 70%, were recovered from this site. The rest of the wild mammals were recorded in small numbers from one or two sites. Hares are the exception, eight being recovered from four different sites, two from Owslebury (Maltby 1987c), three from Oakridge Well (Maltby 1988; 1993), one from Little Somborne (Maltby 1978b) and two from Greyhound Yard, Dorchester (Maltby 1990f; 1993). As in the Iron Age assemblage the majority of the wild mammals are small herbivores (hare) and carnivores. Large ungulates such as deer are not common, with only two each recorded from Greyhound Yard and Oakridge Well. Wild birds make up a larger proportion of the assemblage than wild mammals. Swallows are the most common species, due to a large deposit close to the top of the Oakridge Well (see 6.5). The deposit consisted of 224 elements, but due to post-depositional movement and bone preservation it was not possible to assign each element to an individual, therefore a MNI of 30 was recorded. The deposit appears to have occurred after the well was abandoned, and may be due to swallows nesting in the walls of the feature. The other wild bird deposits are of a very different nature, and appear to have been deposited due to human activity. Ravens are the second most prevalent wild bird, followed by other corvids (Table 6.1). The deposition of ravens occurs in all phases of the Romano-British period, and may have been a continuation of a trend seen in the Iron Age (Serjeantson and Morris 2011). The raven ABGs are found on five different sites, eight from Greyhound Yard, Dorchester (Maltby 1993), one from Northern Suburbs, Winchester (Maltby 1987d), one from Oakridge Well (Maltby 1993), seven from Owslebury (Maltby 1987c) and one from Silchester Insula IX (1) (Ingrem 2006). Therefore, the majority of the raven ABGs are from two sites, Greyhound Yard and Owslebury. The majority of the other corvids are also from these two sites. Three jackdaw and two crow/rook are recorded from Greyhound Yard, and three crow/rook ABGs are present

Chapter 6. Romano-British Southern England Table 6.1 Number of ABGs per species for the Romano-British period. The Sheep/Goat, (S/G) counts include all ABGs identified to either species Early Romano- Middle RomanoLate RomanoSpecies British British British Total Domestic Cattle 22 10 47 79 Mammal S/G 38 11 46 95 Sheep 20 9 28 58 Goat 1 6 7 Pig 18 9 33 60 Horse 5 3 16 24 Dog 50 64 255 369 Cat 5 7 30 42 Wild Red Deer 1 1 2 Mammal Roe deer 1 4 5 Hare 8 8 Badger 1 1 Fox 2 2 Pine marten 1 1 Polecat 13 13 Weasel 1 1 Wild cat 1 1 Domestic Domestic Bird Fowl 7 1 38 46 Domestic goose 1 1 Wild Crow/Rook 2 4 2 8 Bird Raven 6 5 7 18 Jackdaw 2 1 1 4 Buzzard 3 3 Red kite 1 1 Pigeon 3 1 4 Quail 1 1 Swallow 30 30 Fish Sea bream 1 1 Total 158 118 544 820 Unlike the preceding periods, domestic bird remains, in the form of domestic fowl and one domestic goose, make up 5.9% of the ABG assemblage. The majority of the remains are from late Romano-British contexts. This corresponds with the increase in their utilisation during the Romano-British period (Maltby 1997). The one domestic goose ABG comes from a late Romano-British context at the Victoria Road excavations, Winchester (Maltby 1987d). Albarella (2005a) has recently argued persuasively that there is no clear evidence that goose was domestic in Romano-British Britain. It is therefore possible the Victoria Road goose ABG is from a wild rather than domestic animal.

in the Owslebury assemblage. The other jackdaw is from Silchester Insula IX and the other crow/rook ABGs are from Little Somborne with two present (Maltby 1978b) and one from Butterfield Down (Egerton 1996). In contrast to the interpretation of the Iron Age deposits, the majority of the Romano-British corvid ABGs have been interpreted as natural accumulations. However, more recent reports have seen corvid and especially raven ABGs as possible votive deposits (Ingrem 2006). Compared with the previous period, there is an increase in the variety of wild bird species found as ABGs. This trend has also been noted in the ‘normal’ faunal assemblages (King 1991; Parker 1988). However, it should be noted that the majority of the wild bird ABGs are recorded from Greyhound Yard and Owslebury (see below). 6.2.2

The most common domestic mammals are dog, followed by sheep/goat, cattle and pig. Dog dominate the ABG assemblage from this period, which represents a change from the Iron Age. However, differences are apparent between the early and later Romano-British periods. In the early Romano-British period dog and sheep/goat

Domestic species

67

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 6.1 Percentages of the most common ABG animals in the Romano-British periods. Sample size in brackets

Figure 6.2 Percentages of the most common ABG animals, excluding dogs, in the Romano-British periods. Sample size in brackets

Figure 6.3 Percentage of ABGs per species recorded from middle Romano-British context, excluding the Grayhound Yard data, sample size 32

68

Chapter 6. Romano-British Southern England combined constitute 57% of the ABG assemblage, with cattle and pig ABGs making up around 12% of the assemblage each. The middle Romano-British assemblage shows a dramatic change with 54.2% of the ABGs from dogs. Sheep/goat are still the second most common species, but only represent 9.3% of the assemblage.

differ to those seen in the ‘normal’ faunal assemblage where cattle is the most common species. Therefore, even with the large dog sample excluded from the data, the ABG species proportions still differ from the disarticulated faunal assemblage, especially in the later periods.

This change in the species composition continues into the late Romano-British period when there is a slight drop, with dog making up 46.8% of the assemblage. However, the second most common species is cattle, at 9%, closely followed by sheep/goat at 8.4%. The slight drop in dog representation is likely to be due to the increase in species variability, especially wild mammals and birds, compared to the earlier sub-periods. The large number of dog ABGs compared to other species may be due to the high proportion of neonatal and juvenile puppies being recovered on some sites.

6.2.3

Greyhound Yard and the middle Romano-British assemblage

The overall Romano-British assemblage changes to a dog-dominated-one by the middle Romano-British period. However, the large Greyhound Yard sample size needs to be taken into account regarding this pattern. In total, 118 ABGs are recorded from the middle RomanoBritish period (Table 6.1), of which 86 (72%) are from Greyhound Yard. Therefore the overall species pattern seen in this period is largely a reflection of the Greyhound Yard assemblage. If we exclude the Greyhound Yard assemblage the overall species proportions pattern for the middle Romano-British period is very different. The highest proportion of ABGs are still from dogs, but they only make up 25% of the sample, with cattle, pig and cat ABGs making up just over 15% each (Figure 6.3). Without the Greyhound Yard assemblage there is a more even distribution of species, illustrating the dangers of examining a multi-site assemblage without considering the sample size of the different sites.

As with the preceding periods the majority of sheep/goat ABGs are not identified specifically to either sheep or goat. A small number are identified as being from goats, but most of the ABGs that are identified to a single species are recorded as sheep. Therefore, the majority of ABGs recorded as sheep/goat are probably from sheep. This corresponds with the pattern seen in whole faunal assemblages, in which only a small number of goat bones are present, indicating goats were of only minor importance (Coy and Maltby 1987).

Excluding Greyhound Yard, the middle Romano-British assemblage comes from nine sites, five rural settlements (including Owslebury), two towns, one military (Portchester) and the other funerary (Poundbury Cemetery). The largest assemblages are Owslebury (10) and Porchester (11), with the other sites contributing between one and three ABGs. Interestingly, the dog ABGs recorded are only from Owslebury, Oakridge Well and Porchester. The other sites did not have dog ABGs recorded for the middle Romano-British period. Although the sample size is small, these results do indicate that there is no standard ABG species pattern for this period.

The proportion of cattle and pig ABGs remain relatively low throughout the Romano-British period. This, and the high number of dog ABGs, is in contrast to the results from ‘normal’ faunal assemblages, where cattle and sheep/goat are the most common species. There is some variation, with more ‘Romanized’ areas such as towns and military sites having a high proportion of cattle and pig remains present, compared to rural settlements which tend to have a lower proportion and more sheep/goat present (King 1978; 1984; Maltby 1994b). King (1999b) has suggested that the inhabitants of urban and military sites set a dietary pattern that was emulated by groups seeking to become more ‘Romanized’, hence the high cattle/high pig pattern seen on most sites in the later Romano-British period. This contrasts to the dogdominated pattern seen in the ABG assemblage.

6.3

Nature of the assemblage

As discussed above and in previous periods, some sites dominate the overall ABG assemblage due to the large numbers recorded from them. For the Romano-British period just over 75% of the ABGs recorded are from five sites (Table 6.2). Three sites, Owslebury, Oakridge Well and Greyhound Yard, have very large assemblages of over 150 ABGs each. Owslebury has the largest assemblage with 187 ABGs recorded from this period. Combined with the Iron Age ABGs, Owslebury with 232 ABGs, has produced the largest assemblage of any site recorded for this project. The second largest, Oakridge Well, is in some ways even more impressive, as all the ABGs recorded come from just one feature (see 6.5).

If we exclude dog, sheep/goat are the most common ABG species, followed by cattle and pig (Figure 6.2). The proportion of sheep/goat and cattle ABGs is similar to the patterns seen in the ‘normal’ faunal assemblages from early Romano-British sites. For example, the ‘normal’ faunal assemblage from the earliest phase of Dorchester Greyhound Yard, produced 19% cattle, 46% sheep/goat and 22% pig. The ABG pattern changes in the middle Romano-British period, the proportion of sheep/goat decreases to similar levels as cattle and pig. The late Romano-British period sees an increase in the number of domestic fowl ABGs, which overtake pig as the third most common species. However, these patterns greatly 69

Investigating Animal Burials; Ritual, Mundane and Beyond

Table 6.2 The five largest Romano-British ABG assemblages Percentage of Site Site Type No. ABGs total assemblage Owslebury Settlement 187 22.8% Oakridge Well Settlement (well) 177 21.6% Greyhound Yard, Dorchester Town 163 19.9% Northern Suburbs, Winchester Town 49 6.0% Portchester Castle Military 44 5.4% . Table 6.3 Number of sites with ABGs present and the number of ABGs recorded from them. Percentage total sites indicates the proportion of sites with that number of ABGs present. %No. ABGs indicates what proportion of the total ABG assemblage come from the sites. Site Type 1 2 to 4 5 to 9 10 to 29 30 to 49 50 to 99 100 to 150 150< Town 2 5 2 1 1 Villa 1 2 1 Military 1 Settlement 7 8 3 3 2 Funerary 2 2 Total Percentage total sites

10

15

6

7

2

-

-

3

23.3%

34.9%

14%

16.3%

4.7%

-

-

7.0%

Percentage No. ABGs

1.2%

5.0%

5.2%

12.9%

11.3%

-

-

64.4%

Table 6.4 Total number of ABGs recorded for each site type Site type Town Villa Military Romano-British Settlement Funerary

No. ABGs 258 12 44 474 32

Percentage of ABG total assemblage 31.4% 1.5% 5.4% 57.8% 3.9%

Table 6.5 Number of sites recorded with no ABGs present Period Town Villa Military Settlement Iron Age - Romano-British 9 Romano-British 6 2 2 15

Funerary -

Winchester (Maltby 1986b), have between two and four ABGs. The two sites with between 10 and 29 ABGs are Neatham (Done 1986) and Silchester Insula IX (Clark 2006; Ingrem 2006). Although 11 urban sites are recorded, the results come from only three major Romano-British towns, Dorchester, Winchester and Silchester and one smaller town, Neatham. At present no ABGs have been recorded from Clausentum (Southampton), but a very small area of the town has been excavated compared to the others. However, unpublished data does indicate that dog ABGs are present from the town (pers comm. S, Hamilton-Dyer).

The assemblage from the Greyhound Yard site in Dorchester is the third largest, and is the biggest from an urban context. The next largest assemblage also comes from a town site, the Northern Suburbs excavations at Winchester, but it is much smaller with 49 recorded ABGs. In total, 11 of the sites recorded with ABGs present are urban in nature (Table 6.3). The majority of the sites produced only small assemblages. Two sites, Dorchester Prison, Dorchester (Draper and Chaplin 1982) and Hyde Street, Winchester (Birbeck and Moore 2004) each had only one ABG recorded from them. Also, five town sites, Colliton Park, Dorchester (Aitken and Aitken 1982), Silchester North Gate (Area 4) (Hamilton-Dyer 1997b), Silchester Forum-Basilica (Grant 2000), South Grove Cottage, Dorchester (Startin 1981) and Staple Gardens,

Romano-British rural settlements are the most common site type with 24 recorded with ABGs present. Also, the two largest ABG assemblages come from rural settlement sites, Oweslebury and Oakridge Well. However, the 70

Chapter 6. Romano-British Southern England Owslebury is a multi-period site, consisting of an Iron Age and Romano-British settlement associated with ditch systems (Collis 1968). Unfortunately the full excavation report has yet to published, but the faunal remains report is in the public domain (Maltby 1987c). The majority of the faunal remains recorded are from the associated ditches and pits. However, due to the nature of the available data it was not possible to record associations between ABGs and other material types.

assemblages from such sites follow a similar pattern as those from urban contexts. After Owslebury and Oakridge Well, the next largest ‘rural’ assemblages come from four sites, Poundbury settlement (Buckland-Wright 1987), Alington Avenue (Maltby 2002a), Maddington Farm (Hamilton-Dyer 1996a) and Maiden Castle Road (Bullock and Allen 1997). Although large assemblages are recorded, the majority of rural settlements have between one and four ABGs recorded from them (Table 6.3).

Owslebury produced one of the largest faunal assemblages from this type of site, around 110,000 fragments (including 187 ABGs) from the 1962-1972 excavations. A large proportion of the faunal remains were recovered from the ditches, with almost 12,000 fragments recorded from ditches F133 and F642 alone (Maltby 1987c).

A small number of military and villa sites were recorded with ABGs present. The only military site is Porchester Castle, with 44 ABGs (Grant 1975). The four villa sites, Barton Field (Hicklin 2006), Castle Copse (Payne 1997), Downton Villa (Rahtz 1963) and Braishfield (Maltby 1979a) all have small assemblages of ABGs. Therefore the majority of the ABGs examined in this study come from either rural settlements or urban sites. Those from rural settlements account for 57.6% of the total Romano-British assemblage with 31.74% coming from towns (Table 6.4). However, the dominance of rural settlements is mainly due to the two very large assemblages discussed above. Owslebury and Oakridge Well combined account for 77% of all ABGs recorded from rural settlements. Also, the majority of the urban assemblage, 62%, comes from just Greyhound Yard, Dorchester.

6.4.1

The majority, 74% (138), of the ABGs from the site date to the later part of the Romano-British period, with only 39 and 10 recorded dating to the early and middle Romano-British periods respectively. This corresponds with the ‘normal’ faunal assemblage data, where a large proportion of the material dates to the 3rd and 4th centuries (for this analysis, as with the ABGs, the latest possible date has been utilised).

The majority of negative results come from ‘rural’ settlements. Of the 43 sites recorded with faunal assemblages, but with no ABGs, 24 (51%) are from rural settlements (Table 6.5). The majority of these settlements date to the Romano-British period, but there is a small number where occupation continues from the Iron Age. In comparison only six sites, 33%, with negative results are from an urban context. However all of these excavations, with the exception of two sites from Clausentum, are from towns, for which other excavations have produced positive results, mainly Silchester. Also a small number of villa and military sites are recorded with no ABGs present. Of the six villa sites recorded, the majority (4) have ABGs present.

Dog is the most common species recorded for the early and late periods (Figure 6.4). In the early Romano-British period, ABGs of pig and cattle are respectively the second and third most common from the site. In the middle Romano-British sample, dog and cattle ABGs are the most common, although the sample size for this period is very small. By the late Romano-British period the proportion of cattle falls, becoming the third most common species represented by ABGs. The proportion of pig ABGs declines throughout the Romano-British period, and by the later part is less common than domestic fowl ABGs. The results from the late Romano-British period are similar to those seen from the overall ABG assemblage from the southern England sample, with the pattern dominated by dog ABGs (Figure 6.1 and Figure 6.4). Also recorded from Owslebury are ABGs from seven ravens, four domestic cats, three crows/rooks and two hares.

The only site type to always have ABGs present in the faunal assemblage are those of a funerary nature. Therefore, although over half the ABGs recorded come from rural settlements ABGs are more likely to be encountered on funerary, urban and villa excavations. However, we must take into account that only a small number of purely funerary sites exist. Human remains without ABGs found in association are present on a number of sites such as Owslebury. The small number of villas in the sample may account for the lack of ABGs from this type of site. 6.4

Owslebury ABG species proportions

As with other sites and periods examined the ABG species proportions differ from those of the overall faunal assemblages. The overall faunal assemblage from Owslebury is dominated by sheep/goat and cattle remains. Dog may be the most numerous ABG recorded from the site, but the remains of dogs only make up a small proportion of the overall faunal assemblage (Figure 6.5). It is only in the late Romano-British period that the proportion of dog remains in the overall faunal assemblage rises over 5%. This is because of the inclusion of ABGs in the overall NISP counts for the site.

Owslebury

71

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 6.4 Percentages of the main ABG species from Owselbury for the Romano-British sub-periods

Figure 6.5 Percentages of the main species NISP per period for the overall faunal assemblage from Owslebury (including ABGs) data from (Maltby 1987c)

Figure 6.6 Percentage of each species per feature type from Owslebury. Number of ABGs in brackets

72

Chapter 6. Romano-British Southern England As a large proportion of the dog ABGs consisted of complete or close to complete skeletons, and the majority dated to the late Romano-British period, this causes the dog NISP count to be exponentially high compared to the number of individual animals. However, the number of dog remains recorded in the overall NISP counts is still smaller than the number of cattle and sheep/goat. Therefore a higher proportion of the dog population of the settlement and surrounding area was deposited as an ABG compared with the other domestic animals. 6.4.2

The majority of ABGs recovered from quarry features are from cattle, followed by a small proportion of horse and dog. Only 20% of the cattle ABGs recovered are from pit contexts. In comparison the majority of sheep/goat and dog ABGs are from pits. A large proportion (46 %, 42) of the dog remains were recovered from one pit, feature 664. Also, around half of the domestic fowl recorded are from pit deposits, the other half are from ditches. All the domestic fowl ABGs recorded from pits date to the late Romano-British period. In comparison, all but one domestic fowl recorded from ditches date to the early Romano-British period.

Owslebury ABG placement

As already stated the majority of the ‘normal’ faunal material was recovered from the ditch sections. Unlike the Iron Age assemblage from the site, there is a limited amount of intra-site variability from the Romano-British total faunal assemblage (Maltby 1987c). With the exception of dogs, a high proportion of the ‘normal’ domestic mammal faunal remains was recovered from ditches (Figure 6.7). Although a slightly higher proportion of sheep/goat and pig remains was recovered from pits, compared to ditches. The correspondence in the placement of ABGs and the overall faunal assemblage is due to the majority of the dog population being deposited as ABGs.

Although there are problems with some aspects of the data from Owslebury, the detailed faunal report does indicate from which features most of the ABGs were recovered (six dog ABGs are from unknown feature types). The majority of the ABGs are from either ditches/gullies, pits or quarry features. One cat ABG was recorded deposited within the backfill of an oven. Interestingly all the ABGs recorded from the early Romano-British period are from either ditches or gullies. It is not until the middle Romano-British period that ABGs are recorded from pits. However, this is probably due to the nature of the archaeological record, as a number of pits are dated as ‘early to middle’ RomanoBritish, and therefore counted as part of the middle Romano-British sample according to the conventions of this study. Also, the majority of the Romano-British features excavated consisted of ditches.

Cattle, pig and horse ABGs appear to correspond with the overall ditch-deposition-pattern. Interestingly around one third of cattle ABGs were recovered from quarry features (although we must bear in mind the small sample size), but only a small proportion of overall cattle remains were recovered from such features. Therefore a higher proportion of the cattle remains being deposited in quarry features were as ABGs. The majority of sheep/goat remains are recovered from the ditch sections, However, the majority of sheep/goat ABGs were found in pit deposits, indicating a possible difference in the utilisation of areas within the settlement as well as subsequent treatment of individual sheep/goat. Further interpretation is dependent on the nature of the ABGs.

Differences in feature placement are apparent between species. The majority of cattle, pig, horse and raven ABGs are recorded as coming from ditches/gullies (Figure 6.6). The high proportion of raven ABGs from the ditches is due to four of the seven dating to the early Romano-British period. Most of the later-dating raven ABGs are from pit contexts. Horse ABGs are the only species not found in pits. All of them are from either ditches/gullies or quarry features. Only one late RomanoBritish pig ABG is recorded as being found in a pit, all the rest are found in ditches/gullies.

Figure 6.7 Percentage of each species NISP per feature type for the overall faunal assemblage from Owslebury (including ABGs). Data from (Maltby 1987c) 73

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 6.8 Attrition of species population, for individuals that became ABGs, at Owslebury

Figure 6.9 Attrition of species population, using data from the whole Romano-British faunal assemblage from Owslebury

Figure 6.10 Attrition of ABG and non-ABG pig population from Owslebury

74

Chapter 6. Romano-British Southern England 6.4.3

Owslebury ABG ageing data

deposited as ABGs at Owslebury were younger than those from the ‘normal’ faunal assemblage. Unfortunately the small ABG sample size means this is not statistically testable.

Ageing data are also available for a large proportion of the ABGs from Owslebury, enabling us to investigate the age at death of the animals that became ABGs, as well as the attrition of the cohort. The data show that species died or were killed at different ages. The majority of cattle and sheep/goat ABGs with ageing data present were subadults or young adults. Therefore the majority of individuals of these species had lived almost into adulthood before being killed and transformed into the resulting ABGs (Figure 6.8). In comparison, the majority of the pig ABGs belong to juveniles, and most of the dog ABGs are neonatal. Of the eight pig ABGs with ageing data, six are of juvenile age. Dog ABGs have a much larger sample size, and of the 87 deposits with ageing data, 60% are from neonatal animals. The dog ABG ageing data are very similar to the data from the overall dog assemblage. This is to be expected as a large proportion of the dog population from Owslebury appears to have been deposited as ABGs. Maltby (1987c) refers to the ageing data to provide an explanation for the presence of the dog ABGs;

The sheep/goat ageing data from the ‘normal’ and ABG assemblages are in some ways similar to that from the cattle remains. The majority of sheep/goat ABGs appear to have died when sub-adults, with a large drop in the surviving population from 90% to 30%. By youngadulthood only 20% of the population that became ABGs survived (Figure 6.8). Toothwear data from the whole faunal assemblage show that there was a more gradual attrition of the population during adolescence, without the large number of sub-adult deaths, although as with the ABGs, only 20% of the population survived to be young adults. Therefore, both datasets show a roughly similar pattern of population attrition. The difference in the number of sub-adult deaths may be due to the ABG sample size. Again the small ABG sample size means this is not statistically testable. 6.4.4

‘The spectacular concentrations of neonatal puppies dumped in the 3rd-4th century cess pits demonstrates that litters may have been deliberately put down at birth to control the dog population. In addition the epiphyseal fusion evidence indicated that immature dogs were also represented in some numbers. Since their meat does not appear to have been consumed except in rare instances, the presence of these immature dogs may imply either that they were natural mortalities or that these also were put down to keep the dog numbers under control.’

Owslebury ABG composition

The majority of ABGs at Owslebury were deposited as partial skeletons. Complete skeletons are present for sheep/goat, pig, dog, cat and rook/crow. Cat ABGs have the highest proportion of complete ABGs, although the sample size is very small (Table 6.6). The highest number of complete ABGs recorded are from dogs, with 14 recorded, providing 15% of the dog ABG assemblage. The completeness of a large number of dog ABGs is also unknown. This is because a large proportion of the dog ABG assemblage was recovered in disturbed multi-ABG deposits, where elements were admixed due to a series of taphonomic factors. It was therefore not possible to identify each element to an individual ABG deposit, resulting in their composition being recorded as unknown. Maltby (1987c) suggested that it is possible most dogs were deposited on the site as ABGs and that the disarticulated dog material is the result of disturbed ABGs, or secondary deposition.

The majority of the cattle, sheep/goat and pig remains were not deposited as ABGs, and the ageing data show a number of differences and similarities between ABGs and the ‘normal’ faunal assemblage. One of the biggest differences is in the pig age ranges (Figure 6.10). The majority of pig ABGs were from juvenile animals. However, the toothwear data from the ‘normal’ faunal assemblage indicates that there was a large kill-off of young adult pigs, with only 34% of the population living past this stage (Figure 6.9). Because of this difference, Maltby (1987c) suggested that the pigs deposited as ABGs may have been natural mortalities, although he does not rule out a possible ritual connection.

Therefore the dog ABGs of unknown composition may have been originally deposited as complete skeletons. In addition the partial dog ABGs may have originally been deposited as complete skeletons. The body area data does indicate that the majority of partial dog ABGs consist of both the axial and appendicular skeleton body parts, mainly the vertebrae and back limbs. However, only a small proportion of partial dog ABGs include front limb bones (Figure 6.11).

The cattle ABG and ‘normal’ assemblage ageing data also show some differences. The majority of cattle ABGs appear to have died younger than those deposited as part of the ‘normal’ faunal assemblage. The main difference is in the number of adolescent deaths. Around 50% of cattle deposited as ABGs survive to be sub-adults or older, compared to 70% of the ‘normal’ faunal assemblage. The main kill-off of cattle not deposited as ABGs appears to have occurred just before the animals had reached maturity (Figure 6.9). Therefore the majority of cattle

This could indicate that not all dog ABGs were deposited as complete skeletons. If deposited as a partial skeleton the carcass must have been subject to a taphonomic process prior to final burial. Butchery in the form of a substantial number of skinning cuts was recorded on one dog ABG from pit 42, showing that butchery of dogs did

75

Investigating Animal Burials; Ritual, Mundane and Beyond sometimes take place. The front limb-elements are also one of the first parts of a body to be disarticulated during exposed decomposition (see 2.4). Although it would appear that the majority of dog ABGs were deposited as complete skeletons, a number may have either been butchered or left for a period of time to decompose before final deposition.

Table 6.6 Composition of Romano-British ABGs from Owslebury Species Cattle S/G Pig Horse Dog Cat Hare Domestic Fowl Raven Rook/ Crow Buzzard

Complete

1 2 14 2

Partial

25 21 7 11 25 2 2 6 6

1

2 2

Unknown

57

1

Percentage complete

5% 22% 15% 50% -

A proportion of the pig ABG assemblage, 22%, also consists of complete skeletons. However, we must bear in mind the very small size of the assemblage (Table 6.6). The partial pig remains show a very different body area composition compared to the dog ABGs. All of the partial pig ABGs consist of elements from the axial skeleton, either vertebrae or ribs. But body area data were only available from five partial ABGs.

33% -

Figure 6.11 Percentage representation of body areas for partial dog ABGs from Romano-British Owslebury (data from 15 ABGs)

Figure 6.12 Percentage representation of body areas for partial cattle ABGs from Romano-British Owslebury (data from 25 ABGs)

76

Chapter 6. Romano-British Southern England

Figure 6.13 Percentage representation of body areas for partial horse ABGs from Romano-British Owslebury (data from 11 ABGs)

Figure 6.14 Percentage representation of body areas for partial sheep/goat ABGs from Romano-British Owslebury (data from 19 ABGs)

Figure 6.15 Percentage representation of body areas for partial sheep/goat ABGs from Iron Age Owslebury (data from 9 ABGs)

77

Investigating Animal Burials; Ritual, Mundane and Beyond from the feature (Maltby 1994a). The well was associated with a Romano- British rural settlement and appears to have been constructed in the late 1st century AD. The well appears to have been used for a source of water until the late 2nd century AD, after which time material was deposited or accumulated within the feature. The abandonment of the well as a source of water may have been due to a shift in the centre of activity away from the well, leaving it on the periphery of the settlement’s activities and a useful place for the deposition of unwanted material (Oliver 1992, 74). A number of possibly rapid filling events occurred after the abandonment of the well, first in the late 2nd century, then in the late 3rd and early 4th centuries, with the largest event occurring towards the end of the 4th century. During these events a large number of ABGs were deposited within the well (Figure 6.16). During these rapid filling events a large amount of organic material appears to have been deposited, with the 4th century events possibly filling the whole well. Subsequent decomposition appears to have led to the deposits sinking to 15m and below.

The axial elements, especially the vertebrae, are also the most common body part to be included in partial ABGs for the other main species, cattle, sheep/goat and horse. Vertebrae followed by ribs and the lower hind foot are the most common areas to constitute partial cattle ABGs (Figure 6.12). Surprisingly, elements from the upper front limb are not recorded from any partial cattle ABG. It is possible this is a reflection either of butchery practices, or natural disarticulation, as the forelimb is the first disarticulation event to occur in exposed cattle carcasses (Oliver 1992). The partial horse ABGs show a similar pattern in respect of the vertebrae, with ribs also commonly present (Figure 6.13). Unlike the partial cattle ABGs, the upper forelimb is the most common part of the appendicular skeleton present. In contrast the upper hind limb and pelvis are the least common body areas. However, unlike cattle, in horses the tibia – tarsus joint is one of the first to become naturally disarticulated, possibly explaining the small number of upper hind limb elements present. We cannot completely discount the transportation of meat-bearing elements to other areas of the site or sites or more ritualistic reasons to explain such body area differences, although there is much less butchery evidence for horse than other species.

After this time the well filled in naturally through weathering. A small number of ABGs are present within the weathering material, mainly wild mammals in the form of polecats, deer and a large deposit of swallows. The swallow ABGs are all from fledgings, which indicates they may have been nesting in the well. A large number of shrew, mice, vole and amphibian bones are also present within this fill and others suggesting during several periods it was naturally open to the elements. This process is calculated to have taken place over a period of 50 to 300 years (Oliver 1992, 76).

The majority of partial sheep/goat ABGs also have vertebral elements present. However, unlike the previous species, the body areas for sheep/goat are more evenly represented. The axial skeleton does dominate, with the vertebrae followed by the head and ribs being the most common body areas. A smaller proportion of sheep/goat ABGs have parts of the appendicular skeleton present, but both fore and hind limbs are represented in equal measure, unlike the previous species.

The majority of the ABGs from the well date to the late Romano-British period. Two partial cattle ABGs are present towards the very bottom of the well dating to the early Romano-British period and three dog ABGs, two of them complete, date to the middle Romano-British period. As at Owslebury, the majority of ABGs recorded from the site are those of dog (Figure 6.17). The second most common ABG is that of swallow. Cattle, then polecats, are the third and fourth most common ABG species. The polecats were interpreted as a natural deposit, occurring in the upper fill and all consisting of complete skeletons. However, this is assuming that 13 polecats will fall down a well. Their close association with dog, pig and cat ABGs may indicate that the polecats were deposited by human activity. Cattle, sheep/goat and pig make up only a small proportion of the ABG assemblage, which is similar to the pattern seen at Owslebury (Figure 6.4).

Partial sheep/goat body area data are also available from Owslebury for Iron Age deposits. They show that the composition of partial sheep/goat ABGs changes between the two periods, although the Iron Age sample is small. Unlike the Romano-British deposits which are dominated by axial elements, appendicular elements, in particular the lower hind limb and the forelimb are the most common elements for Iron Age deposits (Figure 6.15). Also, no Iron Age partial ABGs had the skull present. At Owslebury there therefore appears to be a shift from partial sheep/goat ABGs consisting mainly of appendicular elements, to one of axial elements in the Romano-British period. This is also a pattern seen when comparing the total ABG assemblage from both periods. 6.5

Oakridge well

The majority of the non-ABG faunal material recorded from the well consisted of the remains of sheep/goat and cattle (Figure 6.18). There is also a high proportion of dog present, but their presence is inflated by the amount of complete dog ABGs (Maltby 1994a, 48). Therefore, as with other sites examined in detail for the RomanoBritish period, the ABG assemblage is not reflective of the species proportions in the overall faunal assemblage.

The Oakridge well assemblage is very different in nature compared with the other two large ABG assemblages of Greyhound Yard and Owslebury. All of the Oakridge ABGs come from just one feature, a well which was fully excavated down to 26.67m (Oliver 1992). In total 24,426 bone fragments (including 172 ABGs) were identified 78

Chapter 6. Romano-British Southern England

Figure 6.16 Section of Oakridge well with relative position of ABGs indicated. The highest possible position has been used. ABG labels with just the species name, indicates that the completeness of the ABG is unknown. Other finds are noted in italics, all human remains consist of disarticulated fragments, the number indicates the MNI (altered from Oliver 1992, Figure 7) 79

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 6.17 Percentages of the main species from the late Romano-British Oakridge Well deposits. Sample size 172 ABGs

Figure 6.18 Percentage NISP for the main species from the total faunal assemblage Oakridge Well (total sample size 20581 fragments, includes ABGs)

Figure 6.19 Attrition of species population, for individuals that became ABGs, at Oakridge Well

80

Chapter 6. Romano-British Southern England Table 6.7 Composition of Romano-British ABGs from Oakridge Well Species Complete Partial Unknown Percentage Complete Cattle 5 7 6 28% S/G 2 1 67% Pig 8 100% Dog 9 2 75 10% Cat 6 Red Deer 1 100% Roe Deer 2 2 50% Hare 2 1 67% Fox 2 100% Pine Marten 1 Polecat 13 100% Raven 1 Quail 1 Swallow 30 The ABG ageing data show that the majority of the cattle, sheep/goat, pig and dog ABGs are from neonatal or juvenile animals. Only a small proportion of dog and cattle ABGs are from individuals that had reached adulthood (Figure 6.19). This pattern is very different to that observed from the Owslebury data, especially for sheep/goat and cattle ABGs. Owslebury also has a large number of neonatal dog ABGs. However, a greater proportion of the dog ABGs are from adult individuals. The large number of juvenile pig ABGs are also present at Owslebury, but again some adult pig ABGs are recorded, whereas no adult pig ABGs are present from Oakridge.

complete skeletons or are of unknown composition (Table 6.7). There is an especially large number of dog ABGs where the composition of the deposit is unknown. This is due to the large number of dog ABGs buried within the same area of the well and also the natural slumpage that occurred. A number of deposits were recorded where the dog ABG elements had been intermixed, and it was therefore not possible to identify which element went with which ABG. As the majority of dog ABGs found in a multi-ABG deposit were neonatal age, Maltby (Maltby 1988; 1994a, 59) interpreted them as belonging to the same litters, possibly indicating control of the dog population.

The main difference between the two sites is in the ages of the cattle and sheep/goat ABGs, which are mainly juveniles from Oakridge Well. In comparison the majority of cattle ABGs from Owslebury appear to have died when sub-adult or young adults. The majority of sheep/goat ABGs from Owslebury are also from subadult individuals (Figure 6.8). Therefore the majority of the Oakridge Well ABG population is from juvenile animals, whereas older animals are being deposited as ABGs at Owslebury. This could be due to the series of rapid depositions at Oakridge Well. Maltby (Maltby 1994a, 56) suggested the high numbers of young calves within the late 4th century fill could indicate a rapid filling of the well in the late winter to early summer. The nonABG material deposited at Oakridge Well is also from older individuals. The cattle-tooth-wear data indicates that at least 57% of the non-ABG individuals were probably over four years old. The majority of the sheep/goat had also reached late adolescence, with the ageing patterns for non-ABGs being similar to those from Owslebury (Maltby 1994a, 56). However the pig ageing data for non-ABGs is similar to the pattern seen for ABGs, with the majority of bones belonging to juvenile animals.

The relatively high proportion of complete cattle, sheep/goat and pig ABGs is unusual for this time period. Also if we take into account the level of slumpage and mixing of deposits, it is possible that the majority of the ABGs were first deposited as complete skeletons. This indicates that a different ABG deposition strategy/activity is taking place at this feature compared to other sites. The high proportion of complete, wild mammal ABGs is also unusual and may support the argument that these represent pit-fall victims (Maltby 1994a). 6.6

Greyhound Yard, Dorchester

The Greyhound Yard excavations in Dorchester (Durnovaria) produced the third largest ABG assemblage from the Romano-British period (Table 6.4). The excavations revealed 35% of a single insula to the south of the central forum and basilica. The grid-patterned internal streets along with timber buildings and associated ‘backyard’ enclosures were discovered on the site (Woodward et al. 1993). The excavations are one of the largest to take place in a civitas capital in the south of England, and resulted in the collection of a very large faunal assemblage of over 40,000 fragments (including 164 ABGs). Most are from pits within the ‘backyard’ enclosures and a large proportion are of middle to late Romano-British date (Maltby 1990f; 1993).

The composition of the ABGs from Oakridge Well is very different to those from Owslebury and Greyhound Yard. A larger proportion of the ABGs consist of 81

Investigating Animal Burials; Ritual, Mundane and Beyond dogs. However, their numbers are very small in comparison with the large number of dog ABGs. The large number of dog ABGs also have an effect on the overall middle Romano-British species proportions.

All the ABGs recorded from this site come from pit deposits. This, however, is a reflection of the archaeology, as the majority of the features were pits from the ‘backyard’ areas. Therefore ABG placement analysis is not carried out. Also, only limited ABG body part information is available at this time because there are no extant individual records of the bones in the assemblage. 6.6.1

The early Romano-British pattern differs greatly compared with that from Owslebury (Figure 6.4) and the overall ABG assemblage pattern for southern England (Figure 6.1). Although by the late Romano-British period these are dominated by dog ABGs, they have a much more even distribution of species in the early RomanoBritish period. In comparison, the Greyhound Yard assemblage changes little throughout the Romano-British period.

Greyhound Yard species proportions

As in the Owslebury assemblage, dog ABGs are the most common. However, unlike the Owslebury results only small numbers of ABGs of other species are recorded. Dogs provide over 50% of the ABGs, in all sub-periods (Figure 6.20). After dogs, cattle are the second best represented species in the early and late Romano-British periods. During the middle Romano-British period sheep/goat are the second most abundant species after

The ABG species proportions differ greatly compared with the non-ABG faunal assemblage. Cattle, sheep/goat and pig are the most common animals throughout the Romano-British period (Figure 6.21). In comparison, only a very small amount of non-ABG dog remains are present. At least 4050 (89%) of the 4572 dog bones

Figure 6.20 Percentages of the main species from Greyhound Yard for the Romano-British sub-periods

Figure 6.21 Percentages of the main species overall faunal assemblage NISPs from Greyhound Yard for the Romano-British sub-periods (excluding ABGs)

82

Chapter 6. Romano-British Southern England

Figure 6.22 Attrition of species population, for individuals that became ABGs, at Greyhound Yard pattern to that of the ABGs, with most cattle reaching adulthood (64%), although in the middle Romano-British period, a high proportion of calves and mature adult cattle were recorded. The majority of the non-ABG cattle appear to have been slaughtered prior to very old age, suggesting a policy of selecting fully grown animals not required for breeding or traction (Maltby 1993, 320). As the cattle ABGs show a similar age pattern, this could indicate they were part of this policy.

recorded from the site belong to ABGs (Maltby 1993, 326). Therefore, as at Owslebury the majority of the dog population present on the site were deposited as ABGs. In comparison only 82 (1.2%), of the 6763 cattle remains and 456 (7%), of the 6910 sheep/goat elements were deposited as ABGs (Maltby 1993, 317, 321). 6.6.2

Greyhound Yard ageing data

The sheep/goat non-ABG ageing data is also similar to that of the ABG data. The non-ABG data show a high kill-off of immature animals around 18-24 months, although slightly older animals are more commonly found in the late Romano-British sample. The ABG data also show a high kill-off of sub-adults, suggesting that the sheep/goat deposited as ABGs were slaughtered at similar ages to those deposited as non-ABGs. The kill-off of sheep/goat at this age range is often interpreted as reflecting a culling strategy concentrated on meat production and is suggested to be typical of faunal samples from this time period (Maltby 1981b; 1984b).

An advantage of the large ABG sample size from Greyhound Yard is the opportunity to investigate the ages at which animals that became ABGs were killed. Ageing data are available for cattle, sheep/goat, pig and dog, although the sample size for ABGs with ageing dating is small for some of these species. The data show two clear patterns. One is that the majority of sheep/goat and pig ABGs were from individuals that had been killed before reaching adulthood. The majorityof sheep/goat were sub-adults (1-2 years), whereas no pig ABGs from individuals older than juvenile (1-14 months) have been recorded. In comparison, the majority of cattle (although the sample size is very small) and dogs that died and became ABGs had reached maturity (Figure 6.22). In some respects this pattern differs to the one seen from Owslebury (Figure 6.8). At Owslebury a higher proportion of the sheep/goat and pig ABGs are from adults. However, the main kill-off events still occur at the same age. At Greyhound Yard the majority of cattle ABGs are from adults, but at Owslebury a higher proportion are from sub-adults and young adults. The main difference is in the dog ABG assemblage, at Greyhound Yard most are from adults, whereas the majority recorded from Owslebury are from neonates.

The pig ABG and non-ABG ageing data are also similar. The majority of non-ABG pig deposits are from animals that were under a year old at time of death. However, unlike the ABG data, older animals are present in the non-ABG assemblage with around 30% of the population reaching adolescence. In the early Romano-British period 26% and 8% in the later Romano-British period reach adulthood (Maltby 1993, 326). Therefore most of the pig population was culled when still immature, which Maltby (1993, 325) suggests reflects that some pigs were being kept in the town and that the urban population had a preference for the consumption of sucking pig. The ABGs present appear to have been part of the pig populations first kill off stage, as no older pig ABGs are recorded. None of the pig ABGs had butchery marks present and the ageing data led Maltby (1993, 325) to suggest that they may be representative of natural mortalities from the towns breeding stock.

As at Owslebury, the majority of the dog remains from Greyhound Yard were recovered as ABGs. Therefore the dog ABG age data represents the whole assemblage from the site. The non-ABG cattle ageing data shows a similar 83

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 6.23 Percentage representation of body areas for partial dog ABGs from Romano-British Greyhound Yard (data from 31 ABGs)

Figure 6.24 Percentage representation of body areas for partial sheep/goat ABGs from Romano-British Greyhound Yard (data from 7 ABGs)

84

Chapter 6. Romano-British Southern England

Figure 6.25 Percentage representation of body areas for partial cattle ABGs from Romano-British Greyhound Yard (data from 13 ABGs) 6.6.3

Greyhound Yard composition

However, this is still a small sample. The vertebraedominated pattern is similar to that seen from the sheep/goat ABGs at Owslebury (Figure 6.14). Unlike Owslebury lower hind limbs are present in smaller numbers, and the upper hind limb appears to be more common than the front limb.

The majority of the ABGs recorded from Greyhound Yard consist of partial skeletons, with complete skeletons only present for sheep/goat, dogs and raven (Table 6.8). As at Owslebury, the majority of complete skeletons are of dogs. The majority of sheep/goat ABGs consist of partial remains. However, unlike the other main domestic mammals, three (30%) complete ABGs are also present, all juveniles, from pit 2310.

The partial cattle ABGs show a very different pattern. Vertebrae are still the most common element present, but there are no occurrences of head or rib elements for the partial ABGs with body-area data (Figure 6.25). Limb elements are present, with the lower hind-limb the most common appendicular body-part encountered. The majority of the partial cattle ABGs consist of either vertebral elements, or appendicular elements, with few deposits of both body areas together, which ties in with the overall pattern.

As with the Owslebury material, Maltby (1993, 326) suggests that the majority of partial dog ABGs from Greyhound Yard may have originally been deposited as complete skeletons, before being disturbed by taphonomic processes. The body area data for partial dog ABGs does suggest that this might be the case. The majority of the body areas are evenly represented in the partial ABG assemblage. The majority of partial dog ABGs, (70% (31) of those with body area data present), have either front or hind upper limb bones present. The remainder of the body areas, (head, vertebrae, ribs, pelvis, and lower limbs) are present in just under half of all ABGs (Figure 6.23).

Table 6.8 Composition of Romano-British ABGs from Greyhound Yard Species Cattle S/G Pig Horse Dog Cat

If partial ABGs were being deposited, we could expect uneven representation of body area data. The higher proportion of limbs present is because three main types of partial dog ABG appear to be present on the site: appendicular elements only; axial elements only; and appendicular and axial elements together. The majority of the partial dog ABGs consist of appendicular and axial elements, followed by just appendicular elements. In comparison with the partial dog ABG remains, those of sheep/goat show a very different pattern (Figure 6.24). The majority of sheep/goat partial ABGs consist of elements from the axial skeleton only, with the head and vertebrae the most common body areas represented. Body area data were present for all partial sheep/goat ABGs.

85

Full 3

42

Partial 13 7 8 1 49 9

Unknown

8

Percentage complete 30% 42% -

Red Deer

1

-

Hare Domestic Fowl Raven Rook/ Crow Jackdaw Pigeon Red kite

2

-

3 7

13%

1

2 2 3 1

1

-

Investigating Animal Burials; Ritual, Mundane and Beyond cattle (14%) (Figure 6.26). As with the urban assemblage we need to consider the extent to which large assemblages might influence the data. Most ABG data come from Owslebury, contributing 42% (39) of the assemblage, with the Poundbury settlement contributing 17% (16) and the remaining 41% (38) coming from the small assemblages of 12 separate sites (Figure 6.27). Therefore this assemblage consists of data from a number of sites, unlike the urban assemblage which is largely a reflection of the Greyhound Yard data.

The Greyhound Yard pattern differs to the one observed from Owslebury where a large portion of the partial cattle ABGs consisted of vertebral and rib elements. Maltby (1993, 318) notes that there is also a low proportion of axial elements in the non-ABG assemblage, possibly due to little primary butchery taking place in the area of Greyhound Yard. Only three partial pig ABGs have body data. All three deposits consist of elements from the head, axial and appendicular skeletons. Unfortunately data are not available for the other partial pig ABGs. The one partial horse ABG also consists of elements from the head, axial and appendicular skeleton.

6.7

There is some variation between sites in the early Romano-British rural settlement ABG assemblage (Figure 6.27). As discussed above, dogs are the most common species deposited at Owslebury. However, dog ABGs are not as common on the other rural settlements in this period. The majority, 68% (11), of ABGs recorded from Poundbury are sheep/goat. No other rural settlements have such extreme species proportions. The other rural settlement have very small sample sizes ranging from one to four ABGs. The combined results from these sites indicate that sheep/goat, cattle and pig ABGs are all more common than those of dog. The reason the overall results from rural settlements show a high proportion of sheep/goat and dog ABGs is due to the influence of the Owslebury and Poundbury samples on the assemblage (Figure 6.26). Therefore the high proportion of dog ABGs on Owslebury is unusual for the time period.

Site type differences

The analysis has shown that differences in the ABG assemblage composition are apparent between samples from Owslebury, Greyhound Yard and Oakridge Well. Such differences are possibly due to the nature of the sites and contexts. This section investigates if these differences are apparent for the total ABG assemblage from the Romano-British period. As in the preceding periods, such an investigation requires some generalisation. As discussed above, differences in the faunal assemblages are apparent between urban centres and rural settlements. This is possibly due to a continuation of an Iron Age diet, compared to the ‘Romanized’ one practised in towns and military sites. To investigate if differences are apparent in the ABG assemblage, sites were placed into four categories, rural, town, military and funerary sites. 6.7.1

The majority of the dog ABGs from Owslebury consist of the remains of neonates. In comparison, the dog remains from the other sites consist of one juvenile from Poundbury (Buckland-Wright 1987), one sub-adult from Quarry Field (Clark 2002), four adults from Brighton Hill South (3) (Maltby 1987a) and Houghton Down (1) (Poole 2000c), the remains of two dog ABGs are of unknown age. Also a number of the dog ABGs from Owslebury were discovered in association, suggesting that the individuals may have been from the same litters. No such deposits were encountered from the other rural settlements. If the dog ABGs from Owslebury were of litters, but the deposits from other sites were older individuals, this would explain the higher proportion of dog ABGs recorded from Owslebury.

Town and rural settlements

The majority of the ABGs recorded come from either rural or urban sites. The most apparent differences between these site types occur in the early RomanoBritish assemblages (Figure 6.26). Dog is the most common species from town sites, making up 56% of the ABG assemblage from the early RomanoBritish period (Figure 6.26). Cattle are the second most common species, but only make up 14% of the assemblage, with pig and cat ABGs the joint third most common making up 8% each. However, we must consider that the majority of the data comes from Greyhound Yard. Only four early Romano-British ABGs are recorded from urban sites other than Greyhound Yard, two from Hyde Street (Birbeck and Moore 2004) and one each from Silchester Forum-Basilica (Grant 2000) and Silchester North Gate (Hamilton-Dyer 1997b).

The large number of sheep/goat ABGs from Poundbury in the early Romano-British period are also unusual compared with the other rural settlement assemblages. The majority of the Poundbury sheep/goat data comes from a deposit in pit D607, which contains seven complete sub-adult sheep/goat ABGs. The sheep/goat all appear to have died and been deposited at the same time, with no evidence of butchery. Buckland-Wright (1987) suggests that the animals may have all died in the winter period from ‘haxia’ (starvation), and comments that seven complete fully healthy sheep/goat would not have been able to fit in the pit. Therefore the number of sheep/goat ABGs from the site has been distorted by one large multi-ABG deposit.

The early Romano-British rural settlements show a very different pattern. Sheep/goat are the most common species, at 30%, followed by dog (24%), pig (14%) and

86

Chapter 6. Romano-British Southern England

Figure 6.26 Percentage of species represented in the ABG assemblages from rural settlements and towns, for the early (top), middle (middle) and late (bottom) Romano-British periods

87

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 6.27 Proportion of ABGs per species from early Romano-British rural settlements

Figure 6.28 Proportion of ABGs per species from late Romano-British rural settlements

Figure 6.29 Proportion of ABGs per species from late Romano-British towns

88

Chapter 6. Romano-British Southern England

Figure 6.30 ABG species proportions for late Iron Age and early Romano-British rural settlements assemblages being ‘drowned out’ by the large dog ABG dataset.

From the middle Romano-British period onwards the dog-dominated pattern is present on urban and rural settlements. Over half of the middle and late RomanoBritish rural settlement assemblages consist of dog ABGs. In comparison sheep/goat are found in much smaller numbers on rural settlements compared with the early Romano-British period (Figure 6.26). For the middle Romano-British period we must consider the small sample size. For rural settlements only 16 ABGs are recorded, ten of which are from Owslebury, the other six ABGs come from four sites (Oakridge Well, Maiden Castle Road, Bradford Down, Cowdery's Down). The middle Romano-British urban assemblage is also of limited size and as with the early Romano-British assemblage is dominated by the results from Greyhound Yard. Of the 88 ABGs, 86 are from Greyhound Yard, with one ABG from Neatham (Done 1986) and Silchester (Hamilton-Dyer 1997b) each. By the later RomanoBritish period the dog ABGs dominate both the urban and rural settlement assemblages, with both site types showing very similar patterns (Figure 6.26).

A number of large datasets are also present in the late Romano-British urban ABG assemblage. The results from Winchester Northern Suburbs make up 40% (49) of the assemblage, with Greyhound Yard (26%) and Silchester Insula IX (18%) also making up a large proportion. The remaining 16% (19) of the assemblage comes from the small datasets ranging from one to nine ABGs, from seven urban sites, Colliton Park, Dorchester (Aitken and Aitken 1982), Dorchester Prison (Draper and Chaplin 1982), South Grove Cottage, Dorchester (Startin 1981), Staple Gardens, Winchester (Maltby 1986b), Silchester Forum-Basilica (Grant 2000), Silchester North Gate (Hamilton-Dyer 1997b) and Neatham (Done 1986). These urban sites all show similar patterns with a large proportion of dog ABGs present (Figure 6.29). Pig and domestic fowl ABGs are more common from ‘other’ urban sites. The rural and urban sites therefore show different patterns with regard to species proportions. Urban sites from the early Romano-British period onwards produce a dog-dominated assemblage. The rural assemblage appears to change between the early to middle RomanoBritish periods. Sheep/goat are the most common ABG in the early Romano-British period. This changes to a dogdominated pattern by the middle Romano-British period, with rural and urban assemblages showing similar patterns in the late Romano-British period The ruralsettlement, early Romano-British pattern can be viewed as a continuation of the late Iron Age pattern (Figure 6.30), indicating a possible continuation in the processes and choices which resulted in ABG deposition on rural settlements. The differences in the early Romano-British period between rural and urban sites could be viewed as a result of ‘Romanization’, with the ABG assemblages on urban sites changing little throughout the Romano-British

The rural settlement assemblage is dominated by the results from Owslebury and Oakridge Well, which combined account for 86% (310) of the sample. The remaining 40 ABGs come from 10 sites with smaller assemblages. However, when the species, proportions are compared, the most common ABGs are still those of dog (Figure 6.28). The ‘other’ sites do however, have a higher proportion of cattle, sheep/goat and pig. This may be because of the type of dog ABG assemblage present. Only Little Somborne (Maltby 1978b) and Maddington Farm (Hamilton-Dyer 1996a) contain neonatal-dog, multi-ABG deposits. The other-dog ABG deposits are of individual dogs of variable age. In contrast the Owslebury and Oakridge Well dog ABG assemblages are dominated by very large multi-ABG deposits of neonates, possibly from the same litters. This results in the other species

89

Investigating Animal Burials; Ritual, Mundane and Beyond Due to the lack of other military sites, we cannot be certain that the assemblage from Porchester Castle is representative of this type of site. We must also consider that the ABG data were extracted from the excavation descriptions, not the faunal report. Complete skeletons are more likely to be noted during excavations and therefore included in the excavation report. However, partial ABGs may have been present but not deemed worthy of description, or not noted.

period. As discussed above, there is a general trend in the south of Britain indicating a gradual adoption of a Romanized culture. Therefore the changes seen in the ABG assemblage from rural settlements, to a pattern similar to the one observed initially from urban sites, could be viewed as Romanization of the rural ABG assemblage. 6.7.2

Military and villa sites A small number of assemblages were also recorded from villas. Eleven ABGs were recorded from four villa sites (Table 6.3), dating to the early and late Romano-British periods. The current dataset appears to indicate that ABGs are rare from villas. It should be noted that, like villas, a large proportion of the rural settlements recorded also have very small ABG assemblages.

Unfortunately ABG data are only available from one military site, Porchester Castle, a later Romano-British shore fort (Cunliffe 1975). The lack of ABGs on the other Romano-British military sites recorded may be due to the excavations taking place on a small scale on the periphery of the site. The Porchester Castle excavations are the largest to take place on a Roman military site in the study region and concentrated on the interior of the fort.

Only ABGs of sheep/goat and cattle are present from the early Romano-British villa assemblage (Table 6.9). A high abundance of sheep/goat ABGs is also seen on rural sites, but the villa assemblage is too small to make any conclusive comparisons. Seven ABGs are recorded dating to the late Romano-British period, the majority of which are of dogs. One sheep/goat and two horse ABGs are also present. Again the late Romano-British period sees dog ABGs become dominant on most sites, but more data would be required from villa sites to make any firm conclusions.

In total, 44 ABGs were recorded from Porchester Castle, the majority coming from the late Romano-British period. In the middle Romano-British sample the most common ABGs are those of pig and cat, with dog and cat ABGs being the most common in the late Romano-British assemblage (Figure 6.31). The high proportion of cat and pig ABGs is unusual and not seen in other Romanized assemblages, such as those from urban sites. The majority of the ABGs consist of complete, or nearly complete skeletons and most, with the exception of two cats and one late Romano-British dog ABG, are from neonatal or juvenile animals.

Figure 6.31 ABG species proportions from Porchester Castle Table 6.9 Number and proportion of ABGs per species from villa sites Period of deposit Cattle S/G Horse Dog Early Romano-British 25% (1) 75% (3) Late Romano-British 14% (1) 29% (2) 57% (4)

90

Chapter 6. Romano-British Southern England 6.7.3

Funerary sites

contexts (Table 6.11). There is also an increase in the number of complete ABGs, particularly of domestic fowl deposited. Although domestic fowl become more common on urban and rural sites in the later RomanoBritish period (Maltby 1997), the majority of domestic fowl ABGs recorded from this period are from funerary contexts. It therefore appears that domestic fowl ABGs are more likely to be encountered from funerary contexts.

Of the 820 ABGs from the Romano-British period, 49 (5.9%) come from funerary contexts, from four sites, Allington Avenue (Maltby 2002a), Poundbury Cemetery (Buckland-Wright 1993), Lankhills Cemetery, Winchester (Brothwell and Harcourt 1979) and Maiden Castle (war graves) (Jackson 1943). All of the ABGs discussed in this section come from graves and are associated with articulated human remains. Also included in this section are ABGs from a number of late Iron Age – early Romano-British ‘Durotrigian’ burials from Allington Avenue, Poundbury and Maiden Castle. Therefore a number of ABGs, (22), discussed in this section date to the late Iron Age but are included in this section as the burial rite overlaps into the early RomanoBritish period.

The change in species proportions between the early and later parts of the Romano-British period could again be seen as a result of Romanization. The majority of the ABGs from funerary contexts are interpreted as grave goods, gifts of food for the deceased (Bailey 1967; Buckland-Wright 1990; Sykes 2003), apart from the complete dog burials which are seen as companion offerings (Buckland-Wright 1993). The ABG data may therefore be showing a change in food tastes, with the increase in domestic fowl ‘offerings’. However, if this was the case, then we could expect pig ABGs to be also present, as pig was certainly consumed on urban and rural sites during the later Romano-British period. Also Cicero indicates that only when a pig had been sacrificed was a human burial legally a grave under Roman law (Toynbee 1971, 50), although this does not necessary mean the sacrificed pig would always have been included in the grave among the incorporated grave goods. Possibly the ABGs incorporated into the graves were from animals that were suitable for the deceased to consume.

The majority of the late Iron Age – early Romano-British ABGs from funerary contexts consist of partial ABGs, with over half belonging to sheep/goat (Table 6.10). ABGs of domestic fowl and pig are the equal second most common. The partial ABGs consist of either appendicular elements, normally the upper fore or hindlimb bones, or axial elements, normally a number of ribs. Domestic fowl ABGs are not common in this period, and the ones recorded from grave contexts account for over half of the domestic fowl recorded from late Iron Ageearly Romano-British contexts. The ABGs from middle to late Romano-British funerary contexts show different species proportions, with domestic fowl the most common species represented, followed by sheep/goat and then dog. No pig ABGs are recorded from middle to late Romano-British funerary

The ABGs covered in this study are only from inhumations, but cremation burials became prevalent towards the end of the early Romano-British period

Table 6.10 Number of complete and partial ABGs per species from late Iron Age and early Romano-British funerary contexts Total Percentage of total Species Complete Partial Unknown Total percentage assemblages Cattle 2 6.1% 3.2% 2 S/G 1 16 51.5% 13.9% 17 Pig 6 18.2% 18.8% 6 Dog 1 1 6.1% 1.9% 2 Domestic Fowl 5 1 18.2% 54.5% 6 Total 2 30 1 33 Table 6.11 Number of complete and partial ABGs per species from middle and late Romano-British funerary contexts Total Percentage of total Species Complete Partial Unknown Total percentage assemblage Cattle 1 2.6% 1.7% 1 S/G 1 9 26.3% 17.5% 10 Dog 4 2 1 18.4% 2.3% 7 Domestic Fowl 7 12 1 52.6% 51.3% 20 Total 12 24 2 38 91

Investigating Animal Burials; Ritual, Mundane and Beyond partial ABGs due to post-deposition taphonomic activity. If this is the case, it would explain the large proportion of partial ABGs with all body areas represented. It may also indicate that the majority of domestic fowl were also deposited as complete skeletons, that later became partially disarticulated.

(Hope 1999). No ABGs are recorded in this study from cremation burials, but finds from the 1930’s excavations at Milland, Winchester, did indicate that ABGs may have occasionally been deposited with human cremations. Associated with a cremation burial was a terra nigra platter, pig, woodcock and two domestic fowl ABGs, all consisting of appendicular elements (Collis 1978, 103). 6.8

The partial sheep/goat ABGs have a different pattern. The majority of partial ABGs are still composed of either axial or appendicular elements, but 20% also consist of elements from both body areas. This indicates that a proportion of the partial sheep/goat ABGs often are not as disarticulated as those from larger domestic mammals. This suggests that sheep/goat ABGs have either gone through a different butchery process, or if no butchery occurred, they may not have been left to decompose naturally for as long as the larger mammals (cattle and horse).

Composition and butchery data

As discussed already, the majority of the ABGs consist of partial skeletons. The composition of the ABGs from the three largest assemblages has been discussed above. This section is concerned with the composition of the overall ABG assemblage. As shown above the composition of the partial ABGs varies between species. Overall for cattle there appears to be two main type of partial ABG, axial elements or appendicular elements. Only 2% of the partial cattle ABGs recorded (with the available data) had elements from both areas of the body present. Horse ABGs also have a similar pattern, although 10% of the partial ABGs have all body areas present. The rest of the partial horse ABGs are composed of either axial or appendicular elements (Figure 6.32). Many of the partial pig ABGs, (48%), consist of elements from just the appendicular skeleton. Bones from the head are also more common, especially when deposited with axial elements. Around 10% of the partial pig ABGs also consist of elements from all areas of the body.

To investigate this theory, butchery marks were recorded when present. In total, 34 butchery marks were recorded from 20 ABGs, which represents just 2.4% of the total Romano-British assemblage. ABGs with butchery marks were present from 12 sites, Alington Avenue (2), Castle Copse Roman Villa (2), Greyhound Yard (7), Little Somborne (1), Manor Farm (1), Oakridge Well (2), Old Down Farm (2), Owslebury (1), Poundbury Cemetery (1), Poundbury settlement (1), Quarry Field (1) and Silchester Insula IX (2). Greyhound Yard has the largest number of ABGs with butchery marks present, (7). All the other sites have only one or two ABGs with marks present. Butchery marks were only present on ABGs of cattle, sheep/goat, pig and dog. However, the number of wild mammals and birds present within the assemblage is small. Butchery marks are present on both complete and partial ABGs, although the majority are observed on partial ABGs (Table 6.12). Sheep/goat followed by cattle most commonly have butchery marks present. The data also show around 10% of partial cattle and sheep/goat ABGs to have butchery marks present.

The dog and domestic fowl partial ABGs have a similar pattern and, like the pig ABGs, a high proportion of them consist of appendicular elements. Unlike the other species, over 30% of dog and domestic fowl ABGs consist of partial skeletons with all elements of the body present. As discussed above, a large proportion of both species are deposited as complete skeletons. The results from a number of sites in this study have led Maltby (1987c; 1987d; 1988; 1990f; 1993) to suggest that most dogs were deposited as complete skeletons, but became

Figure 6.32 Body area percentage for Romano-British partial ABGs (mixed indicates all body areas represented)

92

Chapter 6. Romano-British Southern England Table 6.12 Number and composition of ABGs with butchery marks present

ABG composition Full Partial

Species S/G Dog Cattle S/G Pig Dog

Total Percentage of whole assemblage

Early RomanoBritish 2 1 4 4 1 1 13 8.3%

Middle RomanoBritish

Late RomanoBritish

1 1

2 2

2 1.7%

1 5 0.9%

Percentage of whole species & type assemblage 9.1% 1.0% 10.3% 9.5% 2.4% 1.9%

Total 2 1 7 7 1 2 20 2.4%

Table 6.13 Number of butchery marks recorded per species and body area Upper Lower Upper front front back Period of deposit Species Head Trunk Pelvis limb limb limb Early Romano-British

Cattle S/G Pig Dog

2

3

1

Middle

Cattle

1

Romano-British Late Romano-British

S/G Cattle S/G Dog

1 2 1 1 9 26.5%

Total Percentage Total

1

1 1 5 14.7%

2 2 2 1

2 1 2

Lower back limb 2 1 1 1

1 1 1 2.9%

8 23.5%

1 2.9%

5 14.7%

5 14.7%

Table 6.14 Type of butchery marks observed on ABGs in the Romano-British period Period of deposit Species Cleaver Knife Unknown Early Romano-British Cattle 1 3 S/G 1 10 Pig 3 Dog 5 Middle Romano-British Cattle 1 S/G 1 1 Late Romano-British Cattle 2 1 S/G 1 2 Dog 2 Total 11 22 1 in dog ABGs in the middle and late Romano-British period, we would expect to see a decrease in the number of butchery marks recorded. Also, importantly, if dogs were not butchered, we would expect them to have been deposited as whole carcasses and therefore be more likely to survive as an ABG.

The majority of the ABGs with butchery data are from early Romano-British contexts (Table 6.12). This could be because of a change in Romano-British butchery methods with the possible introduction of professional butchers (see below). Also the data do correspond with the rise in the number of dog ABGs. If the dogs were not being consumed then we would expect only a small amount of butchery marks present. There is evidence that some dogs were being skinned (Clark 2002; Maltby 1987c). If dogs are not subject to as extensive butchery compared to other domestic mammals, then with the rise

Butchery marks were recorded as present from all areas of the body (Table 6.13). Marks appear to be most prevalent on the upper fore and hind limbs in the early Romano-British period, and on the vertebra and ribs in 93

Investigating Animal Burials; Ritual, Mundane and Beyond the late Romano-British. However, we need to consider that the areas butchery marks are recorded from are dependent on which body parts are deposited as partial ABGs. Therefore the pattern is more a reflection of the ABG composition, especially considering the small butchery mark sample size. Unlike the Iron Age, where the majority of the butchery marks appear to have been made using knives (Maltby 1981b; Wilson 1978), during the Romano-British period the cleaver starts to be used with more frequency, possibly due to the introduction of professional butchers (Maltby 1989d; 2007; Seetah 2006). This change is also seen from the ABG data, as all the butchery marks from Iron Age ABGs were from knifes. However one third of the Romano-British observed-butchery marks are made by a cleaver (Table 6.14). The majority of the ABGs observed with knife marks are from the early Romano-British period, with more cleaver marks observed on late Romano-British ABGs. This may be a reflection of the site type data, as the majority of the early Romano-British butchery data are from rural settlements. However, the majority of the late RomanoBritish data is from Oakridge Well. This may indicate that the use of the cleaver became more widespread during the later part of this period. However, one must be cautious in drawing such conclusions based on such a small set of butchery data. 6.9

Figure 6.33 Associated dog ABGs, allegedly tied together at the throat from early Romano-British pit 2310 (Woodward et al. 1993, Plate 25).

Associations

Of the 820 ABGs recorded from the Romano-British period, 177 (21.5%) are from only 46 multi-ABG deposits, from 14 different sites. The largest assemblage comes from Greyhound Yard. A total of 119 (73%) of the 163 ABGs from the site are from 26 multi-ABG deposits. The majority of the multi-ABGs are those of dog, with 77 of the 99 dog ABGs recorded from Greyhound Yard occurring in such deposits. They include two adult dogs from the early Romano-British period that were possibly tied together at the throat (Figure 6.33).

species composition seen in the overall ABG assemblage and could also be seen as a continuation of a late Iron Age trend for multi-ABG sheep/goat deposits. The majority of the sheep/goat multi-ABGs deposits in the early Romano-British period are from the Poundbury settlement. As discussed above, one pit contains the remains of seven sheep/goat, all deposited together. The proportion of dog multi-ABGs increases in the middle Romano-British period. This corresponds with the overall ABG species proportion data, especially considering that the majority of the data come from Greyhound Yard.

The majority of multi-ABG deposits occur in pits and wells (Table 6.15). However, this is to be expected as the majority of the ABG data comes from pit deposits. Also most multi-ABGs are recorded from urban sites. We must also consider that a large proportion of such data is from Greyhound Yard. This is likely to affect the species proportions with dogs followed by sheep/goat and then pig the most common (Table 6.16). Dog remains the most common species in multi-ABG deposits throughout the different Romano-British sub-periods, but the proportions do vary.

The picture from the late Romano-British period is less clear, as the proportion of dog multi-ABG deposits drops. The majority of the data in this sub-period comes from rural settlements, although all of the urban assemblage still comes from Greyhound Yard. The rural settlement multi-ABG data is probably under-represented. This is because only a limited number of ABGs from Oakridge Well were recorded as multi-ABGs. This is because the ABGs were often spread over a number of contexts and it was not possible in most cases to ascertain if ABGs were deposited in direct association. Figure 6.16, shows the relative position of the ABG deposits within the well. For this diagram the highest possible ABG position within the well was utilised. Therefore, some of the ABGs may not be in exactly the right position, but the diagram does indicate the volume of ABGs recovered from the well. It

Sheep/goat multi-ABG deposits are most common in the early Romano-British period. This corresponds with the

94

Chapter 6. Romano-British Southern England The majority of the ADGs recorded are in association with human remains within a funerary context, which are discussed above. Only three other ADGs are recorded. Two from Houghton Down (Poole 2000c), which consist of a dog and cat ABG both associated with flint nodules. The other recorded ADG consists of a cat ABG from Silchester Northgate, which was deposited within what is described as a tile-cist burial. The deposit as interpreted as the burial of a pet (Hamilton-Dyer 1997b).

is also likely that many were first deposited as multiABG deposits. Dog is still the most common species to be included in multi-ABGs in the late Romano- British period. Pigs are the most common species on rural sites. However, four of the rural pig deposits come from one deposit of neonatal pigs at Maddington Farm (Hamilton-Dyer 1996a). A number of features have been described in this chapter that contain a large multi-ABG deposit. However, the majority of such deposits consist of only two ABGs in association (Table 6.17). Only the middle RomanoBritish examples contain large numbers, with four deposits from Greyhound Yard consisting of between 10 and 15 ABGs.

The Silchester excavations have also revealed a number of other possible ADGs, which highlight the problems of recording associations. For example, pit 3251 contained a dog ABG, but also deposited in the pit was a complete pot in the context below the dog and a human infant burial with a coin in the context above. This deposit was not recorded as an ADG, as the dog was deposited in a separate context. However, the passage of time between all three deposits is unknown. Therefore we need to consider if the deposits were related in some way.

The majority of the multi-ABG deposits also consist of just a single species (Table 6.18). However two, both from Greyhound Yard, consist of six species. One early Romano-British multi-ABG from pit 4161 contains seven dogs as well as one cat, pig, sheep/goat, jackdaw and raven ABG. The middle Romano-British deposit also consists mainly of dog, with a number of wild birds. Therefore, although most multi-ABG deposits consist of a small number of same species, there are some examples with large numbers of ABGs from different species.

A number of possible ADGs are also present from Oakridge Well. An almost complete and largely unbroken 1st century AD flagon was recovered close to the base of the well. Oliver (1992, 74) suggests that the watercarrying vessel had probably been placed at the bottom of the newly dug well, as a propitiatory offering. Two, virtually undamaged, Antonine flagons, were also recovered from within the late 3rd/early 4th century deposits (Figure 6.16). The flagons are a unusual samian form and were thought to have been around 150 years old when deposited within the well (Oliver 1992, 75).

Associations between ABGs and other deposits of material the excavators determined to be ‘special’ were also recorded. However, only a limited amount of data regarding Associated Deposit Groups (ADGs) was available. The main reason for this is the reliance on English Hertiage AML reports for the ABG data. As these reports only deal with the faunal remains, it was not possible to ascertain if other material was deposited in association with the ABGs. Also such an investigation is reliant upon the excavation report recording such associations. For example, no ADGs were originally recorded from Greyhound Yard (Woodward et al. 1993). However, reinterpretations have indicated that they were present (Woodward and Woodward 2004).

Further up the well two complete imitation samian Oxfordshire bowls were found, possibly connected to the deposition of human remains of at least 14 individuals (Oliver 1992, 76). In total the human remains came from a minimum of 27 individuals. None of the human remains were found in articulation within the well and some of the individuals are represented by only a few bones.

Table 6.15 Number of multi-ABG deposits per period/site type and feature. The number in brackets indicates the number of ABGs from the deposits Period of deposit Site type Grave layer Pit Quarry Well Early RomanoFunerary 4 (13) British Settlement 2 (9) Town 1 (2) 5 (23) 1 (2) Middle RomanoSettlement 1 British Town 1 (9) 13 (68) Late RomanoFunerary 3 (6) British Settlement 3 (9) 1 (2) 5 (13) Town 4 (14) 1 (2) Villa 1 (3) Total 7 (19) 2 (11) 28 (126) 1 (2) 8 (19)

95

Table 6.16 Species inclusion in multi-ABG deposits per site type for the Romano-British periods

Investigating Animal Burials; Ritual, Mundane and Beyond

96

Chapter 6. Romano-British Southern England Table 6.17 Number of ABGs within multi-ABG deposits Period/ No. ABGs 2 3 to 5 6 to 10 10 to 15 Early Romano-British 7 (54%) 4 (31%) 1 (8%) 1 (8%) Middle Romano-British 7 (47%) 1 (7%) 3 (20%) 4 (27%) Late Romano-British 11 (61% 6 (33%) 1 (6%) Table 6.18 Number of different species present multi-ABG deposits Period 1 species 2 species 3 species 4 species 5 species Early Romano-British 6 4 2 Middle Romano-British 6 6 1 1 Late Romano-British 11 7 Total 23 (50%) 17(37%) 3 (7%) 1 (2%) -

2 (4%)

deposits, that more ABGs have been recorded from the southern England Romano-British period, than the Iron Age. As with the previous Iron Age assemblage, a wide variety of wild bird and mammal ABGs are present in the assemblage. However, a large proportion of these are from the Oakridge Well site.

The human remains material was often fragmentary with some joining fragments located several feet in depth from one another (Mays 1992, 88). Non-metric trait evidence does indicate that some of the individuals deposited around 54ft depth may have been related. Some of the human remains, especially towards the top fills of the well appear to have been deposited with no ABGs in close proximity (Figure 6.16). However, the 14 individuals, as well as some of the complete pottery were deposited in fills which also contained ABGs. However, the associations between the ABGs and the ‘other’ material types are not alluded to in the report. Therefore it was not possible to record any ADGs from Oakridge Well. The excavation report struggles with the apparent disparity between the deposition of complete pots and human remains, with the deposition of a large amount of animal bone ‘butchery waste’. The large grouping of animal and human remains leads Oliver (1992, 92) to suggest;

Overall the most common ABGs are those of dogs, with the proportion of dog ABGs increasing from the early Romano-British period. In this respect, the RomanoBritish ABG assemblage differs to those from earlier periods, in that the most common ABG species is not the most common non-ABG species. However, there are a number of differences between site types. Sheep/goat rather than dog ABGs are the most common species from rural early Romano-British sites, in fact the rural Romano-British species proportions are similar to those found on Iron Age sites. In comparison, ABGs of dogs from the early Romano-British period are the most common species recovered from town sites. This pattern changes in the middle Romano-British period when dog ABGs become common on rural as well as town sites. Dog ABGs are also the most common species found in multi-ABG deposits. Such a change in species proportion could be a sign of the gradual adoption of Roman practices.

‘It seems likely that some kind of plague affected both sheep and cattle [due to the large number of juvenile remains]…the large number of human burials…probably also reflects some kind of disaster. The finds suggest that this cataclysmic year occurred round the last quarter of the 4th century. It is possible that the disaster was the result of some kind of raid…It is more likely that both human and stock fell victim to some disease, possibly one, such as anthrax, which affects both humans and animals.’

We must however, be aware that the Romano-British assemblage is dominated by three sites with over 150 ABGs recorded from them, Greyhound Yard, Oakridge Well and Owslebury. These sites allow us to investigate the composition of ABGs in detail. They show that the animals which were recovered as ABGs were killed/died at a younger age than the animals which compose the non-ABG faunal material. The Owslebury material shows some variation in the features ABGs are recovered from. At the site, a large proportion of sheep/goat, dogs and domestic fowl ABGs are recovered from pits whereas the other species are recovered from ditches. This may correspond with patterning seen in the non-ABG faunal assemblage.

What is certain is that far more Romano-British ADGs exist than have been recorded in this study, but at present the published reports do not aid in their identification. 6.10

6 species 1 1

Summary

The southern England Romano-British dataset represents the largest and most detailed assemblage recorded for the study. ABGs appear to be common phenomena with 55% of sites recorded with faunal remains also having ABGs present. It is noteworthy, considering that the majority of the previous literature on ABGs concerns Iron Age

A number of ABGs are also recorded from funerary contexts. The data indicates there is a change in the species which are deposited in this manner. Partial ABGs of sheep/goat and pig are the most common types 97

Investigating Animal Burials; Ritual, Mundane and Beyond recorded from late Iron Age and early Romano-British funerary sites around the Dorchester region. By the middle and late Romano-British periods both complete and partial domestic fowl ABGs are the most common deposits, followed by partial sheep/goat ABGs. Pig ABGs do not appear to be deposited in funerary contexts at this time. As with the change in ABGs on non-funerary sites, a change in species proportion may be the result of Roman influence and changing practices within society. Overall the Romano-British assemblage shows that the composition of ABGs is not static, it changes as society does. It also shows the value of moving beyond the Iron Age in discussing these types of deposits.

98

Chapter 7. Romano-British Yorkshire 7.1

Introduction

Roman conquest and occupation of the Yorkshire region occurred in the late 1st and early 2nd centuries AD. Although most of the Romano-British site types and material culture are represented in the archaeology from Yorkshire, a number of differences between the north and south of England do occur. Villa sites are not as common and a military presence is more evident in the north. Until recently the majority of the archaeological evidence came from military and town sites, in particular York and Catterick. Faull (1981) suggested that, compared to much of southern and eastern Britain, Roman (west) Yorkshire is very much an unknown quantity. However, Ottaway (2003) has shown that much work has been carried out to address this since Faull’s comment. The overall faunal evidence shows a similar cattledominated Romano-British pattern to that found further south, but the ratios of cattle to sheep/goat and pig, tend to be greater than 70% even in the 1st century AD (Stallibrass 1995b; 2000). Stallibrass (2000) notes that both military and non-military sites show a heavy reliance on cattle, probably due to the environmental constraints and potential of the region. Compared with the southern England Romano-British assemblage, the data from Yorkshire are limited. In total, 36 sites were recorded, with ABGs present on 21 (58%). Therefore, as with the Yorkshire Iron Age data, around half of the sites recorded have ABGs present. A similar pattern is evident in the southern England RomanoBritish dataset (see 6.1). The majority of the sites utilised are from the Yorkshire Wolds and Permian Ridge regions. Two clusters of sites are present around York and Catterick, which reflects the large quantity of archaeological work that has taken place in these areas. In total, 89 ABGs were recorded from Yorkshire for this period. This represents 43% of the overall Yorkshire dataset and is the largest period sample from this region. However, compared with the southern England assemblage, this sample is still very small. In fact, three individual sites from southern England have larger ABG assemblages than the total Romano-British Yorkshire assemblage. 7.2

Species proportions and context

The majority of the ABGs dating to the Romano-British period from Yorkshire are from domestic mammals (81%), with domestic fowl (14%), wild mammals (3%) and wild bird (2%) completing the assemblage. The proportions of species are similar to those from the southern England Romano-British assemblage (see 6.2). The majority of the ABGs date to the middle or late Romano-British period, and all the domestic fowl, wild

mammal and bird ABGs are from these sub-periods (Table 7.1). All the wild mammal ABGs are from the Rudston villa (Chaplin and Barnetson 1976; 1980; 1981), and all the domestic fowl from Trentholme Drive, York (Fraser and Ryder 1968). The other bird remains consist of a partial crow ABG, represented by 41 elements, and a pigeon, which could possibly be domestic, represented by 12 elements, both from Bainesse Farm, Catterick (Meddens 1990a) (unfortunately the body area information is not available). Therefore, although ABGs from species other than domestic mammals are present they have been found on only a few sites. Only domestic mammals are present in the early Romano-British assemblage, with sheep/goat the most common species deposited. Cattle are the second most common. The rest of the assemblage consists of a small number of pig, dog and horse deposits (Table 7.1, Figure 7.1). However, 12 of the 15 ABGs recorded from this period are from Rudston (see below). Sheep/goat remain the most common ABG in the middle Romano-British sample, but the number of dog deposits rises to second place. However, this assemblage is dominated by two sites, Bainesse Farm (Meddens 1990a) and Shiptonthorpe (Mainland 2006). Remains from these two sites make up 78% of the middle Romano-British assemblage. All the cattle ABGs are from Shiptonthorpe and most of the dogs are from Bainesse Farm. Sheep/goat ABGs were found on all of the rural sites (4) dating to this period. In the late Romano-British sample, the species proportions change again with domestic fowl being the most common species followed by dog (Figure 7.1). Pig are the third most common ABG, after being absent from middle Romano-British assemblages. In comparison with the earlier periods, the proportion of cattle and sheep/goat ABGs decrease markedly. Overall, the early and middle Romano-British species proportions are similar in some respects to those seen in the non-ABG faunal material. The non-ABG early Romano-British NISP figures are only available from two sites with ABGs present, Parlington Hollins (Richardson 2001) and Garton and Wetwang Slack (Noddle 1979). They have different species proportions with cattle and sheep/goat the dominant animals respectively. However, on both sites cattle and sheep/goat are the two most common animals. Therefore the species deposited as ABGs in the early Romano-British period are also the most abundant domesticates in the non-ABG assemblage. A similar pattern is evident in the middle Romano-British dataset (Figure 7.2). Sheep/goat are the most abundant species in both the ABG and non-ABG faunal data. The proportion of non-ABG species from late RomanoBritish sites shows a similar pattern, but with cattle and

Investigating Animal Burials; Ritual, Mundane and Beyond Table 7.1 Number of ABGs per species for the Romano-British period, Yorkshire Late Middle Early RomanoRomanoRomanoBritish Total British British Species Domestic Cattle 4 6 4 14 Mammal S/G 6 11 3 20 Pig 2 6 8 Horse 2 3 5 10 Dog 1 10 8 19 Wild Mammal Red Deer 2 2 Badger 1 1 Domestic Bird Domestic Fowl 12 12 Wild Bird Crow 1 1 Bird Pigeon 1 1 Total 15 32 41 88

Figure 7.1 Percentages of the most common animals to constitute ABGs in the Romano-British period from Yorkshire. Sample size in brackets pig becoming more common on some sites (Figure 7.2). The majority of pig ABGs come from the late RomanoBritish period and this may be linked with the general increase in pig remains from faunal assemblages. However, cattle and, in most cases, sheep/goat are still the most common species present in the overall faunal assemblage. Therefore, unlike the previous RomanoBritish sub-periods, the ABG data are significantly different from the non-ABG assemblage, with the most common animals no longer being the most common ABGs as well.

villa sites are included in the 14 late Romano-British sites.

Such a shift in species proportions may be due to changes in the nature of the sites from which ABGs have been recovered. The majority of the early and middle RomanoBritish ABG deposits are from rural settlements (Figure 7.3), with only one from an early Romano-British fort at Castleford (Berg 1999) and two from middle RomanoBritish urban sites at 9 Blake Street, York (Bond and O'Connor 1999; O'Connor 1987) and Chapel Hill, Aldborough (Jones 1971). By contrast, six town and three

Another factor that needs to be taken into consideration is that a small proportion of the sites recorded account for a large percentage of the ABGs (Table 7.2). Four sites, Shiptonthorpe (Mainland 2006), Trentholme Drive, York (Fraser and Ryder 1968), Bainesse Farm (Meddens 1990a; 2002a) and Rudston (Chaplin and Barnetson 1976; 1980; 1981) each produced ten or more ABGs. The deposits from these sites account for 70.5% of the total Yorkshire Romano-British assemblage (Table 7.2).

As discussed previously, the large proportion of dog ABGs may be due to Roman influences, as they are especially present in urban contexts. Therefore, the increase in town sites could explain the increase in the proportion of dog ABGs (Figure 7.4). However, this does not explain the large proportion of middle RomanoBritish dog ABGs. The majority of these are from Bainesse Farm, close to the town of Cataractanium.

100

Chapter 7. Romano-British Yorkshire

Figure 7.2 Triplot of NISP counts for cattle, sheep/goat and pig for Romano-British sites from Yorkshire, with ABGs present. Triangles = early Romano-British, diamonds = middle Romano-British and squares = late Romano-British

Figure 7.3 Percentages of the most common domestic mammals ABGs from middle and late Romano-British period rural settlements, Yorkshire. Sample size in brackets 7.3

Two of these sites, Shiptonthorpe and Bainesse Farm, are rural settlements and the early Romano-British phases of Rudston are also considered to be a ‘native’ rural settlement (Stead 1980, 35). All the ABGs from Trentholme Drive, York come from funerary contexts. In comparison, all the urban sites produced only small numbers of ABGs. This contrasts with the evidence from southern Britain, but may only be a reflection of the size of the excavations, as the investigated areas of York and Catterick were small in comparison to those from Dorchester and Winchester.

Shiptonthorpe

Shiptonthorpe consists of a rural settlement established in the early-to-mid-second century AD next to a Roman road leading from Brough to Stamford Bridge/York (Millett 2006, 306). The 18 ABGs recorded from the site represent the largest assemblage from Romano-British Yorkshire. However, it is very small in comparison with the largest southern England assemblage. The ABGs from Shiptonthorpe date to the middle and late Romano-

101

Investigating Animal Burials; Ritual, Mundane and Beyond sample. Pig constitute over half of the assemblage, with cattle and sheep/goat making up the rest (Figure 7.5). This pattern differs to that seen from the overall late Romano-British assemblage, where dog and domestic fowl are the most common species, neither of which were recovered from Shiptonthorpe. Pig ABGs are not recorded from earlier Romano-British periods, and the Shiptonthorpe assemblage accounts for over half of all pig ABGs recorded from late Romano-British Yorkshire.

British periods, with 11 and seven ABGs from the respective sub-periods. 7.3.1

Shiptonthorpe species proportion and assemblage composition

The middle Romano-British ABG assemblage from Shiptonthorpe corresponds with the overall trend discussed above, with cattle and sheep/goat being the most common species (Figure 7.5). The two middle Romano-British horse ABGs recorded (Table 7.1) also come from Shiptonthorpe. Both are partial and from adult animals. The first consists of three elements from the right upper forelimb deposited within a ditch (context 649). The second, of four elements from the right upper forelimb and was recovered from a large pit feature referred to as a ‘watering hole’. A site plan indicates that this ABG may have been associated with a horse skull (Figure 7.6).

The middle Romano-British assemblage from Shiptonthorpe does correspond with the non-ABG faunal assemblage, in that cattle and sheep/goat are the two most common species. However, sheep/goat are the most common non-ABG species consisting of 49% of the assemblage, with cattle 38%. We must at this point bear in mind that the ABG sample is very small, which may affect the results. The late Romano-British ABG assemblage differs greatly from the non-ABG assemblage. Although the most abundant ABG species is pig, the majority of the non-ABG faunal remains are still from cattle or sheep/goat, with pig constituting only 10.5% of the non-ABG assemblage.

The species in the late Romano-British ABG assemblage from Shiptonthorpe differ to the middle Romano-British

Figure 7.4 Percentages of the most common domestic mammals ABGs from late Romano-British town and villa sites, Yorkshire. Sample size in brackets Table 7.2 Number of ABGs found on different types of Romano-British site in Yorkshire Site Type 1 2 to 4 5 to 9 10 to 20 Town 4 4 Villa 1 1 1 Military 1 Rural settlement 3 2 2 Funerary 1 Total 9 6 1 4 Percentage total sites 45.0% 30.0% 5.0% 20.0% Percentage total ABGs 10.2% 13.6% 5.7% 70.5%

102

Chapter 7. Romano-British Yorkshire

Figure 7.5 Percentages of ABGs per species from Shiptonthorpe Table 7.3 Number of complete and partial ABGs per period and species from Shiptonthorpe Percentage Period of deposit Species Complete Partial Complete Middle Romano-British Cattle 2 3 40% Sheep 4 Horse 2 Late Romano-British Cattle 2 100% Sheep 1 Pig 2 2 50% Total 6 12 33% The two late Romano-British complete pig ABGs were from a young adult and a juvenile, which were deposited in adjacent shallow pits. The two partial pig ABGs were from neonatal and juvenile individuals and were deposited in the same area as the compete pig ABGs.

An interesting aspect of the Shiptonthorpe ABG assemblage is that a high proportion (33%) consists of complete skeletons, most from cattle (Table 7.3). Three of the four complete cattle ABGs are from neonatal animals, the fourth is a late Romano-British young adult (animal burial 3.9). All the partial cattle ABGs are also from neonatal or possibly foetal animals.

7.3.2 All five sheep/goat ABGs are partial skeletons. Four date to the middle Romano-British period, and consist of three juvenile or young adults and one adult female. The adult female was deposited in pit 981, during ‘interior phase 2’ of a possible round house structure. The lower limb bones of the adult female sheep/goat ABG were burnt, but had remained in articulation. This is the only ABG from the site with butchery marks present, consisting of knife marks on the proximal aspects of a humerus and femur, possibly due to disarticulation of the limbs from the trunk. Unfortunately due to the structure of the faunal report, limited body area information is available for the partial ABGs. Body area information is only available for one other middle Romano-British partial sheep/goat ABG, which consisted of three elements from the right upper hind limb. Another middle Romano-British partial sheep/goat ABG also consisted of limb elements, although it is unknown from which limb/s. However, its lower limb elements were also burnt, yet remained in articulation. This ABG was deposited within a posthole, 404, associated with a timber building.

Shiptonthorpe ABG placement

As discussed above, two of the middle Romano-British sheep/goat ABGs were recovered in association with structures, both skeletons displaying signs of burning. One of these deposits was associated with phase 3 of a timber building (site phase 4). Two other ABGs, a partial juvenile sheep/goat and a partial neonatal calf were also deposited in association with the timber structure. As well as the ABGs, seven infant burials were discovered in association with this structure (Figure 7.6). It is possible some ABGs were also deposited with the human infant burials. Mainland (2006, 276) suggests a partial juvenile sheep/goat ABG was deposited with human infant burial 3.6. However, only fragments of the sheep/goats femur, tibia and mandible are present. In addition, a partial foetal cattle ABG is described as being deposited with human infant burial 3.9, but only four bones are present (a mandible, distal metapodial, distal radius and a vertebra). Under the methodology utilised for this study such deposits are not viewed as ABGs and have therefore not been included in this study. 103

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 7.6 Site plan for phase 4 (middle Romano-British) Shiptonthorpe. AB indicates animal burial (Millett 2006, 61). concludes that the similarity in placement between the human infant and the ABG deposits points to a nonfunctional explanation. She suggests that the partial sheep/goat remains had been processed for meat, and may represent the remains from meals, but their position within the site and their ‘careful burial’ suggest they were of a special nature.

Two middle Romano-British ABGs were also deposited in a large pit, interpreted as a possible water hole (Figure 7.6). These consisted of a cattle skull with associated cervical vertebrae and a horse limb bone. In the late Romano-British period the ‘water hole’ feature becomes the main area associated with ABG deposition. This is also the area where human infant burials were placed, suggesting a possible association. Mainland (2006, 276278) puts forward a number of balanced arguments regarding the nature of the ABG depositions. She 104

Chapter 7. Romano-British Yorkshire 7.4

Rudston Roman villa

being due to either a ‘ritual’ activity, or representative of animals that died from disease/natural death.

In total, 16 ABGs were recorded from Rudston, making it the second largest assemblage from Romano-British Yorkshire. Although recorded as a villa, the nature of the site changes from period to period. The villa was not constructed until the 3rd century AD, but the site had already been occupied with a ‘native’ settlement during the late Iron Age to early Romano-British period (Stead 1980, 35). Most of the ABGs, (12) were recovered from early Romano-British contexts, although it is possible some date to the late Iron Age.

The two pig ABGs were also deposited in the eastern area. One consists of a complete juvenile and was found within a pit associated with roundhouse gully 51. The other, of unknown composition, was recovered from pit A4, which was also associated with a roundhouse gully. As at Shiptonthorpe, the early Romano-British ABGs from Rudston are in close association with occupation structures. The authors suggest that the ABGs may represent the remains of meals, especially the burnt sheep/goat ABGs, or that they were deposited for possible ‘ritual’ reasons (Chaplin and Barnetson 1981). They did not appear to consider that both explanations could be applied.

The majority of the early Romano-British ABGs are from sheep/goat followed by cattle (Table 7.4). All of the sheep/goat ABGs consist of partial remains. Body area information is limited, but is available for some of the deposits.

The late Romano-British ABGs associated with the villa show a very different pattern. Only a small number were recovered, including the unusual deposit of two red deer and a badger from the infill of a well. All three consist of complete skeletons and were found in the top fill of the well. The well was just under 100ft (around 28 metres) deep, similar in depth to the Oakridge well in southern England (see 6.5). However, only a small faunal sample of 1,047 identified fragments (not including ABGs or small mammals) was recovered from Rudston, compared with the 24,426 bone fragments from Oakridge. The only similarity is in the positioning of the wild mammal ABGs, which were found in the upper fills of both wells. The deer and badger ABGs from Rudston are interpreted as being fall victims (Chaplin and Barnetson 1976; 1980).

Five of the partial sheep/goat ABGs consist of mixed remains, meaning elements from the head, axial and appendicular areas are all present (one from pit/ditch 14, two from pit 28, one from pit 48 and one from an unknown feature). Another is comprised of just appendicular elements from a sub-adult sheep/goat, from an unknown feature. Three of the early Romano-British sheep/goat ABGs (mixed ABGs from pits 14 and 47 and the appendicular element ABG) also show signs of being burnt. The left lower fore limb and left tibia of the ABG from pit 47 were burnt, its lower hind limbs were missing and the rest of the elements were not burnt. These and the other sheep/goat ABGs were also deposited in close association with round house structures in the eastern area of the site.

The other late Romano-British ABG consists of the appendicular elements of a sheep/goat of unknown age, recovered from pit 8, under the floor of building 7, a room within the villa complex. The bones had been burnt and were deposited within a charcoal mix. Two infant burials were also uncovered from under the building’s floor. This ABG has been given a tentative late RomanoBritish date. The phasing of the site is unclear and the remains of a late Iron Age – early Romano-British roundhouse gully was also present under the floor. Therefore the ABG may have been associated with the early Romano-British phase of the site, which would make it contemporary with the other burnt sheep/goat ABGs discussed above.

The deposition of ABGs in association with structures was also recorded from the late Iron Age site of Garton and Wetwang Slack (see 5.5.3). It is also a feature of the Shiptonthorpe assemblage, where some of the sheep/goat ABGs are also burnt. This could point to a late Iron Age early Romano-British trend in Yorkshire. Burning was reported on only a few ABGs from southern England none of which were in association with structures. In contrast, the cattle ABGs all consist of complete skeletons deposited in pits within the eastern excavation area. Three are juvenile and one, from pit 19, neonatal. Chaplin and Barnetson (1981) discuss their presence as

Table 7.4 Number of ABGs per species and period from Rudston Roman villa Red Period of deposit Cattle S/G Pig Deer Badger Total Early Romano-British Late Romano-British

4

6 1

105

2

12 2

1

4

Investigating Animal Burials; Ritual, Mundane and Beyond 7.5

Trentholme Drive, York

with human remains are present from this site, compared to others in this study, we must consider that of the 120 inhumations, only six had ABG deposits present. The dominance of domestic fowl does correspond with the southern England data. However, the direct association between human remains and horse ABGs is only recorded from this site.

The ABGs from Trentholme Drive, York (Fraser and Ryder 1968) represent the third largest assemblage from Romano-British Yorkshire. It consists of 14 ABGs of middle to late Romano-British date (in line with the study’s methodology, all the ABGs are assigned to the late Romano-British period). Compared with Shiptonthorpe, the ABGs come from a very different site type and contexts, although the association with human remains is a common feature. The Trentholme Drive site is part of the Mount Cemetery located to the south of York, next to the York-Tadcaster road (Wenham 1968, 5). The excavation uncovered the inhumations of at least 120 people (Wenham 1968, 146). The majority of the ABGs (11) are associated with human remains.

7.6

Bainesse Farm, Catterick

The site of Bainesse Farm is a rural roadside settlement in the hinterland of the Roman town of Cataractanium. It produced one of the largest faunal assemblages (over 25,000 fragments) from the Catterick and its hinterland project (Meddens 1990a; 2002a). The faunal analysis took place before the final phasing of the site, meaning all the material was dated to late 1st to late 3rd century (Stallibrass 2002a, 392). Therefore, for this study the material was assigned to the middle Romano-British. The site produced 14 ABGs and, as discussed above, is the largest middle Romano-British assemblage. Dog ABGs make up over half of the assemblage, with three sheep/goat, one horse, one crow and one pigeon ABG also present (Figure 7.7). These species proportions do not reflect the pattern seen in the non-ABG faunal assemblage, which is dominated by cattle and sheep/goat.

As discussed above, 12 of the 14 ABGs from this site are from domestic fowl. The remaining two consist of partial horse deposits. Of the 12 domestic fowl ABGs, all but three are in association with human burials. Two of these three (animal remains 43 and 44) are complete and were found deposited together in pit 4X. The other consists of the partial remains of what is described as a ‘fighting cock’ from an unknown context. With one exception the rest of the domestic fowl ABGs recovered in association with human remains consisted of complete skeletons. One grave (4 VII), contained two complete domestic fowl ABGs, which had been deposited next to the feet of a 30-35 year old male inhumation. Only two of the human inhumations had other forms of material culture also deposited with them. A broken cooking pot and complete domestic fowl were present in grave 2B XII. The inhumation in grave 5A II/V was accompanied by a complete, fragmented jar, in which the complete domestic fowl ABG was positioned. The faunal report discusses that the bones of ‘game birds’ were also found within five pots associated with human inhumations, but it seems to imply that the domestic fowl from 5A II/V was the only ABG (Fraser and Ryder 1968). Egg shells were also recovered from within four pots showing that not only birds but also their eggs were being deposited with human inhumations.

The only wild bird ABGs from Romano-British Yorkshire are recorded from the site. Both consist of partial remains. Unfortunately, the available data on the ABGs from the AML report (Meddens 1990a) and the published report (Meddens 2002a) are limited. Details of the area of deposition, ageing, body part information and associations with other materials are not given. It is reported that the two ‘bird skeletons’ were recovered from layers and the remaining ABGs were found in either pits or gullies. The site also has the largest collection of dog ABGs from the region, making up 42% of the total number from Romano-British Yorkshire. Of the 358 dog bones recovered from the site, 235 (65%) came from the recorded ABGs. All are partial skeletons, ranging from four to 69 elements. This is the only Yorkshire site with a dog-dominated pattern more commonly observed on the southern England sites.

The two partial horse ABGs were present in grave contexts. Both consist of upper front limb elements and pelvis. The horse ABG recovered from grave 5AVI was also found in association with a skull, possibly from the same animal, although this grave is recorded as being disturbed by robbing.

Little information is provided about the other ABGs from Catterick apart from their interpretation. Meddens (2002a, 425) argued that the presence of young animals and ‘whole skeletons’ both suggest that Bainesse Farm was a producer site. The use of the term ‘whole skeletons’ is confusing, as the three sheep/goat ABGs are made up of 53, 28 and 4 elements each. It may be that a large proportion of the body areas are represented by the two largest sheep/goat ABGs but this cannot be determined from the published report.

There appears to be no correlation between the age of the human inhumations and the ABGs deposited with them. More of the human inhumations are male (3), than female (1), however sex is not given for six of the skeletons. This also is similar to the overall male to female ratio of 4:1 for the cemetery as a whole (Wenham 1968, 147). Although a relatively large number of ABGs associated

106

Chapter 7. Romano-British Yorkshire

Figure 7.7 Percentage of ABGs per species from Bainesse Farm, Catterick 7.7

Overall composition

shows signs of spavin on the right hind leg. This, along with the bit-wear evidence, suggested to Legge (1991a) that both animals were utilised for riding. The horses were deposited in pits next to each other, to the east of a square Iron Age enclosure, and within a Neolithic enclosure, but were radiocarbon dated to the early 2nd century AD. The closest human burial is radiocarbon dated to 1740-1620 cal BC and the horses appear to have been deposited two to three hundred years later than the Iron Age burials, but with no other Romano-British activity close by (Stead 1991, 27). Most of the partial horse ABGs recorded from Romano-British Yorkshire consist of limb elements (Table 7.6).

As discussed for the separate sites above, a large number of the ABGs from Romano-British Yorkshire are complete skeletons. Over half of the cattle ABGs are complete, all neonatal or juvenile individuals. This pattern differs to the one seen from southern England were most of the cattle consist of partial remains. Almost all the domestic fowl ABGs were also complete, although all of these come from Trentholme Drive, York. Although domestic fowl are also common in southern England funerary sites, the majority are partial remains. Most of the sheep/goat ABGs are partial (Table 7.5), which does correspond with the southern England pattern. Body area data are limited for the partial ABGs, but most seem to consist of mixed elements, meaning that the head, axial and appendicular body areas are all represented (Table 7.6). However, the majority of these come from Rudston.

Table 7.5 Numbers of complete or partial ABGs per species from Romano-British Yorkshire Species Complete Partial Unknown Cattle 9 5 S/G 1 19 Pig 3 4 1 Horse 2 8 Dog 3 16 Red Deer 2 Badger 1 Domestic Fowl 10 2 Crow 1 Pigeon 1 Total 31 56 1

Interestingly, most of the dog ABGs are also partial skeletons, with only three complete deposits recorded from Yorkshire in this study. In comparison, around half of the dog ABGs from southern England were found complete (see 6.8). Also, the majority of the partial remains from southern England contain bones from most body areas, indicating they may also have been deposited complete. Unfortunately the body area information for partial dog ABGs from Yorkshire is limited (Table 7.6). Three from Dalton Parlours (Berg 1990b) consist of just axial elements. One, from 58-59, Skeldergate, York, (Berg 1987) has all body areas represented and may have been deposited as a complete carcass.

Table 7.6 Body area information per species, for partial Romano-British ABGs from Yorkshire Species Axis Leg Mixed Cattle 3 1 S/G 3 6 Pig 2 Dog 3 1 Horse 2 4 1 Domestic Fowl 1

At least three of the eight pig ABGs are complete skeletons, one from Rudston and two from Shiptonthorpe. Those from southern England are mostly partial. Most of the horse ABGs from Romano-British Yorkshire are partial. The only two complete horse ABGs are from Kirkburn. Both are full adults, one aged around 13 years, the other around seven, based on the height of maxillary teeth (Legge 1991a). The older horse also shows signs of bit-wear on the teeth. The younger horse 107

Investigating Animal Burials; Ritual, Mundane and Beyond Table 7.7 Butchery marks recorded on Romano-British Yorkshire ABGs Period of deposit Site Name Species Area of butchery Middle Romano- 9 Blake Street, York Dog Lower back limb British Parlington Hollins S/G Lower back limb Shiptonthorpe Horse Upper front limb Sheep Upper front limb Upper back limb Late RomanoBritish Dalton Parlours Horse trunk

Knife

Bainesse Farm. However, each has very different characteristics. Shiptonthorpe and Rudston are rural in nature and the majority of the ABGs are from the earlier Romano-British period. On both sites ABGs appear to be deposited in close association with buildings. Also, both produced burnt sheep/goat ABGs. In comparison Bainesse Farm is on the hinterland of Cataractanium and has a dog-dominated assemblage. Trentholme, York, is a funerary site where all but two of the ABGs are from domestic fowl. These results show how variable ABGs can be between different sites and contexts. The Yorkshire Romano-British assemblage further demonstrates that ABGs are not merely a ‘Wessex’ or prehistoric phenomenon.

As discussed in previous chapters, the most likely human event for creating partial ABGs is butchery. However, butchery marks were not noted on most ABGs; only five ABGs have butchery marks reported (Table 7.7). Most were made using a knife and all but one were recorded on limb elements, occuring either during skinning (for the dog ABG) or dismemberment. One horse ABG from Dalton Parlours did have knife marks present on vertebral elements, indicating possible stripping of the meat (Berg 1990b). It is likely that the other partial ABGs may have been subjected to some form of butchery, which either left no mark, or was not reported. 7.8

Type of butchery Knife Knife Chop Knife Knife

Summary

In comparison with the southern England Romano-British dataset, the one from Yorkshire is very small. However, it is still the largest ABG assemblage from Yorkshire and constitutes 43% of the total number of recorded deposits from the region. Over half of the Romano-British sites from Yorkshire with faunal remains have ABGs present, which is similar to the pattern seen from southern England and suggests that deposition of ABGs in the Romano-British period was relativity common. The majority of the ABG assemblage consists of domestic mammal species, and only five wild species deposits were recorded. Also, domestic fowl provide a significant proportion of the assemblage. In total, sheep/goat, followed by dog, cattle and domestic fowl are the most common species. In this respect the species composition of the Yorkshire Romano-British assemblage differs greatly to the one observed from southern England. The main difference between the two regions is that dog ABGs do not dominate the Yorkshire assemblage. There is a difference in species proportions between rural and urban sites, with sheep/goat ABGs the most common from the former and dog from the latter. This pattern appears to be present in the Yorkshire assemblage throughout the Romano-British period and may indicate a longer continuation of ‘native’ practices in the north of England. As with other samples, a small number of sites produce a large proportion of the assemblage, 70.5% derives from four sites, Shiptonthorpe, Rudston, Trentholme and 108

Chapter 8. Medieval Southern England 8.1

Introduction

The end of the Romano-British period was a time of great social change. By the beginning of the 5th century AD the Roman army was recalled from Britain, in effect leaving Britain seceded from the Roman Empire. However, this should be viewed within the context of the gradual decline of the Empire, with the last quarter of the fourth century marking a recession in activity (Esmonde Cleary 1989, 131). At this time there was an influx of migrations from Germanic Europe, either in large numbers imposing their society upon the ‘native’ Britons (Esmonde Cleary 1989, 204), or in small numbers becoming a new aristocracy (Scull 1993, 70). Alternatively the ‘native’ Britons started to adopt Germanic cultural traits. Hines (1992) argues that there was a relatively intense forging of new identities among mixed groups, with no obvious moving frontier of migration from coastal enclaves. During this time there was also an abandonment of most towns and the breaking up of England and Wales into a number of different kingdoms. This period is often referred to in southern England as either the early Medieval and/or the Anglo-Saxon period and can be divided into the early Anglo-Saxon (AD 450-600), middle Anglo-Saxon (AD 600 to 850) and late AngloSaxon (AD 850 to 1050). For these periods the archaeological sites take a number of forms including farmsteads and villages, high status manors, fortifications, wics and, later, towns, minsters and monasteries. By the middle Anglo-Saxon period we also start to see evidence for the conversion to Christianity in the archaeological and written records. The Anglo-Saxon period ended in 1066 with the Norman invasion, by which time the majority of the separate kingdoms had been amalgamated into England and urban centres had developed. Within this study the postNorman Medieval period is split into the high Medieval (AD 1066-1300) and late Medieval (AD 1300-1550). The faunal data from the Anglo-Saxon period indicate a relative continuity of species use along Romano-British lines, with cattle remaining the dominant species (Bourdillon 1980b; Fowler 2002, 230; Maltby 1981b). Biddick (1984) studied a number of later prehistoric, Romano-British and Anglo-Saxon animal bone assemblages and suggested there was a continuum of pastoral farming from later prehistory until the seventh or eighth century AD. Pigs appear to be utilised to a greater extent in the early and middle parts of the Anglo-Saxon period on some sites (Crabtree 1989a;1989b), and there is an increase in the proportion of sheep/goat towards the end of the period (Coy and Maltby 1987). The majority of the faunal data for this period come from a small number of large assemblages mainly from towns (for the Anglo-Saxon period the term ‘town’ is used in a loose sense to include ‘proto-towns’ and wics) such as Hamwih

and Winchester (Bourdillon 1980b; Coy and Maltby 1987). In all periods there are concentrations of sites from Southampton, Winchester and Dorchester. Interestingly, all the sites around Dorchester have ABGs present . In the late and high Medieval periods, there are also concentrations of sites around the Christchurch, Poole and Salisbury areas. These concentrations are the result of modern day development in the towns producing a large number of excavations, each of which is recorded as a separate site. Compared with the previous two periods covered in this study (Iron Age and Romano-British), the Medieval ABG assemblage from this region is very small with only 182 recorded. Seventy nine of these ABGs are from the early Medieval period, 57 from the high Medieval and 24 from the late Medieval period. The deposits come from 30 separate sites, most dating from the middle Anglo-Saxon to high Medieval periods. 8.2

Species proportions

As with assemblages from previous periods, the majority of the ABGs (72%) are from domestic mammals. However, this is a much smaller proportion compared with earlier periods. This is because domestic bird remains make up 22% of the assemblage. Very few ABGs of wild mammals or birds were recorded and they constitute only 4% and 1% of the assemblage respectively. 8.2.1

Birds

Only two ABGs from wild birds have been recorded (Table 8.1), both from early Anglo-Saxon contexts at Greyhound Yard, Dorchester (Maltby 1993). One consists of five elements from a medium-sized duck from the top infill of well 5145. The size of the bones indicates that they belonged to a species of duck smaller than mallard or domestic varieties. The location of the bones, however, makes it possible they are of a residual late Romano-British date. The other ABG is a partial raven, consisting of seven elements, from the same context and this again may be residual Romano-British in origin. Compared with earlier periods there is a larger proportion and variety of domestic bird ABGs recorded. Domestic fowl are the most common species. Only two domestic fowl ABGs (both partial) were recorded from the AngloSaxon period, both from late Anglo-Saxon contexts at Sussex Street, Winchester (Coy 1984b). However, the numbers of domestic fowl ABGs increase in the high and late Medieval periods (see below).

Investigating Animal Burials; Ritual, Mundane and Beyond

Domestic Mammal

Wild Mammal

Domestic Bird

Wild Bird Total

Table 8.1 Number of ABGs per species from southern England Medieval sites Early Middle Late AngloAngloAngloHigh Late Species Saxon Saxon Saxon Medieval Medieval Cattle 2 17 5 2 S/G 2 4 15 Pig 6 9 4 Horse 2 4 6 2 Dog 4 11 16 8 6 Cat 2 3 4 1 Roe Deer 1 Fox 4 Badger 1 Polecat 1 Domestic Fowl 2 18 8 Domestic Goose 1 2 Peregrine Falcon 1 Sparrowhawk 1 Goshawk 3 Raven 1 Crow/Rook 2 Duck 1 6 20 52 80 24

Total 26 21 19 14 45 10 1 4 1 1 28 3 1 1 3 1 2 1 182

from its associated material culture, or the assumption that birds of prey are not normally consumed. Therefore high proportions of bird of prey remains may be a indication that they were used for hunting (Prummel 1997, 336).

In comparison with domestic fowl the rest of the bird species are represented by only a small number of ABGs. Three complete goose have been recorded, one of which is from the middle Anglo-Saxon period from High Street, Ramsbury (Coy 1977a; 1980a). The other two are from high Medieval contexts at Faccombe Netherton (Sadler 1990). This is the first time domestic goose has been encountered in the ABG record. Domestic goose were possibly present in Britain during the Romano-British period and perhaps even before, but were not commonly utilised (Albarella 2005a). The appearance of domestic goose ABGs corresponds with a general increase in the utilisation of this species from the Anglo-Saxon to the high Medieval period (Grant 1988; Serjeantson 2002).

The authors describing the raptors in this study have all suggested that they are the remains of birds used for hawking, therefore resulting in their categorisation as domestic birds in this study. In addition, the ecology of the birds suggests that they would not have acted as scavengers. Therefore their presence is most likely due to their being used as tamed hunting birds (Mulkeen and O'Connor 1997).

The other domestic bird ABGs are all from raptors, all from Faccombe Netherton apart from one goshawk from Portchester Castle, which was possibly associated with a horse ABG (Eastham 1977). All the raptor ABGs consist of complete skeletons. The peregrine falcon and a goshawk from Faccombe Netherton were males. The sparrowhawk and the other goshawk were both female. The female goshawk also had pathology present; a false joint between the coracoid and the proximal humerus had formed due to a fracture of the coracoid. In addition, exostoses were present on the distal end of both the right and left tarsometatarsi, which Sadler (1990, 506) suggests could have been caused by the jesses, if the individual was a tamed hunting bird.

8.2.2

Wild mammals

The majority of the wild mammal species ABGs are recorded from the high Medieval period (Table 8.1). The fox, badger and polecat ABGs all come from Faccombe Manor (Sadler 1990). The four fox ABGs all consist of partial remains, with three from pit contexts and one from a layer. Unfortunately, little information has been published regarding them, but they are all interpreted as waste from skinning. The badger also consists of a partial ABG, found within a ditch fill. Again little zoological or contextual information has been given, but the animal’s death was interpreted to be the result of a dog attack. The polecat ABG was recorded from a garderobe pit, and consists of a complete female skeleton. Sadler (1990) suggested that it may have actually been a domestic pet (ferret) and not wild, and possibly used for hunting rabbits, but this conclusion seems to be based solely on

Hawking was introduced to Europe in the third to fourth centuries AD, becoming popular in the high Medieval period (Cherryson 2002; Epstein 1943). Archaeologically, the evidence of hawking comes either

110

Chapter 8. Medieval Southern England faunal assemblages, in which cattle and sheep/goat tend to be the most common species.

the completeness of the ABG. The only other wild mammal ABG consists of a partial roe deer, consisting of the upper forelimb elements from a late Medieval demolition layer at Sussex Street, Winchester. This was interpreted as butchery waste (Coy 1984b).

Sheep/goat are the second most common species among ABGs from the high Medieval period. However, most (87%) of the 13 sheep/goat ABGs are from Easton Lane (Maltby 1989b). These are all from one feature (pit 5265), which contained eight largely complete adult ewes, the partial skeleton of a immature male and four foetal/newborn partial lambs. Maltby (1989b, 128) suggested this was probably a deposit of animals that died of natural causes either through disease or during severe weather conditions in winter or early spring. The feature was not closely dated and it is possible the ABGs are from as early as the late Anglo-Saxon period,.

The small proportion of wild species ABGs corresponds with the majority of faunal assemblages from the period, which have also produced small percentages of wild faunal remains (Coy and Maltby 1987). There are, however, a small number of exceptions, for example, red deer are the most common species from high Medieval contexts at Faccombe Netherton but none were found as ABGs (Sadler 1990). 8.2.3

Domestic mammals

Dog ABGs are still represented in the high Medieval period, but in much smaller numbers. There is also a higher proportion of pig present. However, six pig ABGs are from a single deposit within a cess pit at Faccombe Netherton, all consisting of near complete skeletons from ‘very immature’ individuals (Sadler 1990, 481). The proportion of dog ABGs increases in the late Medieval period, with four of the six dog deposits recorded from New Road, Winchester (Coy 1984b). However, compared with the previous Medieval periods, the sample size for the late Medieval period is small, possibly indicating that the deposition of ABGs was becoming uncommon.

The majority of the ABGs recorded from Anglo-Saxon contexts (early Medieval) are from domestic mammals. As in the Romano-British period, dogs are the most common species (Figure 8.1). Investigation of the early Medieval assemblage reveals that dog is the most common species among ABGs in both the early and middle Anglo-Saxon samples. In the late Anglo-Saxon period cattle, followed very closely by dog, is the most common species (Table 8.1). There is therefore a marked increase in the number of cattle ABGs in the late AngloSaxon period. However, the majority of these are from only two sites; eight are from Poundbury (BucklandWright 1987), which has a long tradition of ABG deposits, and another eight are from Sussex Street, Winchester (Coy 1984b). The Poundbury cattle ABGs may be earlier than the late Anglo-Saxon period, as chronological information for them is very limited and they were recorded as dating from the early to late AngloSaxon period. Therefore, within the conventions of this study they have been assigned to the later period.

8.3

Nature of the assemblage

The majority of the Medieval sites have very small ABG assemblages. Of the 29 sites, 13 (44.8%) have only one ABG present. The two sites with the largest assemblages are Faccombe Netherton, which has 48 (Sadler 1990), and Sussex Street, Winchester with 33 (Coy 1984b). The deposits from these two sites account for 44.5% of the Medieval ABG assemblage (Table 8.2). The majority of the Sussex Street remains (23) date to the late AngloSaxon period and most of the Faccombe Netherton ones (35) to the high Medieval period.

The high percentage of cattle ABGs would correspond with the non-ABG faunal assemblage. However, dog remains make up only 1-2% of most faunal assemblages, and we would also expect there to be a higher number of sheep/goat and pig ABGs. Therefore the early Medieval ABG species proportions do not reflect those seen in complete faunal assemblages.

The remains from Sussex Street account for 29% of the Anglo-Saxon assemblage. The next largest assemblage consists of 10 ABGs from Clifford Street, Southampton (Bourdillon 1990b). The rest of the Anglo-Saxon data come from 15 other sites with assemblages of between one and nine ABGs.

In the high Medieval period the overall proportion of cattle drops to only 6% of the ABG assemblage. Domestic fowl are the most common species in the high and late Medieval periods (Figure 8.1). However, the majority of the bird ABGs come from Faccombe Netherton (Sadler 1990). Of the 17 high Medieval domestic fowl ABGs, 14 are from Faccombe Netherton. Seven of the eight late Medieval domestic fowl deposits also come from this site. The other high and late Medieval domestic fowl ABGs are from the Western Suburbs excavations at Winchester.

The assemblage from Faccombe Netherton constitutes a large proportion of the high and late Medieval assemblage. Of the 79 high Medieval ABGs, 35 (44%) come from this site. The second largest assemblage consists of 21 ABGs from Easton Lane (Maltby 1989b), meaning that combined these sites account for 70% of the high Medieval dataset. However, as noted above, some of the Easton Lane deposits may be late Anglo-Saxon in date. The rest of the high Medieval data comes from eight other sites with assemblages of between one and six ABGs.

Therefore, as in the early Medieval sample, the ABG species proportions are very different to the non-ABG 111

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 8.1 Percentages of the most common animals to constitute Medieval ABGs, southern England. Sample size in brackets high and late Medieval date have ABGs present. However, the largest ABG assemblage in this period comes from a manorial site. Again, we must be careful in drawing too many conclusions, as overall only seven manorial sites with ABGs have been recorded.

The largest late Medieval assemblage also comes from Faccombe Netherton, but consists of only seven ABGs. However due to the overall small size of the late Medieval sample, these seven still account for 29% of the dataset. The second largest assemblage (six ABGs) is from New Road, Winchester (Coy 1984b). The rest of the dataset comes from five other sites with between one and four ABGs present. Therefore when examining the Medieval dataset we must again be conscious that any patterns present may be strongly affected by the assemblages from a small number of individual sites.

8.4

Site type

As discussed above, the ABGs come from a number of different site types and as with previous periods there is variation in the species relative abundance. The majority of the Anglo-Saxon ABGs are either from rural or urban settlements. The early Anglo-Saxon assemblage is very small, with only one dog ABG recorded from a rural settlement, Grove Farm (Bourdillon 2006), and five ABGs (three dog, one duck and one raven) from urban contexts at Greyhound Yard, Dorchester, all of which were probably residual from Romano-British deposits (Maltby 1993).

The ABG data also come from a number of different site types. Overall, town and rural settlements have been the most common types recorded with ABGs. The majority of the Anglo-Saxon data are from rural settlements whereas the high and late Medieval data mainly come from towns. This is a reflection of the nature of the archaeology recorded, with biases towards areas of most frequent excavation within southern England. As mentioned above, in the early Anglo-Saxon period there was a general abandonment of urban centres. The large proportion of ABGs from urban sites in the high and late Medieval period is due to the large number of urban sites excavated, especially in Southampton and Winchester.

The middle Anglo-Saxon assemblage is larger and over half of the ABGs are from urban contexts. A large proportion of these are from dogs (Table 8.4). However, all the dog ABGs are from the upper fill of pit 56 at Clifford Street, Southampton (Bourdillon 1990b). This deposit consists of eight complete neonates, and a complete adult female and adult male. The neonates are interpreted as possibly being the female’s puppies. In comparison, sheep/goat and cat are the most common species from rural settlements. However, only seven ABGs have been recorded from middle Anglo-Saxon rural settlements, so the sample size is too small for any firm conclusions.

In the Anglo-Saxon period six assemblages from rural settlements and five urban centres have ABGs present. However, compared with the overall counts of sites with faunal material present, a higher proportion, (71%), of urban sites had ABGs present compared to rural sites (35%) (Table 8.3). Such a striking difference does not occur in the high and late Medieval periods, with 21% of rural and 16% of urban sites having ABGs present. ABGs appear to become less common on both site types in the later parts of the Medieval period. But with such a small sample size, extreme variability in percentage calculations may be expected. Manorial sites show a different trend with ABGs present on all such Anglo-Saxon sites. Fewer manorial sites of

Cattle are the most common species from rural and urban settlements in the late Anglo-Saxon sample (Table 8.4). All the cattle ABGs from rural settlements are from Poundbury (Buckland-Wright 1987) and the majority (8) of the urban deposits are from Sussex Street, Winchester (Coy 1984b). However, unlike the dog ABGs discussed 112

Chapter 8. Medieval Southern England Table 8.2 Number of Medieval sites from southern England with ABGs. Percentage total sites indicates the proportion of sites with that number of ABGs present. Percentage No. ABGs indicates what proportion of the total ABG assemblage come from the sites Site Type 1 2 to 4 5 to 9 10 to 29 30 to 49 Town 5 3 3 1 1 Manorial 2 1 1 Military 1 Rural Settlement 6 3 1 1 Total 13 7 5 2 2 Percentage total sites

44.8%

24.1%

17.2%

6.9%

6.9%

Percentage No. ABGs

7.1%

11.5%

19.7%

17.0%

44.5%

Table 8.3 Total numbers of Medieval site types from southern England recorded by period. The number in brackets indicates the number of sites with ABGs present Anglo-Saxon – High/late Period Anglo-Saxon later Medieval Medieval Total Funerary 2 2 Industrial 2 2 Manorial 1 (1) 1 (1) 5 (2) 7 Military 1 (1) 4 5 Monastic 1 1 Rural Settlement 17 (6) 1 (1) 14 (3) 32 Town 7 (5) 3 (2) 42 (7) 51 Table 8.4 Number of ABGs per species and site type for the Anglo-Saxon period, southern England Domestic Period Site type Cattle S/G Pig Horse Dog Cat Fowl Early Anglo- Rural Settlement (1) 1 (100%) Saxon Town (5) 3 (60%) Middle Anglo- Rural Settlement (7) 1 (14%) 2 (29%) 1 (14%) 2 (29%) Saxon Town (13) 1 (7%) 2 (14%) 10 (71%) Late Anglo- Rural Settlement (10) 8 (80%) 1 (10%) 1 (10%) Saxon Manorial (7) 7 (100%) Town (35) 9 (25%) 4 (11%) 6 (17%) 4 (11%) 8 (22%) 2 (6%) 2 (6%) The majority of the high Medieval ABGs are from manorial sites, in particular Faccombe Netherton, with domestic fowl the most commonly represented species, followed by pig (Table 8.5). Domestic fowl along with cattle are the most common species from high Medieval urban contexts, but this is from an assemblage of only eight ABGs. In comparison, domestic fowl are rare on rural settlements, where over half of the ABGs are from sheep/goat. This may be a reflection of the species utilised on these site types. The remains of cattle are generally more commonly encountered in non-ABG urban assemblages, whereas sheep/goat are generally more common on rural settlements. Also, domestic fowl and pig are better represented on ‘high status’ manorial sites compared with urban and rural sites (Albarella 2005b; Grant 1988; Thomas 2005). The Portchester

above, none of the cattle ABGs were recovered in association with each other. Overall, the Anglo-Saxon urban and rural ABG assemblages have different proportions of species . Cattle are the most common species from rural sites, albeit largely due to assemblages from two sites, whereas dog ABGs dominate the urban sample (Figure 8.2). The urban pattern is similar to the one observed from RomanoBritish towns. However, the middle Anglo-Saxon multiple dog deposit from Clifford Street, Southampton, accounts for a large proportion of the assemblage. Dog ABGs are not as common from rural settlements. Also, there is a greater amount of species variability in the urban ABG assemblage compared with the rural sample. However, this may be due to the larger sample size.

113

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 8.2 Species percentages of ABGs from rural and urban Anglo-Saxon sites, southern England. Sample size in brackets Table 8.5 Number of ABGs per species and site type for the high and late Medieval periods, southern England Domestic Period Site type Cattle S/G Pig Horse Dog Cat Fowl High Castle (6) 1 (17%) 4 (67%) Medieval Rural Settlement (25) 1 (4%) 14 (56%) 3 (12%) 3 (12%) 1 (4%) 1 (4%) Manorial (41) 1 (2%) 6 (15%) 2 (5%) 2 (5%) 4 (10%) 14 (34%) Town (8) 3 (38%) 1 (13%) 1 (13%) 3 (38%) Late Rural Settlement (3) 2 (67%) 1 (33%) Medieval Manorial (7) 7 (100%) Town (14) 2 (14%) 2 (14%) 1 (7%) 6 (43%) 1 (7%) 1 (7%) Castle (Grant 1977; 1985) excavations also produced a small sample of ABGs with dog the most common species, which is a continuation of the Romano-British pattern observed from the site.

feet are gnawed, yet were still found in articulation, indicating this ABG must have been exposed for a short period of time before burial.

Overall, domestic fowl is the most common ABG species recorded from the late Medieval period (Figure 8.1), but this is due mainly to the large number of domestic fowl deposits from Faccombe Netherton. This is why 100% of the ABGs from late Medieval manorial sites are domestic fowl (Table 8.5). Domestic fowl ABGs are not recorded from any lower status rural settlements, although the sample size is very small. However, only one domestic fowl ABG is recorded from an urban context, at Crowder Terrace, Winchester (Coy 1984b). The majority of the ABGs from late Medieval urban contexts are from dogs, which come from two sites, New Road, Winchester (Coy 1984b) and the Staggs site, Christchurch (Coy 1978a). All these dog ABGs were recovered as individual deposits, rather than associated groups.

8.5

Feature type

As in the previous periods, the majority of Medieval ABGs have been recovered from pits (Table 8.6). With ditch fills and layers the second and third most commonly recorded features. There appears to be little species variation in different types of features. The majority of all species were deposited within pits. This evidence is in contrast to the arguments put forward by Hamerow (2006), who has suggested that there is a trend for the placement of ‘termination deposits’, including ABGs, in Anglo-Saxon ditches and particularly in sunken feature buildings. Only one ABG has been recorded in this study associated with a sunken feature building, a partial cattle ABG from Cowdery's Down (Hamerow 2006; Maltby 1982c; 1983a).

All the late Medieval rural settlement ABGs are from one site, West Mead, Dorset (Hamilton-Dyer 1999b). Both pig ABGs consist of partial skeletons. In comparison the horse ABG is a complete skeleton of a male, around 14 years old, with a mean height of 1.3 metres. Vertebrae and lower foot pathology suggests it was a working animal (Hamilton-Dyer 1999b). Interestingly the lower

Of the 78 Anglo-Saxon ABGs recorded in this study, 53 (70%) are from pits, whereas eight (10%) are recorded from ditches. There are a number of reasons for the discrepancy between this study and that of Hamerow (2006). The majority of Anglo-Saxon ABGs recorded in this study are from urban contexts, mainly Southampton

114

Chapter 8. Medieval Southern England Four complete horse ABGs are present from three different sites. The horse ABGs from St Georges Road (Bullock and Allen 1997) and West Mead (HamiltonDyer 1999b) are both from old adults. The complete horse ABG from Faccombe Netherton (Sadler 1990) is from a foetal individual, the long bone length suggesting the pregnancy had lasted six to seven months. This is the only young horse ABG recorded from the Medieval period, all the partial horse remains being either from adults or old adults. Unfortunately ageing data are not provided for the complete horse ABG from Portchester Castle (Grant 1976b).

(Hamwic) and Winchester. Pits were the most common feature type excavated from these sites. Hamerow’s (2006) study instead selected a small number of early Anglo-Saxon settlements from across Britain, none of which were urban in nature. Also, non-ABGs such as a number of single skull deposits and ‘piles of disarticulated bones’, were included in Hamerow’s study. Indeed, the majority of deposits that could be described as ABGs are in fact from pits. We can therefore question Hamerow’s (2006, 28) suggestion that ‘termination deposits’ were widespread in the Anglo-Saxon period. They may be present on some sites outside of this project’s study area, pointing to possible regional variation, but this is not confirmed by the Yorkshire results. 8.6

Cattle ABGs show a different pattern to the other domestic mammals, as only one survives as a complete skeleton. This ABG is from the high Medieval pit, 537, Easton Lane (Maltby 1989b) and consists of 151 bones from an adult steer.

ABG Composition

All the complete domestic fowl and pig ABGs are from one manorial site, Faccombe Netherton. Therefore it would be unwise to consider this to be indicative of differences between site types. All the ABGs from Portchester Castle are also complete, but this is likely to be due to the recording and reporting strategies used for the site. As the majority of the Portchester ABG information was obtained from the feature descriptions rather than the faunal report, it is likely that partial ABGs may have been missed or not deemed worthy of reporting.

A number of differences in ABG composition are evident between species. The majority of domestic bird ABGs consist of complete skeletons, the only exception being domestic fowl. All the domestic fowl from Faccombe Netherton consist of complete skeletons, whereas all the domestic fowl from other sites are partial (Sussex Street, Crowder Terrace, Winchester and Easton Lane). All the raptor ABGs also consist of complete skeletons. Their completeness is one of the reasons Sadler (1990) suggests they were tame hunting birds. In contrast, all the wild bird ABGs consist of partial remains. The majority of wild mammal ABGs are also partial, or the data are not available. The one polecat ABG is the exception, if indeed this animal was wild.

For some of the partial ABGs, body-area data were available. The cattle and sheep/goat partial ABGs have a similar pattern, with the ABGs composed of either axial or appendicular elements (Figure 8.3). A small number of partial-cattle ABGs contain both axial and appendicular elements. Also, more partial-sheep/goat ABGs consist of just appendicular elements. However, few sheep/goat ABGs have data available.

Composition data are available for all domestic mammal ABGs. As in previous periods, a high proportion of the dog deposits consist of complete skeletons. Ten (33%) of the complete skeletons are from Clifford Street, Southampton, pit F56 (Bourdillon 1990b). The rest of the deposits consist of isolated complete ABGs. In contrast, the majority of cat ABGs are partial. However, all five of the partial cat ABGs with data available are from neonatal individuals, whereas the two complete cat ABGs are from young adults. It has been suggested that the neonatal individuals were deposited as complete skeletons, but became partial due to the fragile nature of the neonatal bones, post-depositional movement and possible recovery biases (Coy 1986).

The majority of partial-pig ABGs consist of elements from just the axial skeleton (Figure 8.3). One pig ABG from West Mead (Hamilton-Dyer 1999b) is comprised of appendicular elements, and another from Manor Farm (Sykes 2003), consists of both axial and appendicular elements. Again, like the sheep/goat, only a limited number of pig ABGs have body area data available. The partial-horse ABG data show a different pattern. Partial ABGs consisting of just the appendicular elements are the most common, but unlike the other species a number of partial ABGs have both appendicular and axial elements or head and axial elements present (Figure 8.3). Partial horse ABGs appear to have been deposited with a greater degree of different body area articulations, possibly because they have not been as intensively butchered, or have not been left exposed to predepositional taphonomic activity for as long as other species.

Sheep/goat is the second most common domestic mammal to consist of complete skeletons (Table 8.7). However, all the complete sheep/goat ABGs are recorded from the Easton Lane high Medieval pit, 5265 (Maltby 1989b) (see above). The high percentage of complete pig ABGs is also due to the results from one deposit. All the complete pig ABGs are from one cess pit (context unknown) from Faccombe Netherton. The deposit consisted of six individuals all between four and ten weeks old, interpreted as the victims of disease, possibly ‘murraine’ (Sadler 1990, 481). The pig ABGs from all other sites are partial. 115

Investigating Animal Burials; Ritual, Mundane and Beyond Table 8.6 Number of ABGs per species and feature type for the Medieval period, southern England PostSpecies Ditch Layer Midden Pit hole Well Unknown Domestic Cattle 3 2 19 1 1 Mammal S/G 2 19 Pig 4 1 2 12 Horse 2 2 7 2 1 Dog 7 2 32 1 3 Cat 2 8 Wild Roe deer 1 Mammal Fox 1 3 Badger 1 Polecat 1 Domestic Domestic Fowl 28 Bird Domestic goose 3 Peregrine Falcon 1 Sparrow Hawk 1 Goshawk 3 Wild Raven 1 Bird Crow/Rook 2 Duck 1 Total 17 13 2 139 1 5 5 Percentage of total 9.3% 7.1% 1.1% 76.3% 0.5% 2.7% 2.7% Table 8.7 Composition of Medieval ABGs, southern England. Sample size in brackets Species Complete Partial Unknown Domestic Cattle (26) 4% 96% Mammal S/G (21) 43% 57% Pig (19) 32% 68% Horse (14) 29% 71% Dog (45) 67% 33% Cat (10) 20% 80% Wild Roe Deer (1) 100% Mammal Fox (4) 25% 75% Badger (1) 100% Polecat (1) 100% Domestic Domestic Fowl (28) 75% 25% Bird Domestic Goose (3) 100% Peregrine Falcon (1) 100% Sparrowhawk (1) 100% Goshawk (3) 100% Wild Raven (1) 100% Bird Rook/Crow (2) 100% Duck (1) 100% depositions. However, we must be cautious in drawing such conclusions, as only a limited number of deposits have body area data available.

Over half of the partial-dog ABGs are comprised of just appendicular elements (Figure 8.3). The remainder consist of either axial, head and axial, or axial and appendicular elements. In contrast to previous periods, none of the partial-dog ABGs have portions of all body areas present. The lack of complete dog ABGs could suggest the carcasses were disarticulated before deposition. This would indicate that the dog remains were subject to some form of pre-depositional taphonomic activity such as butchery, or that they represent secondary

8.7

Butchery

Butchery marks were only noted on one partial dog ABG from Emwell Street, Warminster (Table 8.8). They were 116

Chapter 8. Medieval Southern England Evidence of disarticulation was also present on a partial sheep/goat at Sussex Street, Winchester, with chop marks on the cervical vertebrae, (Coy 1984b). These marks would have been created during the dismemberment of the neck from the head.

present on elements from the lower front limb and were interpreted as skinning marks. Although no other dog ABGs have had butchery marks recorded, this does not mean that others were not subjected to processing, as butchery can be carried out on a carcass and leave no trace.

Although limited, the butchery mark evidence does indicate partial ABGs were often created as a result of butchery and that meat appears to have been removed from some prior to deposition. Although butchery marks are only present on a small number of ABGs, it is likely that a larger proportion of the ABGs were created by butchery, but no marks were left or were not recorded.

The majority of the butchery marks recorded are from partial cattle ABGs (Table 8.8). Both knives and cleavers appear to have been used. Most of the butchery marks were observed on the axial skeleton, although one deposit from Cowdery’s Down had butchery marks present on the skull, associated with skinning, and on the upper front limbs, due to disarticulation (Maltby 1983a). Butchery marks were also present on a partial cattle ABG from Cook Street, Southampton, indicating dismemberment of the ribs from the vertebrae (Bourdillon 1993b). Knife marks associated with filleting of the meat from the vertebrae are present on a partial-cow from Poundbury (Buckland-Wright 1987). This information indicates that meat has been removed from some of the cattle ABGs before their deposition.

8.8

Associated ABGs

A number of ABGs were deposited in association with each other. In total, 47, representing 26% of the total Medieval assemblage, were recovered from nine multiple ABG deposits. Only one early Anglo-Saxon multiple deposit is present, recorded from Greyhound Yard, Dorchester (Maltby 1993). It consists of two partial dog ABGs, discovered in close association within the demolition layer of a Romano-British building and may be residual.

Filleting marks are also present on the vertebrae of a middle-Anglo-Saxon horse ABG from Cook Street, Southampton (Bourdillon 1993b), again indicating that meat was removed before deposition. Such butchery was regarded as evidence of knackering, processing the meat for consumption by dogs. The eating of horses may have been a taboo at the time, as well as being banned by Pope Gregory III in AD 732 (Grant 1988). Chop marks were also present on the ischium and left femur of a complete horse ABG from St George’s Road, Dorchester. The marks were made during the disarticulation of the back left limb, but there is no evidence of further processing. The left back leg was deposited with the rest of the skeleton, next to the head. The disarticulation was argued to have occurred so that the horse would fit into the pit (Bullock and Allen 1997).

All the other multiple deposits are from pits. As discussed above a middle Anglo-Saxon deposit from Clifford Street, Southampton (Bourdillon 1990b), consisted of ten dog ABGs. At Faccombe Netherton a late Anglo-Saxon deposit consisting of six complete neonatal puppies was recovered (Sadler 1990). Another late Anglo-Saxon multiple ABG deposit was found at Poundbury (Buckland-Wright 1987). It consists of two partial cattle ABGs, both made up of vertebrae, one from an adult and the other from a young adult.

Figure 8.3 Body area percentages for Medieval partial ABGs, southern England. Sample size in brackets

117

Investigating Animal Burials; Ritual, Mundane and Beyond

Period Middle Anglo-Saxon

Late Anglo-Saxon

High Medieval

Table 8.8 Butchery information from Medieval ABGs, southern England Complete/ Area of Type of Authors Site Name Species Partial butchery butchery Interpretation Cook Street, Southampton Cattle Partial Axial Chop Dismemberment Upper Cowdery's Down Cattle Partial front limb Knife Disarticulation Cook Street, Southampton Poundbury Sussex Street, Winchester Emwell Street, Warminster St Georges Road, Dorchester Sussex Street, Winchester

Head

Knife

Skinning

Horse Cattle

Partial Partial

Axial Axial

Knife Knife

Filleting Filleting

Cattle

Partial

Unknown

Unknown

Unknown

Cattle

Partial

Unknown

Unknown

Dog

Partial

Unknown

Skinning

Horse

Complete

Axial Lower front limb Upper back limb

Chop

Disarticulation

S/G

Partial

Axial

Chop

Dismemberment

Table 8.9 ABGs per species deposited in multiple ABG deposits from Medieval sites, southern England. The number in brackets indicates the number of multiple ABG deposits with an ABG of the species present Early Middle AngloLate AngloHigh Species Anglo-Saxon Saxon Saxon Medieval Cattle (1) 2 S/G (1) 9 Pig (1) 6 Dog (1) 2 (1) 10 (1) 6 Cat (1) 4 Domestic Goose (1) 2 Goshawk (1) 1 Peregrine Falcon (1) 1 Sparrowhawk (1) 1 The majority of the multiple ABG deposits date to the high Medieval period (Table 8.9). Three of these are from Faccombe Netherton (Sadler 1990). One, discussed above, is composed of six complete neonatal pig ABGs. A second contained two complete domestic geese. The third is the only deposit to contain more than one species. It produced a goshawk, a peregrine falcon and a sparrowhawk and was interpreted as the burial of hunting birds which had died of natural causes. The other high Medieval multiple deposits include four neonatal partial cat ABGs from within the fill of a cess pit at Osborne House, Romsey (Coy 1986), and the nine sheep/goat, discussed above, from Easton Lane (Maltby 1989b).

majority of the multiple deposits contain only one species and consist of complete skeletons. This may add credence to the suggestions of other authors that multiple ABGs represent culling of a population (in the cases of dogs and cat), or the death of a section of the population due to disease (sheep/goat and pigs).

Overall, dogs appear more likely to have been deposited as multiple ABGs in the Anglo-Saxon period. There are no dog multiple ABG deposits in the high Medieval period, and no late Medieval multiple ABG deposits at all. Sheep/goat and pig are the most common species recovered from multiple deposits in the high Medieval period. However, this is due to the presence of a single large multiple ABG deposit for each species. The

In comparison with the Iron Age and Romano-British periods, only a small number of ABGs have been recorded from the Medieval period in southern England. It would appear that from the Anglo-Saxon period onwards ABGs become rarer, with the majority of the sites producing negative results especially in the high and later Medieval periods.

No ABGs were recorded in association with ‘special deposits’ of other material types. However, this may be a reflection that such terminology and definitions are uncommon in Medieval archaeology. 8.9

118

Summary

Chapter 8. Medieval Southern England As in other periods, the majority of ABGs are from domestic mammals, although the later Medieval assemblages have produced higher percentages of domestic bird ABGs than any other species. There is diachronic variation in the proportions of species which have been recovered. Dog ABGs are still the most common in the Anglo-Saxon period, which may be a continuation of the previous Romano-British pattern. However, there is an increase in the proportion of cattle. The cattle and dog ABGs are recovered from different site types, with dogs common on urban sites, often as multiple deposits, and cattle on rural sites. This pattern changes in the high and late Medieval periods when domestic fowl are the most common species. Following the trend seen in the Romano-British sample, the species composition of the ABG and non-ABG faunal assemblages differs greatly. As in previous periods, a small number of sites produced a large number of ABGs. In particular, the high and late Medieval data are dominated by the assemblage from Faccombe Netherton. The majority of the domestic bird ABGs were recovered from this site. Compared to previous periods, more of the ABGs recorded from the Medieval period consist of complete skeletons. This may be a reflection of archaeologist’s preconceptions regarding recording faunal data from Medieval deposits. Only very recently have archaeologists started to consider the presence of such deposits in these historical periods. Partial ABGs may not have been deemed worthy of note in the majority of faunal and excavation reports. However, this study has shown that ABGs are present in these periods, possibly in greater numbers and complexity than would have been expected.

119

Chapter 9. Medieval Yorkshire 9.1

Introduction

As with the previous periods examined in this study, the Yorkshire data come from a different archaeological background to that from southern England, where the emergence of Anglo-Saxon culture can be seen in the early Medieval period. However, in parts of Yorkshire as well as northern and western England, societies emerged with characteristics rooted in the ‘indigenous’ RomanoBritish past. During the 5th century Anglo-Saxon societies developed in the Humber region, with parts of the Pennines not incorporated within the Anglo-Saxon kingdom of Northumbria until the mid 7th century (Loveluck 2003). However, similar settlement trends to those in southern England, including a move away from urban centres in the early parts of the pre-Norman Medieval period, can be seen in the archaeological record (Hall 1996, 34; Loveluck 1999; 2003). Excavations in York appear to indicate that the site became reoccupied as a proto-urban centre by around AD 800, with a possible wic established next to the river Foss (Carver 1995; Hall 2003). In AD 866 Viking invaders captured York and seized control of the surrounding area. At present, the area of Viking control is undefined, but finds from a number of sites in the Wolds of Eastern Yorkshire, including Wharram Percy and Cottam indicate a Scandinavian influence. Viking control lasted until AD 954 when Yorkshire became part of the newly created English Kingdom, although the region maintained a Scandinavian influence (Hall 2003). As in southern England, AD 1066 marks (for this project) the beginning of the high and late Medieval periods, which lasted until AD 1550.

later Medieval periods. Of the 39 with faunal assemblages dating to only the high and late Medieval periods, 14 (36%) have ABGs recorded as present. With the large number of post-Norman Medieval sites present in the dataset, it is unsurprising that the highest number of ABGs come from the high and late Medieval periods. In total, 72 ABGs were recorded from the 20 sites, of which 14 date to the Anglo-Scandinavian period, and 58 from the later Medieval periods. Compared with the southern England results this is a small sample. Nonetheless, this represents the second largest Yorkshire ABG assemblage behind the Romano-British one. In fact, the 58 later Medieval ABGs account for 29.1% of the Yorkshire assemblage, a higher percentage than the assemblages from prehistoric periods such as the Iron Age. As with the assemblages from previous periods, there are concentrations of sites in certain areas of Yorkshire. A large proportion of the sites are grouped around York due to the substantial excavations that have taken place within the city. Also, a large proportion of the sites with no ABGs present are grouped around the Wolds and Hull. The specific location of the sites appears to be influenced primarily by preservational factors and areas of modern urban development. A similar trend is seen in the southern England data, where the majority of the Medieval faunal assemblages are from towns, which have received a large amount of archaeological work. 9.2

Species proportions and site type

Compared to southern England the faunal data are limited. The majority of the information from Yorkshire is drawn from the large urban excavations in York. During the Anglo-Scandinavian period cattle appear to have been the most common domestic animal, with the proportion higher than seen on southern English AngloSaxon settlements, possibly pointing to a distinctively Scandinavian pattern. However, O’Connor (1989, 199201) suggests that the local environs are more conducive to cattle husbandry compared to sheep/goat. The Medieval faunal remains appear to follow the overall trends discussed for southern England.

Compared with previous assemblages investigated in this study, the ABGs from Medieval Yorkshire have the lowest percentage of domestic mammals (65%). A substantial proportion of the assemblage consists of domestic fowl (23%). Only one wild mammal is recorded, a complete rabbit from an unknown late Medieval context at Fishergate, York (Bond and O'Connor 1999). The rest of the wild species are birds; a complete adult jackdaw and a complete chaffinch were recovered from late Anglo-Scandinavian pits at Coppergate, York (Bond and O'Connor 1999). Also four, possibly complete and associated, duck ABGs were recorded from a late Medieval garderobe pit at Scale Lane/Lowergate, Hull (Phillips 1980).

In total, data were recorded from 49 Medieval sites, the majority (39), dating to the high or late Medieval periods, with only seven sites dating to the Anglo-Scandinavian period. Three contained assemblages covering multiple sub-periods. On 20 (40%) of the sites ABGs were recorded as present, a higher proportion compared to the Medieval sample from southern England. Four of the seven sites dating only to the Anglo-Scandinavian period had ABGs present. In addition, Coppergate, York (Bond and O'Connor 1999) and Addingham (Keith 1997) had assemblages dating to both the Anglo-Scandinavian and

Overall cat, dog and horse are the most common domestic mammals, respectively (Table 9.1). This differs to the species proportions in the Yorkshire RomanoBritish sample (see 7.2) and also from the southern England Medieval assemblage (see 8.2). However, the majority of the Yorkshire Medieval assemblage dates from the late Medieval period, whereas the southern England assemblage dates predominantly to the high Medieval period.

Chapter 9. Medieval Yorkshire Table 9.1 Number of ABGs from Medieval sub-periods, Yorkshire. ULM=unidentified large mammal Middle Early AngloAngloLate AngloHigh Late Species Scandinavian Scandinavian Scandinavian Medieval Medieval Total Cattle 1 6 7 S/G 1 2 3 Pig 1 1 2 Horse 1 1 6 8 Dog 1 1 8 10 Cat 3 14 17 Rabbit 1 1 ULM 1 1 Domestic Fowl 2 15 17 Duck 4 4 Jackdaw 1 1 Chaffinch 1 1 Total 3 3 8 2 56 72 Table 9.2 Number of Medieval site types with ABGs from Yorkshire. Numbers in brackets indicates the number of sites with ABGs. Other number indicate the number of sites without ABGs AngloSite Type Scandinavian High Medieval Late Medieval Cemetery (2) Town (1) (1) 11 (10) Manorial/Castle 2 4 (2) Monastic 1 (1) 5 (2) Rural Settlement 3 (3) 3 This is the only ABG recorded from early Medieval Yorkshire in association with a building. As discussed previously, Hamerow (2006) has suggested that ABGs may represent ‘termination deposits’ associated with buildings, but there appears to be little evidence of this from Yorkshire.

The majority of sites with no ABGs present are also from the late Medieval period (Table 9.2). Only three sites with faunal remains present, but no ABGs, are recorded for the Anglo-Scandinavian and high Medieval periods. This would suggest that, although only a small number of ABGs are present from these sub-periods, when bone survives they may be relatively common. In comparison only 37% of the late Medieval sites recorded with animal bones present also have ABGs. 9.3

The third early Anglo-Scandinavian ABG consists of a complete small dog recovered from enclosure ditch D at Ferrybridge (Richardson 2005). The enclosure is of Romano-British date, but the excavation report suggested the ABG was deposited three centuries after the abandonment of the settlement, which would give it an Anglo-Scandinavian date. The dog was deposited close to the skull of a mature female inhumation, which led Richardson (2005) to suggest the dog carcass may have been a pillow for the inhumation.

Anglo-Scandinavian

Only 14 ABGs are recorded from the AngloScandinavian period, most from the later parts of the period. They are recorded from six separate sites, with all of the later ABGs recorded from Coppergate, York (O'Connor 1989).

All of the recorded middle Anglo-Scandinavian ABGs were recovered in association with human remains. A partial sheep/goat ABG, consisting of axial and upper back limb elements, was found alongside human infant bones at Wharram (Pinter-Bellows 1992). The ABG is described as being in close association with the human remains, which are from an individual aged 39-41 weeks and may have been stillborn. Not all elements of the ABG were in articulation, and Pinter-Bellows (1992) described the deposit as a dump of butchered bones. Evidence of the butchery process is present, with knife cuts on the

Three ABGs have been recorded from the early AngloScandinavian period. Two of these are from Parlington Hollins (Richardson 2001). A partial horse ABG from an old stallion or gelding was recovered from a pit, and a partial ABG of a sub-adult female pig was present in the secondary backfill of a sunken feature building. Unfortunately, body part information was not published for these deposits, but butchery evidence in the form of a chop mark to the left scapula of the pig was noted (the report does not state which other elements are present). 121

Investigating Animal Burials; Ritual, Mundane and Beyond vertebrae, pelvis and femur, all indicative of disarticulation. As at Ferrybridge, the ABG and human remains were deposited in the early Medieval fill of a Romano-British ditch. At Wharram there is no trace of a grave cut and the remains appear to have been deposited within the ditch fill, possibly covered over by a large stone. This led Pinter-Bellows (1992) to suggest the ABG may just be ‘normal’ waste.

composition of the ABGs together with the butchery marks and the high proportion of foot elements in the non-ABG assemblage led O’Connor (1989, 186) to suggest the cat remains, including the ABGs, represent a collection of cat skins.

The other middle Anglo-Scandinavian ABGs were recovered from formal cemetery sites. A horse ABG (Table 9.1), was found in association with human remains at the Addingham cemetery (Keith 1997). It consisted of a pelvis and left, upper-back limb but no other composition information is given. The horse was recovered from the grave of a young adult male, which also contained the secondary interment of a female with sword wounds to the head. Non-ABG faunal remains were recovered from two other graves at the site, although one deposit consists of just horse teeth. Therefore, of the 55 graves present, only three contained animal remains and only one of these an ABG.

Only two ABGs have been recorded dating to the high Medieval period (Table 9.1). One complete juvenile pig was found in the excavations at West Street, Gargrave (Dobney 2005). Unfortunately no composition or contextual information was given. A complete dog was recovered from a ditch fill at Addingham (Keith 1997). The ditch post-dates the cemetery discussed previously and is associated with a Medieval manor. No further information was provided on the deposit.

9.4

High Medieval

This assemblage differs greatly to the southern England sample, where the majority of Medieval ABGs come from the high Medieval period (see 8.2). However, the southern England assemblage is dominated by a large assemblage from Faccombe Netherton. It may also be a general reflection of the archaeology, with the majority of the southern England, high Medieval assemblage coming from rural settlements. In contrast, very little is known regarding Medieval rural settlement in Yorkshire beyond a few key sites (Hall 2003; Moorhouse 2003).

An unidentified large mammal ABG, consisting of ribs, was recovered from a grave context at Sewerby (Hirst 1985, 94). This is the only mention of faunal remains in the report. The ABG was recovered in association with what was described as the ‘richest grave on the site’, an inhumation of a female, 17 to 25 years old. The grave also contained a bronze cauldron, a wooden box, a number of bronze brooches and glass beads. The ABG, combined with the other finds, is interpreted as an offering. ABGs were not present in the other 58 graves from the site. Animal remains are known from other early Medieval cemeteries, most consisting of single elements or partial ABGs and are interpreted as offerings of food (Crabtree 1995), or the remains of a feast (Lee 2007). Although no cemetery sites were recorded without ABGs present (Table 9.2), it is noteworthy that only one grave from each formal cemetery site had ABGs present which supports Williams (2005) suggestion that, although present, animal remains were rarely interred in graves.

9.5

Late Medieval

In contrast, the majority of the Medieval Yorkshire ABG assemblage comes from the late Medieval period. Overall, 56 ABGs have been recorded from this period from a total of 14 sites (Table 9.3). However, there are a number of concentrations, with York excavations providing seven of the sites and Hull three. The majority (10) of the sites are urban in character, with the rest either manorial or monastic. This is similar to the pattern seen in the southern England late Medieval assemblage (see 8.4), as archaeological excavations concentrate on urban areas due to redevelopment.

In comparison, the late Anglo-Scandinavian ABGs were all recovered from the urban site at Coppergate, York (O'Connor 1985; 1989). This is the only early Medieval site with bird ABGs present in the form of a jackdaw and chaffinch and two domestic fowl. The domestic fowl consist of a complete immature ABG from pit 26900 and a partial deposit (elements unknown) from pit 18602. The domestic mammal ABGs are comprised of a partial neonatal cow (head and cervical vertebrae present) and three cats. The cat ABGs are all from the fills of separate pits, one is a complete articulated skeleton with knife marks to the skull. Another consists of just the skull and cervical vertebrae and also has knife marks present on the skull. The final cat ABG is also partial, although it is unknown exactly which elements are present. However, the skull is present, and also bears knife marks. The butchery marks on all these skulls consist of parallel knife cuts immediately above and between the eyes. The

9.5.1

Manorial

A number of differences are present between the site types. The majority of cattle ABGs are recovered from manorial sites. However, all are from one site, Higher Land, Gargrave (O'Connor 1983b). The report refers to the ‘burial’ of three articulated cattle limbs, but no further composition or contextual information is given. They were interpreted as the product of butchery waste, disease or natural death. The other cattle ABG from the site consists of a partial adult, with elements from the head, axial and fore limbs present. Although it was deposited within the backfill of a building foundation ditch the interpretation for the deposit was the same as for the other partial cattle ABGs from the site. 122

Chapter 9. Medieval Yorkshire Table 9.3 Number of ABGs per species recorded from late Medieval site types, Yorkshire. The number in brackets indicates the number of sites Domestic Site type Cattle S/G Horse Dog Cat Rabbit Fowl Duck Total Manorial/Castle (2) Monastic (2)

4

Town, Bawtry (1) Town, Hull (3) Town, York (6) Total (15)

1 1

1 1

6

1 1 2

2 3 6

2 1 1 5 8

Monastic

The only wild mammal recorded from the period comes from the Gilbertine Order priory of St Andrews at Fishergate, York (Bond and O'Connor 1999). It consists of a complete rabbit recovered from a pit, although no further information is given. Two complete cats from a stone-lined pit, 1387, were also recovered from the site. This corresponds with the evidence from other excavations at York where cats have been the most common ABG species recovered (Table 9.3).

4 1

15 15

4

7 10 30 56

The ABGs from Hull are comprised of similar species, although they come from three separate sites. The Ousefleet Property, High Street and the Hotham Property, Blackfriargate are situated next to each other (Armstrong and Ayers 1987). Cat ABGs were recovered from pits within the backyards of each tenement (Berg 1987). Each consisted of a partial ABG from a juvenile individual. A partial dog was also recovered from a pit in the backyard of the Ousefleet Property. It was a mixed deposit, consisting of elements from the axial skeleton, and upper fore and hind limbs. Chop marks, possibly associated with disarticulation, were present on the pelvis and upper back limbs. This is the only ABG from the Yorkshire late Medieval period with butchery marks observed. The reporting authors did not interpret these ABGs.

A horse ABG, referred to as ‘horse 1955’, but with no compositional information was recovered from Kirkstall Abbey (Ryder 1961). It was found in an area of the site called the ‘meat kitchen’. As the ABG was recovered from the 1955 excavations on the site it was probably a substantial deposit, as small partial ABGs were not often recorded at this time. Five other monastic sites have faunal remains present, but no ABGs (Table 9.2). Overall, ABGs do not appear to be common on monastic sites, which corresponds with the southern England data. 9.5.3

1

juvenile, the other from a young adult. Unfortunately, no further information regarding these deposits is available. The ABGs from the well were interpreted as ‘waste’, the well being described as a convenient, neighbourhoodburial location, analogous to the filling of a modern skip (Mounteney and Cumberpatch 1996). Interestingly, the very top most fills of the well date from the 17th and 20th centuries and also contain both cat and dog ABGs.

A dog ABG consisting of a partial adult was recovered from a layer at Pontefract Castle (Richardson 2002). Further information regarding the deposit was not present. 9.5.2

2 3 7 14

5 4

The other ABGs from Hull come from the Scale Lane/Lowergate site (Phillips 1980). The species present differ to the ones discussed above. A complete adult sheep/goat was found in a well and a complete neonatal calf in a cess pit. This site also has the only examples of wild bird ABGs from later Medieval Yorkshire, with four partial duck deposits recovered from a garderobe pit. Little further information was given for these ABGs.

Towns

As discussed above, the majority of the ABGs from this period come from three urban sites, Bawtry, Hull and York.

Faunal remains were noted on five other Hull Medieval sites but none had ABGs present. However, partial horse and cattle ABGs dating to the 18/19th century were found in Sewer Lane (Armstrong 1977), demonstrating that ABGs are present beyond the periods examined in this study.

The ABGs from Bawtry, South Yorkshire were all recovered from the excavations at 16-20, Church Street, in fact all were found in one feature, well 482 (Mounteney and Cumberpatch 1996). The ABGs formed two groups. One group was recovered from the lower fill, 610-2. Within this were two partial horse deposits, both consisting of elements from upper back limbs, although it is possible they may have been from the same individual. Also within this context was a partial sheep/goat, comprised of the head and cervical vertebrae. The other ABGs from the Bawtry well are from the upper fill, context 639. Within this context were a partial neonatal calf, a partial dog, and two partial cats, one from a

The majority of the ABGs from late Medieval York are from cats or dogs. A complete adult dog was recorded from a layer at 118-126 Walmgate (O'Connor 1984b). The individual had at some point in its life suffered a blow to the tail, resulting in the fusing of the caudal vertebrae. A partial dog consisting of 23 unknown elements was present at 1-5 Aldwark (O'Connor 1984a). It had been deposited within a ‘dump’ over a disused 123

Investigating Animal Burials; Ritual, Mundane and Beyond 9.6

floor within a building. The other three dog ABGs were all recorded from 16-22 Coppergate (O'Connor 1983a). All are complete and are deposited within different layers of what was also described as a ‘dump’. Due to the nature of the archaeological context, these ABGs were interpreted as ‘waste’ deposits.

Summary

Although the ABG assemblage from Medieval Yorkshire is relatively large in comparison to the prehistoric assemblages from this region, the lack of detailed information has resulted in less analytical analysis compared with other periods. However, the above descriptions have shown that a number of characteristics are present in the assemblage. It has also shown, along with the southern England assemblage, that ABGs are present from a varied number of Medieval sites and contexts and deserve further attention.

A larger number of cat ABGs are present in the late Medieval York assemblage (Table 9.3). A partial cat was recovered from pit 76 at 1-5, Aldwark (O'Connor 1984a). Five cat ABGs were recovered from 58-59, Skeldergate (O'Connor 1984b). All appear to have been recovered in association within pit 676. Three consist of partial juvenile remains; the other two are complete ABGs from adult animals. Ten complete domestic fowl were also recorded from this feature, although it is unknown whether these were in close association with each other or the cat ABGs. Due to the nature of the publication no further information is available nor interpretation offered.

Only a small number of Anglo-Scandinavian ABGs are present, all of which were associated with human remains in the middle Anglo-Scandinavian period. In this respect, the assemblage differs greatly to the southern England one. At present, it is unknown whether this is a true regional variation as only two early Medieval cemeteries with faunal remains present were recorded from southern England.

Cat and domestic fowl ABGs are also present in the Bedern Foundry assemblage (Bond and O'Connor 1999). A complete cat was found in post-hole M5a. Three complete domestic fowl, possibly females, were also recovered from the same post-hole, although it should be noted that the report is unclear on the exact nature of these ABGs. Two complete domestic fowl were also recovered from foundation trench 2682. The ABGs appear to be deposited within the foundations of one of the workshops built on the site. Bond and O’Connor (1999, 368), drawing parallels to sites in Lincoln, tentatively suggested there may be an association between domestic fowl and cats, and that these ABGs may represent foundation offerings.

The lack of ABGs from the high Medieval period is also in stark contrast to the southern England assemblage. In comparison, the majority of the Yorkshire Medieval ABGs dated to the later part of the period. Domestic fowl were the most common species from the late Medieval period, which does correspond with the southern England data. However, the domestic fowl came from only two York sites. In comparison, cat ABGs have been found on seven sites. The majority of the ABGs were recovered from urban contexts. However, this may simply be a reflection of excavation and publication priorities in Yorkshire. The majority of the sites recorded are from the city of York, largely due to the large number of extensive, published excavations that took place within the city during the 1970s and 1980s. Although similar species were deposited as ABGs on most of the sites, the nature and context of the deposits differs between the sites. This indicates that the ABGs had undergone different processes in their creation.

The horse ABGs recorded from York (Table 9.3) are all recorded from Fox Inn, Low Petergate (Ryder 1970). These are the only ABGs recorded from this site. All were found in layer 8, which is described as a deep organic layer with an abundance of finds. Although found in association, they are all different types of ABG. The deposit described in the report as ‘horse burial 1’ is from an adult animal and consists of just eleven vertebrae (it is unknown which type of vertebra). ‘Horse burial 2’, has 42 elements including vertebrae, ribs, pelvis and the left and right upper hind limbs from an old adult. ‘Horse burial 3’, is the most complete with 81 elements present, including the head, vertebrae, ribs, and an upper front limb from an adult individual. It is also possible that further elements from this ABG have been disturbed by later wooden piles. ‘Horse burial 3’ had suffered inflammation of the maxillary plate, and bit-wear was present. No butchery marks were reported on the ABGs and no interpretation was offered. Although the sites discussed above were all from the same town, the nature and context of the ABGs differs on each site.

124

Chapter 10. Patterns and Trends 10.1

Introduction

In the previous sections a detailed review of the available quantifiable ABG data from each archaeologically defined period has been conducted. The analysis has shown that the nature of ABGs as well as their depositional context is extremely variable. Despite this, patterns are evident in different periods. This chapter aims to draw together the previous discussions and examine the overarching trends which are evident within the dataset. As these trends are affected by the nature of the assemblage, the possible biasing effects are firstly investigated. 10.2

Are the patterns real?

The period analyses have shown that some of the variations may be due to small sample sizes and the dominance of ABGs from a small number of sites, although this is only the case for some of the trends. We must therefore ask if the general trends and patterns seen are due to past human action, or biasing effects on the data. 10.2.1 Sample size and ABGs In some periods there appears to be little positive correlation between the ABG and total NISP species proportions. One factor we need to also consider is the relationship between a site’s sample size (number of faunal elements present) and the number of ABGs present. In a number of periods examined, the sites with the largest faunal assemblages such as Windmill Hill (see 3.4) and Danebury, (see 4.3) also have the largest sample of ABGs present. To investigate this, NISP counts for identified elements was compared against the number of ABGs present. The available data did restrict the number of sites that could be compared. The majority of sites do not give NISP counts excluding ABGs and the number of elements belonging to ABGs was not often given. Also NISP counts were not available for all sites. The majority of the sites utilised within this study have total NISP counts of less than 2,000, and have fewer than 25 ABGs present. In fact, 63 of the sites have five ABGs or less. A number of sites have larger ABG assemblages but only Owslebury and Danebury have very large total NISP counts. The use of non-linear regression (see Shennan 1997, 161), indicates there is a modest positive relationship between the number of ABGs and the total NISP counts, but this is not statistically significant. The coefficient of determination (r2) has a value of only 0.58 (Figure 10.1). To investigate this further, the number of ABGs against the total NISP count was plotted for sites with less than

60 ABGs (Figure 10.2). This also showed there is only a very limited relationship. The r2 value of 0.062, indicates that the relationship between the number of ABGs present and the total NISP is a very weak one. Therefore, there appears to be a high degree of variability in the number of ABGs present on a site, compared to the total NISP count. This indicates that factors other than sample size are influencing the number of ABGs deposited and recovered. 10.2.2 Taphonomy One important factor to consider is taphonomy, which has been discussed in greater detail in chapter two. Variations in the frequency and nature of ABGs between sites may be due to taphonomic bias. One of the main mechanisms involved is post-depositional sediment movement. This can be caused by two separate events; the decomposition of organic matter and hence the movement of the sediment matrix, or subsequent activity on the site. The latter has been noted on some sites to have disturbed ABGs, but does not appear to be a major factor in their destruction. The amount of post-depositional sediment movement due to decomposition is dependant upon the amount of organic material incorporated within the fill matrix and upon the placement of ABGs within the features. This differs between periods and sites. The majority of Neolithic material is found in the basal layers of features. A recurring feature of Neolithic pits is that the sides show little evidence of weathering and the pits appear to have been backfilled soon after digging and the deposition of the primary fill (Thomas 1999, 64-65). The backfilling layers above the primary fills contain only small amounts of material culture and mainly consist of a soil matrix. Causewayed enclosure ditches show a different pattern. Unlike the pits, they were left open for a period of time, as shown by the silting layers found at the bottom of the ditches, before the primary fills, which contain the majority of the material culture, were deposited. At Windmill Hill, some ditches also contained secondary fill deposits with ABGs, after which the ditches were generally filled by a natural silting process with a small amount of re-cutting. It appears from Windmill Hill and other causewayed enclosures that there was only limited dumping of material and the filling of the ditches was primarily a natural process (Whittle et al. 1999b, 353). The majority of the deposits were limited to the bases of the ditches. The ABGs deposited in long barrow and late Neolithic henges also appear to follow this pattern. Therefore, the majority of Neolithic cultural material appears to have been deposited in the bottom fills of features. This would limit the extent to which ABGs would have been disturbed by sediment movement

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 10.1 Total NISP plotted against number of ABGs present for each site with available data. The line represents the polynomial non-linear regression to an order of 3

Figure10.2 Total NISP plotted against number of ABGs present for sites with available data and less than 60 ABGs. The line represents the polynomial non-linear regression to an order of 3 only two are complete. Both are from foetal/neonatal individuals that were deposited at the very bottom of the shaft (see 3.7.2). Similar evidence can be found within shallow features such as the deposition of complete cattle and sheep/goat from Down Farm (Legge 1991b) and complete cow and calf ABGs from Crab Farm (Papworth 1992) (see 3.7.3).

caused by decomposition of organics within the soil matrix. As there was, to all intents and purposes, a ‘hard bottom’ to the fill, movement of separate ABG elements would have been limited. The same is also true for other periods, which would explain why complete ABGs are often found in the bottom of features. For example, of the 17 ABGs from the Wilsford shaft (Ashbee et al. 1989), 126

Chapter 10. Patterns and Trends started to become a standard feature of the faunal report. Although there are some exceptions (for example Cornwall 1958; King 1962; Stopes et al. 1952), the majority of reports before the 1970’s consisted of text descriptions of some of the more interesting aspects of the assemblage. However, with the presentation of quantifiable data becoming a standard for zooarchaeologists, ABGs started to pose a problem. As they consist of many elements from a species they exaggerate the NISP counts, the most commonly used quantification method. Therefore, zooarchaeologists started noting their presence, either by excluding them from NISP counts, or noting in the accompanying text that remains from an individual animal were present. Published descriptions and interpretations of ‘animal burials’ from the 1980’s onwards meant archaeologists in the field became more aware of them, and were accordingly more likely to notice and record their presence, reaching the situation we have today where the absence of ABGs on some sites is reported (for example Charles 2002).

Not all complete ABGs are recovered from the bottom of features. A number of complete ABGs have been found, for example, in the Romano-British shafts/wells at Dorchester (see 6.2.3). However, the majority of ABGs from such contexts are not complete. As previously discussed, a significant number of partial dog deposits consist of ‘mixed’ ABGs, which may suggest they were deposited complete but became disarticulated through post-depositional movement within the feature. If partial ABGs are recovered close to the bottom of a feature, then they are unlikely to have been affected by postdepositional movement. This suggests they became partial by human action such as butchery. Alternatively, some partial ABGs may represent secondary deposition of part of a complete carcass of the animal that had originally been deposited elsewhere, perhaps within a midden. This is why detailed investigation of ABG remains is important. Weathering damage is not reported for any of the ABGs, and gnawing was noted on only ten of the 2,062 recorded. This may indicate that most ABGs represent primary depositions. However, such detailed information concerning these deposits is rare and in many cases may not have been reported.

The data collected in this study appear to support the above assumptions. The largest samples of ABGs were recorded from the 1980’s onwards (Table 10.1). In addition, it is from the 1980’s onwards that the majority of sites included in this study were published (Table 10.2). Interestingly, the 1980’s also produced the largest sample of publications. This appears to be due to the publication of a number of assemblages in the South of England, in particular Danebury, Owslebury (AML), Oakridge Well (AML) and Winchester Suburbs (AML). In fact a large proportion of the data was reported in the very detailed AML reports produced by the Faunal Remains Unit of English Heritage, based at Southampton University. The unit was at its most prolific in the 1980’s and it is their attention to detail in describing ABGs that accounts for the large number recorded from this decade.

In effect there is no hard and fast rule. Each feature/deposit needs to be taken as a separate entity. Luckily for very deep features such as the Oakridge Well (see 6.5) and Wilsford Shaft (see 3.7.2) such movement can be seen archaeologically and its impact taken into account by the zooarchaeologist. 10.2.3 Archaeologists Other taphonomic factors affecting the data are excavation and post-excavation methods. For a study of this nature it is not possible to exert much control over the quality of the data available. However, there are a number of assumptions we can make. If we look at the history of archaeology we can see a number of changes in the priorities and goals of excavations as archaeology has developed as a profession. Throughout the last century excavation techniques and, more importantly, recording of archaeological sites have improved. Therefore we could work on the assumption that on more recent excavations ABGs have been spotted, if present. This of course is an immense assumption to make. There are many variables to consider, such as the skill and experience of the individual archaeologist and, more recently, the time scale applied to the excavation.

Therefore it appears that the more recent the excavation and publication, the more likely ABGs are to be recovered and reported upon because the excavators and zooarchaeologists were aware of their presence and possible significance, although the quality of data which is published can still be variable. One factor that shows the increased awareness of ABGs is the explicit note of the deposits completeness. It was not until Grant’s (1984a) publication of ‘special deposit’ types that the incomplete skeletal nature of ABGs was fully realised. Although there is some variability between period types, the majority of pre-1980 publications have a higher percentage of complete ABGs (Table 10.3). The lowering of the proportion of complete ABGs reported upon in the last couple of decades may be taken as an indicator that more incomplete ABGs were now being recognised and/or deemed worthy of inclusion in a publication.

We can expand the argument, with particular concern for the animal remains. The detailed study of faunal remains from archaeological sites was not common until the 1970’s, encouraged in part by the advent of ‘processual’ archaeology (see 1.2.2). From this time the reporting of faunal remains also improved, moving away from being mainly concerned with metrical assessment of species, towards a much more in-depth approach, especially regarding quantification. It was not until the 1970’s that the calculation of the NISP and the MNI data from sites

All the main differences between periods and site types discussed in previous chapters use data primarily from the 1980’s onwards. Therefore it is probably safe to

127

Investigating Animal Burials; Ritual, Mundane and Beyond Table 10.1 Number of ABGs reported per period for each decade when data were collected Bronze Iron RomanoEarly Later Decade Neolithic Age Age British Medieval Medieval Total 1940's 3 6 7 16 1950's 1 3 4 1960's 7 1 9 19 1 37 1970's 1 61 65 2 9 138 1980's 2 34 305 492 46 81 960 1990's 33 17 88 240 35 66 479 2000's 10 7 312 85 9 5 428 Table 10.2 The percentage of sites published by decade for each time period Early Later Neolithic Bronze Age Iron Age RomanoMedieval Medieval Decade (14) (23) (59) British (65) (22) (32) 1940's 14% 3% 2% 1950's 4% 2% 1960's 21% 4% 7% 6% 3% 1970's 4% 15% 12% 9% 6% 1980's 14% 35% 24% 34% 32% 50% 1990's 29% 30% 22% 26% 36% 28% 2000's 21% 22% 27% 20% 23% 13% Table 10.3 Percentages of complete ABGs reported in each decade. Number in brackets indicates the total sample size RomanoEarly Later Decade Neolithic Bronze Age Iron Age British Medieval Medieval Total 1940's 33% (3) 67% (6) 14% (7) 38% (16) 1950's 100% (1) 0% (3) 25% (4) 1960's 0% (7) 0% (1) 44% (9) 74% (19) 0% (1) 49% (37) 1970's 0% (1) 72% (61) 26% (65) 100% (2) 67% (9) 50% (138) 1980's 0% (2) 9% (34) 19% (305) 21% (492) 13% (46) 37% (81) 20% (960) 1990's 15% (5) 52% (17) 7% (88) 28% (240) 57% (35) 73% (66) 32% (479) 2000's 10% (10) 29% (7) 17% (53) 33% (28) 33% (3) 20% (5) 21% (428) In total, 2,062 ABGs were recorded in this study, the vast majority of which (90%) are from sites in southern England (Table 10.4). This, however, is not an indication that ABGs are more common in southern England. It is more a refection of the nature of the archaeological datasets from both regions. Simply, more data are available from southern England and therefore more ABGs have been recorded. Overall the reports from 493 sites were examined for this study, 213 with ABGs present. The number of sites recorded with ABGs present shows that for some periods the proportion of sites with ABGs is similar for both southern England and Yorkshire (Figure 10.3).

assume that the patterns present are not the result of archaeological reporting strategies. This may be, in truth, a leap of faith, but it is also one archaeologists are used to making. We have to assume that the data we are working with is a reasonable reflection of the standards of archaeology. 10.3

How common are ABGs?

The majority of previous literature regarding ABGs has had a predominantly prehistoric focus. In particular, deposits from the Iron Age have received a large amount of attention, probably because of the well known work of Grant (1984a; 1989a; 1991) and Hill (1995; 1996). A survey of the published literature would lead one to believe ABGs are almost a purely prehistoric phenomenon. However, this study has shown that this is not the case with more ABGs recorded from the RomanoBritish period than from any other in both the southern England and the Yorkshire datasets.

As already stated, the largest numbers of ABGs come from Romano-British contexts, and represent 43.9% of the total assemblage. Deposits from Iron Age contexts constitute the second largest group. Together, ABGs from Iron Age and Romano-British contexts make up 81.9% of the assemblage recorded for this study. In addition, the Medieval ABGs constitute a larger proportion of the 128

Chapter 10. Patterns and Trends

Region Southern England Yorkshire Total Total percentage

Table 10.4 Number of ABGs recorded per region and period Bronze Iron RomanoEarly Later Neolithic Age Age British Medieval Medieval 54 61 746 820 78 104 1 38 88 14 58 55 61 784 908 92 162 2.67%

2.96%

38.02%

44.03%

4.46%

Total 1863 199 2062

7.86%

Figure 10.3 Percentage of sites per region and period with ABGs present Finally, the presence and absence data indicate that a high proportion of Iron Age, Romano-British and early Medieval sites, in both southern England and Yorkshire, have ABGs present. Although the total number of ABGs is higher for the Romano-British period compared to the early Medieval period, a similar proportion of sites have ABGs present. This indicates that although ABGs are found in greater concentrations in the Iron Age and Romano-British periods, they are still present on a high proportion of sites in later periods, albeit in smaller numbers.

assemblage, for both southern England and Yorkshire, than Neolithic and Bronze Age ones. In fact a larger proportion of those recorded are from historic, as opposed to prehistoric contexts. Therefore, although a large proportion of the literature is concerned with prehistoric ABGs, they would appear to be just as common from historic contexts. One of the problems with looking at just the total number of ABGs recorded per period is that the sample can be biased by large assemblages from individual sites. This is especially the case for the southern England Iron Age and Romano-British periods. To combat this, we can instead look at presence and absence data, which displays three interesting trends. Firstly, ABGs on Neolithic and Bronze Age sites are rarer from Yorkshire than southern England. However, this may be due to a distinct lack of sites with faunal material present from Yorkshire. Secondly, ABGs are more common from Medieval sites in Yorkshire, compared to southern England (Figure 10.3). This is also indicated by the ABG counts for Yorkshire, where, in contrast to southern England, the second largest assemblage is from the later Medieval period. This difference does not appear to be due to bias in the publications, as the majority of southern England (94.2%) and Yorkshire (94.7%) reports were published from the 1980’s onwards. The difference may be due to the scale and detail of the excavations, with the majority of the later Medieval ABGs from Yorkshire recorded from the excavations in York.

10.4

Changing species

One of the constants in the ABG data is the domination of domestic animals. Overall, 1,679 are from domestic animals which represent 91.1% of the total assemblage. There are slight variations between periods and regions, with the lowest percentage (85.5%) coming from later Medieval Yorkshire and the highest (100%) from Iron Age Yorkshire. Wild animals therefore appear to be rarely deposited as ABGs. This trend matches observations from the total faunal assemblages, with wild animals relatively rare in all periods, apart from some high status sites, particularly of Medieval date (see Crabtree 1996; Grant 1989b; Hambleton 1999; Maltby 1981b; Pollard 2006; Sykes 2006b). Throughout this study the overall species NISP counts were recorded for sites with and without ABGs present. 129

Investigating Animal Burials; Ritual, Mundane and Beyond from late Neolithic sites examined in this study, despite the evidence that the late Neolithic sees a rise in the utilisation of pigs (Thomas 1999). This is not clear in Figure 10.4, as the graph is only designed to show broad inter-period trends. The surprising lack of pig ABGs in the late Neolithic may reflect differences between site types of the early and late Neolithic and the limited size of the sample. The majority of the data producing late Neolithic fauna in southern England comes from henge enclosures, but only two ABGs have been recorded from this site type, a dog and a sea eagle, both from Coneybury Henge (Maltby 1990b) (see 3.4). Durrington Walls (Harcourt 1971b) has produced one of the largest faunal samples from a henge enclosure, however, no ABGs were recorded in that sample from that site.

That evidence is utilised in this section to provide a comparison of the ABG and total faunal assemblage species proportions. Where possible the NISP counts excluding ABGs were recorded. However, for the majority of sites this data was not available. Therefore the NISP counts for the total faunal assemblage also include ABGs. This should not unduly affect the data for most species but it does have an effect on the proportion of dog remains recorded in the total faunal assemblages. Although domestic animals consistently make up a large proportion of the ABG assemblages, there is substantial regional and chronological variation in the relative abundance of different domestic species represented. 10.4.1 Southern England

However, this may be a refection of the date of the excavation and limitations of the original faunal analysis rather than a real absence. Recent excavations on the site indicate some pig ABGs are present (Coy 1993). Only one pig ABG is present in the Bronze Age sample, from the late Bronze Age settlement at Bell Street, Romsey, Hampshire (Coy 1993). The proportion of pig in the overall faunal assemblage also falls in the Bronze Age.

The majority of the southern England Neolithic ABGs are from cattle, which make up 53% of the assemblage. Pigs and dogs are the second and third most common species respectively (see 3.3). However, we must note that of the 55 Neolithic ABGs, 26 (43%) are from Windmill Hill, 17 of which are cattle. However, cattle would still be the most common species if the Windmill Hill data were excluded. Cattle are also the most common species in the total faunal assemblage from the Neolithic. They make up 45.7% of the combined NISP count from the 13 sites included in this study (Figure 10.4), a percentage not very different from the ABGs (Figure 10.5).

The only comparable study into ABGs from the Neolithic and Bronze Age is Behrens’ (1964) broad investigation, which recorded 459 deposits from sites in Europe, Africa and Asia (see 1.2.3). Over 50% were from dogs. In contrast, only 14 dog ABGs (12%) are present in the Neolithic and Bronze Age sample from southern England. This may highlight the need to take a regional approach to such data, although the recording of only complete skeletons by Behrens may increase the percentage of dog deposits compared to other species which are more often found as partial skeletons..

Examination of the southern England Bronze Age ABG data shows a different pattern with sheep/goat the most common species, (45%), followed by cattle (36%) (Figure 10.5). This represents a large rise in the percentage of sheep/goat ABGs. In the Neolithic, only four are recorded from three different sites, representing just 7% of the assemblage. In comparison, there is a drop in the number of pig ABGs with only one present in the Bronze Age assemblage. The increase in sheep/goat ABGs during the Bronze Age mirrors the trends seen in the overall faunal material. Sheep/goat, made up predominantly of sheep, dominate the ‘normal’ faunal assemblages. In the combined NISP count from the 43 Bronze Age sites included in this study (most without ABGs present) sheep/goat make up 51.4% of the assemblage. In contrast sheep/goat represent only 12.8% of the combined Neolithic assemblages from 13 sites (Figure 10.4).

Sheep/goat (35%) remain the most common ABG species in the Iron Age sample (Figure 10.5) (see 4.2). They also remain the most common species (53%) found within the total faunal assemblage, which is dominated by the large datasets from Wessex chalk downland sites. Cattle and pig are the second and third most common animals represented in the total faunal assemblage. However the Iron Age is the first period when there are major differences between the total faunal and the ABG assemblages. Dog followed by cattle are the second and third most common ABG species. Although dogs consistently make up a larger proportion of the ABG assemblage compared to the total assemblage, we must consider that the inclusion of ABGs in the total faunal assemblage counts, exaggerates their presence. Therefore, the difference is likely to be more marked than shown in Figure 10.4 and 10.5.

Therefore the pattern in the proportion of cattle and sheep/goat ABGs appears to follow the trend seen in the overall faunal data. The pig data are slightly different. Pigs are the second most common species found as ABGs in the Neolithic. They are also the second most common species in the total faunal assemblage. However, whereas the percentages for cattle and sheep/goat are similar between the ABG and total faunal assemblages, pig make up a much higher proportion of the non-ABG faunal assemblage. Interestingly the majority of pig ABGs are from early and middle Neolithic sites. None are present

The increase in dog ABGs continues into the RomanoBritish period, where they make up 45% of the assemblage. The proportion of sheep/goat drops significantly to 11.6%. Cattle remain the third most common ABG species, although there is also a drop in the proportion of cattle from 15.1% to 9.6%. Interestingly 130

Chapter 10. Patterns and Trends

Figure 10.4 Total percentage NISP for the most common species per period from southern England sites. Includes data from sites both with and without ABGs. ABGs are included in the NISP counts. Total sample size in brackets

Figure 10.5 Total percentages of ABGs from each period for southern England. Number of ABGs per period in brackets proportion of cattle ABGs, which had decreased in every period since the Neolithic, but in the Anglo-Saxon period are the second most common species (24.4%). It is possible that the small ABG sample size from the AngloSaxon period may affect the results. For example, a large proportion of the dog deposits are from the upper fill of pit 56 at Clifford Street, Southampton (Bourdillon 1990b) (see 8.3). The proportion of dog ABGs drops significantly in the later Medieval period to 13.5%, although this still makes dog the second most common ABG species. For the first time a bird species makes up a significant proportion of the assemblage with the rise of domestic fowl ABGs from 2.6% in the Anglo-Saxon period to 25%. However, this is possibly due to the small and restricted sample with all but three domestic fowl

the proportion of horse ABGs reached its highest level in the Iron Age sample (13%), but drops to its lowest level (2.2%) in the Romano-British assemblage. Significant changes occur to the ABG species representation in the transition from the Iron Age to the Romano-British period. However, these changes did not occur quickly. In the early Romano-British period, sheep/goat remain the most common ABG species (see 7.2). The ABG results are in stark contrast to the species proportion in the total faunal assemblage in which cattle, sheep/goat and pig are the three most common species. The proportion of dog ABGs (39.7%) drops slightly in the Anglo-Saxon assemblage, although, they remain by far the most common species. There is a rise in the 131

Investigating Animal Burials; Ritual, Mundane and Beyond pit 113, Suddern Farm (Poole 2000d) and two from pit 5, Viables Farm (Maltby 1982b).

deposits being recorded from Faccombe Netherton (Sadler 1990). There is also a slight rise in the percentage of sheep/goat (14.4%) and pig ABGs (12.5%), the latter of which had remained around 8% since the Iron Age (Figure 10.5).

The Romano-British period produced the largest ABG assemblage from Yorkshire (see 7.2). Species represented change dramatically compared with the Iron Age. The proportions of sheep/goat (22.7%) and dog (21.6%) ABGs rise. The percentage of cattle (15.9%) falls to only the third most common species, and the proportion of pig falls to only 9.1%. The decrease in the proportion of pig ABGs is probably due to an increase in the amount of data from settlement sites, as well as changes in the ABGs deposited within funerary settings. Domestic fowl and horse ABGs are also present in the Yorkshire assemblage for the first time in this period.

Again the ABG species proportions from the AngloSaxon and later Medieval periods contrast with the overall faunal assemblage data. The proportion of cattle peaks in the Anglo-Saxon period and cattle, sheep/goat and pig remain the three most common species. There is a rise in the total number of domestic fowl in the later Medieval periods. This may be slightly exaggerated by the inclusion of ABGs in the overall faunal assemblage NISPs, as analysis of the faunal assemblages from sites with no ABGs present give the proportion of domestic fowl at 6.3% as opposed to the 10.2% from the combined assemblage utilised in Figure 10.4.

In contrast, the main change in the overall faunal assemblage is a decrease in the proportion of sheep/goat with cattle becoming the most common species (Figure 10.6). This mirrors the change seen in southern England (Figure 10.4). However, the Yorkshire and southern England ABG assemblages show a very different species makeup. The main difference is that the Yorkshire sample does not display the dog-dominated pattern seen in the southern England data. This may be due to differences in the type of site and features excavated. The majority of the southern England Romano-British data comes from large urban excavations, whereas rural settlements provide most of the Romano-British ABG data from Yorkshire. Compared with southern England a limited number of faunal assemblages from urban contexts are available from Yorkshire. The majority of the dog ABGs from southern England are from pit/well deposits within Dorchester, Winchester and Silchester. Maltby (Maltby 2010, 297-304) has discussed the evidence from 16 Romano-British towns, noting that dog ABGs are most often found within deep pits and wells. A large number of similar features from urban contexts have not been excavated in Yorkshire. However, to test this supposition we will have to wait for further Romano-British urban excavations to be carried out in Yorkshire, or extend the comparison by feature type to other regions.

10.4.2 Yorkshire The Yorkshire ABG data differs to the southern England one. Although sample size is an issue for the Yorkshire assemblage, it does show that regional differences need to be taken into account regarding these types of deposits. Only one ABG consisting of a partial fox skeleton from Whitegrounds Barrow 1 (Riggott and Williams 1984) was recorded from the Neolithic (see 3.2). None were recorded from the Bronze Age Yorkshire dataset. However, this is likely to be due to the small amount of faunal material available to study from Yorkshire for these periods. Fortunately a larger dataset is available from the Iron Age, with the majority of the ABGs consisting of either pig (42.1%) or cattle (36.8%) (Figure 10.7). This is in sharp contrast to the total faunal assemblage from the period, which has a similar species pattern to the southern England data, with sheep/goat dominating (Figure 10.6). Pig remains make up a much larger proportion of the ABG assemblage from Iron Age Yorkshire than in the southern England sample. Such a large difference is probably due partly to the small Yorkshire sample size and perhaps more significantly to the dominance of funerary sites (see 5.2). However, even on settlement sites cattle ABGs are more common than those of sheep/goat at a ratio of 2:1,which is a complete reversal of the southern England results.

Moving into the Anglo-Scandinavian period the ABG species proportions change again. Cat, domestic fowl and horse become the most common species, with cattle, sheep/goat, pig and dog being represented by only one ABG each. The change in species proportions is likely to be due to the very restricted sample for this period, with over half the ABGs, including all the cats, coming from the excavations at 16-22 Coppergate, York (O'Connor 1989). However, the trend does continue into the later Medieval period with domestic fowl and cat being the two most common ABG species respectively, followed by dog (Figure 10.7). The higher proportion of domestic fowl and dog ABGs from later Medieval Yorkshire does correspond with the pattern for the same period in southern England (Figure 10.5). However, the high proportion of cat remains is different, with cat ABGs making up 22.6% of the ABG assemblage from Yorkshire, but only 4.8% from Wessex. This could be

Another contrast is that no horse ABGs have been recorded from the Iron Age of Yorkshire, whereas a total of 97 horse ABGs have been recorded from southern England. Horse remains are present in small numbers in the total faunal assemblage from Yorkshire, so they were present. The difference may be due to the high proportion of ABGs from funerary contexts. None of the horse ABGs from southern England were recovered from features which could be defined as funerary and indeed only three of the horse ABGs from southern England are in association with articulated human remains, one from 132

Chapter 10. Patterns and Trends

Figure 10.6 Total percentage NISP for the most common species per period from Yorkshire sites. Includes data from sites both with and without ABGs. ABGs are included in the NISP counts. Total sample size in brackets

Figure 10.7 Total percentage of ABGs from each period for Yorkshire. Number of ABGs per period in brackets For this study, analysis of the composition of ABGs has been carried out at three levels. The first examines whether the ABG consists of a complete skeleton. The second compares which general areas of the body (e.g. hind limbs, head etc) are present. The third level considers in detail which specific body areas are present (e.g. left upper hind limb).

due to the dominance of data from York where excavation and particularly sieving standards were high. As with the southern England data, there is little correspondence between the ABG assemblage and the total faunal assemblage (Figure 10.6 and 10.7). 10.5

Composition; not all ABGs are the same This approach has been employed in the examination of the composition of ABGs in the previous chapters, although the level of detail is dependant on what is available from the published reports utilised. The main difference between the previous classifications and the ones used in this study is the exclusion of skulls. Grant (1984a), Wait (1985) and Hill (1995) all recorded deposits of skulls as a category of ‘special deposits’. As

As well as variation in species, these deposits also vary in form. Previous authors have also noticed such a trend. Grant (1984a), Wait (1985) and Maltby (1985a) distinguish different types of ABGs. Hill (1995, 57) combined Grant’s and Maltby’s classifications and distinguished four types of ABGs (see 1.2.8).

133

Investigating Animal Burials; Ritual, Mundane and Beyond the mortality profile of complete or partial sheep/goat. There is however a noticeable difference for pigs. Only 8.6% of the complete pig ABGs are from individuals that lived beyond the juvenile stage of development. In comparison 18.8% of partial pig ABGs are from individuals older than juvenile.

outlined in chapter 1, individual skull deposits are not counted as ABGs in this study. 10.5.1 Complete ABGs As noted by Hill (1995, 59) the deposition of complete carcasses was rare in the Iron Age, and this appears to be the case for the other periods covered in this study. Overall, the majority of ABGs consist of partial skeletons (Table 10.5), however this does vary between species and periods.

Surprisingly a higher proportion of complete pig ABGs are recorded than dogs. However, the completeness of a large proportion of dog remains is unknown (Table 10.5), the majority from Romano-British contexts (see 6.8). This may be due to taphonomic factors, in particular postdepositional movement and the mixing of multiple depositions within the deep pits/wells where they were often found. Maltby (1987c; 1993, 326; 1990) has suggested that the majority of the dog ABGs would have been originally deposited as complete skeletons. This would explain the even spread of dog elements in the partial ABGs and non-ABG faunal assemblages from many of the Romano-British sites. There also appears to be little overall age difference between complete and partial dog ABGs (Figure 10.8 and 10.9).

The vast majority of the domestic mammal ABGs recorded are incomplete. Cattle and horse are the domestic mammals that are most often found as partial remains. It is probably no coincidence that these are also the two largest mammals represented. This may simply reflect the practicality of depositing a complete cow or horse. The majority of complete cow ABGs encountered in this study have been from neonatal or juvenile individuals (Figure 10.8). In comparison, 42.4% of partial cattle ABGs are from adults (Figure 10.9). This pattern has been noted in a number of periods, in particular the southern England Iron Age assemblage, where many of the complete ABGs have been interpreted as natural deaths (see 11.4.1).

A higher proportion of the wild mammal ABGs recorded also consist of complete skeletons. However, this is due to a small number of Iron Age and Romano-British sites affecting the data (see below). The majority of domestic bird ABGs also consist of complete skeletons, most from the Medieval site of Faccombe Netherton (Sadler 1990).

Horse remains show a very different pattern with the majority of complete and partial ABGs coming from individuals that have reached maturity. This may be a reflection of the differences in status and utilisation of the two animals, with little evidence for the consumption of horse meat in Britain and a low kill-off of immature animals. There are generally very few cases where bones of young horses have been found in non-ABG assemblages (e.g. Maltby 1981a; 1993).

Although the overall assemblage shows some species are more commonly found as complete skeletons, there is also much variation between the periods. The Neolithic and Bronze Age assemblages have a very different pattern to the later Iron Age and Romano-British periods. The highest proportion of complete ABGs in the Neolithic assemblage are sheep/goat (Figure 10.10). But this is due to the very small sample size of only four sheep/goat deposits, two of which are complete.

A higher proportion of sheep/goat and pigs have been recorded as complete ABGs. There is little difference in

Table 10.5 Percentage of complete, partial and unknown ABGs for the total assemblage (southern England and Yorkshire), per species (not including fish or snake) Species Complete Partial Unknown Cattle (303) 16% 82% 2% S/G (437) 20% 77% 3% Pig (181) 35% 61% 4% Horse (155) 8% 92% 1% Dog (593) 30% 39% 31% Cat (77) 35% 57% 8% Domestic Fowl (109) 56% 42% 2% Other Domestic Bird (9) 89% 11% 0% Wild Mammals (76) 59% 32% 9% Corvids (69) 9% 72% 19% Other Wild bird (50) 2% 36% 62% Total (2059) 26% 60.8% 13.2%

134

Chapter 10. Patterns and Trends

Figure 10.8 Mortality profiles of complete ABGs per species (combined results from all periods and regions with ageing data available)

Figure 10.9 Mortality profiles of partial ABG per species (combined results from all periods and regions with ageing data available)

Figure 10.10 Percentage of complete ABGs present for the main species per period 135

Investigating Animal Burials; Ritual, Mundane and Beyond Seven complete sheep/goat ABGs are also present in the Bronze Age sample, the majority from the Crab Farm enclosure (Locker 1992a). The highest proportion (32%) of complete cattle ABGs is also recorded from the Bronze Age. With the exception of the Down Farm Pond Barrow (Legge 1991b), all the complete cattle ABGs are from possible settlement sites (see 3.7.3).

A substantial proportion of dog ABGs in each period consist of complete skeletons, although it is not until the Anglo-Saxon period that the proportion of complete dog ABGs is higher than for any other species. Overall, more complete dog ABGs were recorded than any other species. Also many of the partial dog ABGs may have originally been deposited as complete skeletons.

No complete wild mammal ABGs are present in the Neolithic or Bronze Age assemblages. However, 76% and 65% of the wild mammal ABGs from Iron Age and Romano-British sites respectively are complete. However, both samples are heavily affected by individual sites. Thirteen of the 21 Iron Age wild mammal ABGs (excluding cat) are from a single deposit of 12 foxes and one red deer at Winklebury Camp (Jones 1977). In the Romano-British sample, 20 of the 24 complete wild mammal ABGs are from the Oakridge well (Maltby 1993). It is interesting to note that the Winklebury Camp red deer ABG is the only one encountered in this study that consists of a complete skeleton. The majority of the complete wild mammals are from small carnivores, such as fox, stoat and weasel. There may be a number of explanations for this. There is some evidence that small carnivorous mammals were consumed during the Mesolithic (Charles 1997), but very little evidence for such activity from the Neolithic onwards. Butchery marks are only present on one ABG. The lower front and hind limbs of a complete fox at the Iron Age site of Nettlebank Copse bears knife cuts, which are thought to be indicative of skinning (Poole 2000b). These ABGs may be complete because the carcasses had only been skinned and no further processing took place. This of course assumes that they are the result of human deposition.

In the later Medieval sample domestic fowl are the most commonly complete ABG. As with the Iron Age and Romano-British wild mammal data, this is heavily affected by the data from a small number of sites. As already noted, all the complete domestic fowl ABGs from southern England are from one site, Faccombe Netherton (Sadler 1990). A similar pattern is seen in the Yorkshire data where all the Medieval domestic fowl ABGs consist of complete skeletons, but all are recorded from two sites in York, The Bedern Foundry (Bond and O'Connor 1999) and 58-59, Skeldergate (O'Connor 1984b). In fact, only one domestic fowl from the total Medieval ABG assemblage has been recorded as a partial skeleton. There is generally a high proportion of complete ABGs for most species in the Medieval period, with 59% of sheep/goat and 50% of pig ABGs consisting of complete skeletons. However, these figures are dominated by a restricted number of southern England sites and may be a reflection of attitudes to these types of deposits in Medieval archaeology. 10.5.2 Partial ABGs If the data were available, partial ABGs were placed into categories dependant upon the body areas present. These can broadly be defined as axial, limb or mixed. Unfortunately detailed body area information, the third level of data referred to above, was not often included in the reports examined. Therefore the majority of this discussion has to utilise cruder distinctions.

During the Iron Age and Romano-British periods, domestic fowl are often found as complete ABGs. The Iron Age sample is small. Only six deposits were recorded dating to the Iron Age, three of which are complete. The number of domestic fowl ABGs increases in the Romano-British period to 58. In this period there appears to be a specific pattern in their deposition, with the majority recorded from funerary sites such as Poundbury (Buckland-Wright 1993) and Trentholme Drive (Fraser and Ryder 1968).

The overall body area proportions for partial ABGs appears to correspond with a number of the complete/non-complete species patterns, with cattle and horse ABGs again being similar. These species have the highest proportion of partial ABGs consisting of just axial elements. They also have the lowest proportion of partial ABGs consisting of mixed body areas (a combination of axial, limb and head elements) (Table 10.6). Cattle and horse partial ABGs also include a high proportion of limb deposits. It appears that both cattle and horse partial ABGs have been created after an intensive dismemberment process. This would correspond with the small number of complete deposits present, and would also explain the small number of ‘mixed’ ABGs recorded. The creation of these types of cattle and horse ABGs appears to be a constant theme throughout the periods and regions covered in this study.

A relatively high proportion of the pig ABGs from the Iron Age and Romano-British periods also consist of complete skeletons (Figure 10.10). In this regard the Iron Age assemblage is dominated by the results from Danebury, from which over half the complete pigs are recorded. The majority of these were neonatal and dated to the middle Iron Age. No one site dominates the Romano-British period’s assemblage. Again, all the complete pig ABGs are from neonatal or juvenile individuals, as is the case for partial pig ABGs. Perhaps significantly, of the 22 complete pig ABGs, all except two, one from Silchester (Grant 2000), and one from Portchester (Grant 1975) are from rural settlements.

A high proportion of partial sheep/goat ABGs also consist of limb elements only. Only a small percentage of 136

Chapter 10. Patterns and Trends

Figure 10.11 Triplot of the body area proportions for partial dog ABGs per period (including data from both regions)

Figure 10.12 Triplot of the body area proportions for partial cattle ABGs per period (including data from both regions)

137

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 10.13 Triplot of the body area proportions for partial sheep/goat ABGs per period (including data from both regions) The composition of partial ABGs also varies between periods. With the exception of the Neolithic and Medieval partial-dog-ABG assemblages, the overall trends discussed above appear to hold true. No mixed partial-dog-ABGs are recorded from Neolithic contexts, where the majority consist of axial element deposits (Figure 10.11). At the opposite end of the scale is the Medieval assemblage, where all the partial-dog-ABGs consist of mixed body areas. However, body-area data are only available from four Medieval deposits. Partial-dogABGs from the other periods include 40%-50% mixed body area deposits. As discussed above, this may indicate, they were being deposited as complete ABGs before subsequent disturbance and attrition.

sheep/goat, pig, dog and domestic fowl deposits consist of just the axial skeleton. However, compared to cattle and horse a large proportion of sheep/goat, pig, dog and domestic fowl consist of mixed partial ABGs. This could indicate these species were being less intensively butchered than cattle and horse. The axial and appendicular body areas may have been left in articulation more often, hence the low number of axial partial ABGs. Also, some of these ABGs may have been deposited complete, but became partial ABGs through postdepositional taphonomic action, including redeposition, therefore resulting in a large number of mixed body area ABGs. This argument may be particularly relevant to dog and possibly pig ABGs, which also have a low proportion of partial limb deposits.

The trend for a small proportion of partial-horse-andcattle ABGs to be made up of mixed body areas is also constant across the periods. Appendicular elements dominate the partial-horse ABGs for all periods, except in the Romano-British and Medieval samples. In these periods there is a roughly 50:50 split between axial and appendicular element partial ABGs. However, except for the Iron Age, the datasets are very small.

Table 10.6 Species body area proportions for noncomplete ABGs for all periods and regions Species Axial Limb Mixed Cattle (230) 35.2% 52.6% 12.2% S/G (271) 19.6% 50.6% 29.9% Pig (69) 18.8% 33.3% 47.8% Horse (133) 22.6% 60.9% 16.5% Dog (144) 15.3% 29.9% 54.9% Domestic Fowl (21) 9.5% 42.9% 47.6%

More variation is seen in the partial-cattle ABG assemblages, although mixed deposits are uncommon. Appendicular ABGs dominate the Neolithic, Iron Age and Medieval assemblages (Figure 10.12), and no axialelement, partial-cattle are recorded from the Medieval period. Like horse, there is a roughly 50:50 split between axial and appendicular-element partial-cattle ABGs in the 138

Chapter 10. Patterns and Trends fragmentation, but also because it is not necessary for each bone to be processed separately. Therefore some elements such as those from the lower limbs are less likely to have butchery marks present. We can therefore assume that the percentage of ABGs with butchery marks present probably underestimates the true proportion of ABGs that were processed. The presence of a high percentage of partial ABGs would also suggest some form of carcass processing was often taking place. However, the butchery mark information does help us to understand some of the processes which were taking place to create these deposits.

Romano-British sample. This could possibly indicate similar processing treatment, assuming the ABGs are man-made. The Anglo-Saxon and Bronze Age partialcattle show a different pattern with axial-body-area ABGs the most common. The very high percentage of partialaxial ABGs in the Bronze Age sample is due to the deposits at Poundbury (Buckland-Wright 1987). The trend for a high proportion of appendicular bodyelements in sheep/goat ABGs is present in all periods (Figure 10.13). All of the Neolithic partial-sheep/goat ABGs consist of appendicular elements, but data is only available from one deposit. Mixed partial sheep/goat ABGs are present in the Bronze Age, Iron Age and Romano-British samples, but are not present in the early and later Medieval assemblages. However, these assemblages are again very small, consisting of four and two deposits respectively.

Butchery marks are most commonly reported on cattle ABGs, with 35 deposits, which represents 11% of the total cattle ABG assemblage. None of the Neolithic cattle ABGs have butchery marks recorded as present. The only complete cattle ABG with butchery marks comes from the Bronze Age Crab Farm site. The cuts on the head were interpreted as being made when the animal was killed (Locker 1992b) (Table 10.7).

Overall as with the complete/partial data a number of trends appear in the composition of the partial ABG assemblages for some species. However, there is also variability between periods, and firm conclusions are not helped by the presence of some very small assemblages and the lack of suitable data. 10.6

The majority of the cattle ABGs with butchery evidence are from the Iron Age, with 20 cases representing 15% of the Iron Age cattle ABG assemblage. Most were butchered using a knife, which was the tool most commonly used at the time. The majority of the butchery marks are recorded from lower limbs and are associated either with primary or secondary processing. Three cattle ABGs show signs of filleting; two from Lains Farm have filleting as well as skinning marks on some of their limb bones (Coy 1991), and a deposit from Groundwell West has filleting marks on some of its vertebrae (Hambleton 2001). The lack of filleting marks on the other cattle ABGs could suggest they were being deposited with meat still attached, rather than just connective tissue. The Iron Age cattle ABGs from the Hasholme Logboat, Yorkshire (Stallibrass 1987) have butchery marks present indicating either disarticulation or dismemberment. This would correspond with Stallibrass’ (1987) suggestion that they represent the transportation of meat joints.

Butchery

When available, butchery information was recorded for each ABG. The majority of the butchery data comes from the Iron Age and Romano-British southern England assemblages. However, this should be expected as these are the largest datasets. In total, 107 domestic-mammal ABGs have butchery marks recorded as present, which represents only 6% of the 1,669 domestic mammal ABG assemblage. However, there are a number of factors we need to take into account. Firstly, there is the almost universal absence of reports that explicitly report butchery marks are not present. We therefore have to assume that butchery marks will be reported, if present. This seems unlikely, as many reports do not even give basic body part information for ABGs. In addition, some of the assemblages included in this study were published before the reporting of butchery information became commonplace. Such reporting is also reliant on the remains being discussed by a zooarchaeologist, which did not happen for some of the ABGs (mainly from the Danebury environs sites). Finally, we must consider the possibility that any carcass processing may not have left butchery marks on the bones of the ABG. If we examine a non-ABG faunal assemblage such as the one from Greyhound Yard, Dorchester (Maltby 1993), only 26% of the cattle, 7% of the sheep/goat and 11% of the pig bones have butchery marks present, even though the carcasses of these species were intensively processed. The majority of faunal remains recovered from archaeological sites have undergone carcass processing, but relatively few have butchery marks present. This is partly because of

The Romano-British period saw a change in butchery methods and associated technology, especially for cattle. During the Iron Age (and previous prehistoric periods) large animals were probably not hung (Wilson 1996, 32) and butchery was carried out on the ground using a knife (Maltby 1989d). In the Romano-British period, specialist butchers would have been present in towns (Maltby 2007; Rixson 2000, 69), with carcasses commonly hung and butchery often carried out using a cleaver (Seetah 2006). This new style of butchery produced a more visible pattern of dismemberment, with specific joints separated by cleaving the bone. For example, the femur and pelvis were frequently dismembered by cleaving the femoral head. Other traits included the more systematic breakage of limb bones for marrow (Maltby 2007). This change is linked to an increase in the intensity of exploitation of animals, particularly cattle, for meat (Seetah 2005).

139

Investigating Animal Burials; Ritual, Mundane and Beyond

Period Bronze Age

Iron Age

RomanoBritish

Table 10.7 Description of butchery marks recorded from cattle ABGs Type of Site Complete/Partial Area of butchery butchery Middle Farm (Bullock and Allen 1997) Partial Upper front limb Unknown Crab Farm (Locker 1992a) Complete Head Knife Poundbury (BucklandWright 1987) Partial Vertebra Unknown Wright, 1987) Partial Vertebra Unknown Flagstones (Bullock and Allen 1997) Unknown Unknown Unknown Maiden Castle (Armour-Chelu 1991) Partial Head Knife Chelu, 1991) Partial Head Knife Lains Farm (Coy 1991) Partial Lower back limb Knife Upper back limb Knife Partial Lower back limb Knife Partial Lower front limb Knife Partial Lower front limb Knife Partial Upper front limb Knife Lower front limb Knife Partial Lower front limb Knife Viables Farm (Maltby 1982b) Partial Vertebra Knife 1982c) Partial Pelvis Knife Partial Lower front limb Knife Winnall Down (Maltby 1985a) Partial Lower back limb Chop Hasholme Logboat Partial Lower back limb Knife (Stallibrass 1987) Upper front limb Knife Partial Upper back limb Knife

Groundwell West (Hambleton 2001) Old Down Farm (Maltby 1981a) Greyhound Yard (Maltby 1993)

Oakridge well (Maltby 1993)

Early Medieval

Silchester Insula IX (Ingrem 2006) Poundbury (BucklandWright 1987) Cook Street (Bourdillon 1993b) Sussex Street, Winchester (Coy 1984b) Cowdery's Down (Maltby 1983a)

Interpretation Unknown Killing Filleting Filleting No detail given Skinning Skinning Skinning Filleting Skinning Skinning Skinning Filleting Skinning Skinning Disarticulation Disarticulation Skinning Disarticulation Disarticulation Unknown Unknown

Partial Partial Partial

Pelvis Vertebra Ribs

Knife Unknown Knife

Disarticulation Dismemberment Unknown

Partial

Vertebra

Knife

Filleting

Partial Partial Partial Partial Partial

Upper front limb Upper front limb Lower back limb Lower back limb Vertebra

Knife Chop Knife Knife Unknown

Disarticulation Disarticulation Disarticulation Disarticulation Unknown

Partial

Vertebra/ribs Lower front limb

Chop Knife

Dismemberment Disarticulation

Partial

Vertebra

Chop

Disarticulation

Partial

Vertebra

Knife

Filleting

Partial

Vertebra

Chop

Dismemberment

Partial Partial

Unknown Axial

Unknown Unknown

Unknown Unknown

Partial

Upper front limb Head

Knife Knife

Disarticulation Skinning

140

Chapter 10. Patterns and Trends

Period Iron Age

RomanoBritish

Early Medieval High Medieval

Table 10.8 Description of butchery marks recorded from horse ABGs Area of Type of Site Name Complete/Partial butchery butchery Balksbury Camp (Maltby 2001) Partial Head Knife Maiden Castle (Armour-Chelu 1991) Partial Vertebra Knife Ribs Chop Nettlebank Copse (Poole 2000b) Partial Lower back limb Knife Partial Lower back limb Knife Partial Lower front limb Knife New buildings (Poole 2000a) Partial Lower back limb Knife Poundbury pipeline (ArmourChelu 1986) Partial Lower back limb Knife Vicking Way (Birbeck and Moore 2004) Partial Vertebra Knife Winnall Down (Maltby 1985a) Partial Lower back limb Knife Partial Upper front limb Knife Partial Upper front limb Knife Dalton Parlours well (Berg 1990b) Partial Vertebra Knife Ribs Knife Shiptonthorpe (Mainland 2006) Partial Upper front limb Chop Cook Street (Bourdillon 1993b) Partial Vertebra Knife St Georges Road (Bullock and Allen 1997) Complete Upper back limb Chop

Interpretation Disarticulation Disarticulation Unknown Skinning Skinning Skinning Skinning Skinning Disarticulation Disarticulation Unknown Unknown Unknown Unknown Filleting Unknown Disarticulation

likely to result in disarticulation, with the vertebrae often split in two. Therefore the ABGs with butchery on the vertebrae may represent carcasses that were not processed to the same extent as was the norm in the non-ABG assemblage.

Only six cattle ABGs from the Romano-British period have butchery marks present. However, this represents 12% of the assemblage, as only 50 cattle deposits are recorded from this period. In this period more of the cattle ABGs have been butchered with a cleaver, which corresponds with the increased incidence of chop marks found on the non-ABG assemblage (Maltby 2007). All of the butchery marks observed are associated with the disarticulation of the carcass and most of the marks were reported on vertebrae. This corresponds to the composition pattern of cattle ABGs recorded in the Iron Age and early Romano-British periods, as most partial cattle ABGs from southern England consist of axial skeleton elements, mainly from the vertebral column.

Only a small number (5) of early Medieval cattle ABGs have butchery marks recorded, but they do represent 25% of the cattle ABGs from this period. Unfortunately little information is available for these ABGs, but the butchery marks appear to be mainly associated with dismemberment and disarticulation of the carcass. This pattern can be seen throughout the cattle butchery records, and it would appear in all periods that very few cattle ABGs have clear evidence of the meat being removed from the carcass before deposition as an ABG.

In the middle Romano-British period the pattern changes with axial elements not as common (see 6.8). Such a change in composition may be linked to the change in butchery style (Morris 2008). The butchery methods used in the Iron Age are more likely to leave connective tissue attached to the bone, maintaining articulation. By comparison, the Romanised style of butchery is more

The horse ABGs show a similar pattern, possibly because they are animals of similar size (Table 10.8). The majority of the butchery marks are recorded on the lower limbs, especially in the Iron Age, and appear to be associated either with skinning or disarticulation of the 141

Investigating Animal Burials; Ritual, Mundane and Beyond

Period Bronze Age Iron Age

Table 10.9 Description of butchery marks recorded from sheep/goat ABGs Complete Area of Type of Site Name / Partial butchery butchery Interpretation Wilsford Shaft (Grigson 1989) Partial Vertebra Unknown Unknown Winterbourne Stoke barrow 44 (Green and Rollo-Smith 1984) Partial Upper front limb Knife Dismemberment Flagstones Enclosure (Bullock and Allen 1997) Complete Unknown Unknown Unknown Maiden Castle (Armour-Chelu Partial Head Chop Filleting Vertebra Chop Filleting 1991) Ribs Knife Filleting Lower back limb Knife Skinning Partial Nettlebank Copse (Poole 2000b) Old Down Farm (Maltby 1981a) Owslebury (Maltby 1987c)

Head Vertebra Upper back limb Upper back limb Upper front limb Vertebra Vertebra Upper back limb Lower back limb Lower front limb Pelvis Upper front limb Upper front limb Vertebra Head Vertebra

Chop Chop Unknown Knife Knife Knife Knife Knife Knife Knife Knife Knife Knife Knife Knife Knife

Filleting Disarticulation Disarticulation Skinning Skinning Disarticulation Disarticulation Disarticulation Disarticulation Disarticulation Dismemberment Dismemberment Disarticulation Disarticulation Disarticulation Unknown

Lower back limb Upper back limb Lower back limb Lower back limb Pelvis Upper front limb Vertebra Lower back limb Vertebra Head Vertebra Head Upper back limb Upper front limb Head

Knife Knife Knife Knife Chop Knife Knife Chop Knife Knife Knife Knife Knife Knife Knife

Skinning Disarticulation Skinning Skinning Dismemberment Disarticulation Skinning Disarticulation Disarticulation Disarticulation Disarticulation Disarticulation Disarticulation Disarticulation Disarticulation

Partial

Lower back limb

Knife

Dismemberment

Partial Complete

Vertebra Upper front limb Vertebra Upper back limb Upper front limb Upper back limb Pelvis Vertebra

Chop Knife Knife Knife Knife Knife Knife Knife

Dismemberment Unknown Unknown Dismemberment Dismemberment Disarticulation Disarticulation Disarticulation

Vertebra

Chop

Dismemberment

Complete Partial Partial Partial Partial Partial Complete

Quarry Field (Clark 2002)

Winklebury Camp (Jones 1977)

Winnall Down (Maltby 1985a) RomanoBritish

Castle Copse Roman Villa (Payne 1997) Greyhound Yard. Dorchester (Maltby 1993)

Partial Partial Partial Partial Partial Partial Partial Partial Partial Partial Partial Complete

Little Somborne (Maltby 1978b) Oakridge well (Maltby 1993) Parlington Hollins (Richardson 2001) Poundbury cemetery (Buckland-Wright 1993) Poundbury settlement (Buckland-Wright 1987)

Partial

Partial Shiptonthorpe (Mainland 2006) Early Medieval

Partial Wharram (Pinter-Bellows 1992)

High Medieval

Sussex Street (Coy 1984b)

Partial

142

Chapter 10. Patterns and Trends

Period Iron Age RomanoBritish

Early Medieval

Period Late Bronze Age Early Iron Age

Middle Iron Age

Late Iron Age Early RomanoBritish Middle RomanoBritish Late RomanoBritish High Medieval Late Medieval

Table 10.10 Description of butchery marks recorded from pig ABGs Complete/ Area of Type of Site Name Partial butchery butchery Interpretation Garton Station (Legge 1991a) Partial Upper front limb Knife Disarticulation Kirkburn (Legge 1991a) Partial Upper front limb Knife Disarticulation Alington Avenue (Maltby 2002a) Partial Upper front limb Chop Dismemberment Partial Upper back limb Chop Dismemberment Manor Farm (Sykes 2003) Partial Upper front limb Chop Disarticulation Parlington Hollins (Richardson 2001) Partial Upper front limb Chop Unknown Table 10.11 Description of butchery marks recorded from dog ABGs Complete/ Area of Site Partial butchery Type Potterne (Locker 2000) Partial Upper front limb Knife Nettlebank Copse (Poole 2000b) Complete Upper back limb Knife Complete Upper front limb Knife New Buildings (Poole Vertebra Knife 2000a) Upper back limb Knife Houghton Down (Poole 2000c) Partial Pelvis Knife Nettlebank Copse (Poole 2000b) Complete Upper front limb Knife Partial Axial Knife Old Down Farm Upper front limb Knife (Maltby 1981a) Upper back limb Knife Balksbury Camp Partial Pelvis Knife (Maltby 2001) Complete Upper front limb Knife Partial Upper front limb Knife Partial Pelvis Knife Partial Lower back limb Knife Old Down Farm (Maltby 1981a) Partial Upper front limb Knife Flagstones (Bullock and Allen 1997) Complete Lower back limb Knife Tolpuddle Ball (Hamilton-Dyer 1999b) Partial Lower front limb Knife Quarry Field (Clark Partial Upper back limb Knife 2002) Head Knife Complete Lower back limb Knife Owslebury (Maltby Upper back limb Knife 1987c) Upper front limb Knife 9 Blake Street, York (O'Connor 1987) Partial Lower back limb Knife Silchester Insula IX (Clark 2006) Emwell Street (Smith 1997) Ousefleet Property, Hull (Berg 1987)

Partial

Head Vertebra Lower front limb Lower back limb Upper back limb Pelvis

Partial Partial

143

Chop Chop Unknown Unknown Chop Chop

Authors Interpretation Skinning Skinning Filleting Filleting Filleting Unknown Skinning Disarticulation Disarticulation Disarticulation Dismemberment Skinning Dismemberment Dismemberment Skinning Dismemberment Skinning Skinning Filleting Skinning Skinning Skinning Skinning Skinning Killing Disarticulation Skinning Skinning Disarticulation Disarticulation

Investigating Animal Burials; Ritual, Mundane and Beyond carcass. The majority of the horse ABGs with butchery marks present are from the Iron Age, with 11 examples representing 11.3% of the species assemblage from this period. Only one horse ABG from Shiptonthorpe, Yorkshire (Mainland 2006) displays marks associated with filleting.

two of the Romano-British pig ABGs from Alington Avenue (Maltby 2002a) were also recovered from funerary contexts in association with human remains. The pig butchery marks also show a clear division between knife marks on Iron Age ABGs and ‘chop’ marks on Romano-British and later ABGs.

A larger number (31) of sheep/goat ABGs have butchery marks present but this only represents 7% of the total sheep/goat assemblage (Table 10.9). The majority (17) of the sheep/goat ABGs with butchery are from the Iron Age contexts, meaning 6% from this period have butchery marks. It therefore appears that the larger cattle and horse ABGs are more likely to have butchery marks present. This may be simply due to their larger size and a similar pattern is often seen in the non-ABG faunal assemblage.

Twenty of the dog ABGs have butchery marks present, which represents 3.3% of the total assemblage. Thirteen are from Iron Age contexts with a large proportion present on appendicular elements. The majority of the marks appear to be associated with primary and secondary butchery practices, although deposits from the Iron Age New Buildings (Poole 2000a) and the RomanoBritish Quarry Field (Clark 2002) sites have filleting marks present (Table 10.11). The majority of the Romano-British butchery marks have been interpreted by the reporting authors to be the result of skinning. However, a larger proportion of the marks on Iron Age dog ABGs suggest that dismemberment of the carcass and filleting also took place. This could indicate that dog meat may have been sometimes utilised in some way, although we must consider that of the 133 dog ABGs recorded from the Iron Age, most do not display any evidence of butchery.

The majority of the sheep/goat ABGs with butchery marks are partial skeletons, although five are complete. Unfortunately, detailed information is not given for each one. It is reported that complete ABGs from the Iron Age site of Quarry Field (Clark 2002) and the Romano-British settlement at Little Somborne (Maltby 1978b) have dismemberment butchery marks. This indicates that the animals were processed to some extent, but the bones of the carcasses were still deposited in association. Both display butchery marks on axial and appendicular elements.

Evidence of carcass processing was only present on two of the 141 dog ABGs recorded from Romano-British towns. A partial dog ABG from Silchester Insula IX, pit 2674, has cleaver marks to the skull and trunk, indicating a possible killing blow and subsequent carcass dismemberment (Clark 2006). The other example, a partial deposit from 9 Blake Street, York, pit 8186, has knife marks on the lower back limbs possibly indicating skinning (O'Connor 1987). This low incidence of butchery observations adds weight to the suggestion that the majority of partial dog ABGs were originally deposited as complete skeletons.

As with the other species discussed above, the majority of the butchery marks are from primary or secondary butchery. Only one partial ABG from Maiden Castle (Armour-Chelu 1991) displays marks associated withfilleting. Again, this may indicate that these ABGs were deposited with meat still attached to the bones. A slightly higher proportion, 9.3% (10), of the RomanoBritish sheep/goat assemblage has butchery marks. It is also noticeable that marks on vertebrae are more common in the Romano-British assemblage compared with the Iron Age sample. This may, as in the case of cattle, reflect a change in processing methods. Such a change is also evident in the non-ABG faunal assemblage, and would indicate that animals deposited as ABGs had undergone at least some of the processes seen on the ‘normal’ faunal material. Again, the majority of the Romano-British butchered sheep/goat ABGs display marks associated with dismemberment of the carcass.

10.7

Animal health; pathology

Only 48 (2%) of the 2,062 ABGs have recorded evidence of pathological changes. The majority were either horse or dog from the Iron Age and Romano-British periods. It is assumed, but cannot be proven, that the remainder of the ABG assemblage did not show evidence of pathology. This would therefore suggest that the vast majority of animals deposited as ABGs were healthy, However, one of the main problems in investigating the health of a population, that also exists in the study of human palaeopathology, is that the majority of diseases do not result in skeletal changes (Roberts and Cox 2003; Roberts and Manchester 1997). The osteological paradox (Wood et al. 1992) also applies to animal remains, in that animals that have reached the point that a disease is causing morphological changes to the skeleton would be some of the healthiest individuals. The majority of individuals would not survive a disease to such a point.

Only two Medieval sheep/goat ABGs have butchery marks recorded and both appear to have undergone similar processes to those already described. A small number (6) of the pig ABGs have butchery marks present, representing only 3% of the species assemblage. All the pig ABGs consist of partial skeletons and all the butchery marks are recorded on upper limb bones and are associated with carcass dismemberment (Table 10.10). The two Iron Age pig ABGs with butchery marks are both from Arras culture funerary sites and were recovered in association with human remains. Similarly, 144

Chapter 10. Patterns and Trends the head and broken limbs, possibly indicating the brutal nature of a dog’s life in a Romano-British town (Clark 2006, 18). However, all the ABGs with trauma have healed fractures, indicating some care of the animals may have taken place.

With animals we also have to take into account the human factor, in that animals may be culled when soft tissue changes are evident, but before skeletal ones occur. Therefore, rarely will the cause of death be evident. At present animal palaeopathology could be viewed as an undeveloped area of zooarchaeology, with research sporadic and case-study based, although recent work has done much to improve this (see Bendrey 2007; Thomas and Mainland 2005; Thomas and Mikliková 2008; Vann and Thomas 2006).

10.8

Depositional context

Throughout this study the site and feature type ABGs were recovered from have been discussed. Although the ‘site types’ are academic constructs, a number of patterns are evident.

The most common pathology recorded on ABGs is degenerative joint disease (DJD), which includes osteoarthritis, ankolysis and exostoses changes to joints, as well as more species-specific diseases such as spavin, ringbone and navicular disease (Baker and Brothwell 1980). In total, 24 (50%) of the cases consisted of DJD. In all periods DJD was most frequently observed on horse ABGs (Table 10.12). This corresponds to the ageing data for horse ABGs, which indicate that the majority are from adult animals, as DJD is a degenerative condition that develops in association with age, and possibly work-induced stress involving traction and riding (Baker and Brothwell 1980).

The majority of the previous literature regarding ABGs from the Neolithic and Bronze Age, discuss their presence at funerary monuments in association with human remains. However, even taking into account the large sample from Windmill Hill, in this study they are only present on three Neolithic funerary monuments, accounting for only six of the 55 ABGs from this period (see 3.4). It is notable, however, that all the ABGs from these sites are of cattle. All the other cattle ABGs recorded are from causewayed enclosures. However, these deposits may still be associated in some way with human funerary practices, as human remains are also recovered from causeway enclosure ditches, some of which were possibly used for excarnation (M Smith 2006). However, none of the ABGs were recovered in direct association (as defined by this study) with human remains. Ray and Thomas (2003) have suggested that a close relationship existed between cattle and people in the

The other main pathology observed in the ABG assemblage is trauma. This mainly takes the form of healed fractures and can sometimes occur in association with DJD. This type of trauma is most often recorded on dog ABGs, especially in the Romano-British sample, the majority from the Silchester Insula IX excavations (Clark 2006). A number of the ABGs showed signs of blows to

Table 10.12 Summary of the pathologies recorded on ABGs per species and period. DJD= degenerative joint disease Domestic Wild Period Pathology Cattle S/G Horse Dog Cat bird bird Neolithic Trauma 1 Bronze Age DJD 1 1 Trauma 1 Periodontal 1 Iron Age DJD 1 6 1 DJD/Trauma 1 1 Trauma 3 Non-specific infectious 2 Periodontal 1 RomanoDJD 1 3 1 British DJD/Trauma 1 Trauma 1 6 1 Metabolic 1 Periodontal 1 Early DJD 2 Medieval Trauma 3 Later DJD 1 2 Medieval DJD/Trauma 1 Trauma 1 Periodontal 1 Total 3 4 16 21 1 1 2 145

Investigating Animal Burials; Ritual, Mundane and Beyond There appears to be no consistent patterning in the early Medieval datasets, although the southern England and Yorkshire sample sizes are small. For the later Medieval period the majority of the Yorkshire assemblage comes from town sites, due to the large number of published and well recorded excavations at York. The southern England assemblage shows a different pattern with the majority of the ABGs recorded either from rural settlements or manorial sites. However, the data are biased by the large Faccombe Netherton assemblage.

early Neolithic, with herds of cattle perceived as communities that mirrored those of humans. This therefore led to the similar treatment of cattle remains, as well as the slaughter and eating of cows ‘in concert with other highly ritualised activities’. This does not, however, explain the lack of similarity in the elements of cattle and humans deposited or the presence of ABGs from other species. Only a small number of Bronze Age ABGs have been recovered from funerary contexts, and the majority are from late Bronze Age settlement sites. In fact, of the 46 faunal reports examined from Bronze Age barrows, only three had ABGs present. The presence of ABGs on ‘domestic’ settlement sites is a continuing trend from the late Bronze Age onwards. Only a small proportion of deposits in later periods are from funerary contexts. However, the ABGs from these sites do differ to those recovered from settlements (see below).

Overall, the majority of ABGs recorded in this study were recovered from features defined by the reporting authors as pits. From the Iron Age onwards pits are the dominant feature type. This is not case in the Neolithic and Bronze Age sample, in which a large proportion of the assemblage is from ditch features (Table 10.13). ABGs from Iron Age and Romano-British sites are sometimes recovered from ditches. On most sites the assemblage was not large enough to investigate if intrasite differences between features were present. However, the analysis of the Owslebury data did indicate that sheep/goat, dogs and domestic fowl ABGs were mainly present in pits, whereas those of larger mammals such as cattle and horse were usually present in the ditch fills. This pattern is similar to the one seen in the non-ABG faunal assemblages from some Iron Age and RomanoBritish sites. This may indicate remains of species that were deposited as ABGs and non-ABGs were treated in a similar fashion. Unfortunately, it was not possible to look at intra-site differences on other sites, but it is an area worthy of further research.

No clear patterns are present in the types of ABGs deposited on different ‘type sites’, with the exception of funerary contexts, although these are often in association with domestic settings. However, differences are apparent in the Romano-British period from southern England between what could be defined as ‘rural’ and ‘town’ sites. Dog ABGs are the most commonly recovered from Romano-British towns in southern England from the early part of the period onwards. This pattern was not apparent on rural settlements, and the early RomanoBritish ABG assemblages from these sites are very similar to the Iron Age ones (see 6.7.1). It is not until the later Romano-British period that the assemblages from town and rural sites show the same pattern. This may indicate a delay in the adoption of Roman ideas on ‘native’ rural settlements, and the persistence of traditional practices amongst ‘un-romanised’ populations.

Feature Bank Ditch Foundation Grave Gulley layer Midden Oven Pit Post-hole Quarry Shaft/Well Other Unknown Total

All the Bronze Age ABGs recorded from shaft/well deposits are from Wilsford Shaft (see 3.7.2). A large proportion of the Romano-British assemblage was also recovered from shaft/well deposits. The majority of these are from Greyhound Yard, Dorchester or Oakridge Well (see 6.5 and 6.6).

Table 10.13 Summary of the number of ABGs per feature type per period Bronze Iron RomanoEarly Later Neolithic Age Age British Medieval Medieval Total 1 1 34 15 41 76 10 10 186 4 3 7 1 1 23 76 2 1 104 4 25 29 2 6 3 26 11 12 60 4 1 2 7 1 1 17 17 688 422 62 109 1315 1 4 8 1 4 18 4 12 16 16 252 5 8 281 1 5 6 4 8 5 1 13 31 55 61 784 908 92 162 2062

146

Percentage Total 0.05% 9.02% 0.34% 5.04% 1.41% 2.91% 0.34% 0.05% 63.77% 0.87% 0.78% 13.63% 0.29% 1.50%

Chapter 10. Patterns and Trends Table 10.14 Number of ABGs per species recovered from grave contexts. The number in brackets indicates the number of ABGs from graves but not in association with human remains Bronze Iron RomanoEarly Later Species Neolithic Age Age British Medieval Medieval Total Cattle 1 3 3 S/G 6 22 28 Pig 15 (1) 4 29 Horse 2 3 1 (1) 6 Dog 9 (3) 9 Domestic Fowl 2 33 (10) 35 Fox 1 1 ULM 2 (1) 2 Table 10.15 Number of ABGs per species recorded in association with human remains from feature other than graves. The number in brackets indicates the number of ABGs that are associated with complete human remains Bronze Iron RomanoEarly Feature Species Neolithic Age Age British Medieval Ditch Cattle 1 S/G 1 (1) Dog 1 (1) Pit Cattle 6 (5) S/G 3 (3) 1 Pig 2 (1) 1 (1) Horse 4 (3) Dog 4 (2) Well Cattle 3 (3) ABGs particularly from Trentholme Drive, York (see 7.5). A large proportion of the ABGs recovered from this site are from grave contexts, but do not appear to be in association with human remains, most probably because of disturbance to the human inhumations.

Although the majority of ABGs were recovered from pits, this study has shown the need to look for these deposits elsewhere. We must also consider that pits are the most common type of feature excavated on an archaeological site, due to the majority of archaeological excavations concentration on the interior of settlements. Therefore, we should perhaps not be surprised that most of the ABGs have been found within pit contexts. 10.9

A clear contrast in the species deposited in grave contexts can be seen between the Iron Age and Romano-British periods. The majority of ABGs deposited in Iron Age inhumations from both southern England and Yorkshire are from pigs. In the Romano-British period, this changes with sheep/goat prevalent in the early part of the period, particularly in southern England. However, during the later Romano-British period domestic fowl become the most common species. The deposition of sheep/goat ABGs in Roman graves appears to be a trend from around the Dorchester region, with pig and domestic fowl the most common species recovered from funerary contexts in other regions (Philpott 1991). It has been suggested that there might be a connection between domestic fowl and female graves for continental sites (Wahl and Kokabi 1987), but this study and others have found no such link (Lauwerier 1993).

Human remains

Throughout the study associations between ABGs and other material groups have been recorded. The majority of associations recorded are from the Iron Age and Romano-British periods and have been in the form of multi-ABGs, particularly of dogs in the Romano-British period. However, a number of associations have been recorded with other material types the majority of which are human remains. As would be expected, most associations between ABGs and human remains are found in grave contexts, the majority from the Romano-British period (Table 10.14). Most of the ABGs in these contexts are partial and are often made up of major meat-bearing long-bone elements. The exceptions are the remains of domestic fowl which appear to have been deposited as complete

This changing pattern may be linked to changes in culinary tastes. All the ABGs deposited in grave contexts are interpreted as offerings, and may well have

147

Investigating Animal Burials; Ritual, Mundane and Beyond associations, it has shown that the nature of ABGs is much more complex than the previous literature would suggest. By additionally drawing upon data regarding the archaeological context, this study has also shown the great variety of feature types ABGs are deposited in. This can start to move us away from the purely pit-based models advocated in the previous literature (for example Cunliffe 1992; Wait 1985).

represented the deposition of a food item to be associated with the deceased. A number of ABGs have also been recorded in association with human remains outside of grave contexts. The majority of these are from the Iron Age (Table 10.15), and a number of authors have already discussed such an association (see 1.2.8). The majority of the human remains in association with ABGs consisted of complete skeletons. 10.10

One of the most important objectives of this study was to examine ABGs from periods beyond the Iron Age. In doing so this research has shown we must also move away from ideas that this deposit type is normally found on Iron Age sites. For both southern England and Yorkshire, Romano-British sites produced the largest number of ABGs. Also, more examples were recorded from Medieval sites in Yorkshire than Iron Age ones. Although often present and an important part of this study, ABGs are much more than an Iron Age phenomenon.

Summary

A number of patterns in the nature of the ABG assemblage have been examined within this and the preceding chapters. It was therefore important to investigate whether the observed patterns are real, or due to biasing factors. Examination of the sizes of the ABG and non-ABG assemblages from the same sites has shown that the relationship between the two is not significant. Therefore, other factors must influence the size of the ABG assemblage from individual sites. However, the composition can be affected by taphonomic actions such as bioturbation. Such factors can be taken into account but it requires the examination of the specific features ABGs are recovered from. Throughout this study one issue that has affected our ability to examine the nature of the ABG assemblage is ourselves, the archaeologist. Studies of this nature are reliant upon data being available and the quantification methods utilised by zooarchaeologists reporting these deposits. However, detailed information concerning ABGs is often not present in publications. Examining the published excavation records has shown that complete deposits are often reported. However, it is not until the 1980’s that large numbers of partial ABGs start to be recorded. This is due to the quantification methods used and an increase in literature drawing attention to them. As the majority of the data utilised in this study comes from the 1980’s onwards, in part due to the rise of developerfunded archaeology, it would appear that patterns within the ABG assemblage are archaeological in nature. Overall, 2,062 ABGs were recorded for this study, the majority of which come from sites within the southern England region. Although the southern England assemblage is much larger than that from Yorkshire, this does not affect the significance of the results from the Yorkshire region. The biggest discernable difference between the two regions is the proportion of species deposited as ABGs. These differences validate one of the objects of this study, which was to move beyond the ‘Wessex’-based model. By comparing and contrasting the results from these two regions, this study has shown that regional differences do exist in the ABG assemblage. This study has gone further and looked in depth at the nature of these assemblages. By investigating aspects such as composition, butchery marks, pathology and 148

Chapter 11. Assigned Meaning 11.1

Introduction

This study has to this point been mainly concerned with describing and comparing the nature and context of ABGs from different periods and study areas. This has shown that the nature of these deposits varies within and between periods as well as regions. The emphasis on description has been purposeful, to quote Bruno Latour (2004, 63); ‘For every hundred books of commentaries, arguments, glossaries, there is only one of description’ It is however, time to move beyond description. As archaeologists we need to describe the nature of the evidence we uncover, but it is equally vital that we understand the meaning of such evidence. Archaeology can be viewed as the understanding of past human actions. The materials we uncover, in this case ABGs, are just the tools we use to aid us in this difficult endeavour. At the beginning of this book, previous literature regarding ABGs was discussed. That literature showed a trend in recent years towards a more ‘ritualistic’ interpretation of these deposits from prehistoric, RomanoBritish and most recently Anglo-Saxon contexts (see 1.2.9). This trend is also visible, especially for the prehistoric and Romano-British periods, in the animal bone reports examined. The previous chapters have provided a solid base of descriptive data, by which we can examine the interpretations archaeologists currently use for these deposits. In doing this we can attempt to answer one of the principal questions set by this study, ‘are the current interpretations of ABGs valid?’ To answer this question, we must examine the reasons why certain interpretations are given to these deposits, why the majority of current authors view prehistoric and Romano-British ABGs as ‘ritual’ deposits and later Medieval ABGs as the results of ‘functional’ activity. However, we must also investigate the concepts upon which archaeologists and zooarchaeologists have based their interpretations. The previous chapters have shown that a ritual/functional dichotomy of ABG interpretation exists. This relationship is not unique to ABGs, it affects many aspects of archaeology and other disciplines. To investigate these matters we could look at the individual interpretations offered. However, such an approach does not reflect the variability of the interpretations offered for similar ABG deposits. Throughout this study it has become apparent that certain types of ABGs (e.g. complete/partial dogs) attract specific interpretations. Therefore after discussing the changing nature of the interpretations and the important concept of structured deposition, the following discussion is structured around the different types of ABGs

encountered in the archaeological record; complete and partial ‘food’ domestic mammals, dogs and cats, wild animals and domestic birds. It then goes on to discuss the levels of interpretations offered and possible problems. 11.2

A pantheon of interpretations

Although there are a number of possible biasing effects in action upon ABG assemblages, the dataset appears to be robust enough for us to understand how the majority were deposited and created. Differences in the assemblages appear to be mainly due to human action. This being the case, we need to examine what those actions were and importantly the meanings behind them. As well as information on the nature of the ABG assemblage, the interpretation placed upon it by the original authors was also recorded (see 1.5.2). Many different interpretations were recorded ranging from ‘culling’ to ‘offerings’ (Table 11.1). A ‘functional’ interpretation was recorded when the original author suggested a number of possible ‘functional’ interpretations rather than one specific explanation. Also, a ritual/sacrifice interpretation was recorded when the original author specifically alluded to a form of ritual activity, often mentioning the possibility of sacrifice. The interpretations given by authors fall into four overarching categories: functional, ritual, mixed (where both a functional and a ritual explanation is offered) and unknown (where no interpretation is given). Published literature regarding ritual interpretations of ABGs has until recently been centred on prehistoric examples, with Iron Age interpretations along ritual lines drawing on previous Neolithic and Bronze Age work. Interpretations of ABGs in this dataset appear to be largely dependant upon the period to which they belong. Over 70% of ABGs from a Neolithic context have been interpreted as being the result of a ritual activity. The number of ritual interpretations steadily decreases from period to period (Figure 11.1). By contrast, only 13% of Neolithic ABGs have been interpreted in a functional way. As the proportion of ritual interpretations decreases, functional explanations increase. Functional explanations are given to 63% of the high and late Medieval period ABGs, with only 4% given a ritual interpretation. Also, the proportion of ABGs not given any interpretation is at its highest in the Medieval periods, although this is also linked to publication date. The Iron Age is currently the crossover point where a similar number of functional and ritual interpretations have been offered. This is due to a number of factors. The Iron Age is the latest prehistoric period. Until recently there existed an academic division between those studying the prehistoric and the historic periods, especially regarding interpretation of ritual. The Iron Age

Investigating Animal Burials; Ritual, Mundane and Beyond

Table 11.1 Summary of how ABGs have been interpreted by reporting authors by period

Period of deposit Functional Culling Disease Fall Natural death Waste Mixed Ritual/Sacrifice Feast Foundation Offering Offering Unknown

Neolithic

7 6 24 14 1 3

Bronze Age 9

1 5 6 28

Iron Age 8 17 1 14 82 221 94 237

RomanoBritish 8 238 9 32 95 141 115 64

1 11

1 25 84

2 68 136

Early Medieval 1 10 1 6 33 4 1

High and Late Medieval 8 7 19 12 56 1

6 2 34

53

Figure 11.1 Percentages of different interpretation categories by period is also the first period from which the archaeological record is dominated by settlement evidence. The majority of the Neolithic and early to middle Bronze Age sites are of a funerary or possibly ceremonial nature. Because of this, ritual interpretations were first suggested for ABGs from the Neolithic and Bronze Age. It was not until Grant’s and Hill’s publications that Iron Age ABGs started to be mainly interpreted along ritual lines.

is probably due to the nature of the archaeology the ABGs are associated with. The Iron Age data show a clear build-up of momentum in ritual interpretations. It was not until the 1960’s that ritual interpretations started being applied to ABGs from the Iron Age, and then it was only one pig limb found in association which human remains at Hod Hill (Bunting et al. 1968). The proportion of ritual interpretations for Iron Age ABGs decreased in the 1970’s with only a couple from round houses at Garton and Wetwang Slack (Noddle 1979) interpreted as either a foundation offering or just a ‘ritual’ deposit. The number of Iron Age sites with ABGs interpreted as ritual deposits increased in the 1980’s and continued to do so until the current decade during which analyses of ABGs from 18 (62%) sites have given ABGs ritual interpretations (Figure 11.2).

This can clearly be seen when the interpretations given are compared with the decade of publication. For this analysis the number of sites, not the number of ABGs, has been used. This has been done to ensure that large ABG assemblages such as Danebury do not skew the results. The data indicate that ritual interpretations have been given to Neolithic and Bronze Age ABGs in all the decades this study draws published literature from. This

The Romano-British data show how ritual interpretations of ABGs have now spread into ‘historic’ periods. A 150

Chapter 11. Assigned Meaning number of ritual interpretations were given to ABGs from Romano-British sites in the 1940s, 1960s and 1970s. However, these ABGs were all from funerary contexts. Deposits from a small proportion of sites in the 1980s and 1990s were also given ritual interpretations, but again they were mainly from graves. However, so far during this current decade ABGs have been given a ritual interpretation from 13 (65%) of the sites published, and an interpretation has been offered for every ABG (Figure 11.3). The sharp rise in ritual interpretations during this decade is probably due to the influence of Iron Age interpretations of ABGs and the developments of TRAC (Theoretical Roman Archaeology Conference) (e.g. Clarke 1997; Fulford 2001; Richardson 1997; Woodward and Woodward 2004). There is also a move towards ritual interpretations for ABGs from early Medieval sites, influenced by recent work such as Hamerow (2006). Currently only a very small number of ritual interpretations for ABGs from the

late Medieval period have been given, most associated with building foundations. The interpretations show the current dichotomy that exists between ritual and functional interpretations of ABGs, as well as the preconceived ideas that exist between archaeological periods and the current postprocessual paradigm. 11.3

Structured deposition

Before looking in depth at the interpretations given to ABGs, the interlinked concept of structured deposition should be discussed. This term was first brought to archaeologists’ attention by Richards and Thomas (1984) in their description of deposits found on Neolithic sites, in particular those from the southern circle within Durrington Walls, which had otherwise been described as non-domestic, symbolic or unusual.

Figure 11.2 Percentages of different interpretations given to ABGs from Iron Age sites per decade. Totals in brackets are of number of sites (not total ABGs)

Figure 11.3 Percentages of different interpretations given to ABGs from Romano-British sites per decade. Totals in brackets are of number of sites (not total ABGs) and no 1950’s sites were present 151

Investigating Animal Burials; Ritual, Mundane and Beyond The same calculations have also been made for some large excavations utilised in this study and they show a similar pattern (Table 11.2). Also, if we carry out similar calculations for ABGs, we see that the deposition of ABGs could be considered a rare event. We must also take into account that these calculations are generalised and do not take into account the deposition of multiple ABGs. For example, the overall figures indicate that one ABG was deposited at Greyhound Yard roughly every two years. However, of the 163 ABGs, 119 are from 26 multiple ABG deposits (see 6.9). If we take this into account, there are 70 ABG deposit events at Greyhound Yard, which averages out at one every nine to ten years in the features investigated.

They drew upon work regarding the symbolic nature of material culture, in particular Turner (1967, 20), who suggested that the structure and properties of a symbol become those of a dynamic entity with certain contexts of action. Richards and Thomas (1984) emphasised that the degree of formality for the deposition of artefacts was a defining characteristic of a ritual deposit. Hill (1995, 95) argued that Iron Age ABGs are a result of structured deposition. However, he makes the important distinction that structured deposition is not the same as ritual, suggesting that all it shows is that the deposits contain well preserved material, whatever the origin. He suggests that; ‘All daily refuse maintenance strategies will be structured through deep-rooted cultural norms. It will be structured deposition, even if such patterning will be quickly broken down’ (Hill 1995, 96).

Therefore, ABGs probably represent rare events, but the same could be said for the survival of any archaeological material. This is why zooarchaeologists have long been concerned with the biasing effects of taphonomic processes. What this study shows is that in regards to structured deposition, Hill (1995) is correct. ABGs from all periods are structured deposits, as we could argue all archaeological remains are. Few would argue that the large accumulations of specific types of bone elements, related to carcass processes in Romano-British towns (Levitan 1989; Luff 1993; Maltby 1993; 2007) or Medieval and post-Medieval towns (Armitage 1978; Dobney et al. 1996; O'Connor 1984b), are structured deposits (i.e. subject to careful and deliberate selection) but they are not interpreted as ritual depositions. Although Hill (1995) clearly separated the two concepts of structured deposits and ritual, this appears to have been widely missed. As its inception was related to the identification of ritual deposits, structured deposition and ritual have been and continue to be, linked in the archaeological literature (for example see, Batt and Dockrill 1998; Chapman 2000; Pollard 1995; 2001; Stoddart 2002; Walker 2002).

In essence Hill (1995) is suggesting that the majority of archaeological material recovered from prehistoric sites, is the result of structured deposition. The material is the direct result of human practice in the past. This is best shown by the limited amount of data we have available. Using the simple calculations of the number of pits recorded from an excavation, divided by the probable duration of the site, Hill (1995, 1-2) showed that the material deposited within pits accounted for only a small fraction of the waste produced at Danebury, Gussage and Winnall Down. For example, dividing the overall number of pits by the duration of the site, he estimated that only one pit was filled every five years at Winnall Down (Table 11.2). Hill (1988, 34) had previously shown that only approximately 100 identified bone fragments were deposited a year at Danebury, assuming it had a permanent population of between 250-500 people and ignoring any destruction of bones from taphonomic processes.

Table 11.2 Number of pits and ABGs from Iron Age and Romano-British sites, showing the number of pits open at any one time, the number of ABGs deposited per year.* indicates estimated number of pits. ** the estimated Owslebury counts include ditches as the majority of ABGs were recovered from them. After (Hill 1995, 3) Probable Probable Probable number of number of duration of No. of pits filled No. of ABGs per Period Sites site (yrs) pits per year ABGs year Iron Age Balksbury Camp 500 134 152 0.2 0.3 Iron Age Danebury 450 5000* 102 11.1 0.22 Iron Age Gussage 650 381 7 0.6 0.01 Iron Age Winnall Down 500 110 49 0.26 0.098 Late Iron Age to late RomanoBritish Owslebury 500 70** 187 0.14 0.37 Greyhound Yard, Romano-British Dorchester 400 96* 163 0.24 0.4 Romano-British

Portchester Castle

400

83

152

0.2

44

0.1

Chapter 11. Assigned Meaning (offering and sacrifice). The interpretation offered is often linked to the archaeological paradigm prevalent at the time.

Such assumptions are clearly unhelpful. The concept of structured deposition can help us identify archaeological material that has been produced by culturally specific practices, but it cannot assign meaning. It merely shows us where to look for the dominant structural principles of a society. Perhaps the problem is that the majority of archaeological material is the result of structured deposition. If this is case, then the term may be defunct 11.4

In the majority of cases, it is impossible to assign the cause of death to animal remains. We can infer that the vast majority of faunal remains represent animals killed for human consumption, but rarely can we find direct evidence of their slaughter. The assumption that the ‘normal’ fragmented faunal material results from animals slaughtered for human consumption leaves zooarchaeologists with a problem when complete ABGs of common ‘food’ animals (cattle, sheep/goat and pig) are recovered. The completeness of such deposits suggests little or no processing took place, leaving zooarchaeologists to speculate about how the animals died. Only on rare occasions is it possible to ascertain cause of death, such as the cattle breach-birth at Gussage All Saints (Harcourt 1979b) (see 4.5).

Domestic ‘food’ mammals

In this section we consider the interpretations offered to explain ABGs for cattle, S/G, pig and horse. These are the most common domestic mammals in the total faunal assemblage, and the species, which have been shown to have commonly supplied primary products for human consumption. Although it could be argued that dog and cat were also at times consumed, the evidence for this is very limited. Horse is included in this category because although not often eaten, there is evidence of horse butchery from most periods. These species have also been discussed together because of the similarities in their ABG composition. The major difference in the interpretation of these species is between complete and partial ABGs, and therefore these categories of ABGs are discussed separately.

A number of authors have interpreted complete domestic species ABGs as the deposition of animals that died from disease or trauma. For example, Buckland-Wright (1987) suggests that seven complete sheep/goat ABGs from a single late Iron Age/early Romano-British pit at Poundbury had died as the result of haxia as; ‘It would be impossible to fit seven healthy sheep into a pit of this size’(Buckland-Wright 1987, 131).

11.4.1 Complete ABGs; unfit for consumption, or sacrifices?

The sheep/goat from Poundbury also showed no evidence of having been processed. This is a common factor in the interpretation of ABGs as the result of disease, based on the assumption that diseased animals would not have been eaten. However, our modern perspective about what is fit to eat may be biasing our view and further work is needed on this topic.

As shown previously, complete ABGs are present in the archaeological record from most periods and regions, albeit in much smaller numbers than partial ones. The proportion of complete compared to partial ABGs also varies between species and periods (see 10.5). A number of explanations have been utilised to explain these complete ABGs ranging from the functional (culling, disease, natural deaths and pit falls), to the ritualistic

Table 11.3 Numbers of ABGs per period interpreted as animals that died of disease. Number in brackets indicates the number of partial ABGs Early High to Late Species Romano-British Medieval Medieval Cattle 1 S/G 9 (1) 12 (4) Pig 7 (1) Dog 1 Table 11.4 Numbers of ABGs per period interpreted as natural deaths. Numbers in brackets indicate the number of partial and unknown ABGs High to Late Species Iron Age Romano-British Early Medieval Medieval Cattle 2 9 (7) S/G 28 (21) 5 (3) Pig 6 (5) 7 (1) 2 (2) Horse 1 (1) 1

153

Investigating Animal Burials; Ritual, Mundane and Beyond interpretation. Two very similar deposits from the same region are interpreted slightly differently. It also shows that zooarchaeologists find it very hard to draw distinctions between animals that died of ‘natural’ causes and those that died from disease. For example, osteologically, we would not be able to tell the difference between cattle that died of old age, and old cattle that died of BSE or any other modern disease..

Although the majority of ABGs with pathologies present are from horses and dogs (see 10.7), it is sheep/goat and pig that are most often interpreted as the deposition of diseased animals (Table 11.3). However, none of the ABGs interpreted as diseased animals have pathologies present. The majority of the ABGs interpreted in this manner are from the High Medieval period, most from Easton Lane, Hampshire. All 12 High Medieval sheep/goat are from pit 5265, which Maltby (1989b) suggests may represent members of a breeding flock that died of disease during the winter (based on ageing evidence).

Other authors have viewed these deposits differently. The reason why arguments have often been put forward that an animal died of disease or natural causes is because the ABG is complete (or thought to have been deposited complete) and has not been processed for meat. However, the reason for non-consumption could be purely cultural. Such arguments are often put forward to explain the presence of ABGs of dogs and cats (see 11.5).

Alongside diseases, a number of complete ABG have been interpreted as natural mortalities, in that they belonged to animals that were not killed by humans (Table 11.4). The definition of what constitutes a natural death is unclear. Authors have discussed animals dying from old age, starvation and ‘natural’ juvenile mortalities. As well as the main domestic mammals, natural death has been used as an explanation for some dog and cat ABGs as well as wild animals. We could expect the main criterion for an ABG to be interpreted as a natural death is completeness, but the majority of ABGs interpreted in this manner are either partial, or of unreported completeness (Table 11.4). However, it has been assumed by the majority of the reporting authors that the animals were originally deposited as complete skeletons that later became partially disarticulated due to post-depositional taphonomic process.

Both Grant (1984a) and Wait (1985) recognise that the completeness of the ABG is variable. However, their interpretations of the different ‘special animal deposits’ vary little. Both see the majority of ABGs as the result of a sacrifice. Grant (1984a) views both complete and partial ABGs as the deposition of sacrificial animals, but suggests that partial ABGs were deposited so that the rest of the animal could be utilised, thus minimising loss to the community. Grant does make a distinction between juvenile and older complete ABGs, viewing young animals as possible natural deaths, although her distinctions between old and young animals are completely arbitrary.

Interpretations citing death by natural causes are more widely used by authors than disease. For example, excavations of late Iron Age features at Whitcombe, Dorset revealed pit 4/16, which contained 13 sheep/goat ABGs, all juvenile apart from one adult. BucklandWright (1990) suggested that the animals had died naturally over winter due to starvation. Buckland-Wright used the lack of butchery marks to suggest the animals were not processed, although the report does not state how complete the skeletons were. He also states that the pit would have been too small to contain 13 complete, healthy sheep, and therefore they died from starvation. This is very similar to the explanation Buckland-Wright (1987) made for the seven sheep/goat ABGs, at Poundbury, which he says died of haxia (see above). These two examples show the inconsistency in ABG

Hill (1995, 58) also identifies that different types of ABGs are present in the Iron Age archaeological record. Like Grant, Hill views complete ABGs of the main domestic mammals (cattle, sheep/goat and pig) as similar in origin and purpose to partial ones. Hill makes the important point that feasting and sacrifice are often combined and that; ‘One variable for manipulation within a rite was the proportion of animal reserved by the host/organiser (a greater or lesser proportion of an animal could be eaten by the living participants or offered to the spiritual participants – deposited in the ground)’ (Hill 1995, 103).

Table 11.5 Ritual interpretations of complete ABGs of cattle, sheep/goat and pig, per period High and Authors Bronze Iron RomanoEarly Late interpretation Species Neolithic Age Age British Medieval Medieval Ritual/Sacrifice Cattle 1 6 12 5 S/G 1 4 14 1 Pig 1 15 4 Offering Sheep 2 Pig 1 Foundation Offering Cattle 1 154

Chapter 11. Assigned Meaning mammals are interpreted as offerings: a juvenile pig from the Iron Age Grindale Barrow II (Manby 1980) and juvenile sheep/goat from the early Romano-British sites of Poundbury (Buckland-Wright 1993) and Pins Knoll (Bailey 1967). Each of these three ABGs has been interpreted as an offering because of their close association with human remains. The rest of the ‘food’ domestic mammals are interpreted within the meta-level ritual category, the majority implying that the ABGs represent sacrificed animals (see below).

Therefore, Hill sees both partial and complete ABGs of the main domestic mammals as being produced by the same activity, but their varying level of completeness is due to cultural choice at the specific event. He does, however, view the deposition of complete horse, dog and wild mammals differently, suggesting these species were treated in special ways, and were particularly favoured for deposition (Hill 1995, 103). In the Iron Age, dog and wild mammal ABGs were certainly the most common animals to be deposited as complete ABGs. However, Hill’s inclusion of horse in this category is not confirmed by the data. The Iron Age is the first period in which complete horse ABGs are recorded but only a small proportion of these horse ABGs consist of complete skeletons, and the majority consist of just limb elements (see 4.5). In fact a higher percentage of cattle, sheep/goat and pig are deposited as complete or near complete than horse. In defining a possible Iron Age cultural classification of animal species, Hill (1995, 104) places horses with dogs and wild species. This appears to be based on previous literature, which stressed the close association of these species with humans, their links with ‘Celtic’ deities, and the high proportion of ‘special deposits’ present, rather than similarities in the nature of the ABGs themselves.

11.4.2 Partial ABGs; butchery waste or ritual rubbish? As this study has shown the majority of domestic mammal ABGs recovered from all periods are partial in nature, most consisting of only a small number of elements. Unlike complete ABGs, those of a partial nature have not always been recognised in the archaeological record, with detailed recording only taking place in the last couple of decades. Although the work on the Danebury assemblage raised awareness of partial ABGs, especially for the Iron Age, the majority of zooarchaeologists initially did not concur with Grant’s interpretation (1991) (see 1.2.7). The most common functional interpretation given to partial ABGs is that they represent waste from the ‘normal’ butchery process. This view was prevalent in the 1980’s and early 1990’s (see 11.2). For example in discussing the remains from Owslebury, Maltby (1987c) stated that;

A much higher percentage of complete horse ABGs are present from the Romano-British and later Medieval periods, although the sample size is small (see 10.5.1). Of the four complete Romano-British horse ABGs, two each are from Kirkburn (Legge 1991a) and Barton Field Villa (Hicklin 2006). All were adult animals and are interpreted as ‘ritual’ deposits. Their interpretation is linked to the location of the deposits. At Barton Field, the ABGs were within the entranceway to the villa, and at Kirkburn they were deposited close to an Iron Age funerary site. The Barton Field horse ABGs are also recently published, and fit into the overall trend of recently examined RomanoBritish ABGs being interpreted as ritual deposits. The later Medieval complete horse ABGs are either interpreted as functional deposits, natural death (Sadler 1990), or waste (Hamilton-Dyer 1999b), or no interpretation is offered (Bullock and Allen 1997; Grant 1985).

‘It is thought that the large proportion of articulated bones were not of any particular significance that cannot be explained by the events normally associated with pastoral farming.’ Maltby and other authors were open to the possibility of ritual/symbolic aspects, but viewed the arguments made at the time for ritual deposition of ABGs as unconvincing. These views were largely influenced by the work of processual archaeologists, in particular Binford’s (1978) ethnographic investigation of the Nunamiut. As part of this study Binford (1978; 1983) showed how the Nunamiut process deer into different body units, which are then further processed, resulting in a spread of ‘tossed’ aside elements as well as articulated body parts around a butchery area. Wilson (1996, 32) specifically refers to the similarities between the patterns seen by Binford and the ABGs from Mount Farm, Berinsfield, Oxfordshire. Binford’s (1983) work on the Nunamiut was intended to provide a possible framework within which to investigate small groups of hunter-gatherers. In particular he was interested in the Mousterian culture. It may be unwise, however, to compare such very different cultures as there are few methodologies for cross-cultural comparisons (Roux 2007).

Green (1992, 116, 119, 123) in her investigation of Iron Age animals suggests that pig, cattle and sheep/goat ABGs fall into one of two groups; the animal was slaughtered but not eaten and was buried as a gift to the supernatural powers; or the animal was butchered and the meat placed as a food-offering to the dead or consumed in a ritual feast. Therefore complete ABGs represent uneaten offerings to supernatural powers. In the reports examined very few zooarchaeologists have offered specific ‘ritual’ interpretations for complete domestic ‘food’ species ABGs (Table 11.5). One complete juvenile cow ABG from Iron Age Garton and Wetwang Slack (Noddle 1979) was interpreted as a foundation offering, due to its deposition under a round house (see 5.5.3). Only three complete domestic ‘food’ 155

Investigating Animal Burials; Ritual, Mundane and Beyond complete deposits are viewed as ‘waste’. For example, from the Iron Age sample 125 partial sheep/goat were interpreted as ‘waste’, compared to one complete sheep/goat ABG. However, the interpretations given to the ABGs are also influenced by the archaeological paradigm the zooarchaeologists were working within.

The interpretation of ABGs as waste is a reflection of our understanding of the butchery process. It assumes that the carcass has been dismembered into sections, with the meat then filleted, and the connective tissue left on the bone keeping elements in articulated positions. We could therefore expect the composition of ABGs to change as butchery practices do. This may be the case with the transition from the Iron Age to Romano-British periods. However to complicate matters, Hill (1996) points out that ‘ritual’ activity would have used the same technologies and practices as those in daily life. The introduction of specialist butchers, however, may well have resulted in the separation of ritual/mundane practices in the Romano-British period (Maltby 2007). But the important point to consider with ABGs is that we see the result of the butchery process, not the intention behind it.

The majority of the ‘waste’ interpretations on material from prehistoric and Romano-British contexts were made by authors reporting before the 1990s. Medieval partial ABGs are still, when commented upon, reported as butchery waste (Table 11.6). However, a number of the authors reporting on prehistoric material remain open to a symbolic interpretation of partial ABGs from the main domestic species, often suggesting they may constitute evidence of feasting (Armour-Chelu 1991; Grigson 1999; Maltby 1985a;1985f; 1990c). As Lévi-Strauss (1964) points out, the human communal sharing of food is unusual behaviour. In other species eye contact, opening of mouths and exposure of teeth combined with the presence of food between individuals would normally result in conflict and violence. The anthropologists Douglas (1963), Goody (1982) and Lévi-Strauss (1964) all viewed feasting as an important activity by which status, power and social order are negotiated. There is no set definition of what constitutes a feast. Dietler (2001) see feasting as large-scaled ritual consumption events, whereas Douglas (1984), discussed the American family meal as a feasting event. Parker Pearson (2001, 10) taking an holistic approach, sees feasts as;

The survival of an ABG is dependant upon how the ‘waste’ is deposited. It is often assumed that the majority of ‘normal’ bone waste for most periods is deposited within middens, resulting in the poor preservation of faunal material (Maltby 1985f). Hill (1995, 28) using the later prehistoric midden at Runnymede (Serjeantson 1991b) as an example, suggests there is little evidence of articulated or associated bone deposited in middens, and therefore, for the Iron Age, ABGs in pits must represent primary deposits. Potterne (Lawson 2000) is the only midden site recorded in this study and has only a small number of ABGs present (see 3.7.3). However, the nature of large midden sites is not yet fully understood and the material deposited on them may be unrepresentative of settlement-based midden material which is subsequently incorporated within pits.

‘occasions when large-scale hospitality creates debts and obligations, when reputations are made and lost, when social order is exhibited, challenged and reformulated, when the work of many may be claimed by the few, and when new factions, mobilisations, alliances and other relationships are formed and dissolved.’

Maltby sees primary deposition within features, normally on the periphery of settlement, in the Iron Age and Romano-British periods as representing the removal of more odorous, offensive, waste (Maltby 1985f; 1987c; 1989b). This may be the explanation for the higher proportion of cattle and horse remains in peripheral features, as one would expect them to produce larger quantities of waste. In contrast, the remains of small species such as sheep/goat tend to be recovered closer to the centre of settlements (Maltby 1989b; Wilson 1996). Such a pattern is also present in the ABG data from Owslebury, indicating similar treatment of the ABG and non-ABG material.

The main feature of most definitions of feasting highlights large-scale involvement (Dietler 2001), although authors do not state how many people are required to make a large-scale event. Recently Hayden (2001) has outlined criteria for the differentiation of feasting waste from that created by ‘normal’ day to day consumption. This includes the presence of large items, in large quantities, enhanced architectural arrangements and special locations (such as a funerary context).

Generally, ‘waste’ is the most popular functional interpretation for partial ABGs. Only a small number of

Table 11.6 Total number of domestic mammal ABGs per period interpreted as ‘waste’. Number in brackets indicates the number of complete ABGs Bronze RomanoEarly High and Late Species Neolithic Age Iron Age British Medieval Medieval Cattle 6 2 36 (1) 56 17 6 (1) S/G 3 126 (1) 47 (8) 5 4 (1) Pig 4 18 3 4 Horse 38 4 2 6 (1) 156

Chapter 11. Assigned Meaning

Recently, Ralph (2007) has used an integrated approach combined with the arguments put forward by Dietler (2001) and Hayden (2001) to suggest that feasting deposits are common on Iron Age East Anglian sites. However, such criteria were developed with Mesoamerican cultures in mind and many of Ralph’s conclusions are based on the presence of individual objects, although her arguments regarding the deposit at Baldock are convincing (2007, 99).

consist of partial limb elements. Other authors have alluded to the possibility that the deposits they examined may be the result of feasting, but have tempered such interpretations by also offering functional or other ritual explanations. Armour-Chelu (1991, 151) suggested that some of the partial ABGs from the Iron Age features at Maiden Castle might be interpreted as the residue of ‘special meals’, but felt they were unlikely to represent ritual activity. Mainland (2006), discussing the RomanoBritish Shiptonthorpe ABGs, suggested that some of the deposits with evidence of burning may be associated with feasting, but also argued they may have been sacrificed.

Hill (1995, 62) uses similar ideas to suggest that large quantities of unarticulated faunal material possibly from the same animal may be linked to the formation of ABGs. Using Firth’s (1963) notion of ‘reservation’ in sacrifice he suggests that ABGs may represent the reserved portion of the carcass, with the rest consumed in a feast and deposited in association. Hill (1995, 63) rightly points out that in most ethnographic examples, feasting and sacrifice often occur during the same ritual event.

The fact that feasting did occur in the past is not in contention. Green (1992, 162, 179-171) discusses feasting in the Iron Age, by drawing upon Irish vernacular writings which date to AD 1100 onwards. We also have ample literature and iconographic evidence of feasting from the Romano-British period onwards (Dunbabin 2003; Hammond 1993). Although as Hadley (2005, 117) points out when discussing Medieval feasting;

Parker-Pearson (2006) goes further and suggests that most if not all animal bones from prehistoric archaeological sites may be the result of feasting. A more balanced approach is offered by Serjeantson’s (2006) investigation of the Neolithic and Bronze Age Runnymede midden. Serjeantson (2006, 130) makes the important point that;

‘The stylised and orderly feasts in manuscript illuminations present dining as the elite wished it to be perceived, and they reflect the importance of prestigious display for the maintenance of social distinctions.’

However, the generalisations that ‘large’ quantities of material would be deposited, offer no quantifiable basis.

From faunal remains the best evidence we have of feasting comes from the later Medieval periods, with a wider range of species represented from castles, palaces and religious sites (Grant 1988; 2002). In particular, there is a strong bias in the representation of body parts, particularly the hindquarters of red and fallow deer (Albarella and Davis 1996; Maltby 1982a; Sykes 2005). Both Sykes (2005; 2006b; 2007) and Thomas (2005; 2007) have investigated this phenomenon, and linked it to the ‘ritualised’ unmaking process which was described in a number of hunting manuals. The deer was skinned and gutted at the kill site, the pelvis given to the raven, the left shoulder to the ‘unmaker’, the right to the forester, the haunches to the lord and the rest to the hounds. It was the haunches which were then feasted upon by the elite classes, which can then be detected in the faunal remains found on high status sites. The haunches do not appear to be regularly deposited as ABGs, and none have been recorded in this study. The remains of deer were deposited along with the rest of the household ‘rubbish’. Grant (2002), discussing Medieval swan remains, points out that rich and poor may have lived separate lives but were often in close proximity to each other and therefore some of their food waste would end up on the same rubbish dumps.

‘Though the quantity of bones is very large, this does not in itself indicate feasting.’ Serjeantson (2006) argues that the size of the animals and the cooking methods used may be more useful indicators of feasting. The size of the animals must be considered alongside the community’s ability to store the meat. Fragmentation, butchery, burning and other taphonomic signatures can inform us of the preparation and cooking methods used. Bones that have been heavily fragmented were probably used in soups and stews (Outram 2001). Serjeantson suggests that the ‘normal’ meals in the Neolithic would have consisted of such stews and soups, similar to Medieval pottage of cereals and vegetables in which bones were included. The faunal material from Runnymede shows this pattern. However, some joints of pig and cattle appear to have been cooked over a fire, rather than stewed, which may be indicative of feasting. Although not conclusive proof of feasting, this approach at least moves away from the large sweeping generalisations of other authors and provides additional criteria to assist in interpreting ABGs. Very few ABGs have been interpreted as the specific result of feasting events (Table 11.7). Only the ABGs from the Neolithic sites of the Coneybury Anomaly (Maltby 1990c) and some of the cattle ABGs from the Keiller excavations at Windmill Hill (Grigson 1999; Jope 1965) have been explicitly described by authors as the result of feasting. The ABGs from both sites mainly

Therefore in the Medieval period we have evidence of rituals associated with specific animals and feasting. However, we have no evidence of ‘ritual’ deposition of the remains from the feast. The ritualised acts only take place been people, as a possible enforcer of status, with the food provided by the animals acting as a medium. It 157

Investigating Animal Burials; Ritual, Mundane and Beyond connective tissue. If this is the case, then why would the deposition of meat represent a feast? It may occur because it was not required for the feast, in effect leftovers, but considering the level of processing undertaken on non-ABG faunal remains, for example to extract and consume the marrow, this seems unlikely. It may represent an ‘offering’ that occurs at a feasting event, but if this is the case, the ABG represents the offering activity, not the feasting, and the two may not be exclusive.

appears that once these acts have taken place, the remains of the animals are of no consequence. The main reason we can identify such acts archaeologically in the Medieval period is due to the development of haute cuisine (Goody 1982). The majority of the literature and iconography from the Romano-British period onwards depicts feasting associated with haute cuisine, for example Medieval banquets with rare birds and red deer haunches, which we can identify in the faunal record. It is our ability to identify feasting using basse cuisine which is problematic, as there are no rare species, or specific cuts of meat for us to look for. Therefore, in all likelihood feasting did take place in prehistory, but we need further work in how to distinguish it from ‘normal’ consumption. If we take Serjeantson’s (2006) suggestion of roasting, only five ABGs, all sheep/goat from the Romano-British period, may represent evidence of feasting (see 7.3 and 7.4). We must also consider two points: the act of feasting may be the important factor, not the deposition of the resulting material; also feasting in the prehistoric periods may have used the same food preparation methods as ‘normal’ consumption, therefore also resulting in ‘normal’ waste. Another point of consideration is that we do not know how common meat consumption was in prehistoric periods. If the faunal remains recovered from prehistoric sites are representative of a group’s meat consumption, then meat may have made up only a small proportion of the diet. If this is the case we could argue that all ‘normal’ faunal remains may be the remains of feasting. If this is the case, then ABGs may not be representative of distinct/unusual feasting activity. Without further work in this area, in particular on what everyday meals consisted of, feasting may remain an elusive activity in the archaeological record, although future isototpe analysis of human remains may help indicate levels of protein in the diet.

A much larger proportion of ABGs have been interpreted as offerings of some kind (Table 11.7). Two partial domestic mammal ABGs have been interpreted by the reporting authors as foundation deposits. A pig’s skull and vertebrae found in a pit under a late Romano-British building at Dorchester Prison (Draper and Chaplin 1982) and a sheep/goat skull and vertebrae, in association with a human neonatal burial, under the wall of a RomanoBritish villa at Bradford Down (Rixson 1982). As well as these ABGs, one complete cattle ABG, one cat and five partial domestic fowl ABGs are also interpreted as foundation offerings/deposits. All of these deposits were found in close association with buildings. However, this does not always result in a foundation deposit interpretation. For example, a number of partial ABGs of cattle, sheep/goat and pig were recovered in close association with buildings at Shiptonthorpe, Yorkshire (Mainland 2006) and Rudston Roman Villa, Yorkshire (Chaplin and Barnetson 1980) (see 7.3.2 and 7.4). These deposits were given more generic ‘ritual’ interpretations by the authors. A number of partial ABGs are also interpreted more generically as ‘offerings’ (Table 11.7). Of the 52 partial domestic ‘food’ mammals interpreted in this way, all but four are from formal funerary contexts. Three of those four ABGs were found in close association with human remains. Two ABGs of an Iron Age partial sheep/goat and a pig ABG, both from Pit 15C, Hod Hill, were found in association with an articulated-human-female skeleton (Bunting et al. 1968). A Romano-British pig-lower-leg ABG is recorded from Winchester Northern Suburbs,

We must also consider that, when present, the majority of butchery marks recorded from ABGs are associated with skinning, disarticulation or dismemberment and very few ABGs show signs of filleting (10.6). This suggests that the meat may still have been attached to bones, not just

Table 11.7 Total number of domestic mammal partial ABGs per period interpreted as ritual Authors Bronze Iron RomanoEarly interpretation Species Neolithic Age Age British Medieval Feast Cattle 9 Foundation S/G 1 Offering Pig 1 Offering Cattle 1 3 S/G 7 17 Pig 15 7 Horse 2 Ritual/Sacrifice Cattle 10 2 34 7 1 S/G 2 12 51 7 Pig 4 14 4 Horse 1 43 4 Total 25 16 164 54 1 158

Chapter 11. Assigned Meaning midden material was disposed of within a pit/well feature. Disassociation of elements from the carcass may have occurred through natural biogenic processes. Therefore upon secondary deposition the partial dog ABG was created. Therefore, the partial ABGs may not be reflective of primary depositions of dog carcasses, but rather midden material in which the dog remains had already been incorporated. Another possibility is that the dog ABGs represent primary deposition within the feature and that slumpage caused by the rotting of organic material, resulted in the disassociation of some elements. However, both explanations result in the survival of partial dog ABGs, with the missing body elements also incorporated within the archaeological fill. There is of course the possibility that both activities took place, with partial dog ABGs from a midden being deposited and further disassociation taking place due to slumpage within the feature.

Feature 168, in association with a human neonatal skeleton (Maltby 1987d). Only one partial domestic ‘food’ mammal has been interpreted as an offering without been in association with human remains. This consists of a Bronze Age partial cattle ABG, with limited data available for it, from the enclosure at Thomas Hardye School, Dorchester (Smith 2000). The rest of the major domestic mammal partial ABGs have been interpreted using a generic ritual sacrifice idiom. This is due to the problems with defining a specific interpretation for ritual deposits (see 11.8). Grant (1984a) did comment upon the interpretation of partial ABGs, but suggests both partial and complete ABGs are ritual sacrifices. She does, however, use an economic argument in regards to partial ABGs, by suggesting some of the sacrificed animals had been utilised to minimise the total loss of an important resource. Hill did not focus upon the composition of ABGs in his interpretations and suggested the same interpretation for both complete and partial deposits, namely sacrificial feasting events, with the variability due to cultural choice at the specific event. Despite Hill’s more detailed interpretation, the majority of zooarchaeologists have not interpreted Iron Age deposits as the result of feasting/sacrifice. This is possibly because the two are inseparable (see above), yet still viewed by zooarchaeologists as separate events. The majority have used a more generic category of ‘ritual’. 11.5

If the former scenario did take place, then analysis of the ABGs bones may yield evidence in the form of gnawing, weathering and erosion visible on the bones, indicating a period of above ground exposure. However, no such evidence has been reported for the assemblages included in this study. This may well indicate that the latter scenario is the more likely. The deposition of large numbers of complete dog carcasses corresponds with the general view that dog meat was not consumed in Roman Britain. Only two of the 141 dog ABGs recorded from Romano-British towns had evidence of carcass processing (see 10.6). This pattern is also seen on Romano-British rural settlements, with only two dog ABGs from the early Romano-British period displaying butchery marks. During the RomanoBritish period dogs do not appear to have been consumed, at least not in any great number that is evident archaeologically. Their remains were deposited, possibly as primary depositions, within pits and wells in both rural and town settings, although larger numbers are found in urban contexts. A small number of dog ABGs are present in ditch fills on rural sites, although very few ditches are present in urban contexts. However, at Owslebury, where the majority of faunal remains come from ditches, dog ABGs predominantly came from the pits within the settlement (see 10.8). The fact that a large proportion of these individuals were immature, with several cases where large numbers of newborn puppies were found, has also led zooarchaeologists to interpret them as the result of culling, to control the population.

Domestic dogs and cats; population control or sacred animals

As discussed in the previous chapters dog and cat ABGs often consist of complete skeletons, or are argued to have been deposited complete. Dog and cat ABGs are recorded from the Neolithic onwards, although the earlier prehistoric-cat ABGs are most probably from wild individuals. By the Romano-British period dogs are the most common species recovered as an ABG, and this period also sees an increase in the number of cat deposits. Previous literature on dog ABGs have centred on the Romano-British period, due to the large numbers encountered. Maltby (1986b; 1993; 1995a), in his investigations of the faunal remains from Romano-British Dorchester and Winchester, suggested that groups of dog ABGs may be the result of population culling. Both sites had a large number of deposits recorded from well/pit features. Large percentages of these are complete ABGs and Maltby has argued that the partial remains were originally deposited as complete carcasses, which subsequently became disassociated through taphonomic activity. The body area analysis for Dorchester indicates this may well be the case (see 6.6.3).

A small number of Iron Age dog ABGs have been interpreted as the result of population culling activities (Table 11.8). All of these consist of neonatal dog ABGs from Balksbury Camp (Maltby 1995a; 2001) and eleven are found in associations of two or more deposits. It is these aspects that influenced Maltby (1995a) to interpret them as the result of culling, although it can also been seen as an extension of his arguments regarding RomanoBritish dog ABGs. The majority of the dog ABGs interpreted as the result of culling are from the Romano-

We must, however, consider what form of taphonomic activity took place. Two main scenarios could have occurred. Complete dog carcasses may initially have been deposited within a midden along with other general household material. Then, after a period of time, this 159

Investigating Animal Burials; Ritual, Mundane and Beyond Table 11.8 Number of dog and cat ABGs per period interpreted as the result of culling RomanoEarly High and Late Species Iron Age British Medieval Medieval Dog 17 237 9 2 Cat 4 Table 11.9 Number of dog and cat ABGs per period interpreted as natural mortality High and Late Species Bronze Age Iron Age Romano-British Early Medieval Medieval Dog 1 13 27 3 4 Cat 11 1 Table 11.10 Total number of dog and ABGs per period interpreted as ritual depositions Authors Bronze Iron RomanoEarly High to Late Species interpretation Neolithic Age Age British Medieval Medieval Dog Offering 7 1 Ritual/Sacrifice 5 2 37 18 Cat

Foundation Offering Ritual/Sacrifice

1 3 adult remains are interpreted as the result of natural death and young animals as the result of culling. The exception is the complete female dog ABG from the middle AngloSaxon site at Clifford Street, Southampton, which is interpreted as a diseased animal, although no pathology is present (Bourdillon 1990b).

British period and all are from reports by Maltby (1978b; 1987c; 1987d; 1993). A small number of Medieval dog ABGs are also interpreted as the result of culling activity, although interpretations of the Medieval dog remains are often not given, even when butchery is present. This is likely to be a reflection of the assumptions concerning ABGs from these periods . All the early Medieval dog ABGs interpreted in this manner consist of neonatal individuals from Greyhound Yard (Maltby 1993) and Faccombe Netherton (Sadler 1990). The two later Medieval dog ABGs from Faccombe Netherton (Sadler 1990) and New Road Winchester (Coy 1984b) are also from young individuals.

Natural death is not as popular an explanation for dog ABGs in the Romano-British period, compared to culling. However, the majority of deposits from the Romano-British period, in this study, came from assemblages reported on by Maltby. Maltby (1987c; 1993) did use natural death as an explanation for a small number of dog ABGs from old animals. In comparison, Hamilton-Dyer (1996a) suggests that three neonatal dog ABGs found in association at the Romano-British Maddington Farm were possibly the result of natural deaths. Scale is also important in these interpretations, with individual dog ABGs explained as natural deaths and multiple ABG deposits as culling episodes.

Interestingly the rest of the dog ABGs from New Road, Winchester are from adult individuals but were interpreted by Coy (1984b) as the result of natural deaths. Natural death is another interpretation sometimes given to dog and cat ABGs inferring that they represent the deposition of animals which died of old age (recorded in this study as a natural death). As these species are generally not thought to have been commonly consumed they would therefore be deposited whole, resulting in an ABG, that is usually complete. Such an explanation is also given for the cat ABGs from the Silchester North Gate area (Hamilton-Dyer 1997b).

A limited number of cat ABGs were recorded in this study. Like dogs, some have been interpreted as the result of culling, although in very small numbers (Table 11.8). As with dogs, cat ABGs are often complete, or when partial, have been interpreted as resulting from the taphonomic actions discussed above. The cat ABGs interpreted as the result of culling are a group of four neonatal kitten ABGs found in association at the late Medieval site of Osborne House, Romsey, where it was suggested they were from a unwanted litter (Coy 1986). A more common interpretation for cat ABGs, especially from the Romano-British period is that they are the result of natural deaths. This may be influenced by the age of the individuals. All the Romano-British cat deposits with a reported age are from sub-adult or older individuals.

There are no real differences between the dog ABGs recorded as natural deaths (Table 11.9) and those interpreted as the result of culling. The explanation can be applied to both old and young animals. Old animals either died naturally or were not required anymore and were therefore culled. Young animals either died naturally by not surviving the birth or catching a disease. Alternatively, they represented an unwanted litter and were culled. Often, as shown by the interpretation of the dog remains from New Street Winchester (Coy 1984b), 160

Chapter 11. Assigned Meaning sites, due to ‘ritual’ explanations being accepted for these deposits earlier than for historic sites.

Three cat ABGs from the Anglo-Scandinavian Coppergate site, York, have butchery marks present (see 9.3). The one complete and two partial ABGs all have knife marks present on the skull, which have been interpreted as the result of skinning. Leading O’Connor (1989) to suggest the ABGs were the result of skinning waste.

The Neolithic dog ABGs interpreted as ‘ritual’ deposits are all from Windmill Hill. The two Bronze Age dog deposits are from Coneybury Henge (Maltby 1990c) and Barrow 23, North Down Barn (Grinsell 1959, 142). Reasoning behind such interpretations are not given by the reporting authors. It seems that such an interpretation is given because dogs are rare finds from this period, and archaeologists see the archaeological sites they are from as areas of ‘ritual’ activity. Therefore the assumption is that ABGs from these sites must have a ritual nature.

The interpretations for dog and cat remains discussed above have been functional in their outlook. However, the majority of prehistoric and Romano-British ABGs are now interpreted as ritual. This is especially the case for dogs. Ritual interpretations of dog ABGs are nothing new, but were originally mainly limited to complete skeletons, either from Neolithic and Bronze Age sites (Behrens 1964; Gabalówna 1958; Jackson 1943), or those in association with human remains (Bailey 1967; Bunting et al. 1968; Collins 1953) or both (Grinsell 1959, 142). Behrens (1964) investigation into 459 cases of complete animal skeletons from 268 continental sites found over 50% of the cases were dogs (see 1.2.3). Dog deposits do not make up such a large percentage of the Neolithic to Iron Age assemblages for this study. However, Behrens (1964) ideas that such deposits, which were not associated with human funerary practices, were the result of offerings for spiritual/divine blessing or animals deemed to be ‘special’ feared or worshipped, remain prevalent, especially for dog ABGs.

Such thinking has not traditionally been used for Iron Age sites. However, the majority of recent dog ABGs recovered from Iron Age settlements have been interpreted as ‘ritual’. Hill (1995, 103-104) suggests that dogs can be viewed in a number of different ways; a positive species, a human friend, almost human; or negative, dirty, polluting; or all of the above. He concludes that Iron Age ‘ritual’ ABG deposits reveal something of the Iron Age animal classification in which dogs are close to humans. Hill (1995, 107-108) argues that dogs are culturally close to humans in the Iron Age, because both require ‘training’ and play an active role within the society, such as watchdogs or sheepdogs. A more prosaic approach is that they are treated similarly to humans because both human and dogs are not consumed.

Two main ritual explanations have been given for the dog ABGs recorded in this study; ‘offering’ and ‘ritual/sacrifice’. Only eight of the dog ABGs have been interpreted as offerings by the reporting authors (Table 11.10). Most of these consist of complete individuals found in association with formal human burials and have thus been interpreted as grave good offerings. When found in association with human remains, ABGs are invariably interpreted as ritual depositions (see below). A good example of this are the four dog ABGs from the Romano-British site of Maddington Farm (HamiltonDyer 1996a). Three of the dog ABGs recovered from a pit were interpreted as natural deaths, but the one dog ABG found in association with human remains is interpreted as an offering.

A number of authors have recently discussed the deposition of dog ABGs from Romano-British sites, suggesting a ritual interpretation. Clarke (1997; 1999), in a re-examination of the Newstead pits, suggested that the skeletal remains of humans, dogs, horses, cattle and deer are significant and the ritual character of some large groups of animal bones and artefacts are virtually inescapable. Clarke came to this conclusion because of the strong patterning of materials concentrated together in the bottom quarters of the deepest pits. The work on the Newstead pits also draws upon the previous work of Ross and Feachem (1976), who had already suggested a ritual interpretation for these features. Importantly Clarke (1997; 1999) also draws heavily on Hill’s (1995) work. Clarke (1999) is critical of Hill’s suggestion that the deposition of ABGs was geared to the manipulation of the supernatural. He suggests that this is a westernised viewpoint, and in the ancient world the supernatural was natural and part of the everyday world. However, Clarke appears to have missed Hill’s same argument. Hill (1995, 112) does suggest that the ritual and mundane worlds were not separate in prehistory, but also unlike Clarke goes further to suggest a process by which the ABGs were created (see above). Clarke (1999) simply suggests that these deposits are of a ‘ritual’ nature, giving no context to the actions which created them. Recently Clarke (in press) has re-visted the Newstead pits and shown that there is a great deal of variability in the ‘ritual’ actions identified.

One cat ABG from the The Bedern Foundry, York (Bond and O'Connor 1999), has been described as a foundation offering (Table 11.2). This consists of a complete ABG, deposited within the post-hole of a 15th century building (see 9.5.3). As with the other ABGs suggested to be foundation offerings, it is the area of deposition (in association with a building) which is the defining factor in the interpretation. The rest of the dog ABGs are interpreted as ‘ritual’ deposits of animals which have probably been sacrificed. As with the major domestic mammals discussed above, these deposits are being interpreted under a general heading, but the reasoning behind the interpretations is not being discussed. As expected, the majority of the dog ABGs interpreted in this manner are from prehistoric 161

Investigating Animal Burials; Ritual, Mundane and Beyond Sacrum Canarium (dog sacrifice) was carried out by priests in Rome around the same time as the Robigalia. The point that sacrifices occurred in the Roman period is not a contentious one. However, Smith does not indicate whether the carcasses were deposited in a specific way. She suggests that her study of seven sites in southern England supports the argument for the ritual deposition of dogs, but appears to utilise the assumption that structured deposition means ritual. However, the important point that a practical outcome does not necessarily rule out spiritual motivation is made (K Smith 2006, 24).

Fulford (2001) draws upon Hill’s (1995) and Clarke’s (1999) work in his survey of ‘ritual’ behaviour’ in Roman Britain. In it he shows that ABGs, especially of dogs, are present in the archaeological records of a number of Roman towns. He sees the structured ABG deposits as a continuation of Iron Age practices, which were largely rurally based. He also suggests that the ABGs were ritual deposits, and points to possible sacrifices outside temple contexts. However, he does conclude that; ‘How far we will be able to understand the meaning and significance of the practices – whether to propitiate chthonic deities, or to ensure fertility, for example – which we have begun to identify, remains unclear’ (Fulford 2001, 216).

With the exception of one case (see above), none of the dog ABGs from the Medieval period have been interpreted as the result of ‘ritual’. This reflects the general period-based trends in interpretation. However, Hamerow (2006) has concluded that ‘special deposits’ are present on Anglo-Saxon settlements. She suggests that the prominence of dog and horse in ‘special deposits’ is a feature shared by Iron Age, Romano-British and AngloSaxon settlements and is explained by the close relationship between dogs and humans. This study has shown that dog ABGs are the most common type found in the Romano-British and Anglo-Saxon periods, although the prominence of horse ABGs is very low in all periods (at least in the areas studied). However, Hamerow offers little interpretation beyond the generalised sense that ‘special deposits’ are ritualistic, which is assumed from the outset.

Woodward and Woodward (2004) have no such doubts. They reinterpreted the large pits excavated at Dorchester as being ritually significant, created as part of a foundation ritual on the formation of a new town. Drawing upon Mediterranean evidence, they note that in the initial founding of a town a central pit was dug designed to receive offerings of the first fruit. Therefore these features were suitable for further ritual deposits in the form of ABGs, coins, gaming pieces and complete pots. Relying on Green’s (1992, 198) work, Woodward and Woodward (2004), see dogs as having a ‘special’ place in the rituals and iconography of Iron Age and Roman Britain. They suggest the dog is traditionally associated with healing, fidelity and protection of humans and therefore suitable for ritual sacrifice and deposition within ritual features. However, Woodward and Woodward (2004) make a number of large assumptions. Their argument for the pits being part of a town’s foundation ceremony is based on Italian evidence and they use a feature from Cosa as an example;

11.6

Domestic birds; offerings and waste

Only a small number of domestic bird ABGs were recorded in this study, the majority of which are from domestic fowl (see 10.4). Three geese (identified as domestic by the reporting authors) and four raptors have been offered interpretations. All the domestic goose ABGs are interpreted as ‘waste’, the result of butchery for daily consumption. The raptors are all from the later Medieval deposits at Faccombe Netherton (Sadler 1990) and are interpreted as the deposition of hunting birds after natural death (see 8.2.1).

‘A high rocky point at the southern extremity of the town formed a visual and ritual focus. In front of the rock-cut footings for the earliest square building or enclosure (dating to 273 BC) lay a natural rectangular crevasse, 2 to 2.5m in depth: and traces of its original filling consisted largely of carbonized vegetable material’ (Woodward and Woodward 2004, 69).

A number of varied interpretations have been put forward for the deposition of domestic fowl ABGs (Table 11.11). Only one deposit from High Medieval contexts at Sussex Street, Winchester (Coy 1984b) has been interpreted as a natural death, because it was aged as an old adult. The rest of the domestic fowl ABGs from the site represent younger individuals and were viewed as ‘waste’. All the domestic fowl ABGs from the site consisted of partial skeletons. Although a large number of complete domestic fowl ABGs from Faccombe Netherton are also interpreted as waste, the report only offers limited data and explanation. A small number of Romano-British and early Medieval deposits are also interpreted as waste, although all these consist of partial skeletons.

This is the only archaeological example they have for the presence of a foundation feature, designed to receive offerings of the first fruits, which may indeed be the case. However, at no point is literature or Mediterranean archaeological evidence used by Woodward and Woodward to justify the deposition of sacrificed dogs or other animals within such a feature. Recently, Smith (2006) has reviewed the iconographic and some archaeological evidence for the ‘ritual’ use of dogs in the Iron Age and Romano-British period. Drawing on ancient literature, Smith (2006, 43) demonstrates that dog sacrifice was carried out in Rome. During the festival of Robigalia, rusty-coloured sucking puppies were sacrificed to protect crops, and a ritual

The majority of the ritual interpretations are given to domestic fowl ABGs from the Iron Age and Romano162

Chapter 11. Assigned Meaning Romano-British period onwards, are interpreted using the same criteria as commonly consumed domestic mammals. Raptors, for which there is no evidence of consumption, are interpreted functionally along similar lines as companion mammals such as dogs and cats.

British periods, which is in line with the general observed trends. A large number of the Romano-British deposits are interpreted as offerings. As with other species this interpretation is given because they were recovered in association with human remains from formal funerary contexts. A small number of domestic fowl ABGs from the Iron Age and Romano-British periods are interpreted as generic ritual/sacrifice deposits. Five complete ABGs from the Medieval site at The Bedern Foundry, York (Bond and O'Connor 1999) are interpreted as foundation offerings. As with other species, this is due to the association between the deposits and a structure. Two were recovered from a foundation trench and the other three from a post-hole, which also contained a cat ABG. This is one of the few examples of ritual explanations being utilised for ABGs from this period.

11.7

Wild species; accidents, natural or special deposits?

Wild mammal and bird species are also present as ABGs in the archaeological record and a number of functional and ritual interpretations have been postulated. These interpretations are also dependant on the nature of the ABGs and their associations. The majority of the wild mammal ABGs are recorded along functional lines. The most common interpretation, which is exclusively utilised for wild mammals, is that

Overall, there appears to be a similar pattern in the interpretation of domestic birds and mammals. Birds such as domestic fowl, which were consumed from at least the

Table 11.11 Total number of domestic birds per period and their interpretation. The number in brackets indicates the number of complete ABGs. ABGs interpreted as unknown or mixed are not included RomanoEarly High to Late Species Authors interpretation Iron Age British Medieval Medieval Domestic Fowl Natural Death 1 Waste 11 2 25 (21) Ritual/Sacrifice 2 (1) 3 Foundation Offering 5 (5) Offering 2 30 (16) Domestic Goose Waste 1 2 (2) Raptor Natural Death 4

Table 11.12 Number of wild mammals per interpretation and period. The number in brackets indicates the number of complete ABGs. ABGs interpreted as unknown or mixed are not included Authors RomanoEarly High to Late interpretation Species Neolithic Iron Age British Medieval Medieval Fall Badger 1 (1) Fox 12 (12) 2 (2) Hare 3 (2) Pine Marten 1 Polecat 11 (11) Red Deer 1 (1) 3 (3) Roe Deer 4 (2) Weasel 1 Natural death Badger 1 (1) 1 Polecat 1 (1) Waste Fox 1 4 Hare 2 Rabbit 1 (1) Roe Deer 1 Feast Roe Deer 5 Offering Fox 1 Ritual/Sacrifice Fox 1 Hare 1 (1) Wild Cat 1 163

Investigating Animal Burials; Ritual, Mundane and Beyond Table 11.13 Number of wild birds per interpretation and period. The number in brackets indicates the number of complete ABGs. ABGs interpreted as unknown or mixed are not included Authors RomanoEarly interpretation Species Bronze Age Iron Age British Medieval Functional Corvid 1 Natural death Corvid 21 5 (1) Swallow 30 Cormorant 1 Waste Corvid 3 Duck 1 Pigeon 2 Quail 1 Ritual/Sacrifice Corvid 1 3 (1) Sea Eagle 1 been consumed for meat. Five fox ABGs have also been interpreted as waste, not from consumption, but skinning for their fur.

they are fall victims. This explanation is often used in zooarchaeological literature to account for the presence of rodent and amphibian remains, which have not been recorded in this study. Such an explanation was postulated by Jones (1977) for the complete red deer and twelve complete fox ABGs found in association at Winklebury, suggesting that the deer may have fallen in the pit and been unable to extricate itself. However, he does add that such an event is unlikely, yet it is the only explanation offered. Hill (1995, 29) later used this example to represent the unsuitability of some functional explanations for ABGs.

Only a small number of wild mammal deposits have been interpreted along ritual lines, most from prehistoric sites (Table 11.12). Five roe deer ABGs from the Neolithic Coneybury Anomaly (Maltby 1990c) have been interpreted as feasting deposits. However, this is not an explanation given purely to these ABGs, but to all the faunal material from this feature. The one fox interpreted as an offering, was recovered from Whitegrounds Barrow, Yorkshire (Riggott and Williams 1984) (see 3.2) and like other ABGs interpreted in this way, found in association with human remains. The other wild mammal ABGs have been interpreted as general ritual/sacrifice deposits.

The majority of the ABGs recorded as pit falls are from Romano-British contexts (Table 11.12), primarily Oakridge Well (Maltby 1993) (see 6.5). The main deposit of wild mammals consists of 13 complete polecats in relatively close proximity. The polecat ABGs are present within the natural weathering fill. Dog, pig, cat and roe deer ABGs were also present in the fill. The filling of the well during this natural weathering stage is suggested to have taken between 50 to 300 years. However, the polecats are present in the same layers of the infill, indicating they may have entered the well in close chronological proximity.

As with wild mammals, the majority of wild bird ABGs are recorded from the Iron Age and Romano-British periods, with most interpreted as natural deaths (Table 11.13). The Romano-British wild bird assemblage includes the large collection of swallows from Oakridge Well (Maltby 1988), which were postulated to have been nesting within the disused well. Most of the ABGs interpreted as waste from the butchery process, are partial and from species we assume would have been consumed. This includes corvids, three ravens from the Iron Age site at Boscombe Down West RAF Station were interpreted as the possible remains of raven stew (Platt 1951).

Therefore we must ask the question, would 13 polecats fall down a well around the same time? The presence of a few domestic ABGs in the same part of the well may suggest human activity and unfortunately we do not know enough about the behaviour of polecats. Zoologists do not write papers on whether polecats fall down holes. The area of wild mammal pitfalls is certainly in need of attention.

A small number of ABGs have been given a generalised ritual/sacrifice interpretation. The partial sea eagle ABG from Coneybury Henge, Wiltshire is interpreted as a ‘ritual’ deposit (Maltby 1990c), due to the nature of the archaeological site it was recovered from and the rarity of the species. This is the only example recorded for this study, although a number of sites outside the study region are thought to have ritualist deposits of sea eagle such as Isbister (Hedges 1984) and white-tailed sea eagle remains are present on a number of prehistoric and historical sites (Yalden and Albarella 2009, 148).

Other functional explanations given to wild mammal ABGs are that they are the result of natural deaths close to human settlement. Sadler (1990) suggests that a late Medieval complete badger ABG from Faccombe Netherton was the victim of an attack by dogs, although there appears to be no evidence on the skeleton for this. Other wild mammal ABGs have been interpreted, like domestic mammals, as ‘waste’. These explanations have been suggested for deposits from both the Iron Age and later Medieval periods, mainly for mammals could have

A number of different interpretations have been offered for corvid ABGs. One from the Romano-British Oakridge 164

Chapter 11. Assigned Meaning (1931), reporting on the excavations, described an unusual deposit discovered by Mr H. C. Hanford;

Well (Maltby 1993) was recorded as having a functional interpretation, as the author suggested it may have been killed as a potential threat to livestock. The majority of Iron Age corvids have been interpreted as the result of natural deaths, all of which are from Danebury. The assumption is made that they were either nesting close by, or feeding on the waste within the settlements middens when they died. Coy (1984c) suggested that some of the corvid ABGs found on Iron Age sites such as Danebury may represent a natural accumulation stating that;

‘In the south-east corner of the building, where there was a shaft of a pit, 4 feet by 3 feet, and some 14 feet deep…the sides were daubed with clay and lined with stone roof-slabs. The filling of the shaft was most curious. First came a layer of ashes and charcoal, then a double layer of stone roof slabs, laid flat; these were placed in pairs, and between each pair were the bones of a single bird and a single small bronze coin, forming as it were, a sandwich’ (Drew 1931, 267).

‘Ravens frequent rubbish dumps and may pick at carcasses so that the high frequency of raven finds on Iron Age settlements in Wessex is not surprising’ (Coy 1984c, 530).

In total there were sixteen of these ‘sandwich’ deposits, Green (1992, 126) stating the bones are all of ravens. However, the site publication states the ABGs are a mixture of raven, crow, buzzard and starling, with also some bones of hare (Drew 1931). As with Green’s other suggestions regarding Iron Age and Romano-British ABGs, she is utilising texts written after AD 1100 and, in the case of the Danebury corvids, applying ideas from them to ABGs mainly deposited around 450-350 BC. To put this time gap in context, would we utilise stories and mythologies written today to explain actions that occurred around AD450?

Although Coy (1984c) suggested a ‘functional’ explanation for the ravens at Danebury in the first faunal report for the site, in the later reports Serjeantson (1991a) offers a ritual interpretation Influenced by the work of Ross (1967), she suggests that corvids and ravens were sacred to the Celtic god, Lugus and may have had a ritual importance. Serjeantson (1991a) goes on to suggest that as scavengers ravens would have fulfilled an important role and therefore a combination of reasons for the ravens’ presence cannot be discounted. Recently Serjeantson and Morris (2011) have revisited the corvid remains and suggested that their biological character may predispose them to becoming symbolic creatures.

11.8

Meta-Level categories and the problem with ‘ritual’

As shown above, a large proportion of the ABGs have been interpreted using generalist categories, mainly for ritual explanations. This is related to the nature of ‘ritual’ as a concept. Handelman (2006) has pointed out that there is a meta-level ritual which encompasses all ritual activities. In effect, feasting, sacrifice and offering deposits are all separate ritual acts, which are classified under the general term ritual. There is also a meta-level concerning the functional/practical, with culling, disease, natural death etc, all part of the functional category of activities.

Cunliffe (1992) and Grant (1989a) also suggest that the ravens at Danebury may have a ritual purpose similar to the other animal ABGs. This change in interpretation between the 1984 and the 1991 publication of the Danebury faunal reports shows how attitudes to these deposits developed and changed in the intervening years. It also shows how the work of Anne Ross became more influential. For example in Cunliffe’s (1974) first edition of his Iron Age Communities publication, the work of Ross (1967) is not mentioned. The first time it is mentioned is in the third edition (Cunliffe 1991b) when ABGs are discussed.

The above analysis has shown that the meta-level interpretation of ritual is often given, but the meta-level of functional is hardly used. The rare cases in this study where ‘functional’ has been recorded as an interpretation have occurred when a number of functional/practical interpretations have been offered. Of the 474 ABGs interpreted as ritual in nature, the generic meta-level ritual/sacrifice explanation was utilised for 354 (74%) of them. Sacrifice has been added to the meta-level category because the majority of publications that use the metalevel explanation for ABGs alluded to the possibility that the animal might have been sacrificed. This is probably due to the influence of Grant’s (1984a) Danebury interpretation, in which it is argued that the deposits are ritual, and possibly sacrificial offerings. The majority of the Danebury text is concerned with arguing the deposits are ‘special’, and distinct from the ‘normal’ animal bone material. In some respects Hill (1995) is also guilty of

Green (1992) also sees the deposition of corvids, ravens in particular, in the Iron Age as a ritual act, although given the influence of Ross (1967) on her work, this is not surprising. Green (1992, 126) suggests; ‘Ravens may have been associated with pits and wells because of a perceived chthonic symbolism: ritual shafts penetrate deep underground, forming a line of communication between the living and the dead, the earth and the underworld powers. Ravens and crows, with their black plumage and their habit of feeding off dead things, were clearly seen as messengers from the Otherworld.’ Green also suggests that such a relationship may exist in the Romano-British period, using the example of raven ABGs from the Roman temple at Jordan Hill. Drew 165

Investigating Animal Burials; Ritual, Mundane and Beyond One of the main problems with the current interpretation of ABGs is that no one seems to know what ritual is. We as archaeologists are at ease in using the term, but very few of us have ever defined it, those that do often use vague concepts such as structured, repetitive, placed, purposeful, unusual, non-domestic. We have already seen that structured does not mean ritual (see 11.3), and it is impossible for archaeologists to reach a consensus regarding ritual if it is described in such vague terms.

this type of approach. Although very detailed discussions are present concerning the nature of the deposits and the possible reason for their deposition, in the end Hill (1995, 95) confirms that they are of a ritual nature. In some respects this is doing a disservice to Hill’s work, but we must consider whether most students, or commercial zoo/archaeologists operating to tight deadlines would fully study all of Hill’s arguments, or simply accept his conclusion that these deposits are ritual. It is understandable that a general concept has filtered into Iron Age archaeology that all ABGs are ritual deposits, and in due course this has also influenced the interpretation of deposits from other periods.

One of the main reasons archaeologists have such a problem in defining ritual is that many still associate it exclusively with religious and spiritual beliefs. For example Insoll’s (2004, 11-12) comments that many archaeologists simply substitute the term ritual for religious and he suggests ritual needs to be placed within its wider religious framework. However, social anthropologists have shown there are many different types of rituals. These can be secular, religious, classrelated, sex-related, personal etc (Bell 1992; 1997; Humphrey and Laidlaw 1994; Kreinath et al. 2006). Although rituals are often a part of religious practices, each ritual has a different meaning and purpose and many secular rituals also exist. Therefore we should not instantly equate ritual with religious. The characteristic that most ‘rituals’ examined by social anthropologists share is that the actions are formulaic, there is, in effect, a script (Snoek 2006). Perhaps we should just see ‘ritual’ as framing a formularised action/activity. If this is the case, then we could argue that a large proportion of the archaeological record was created by a ritualised act. However, using the term still keeps us at a meta-level of explanation.

The use of ritual as an interpretation is also related to archaeologists’ concept of it. Archaeologists have used the term ritual for two closely connected reasons, what is observed is non-functional and is not understood (Hodder 1992, 223). Functional is not utilised as an explanation on its own as it is understood. Therefore a sub-category is used, such as butchery waste. As ritual is not understood, this leads many archaeologists to use the meta-level ritual as an explanation in its own right. Hill (1995) suggested that ritual was embedded within everyday activity in the Iron Age. If ritual activity is embedded within prehistoric society then ritual as an independent act cannot exist. What ABGs do represent are specific activities, which have both functional and ritual elements. This point has been noted by other authors. Brück (1999) has argued that many societies have a monist rather than a dualist mode of thought; ritual and functional are not separate concepts. We view such concepts as separate because of our modern western outlook. Bradley (2003; 2005) has also suggested that throughout prehistory ritual and domestic life are intertwined and it is impossible to separate them. Pluskowski (2002) has noted that in the Medieval period the conceptual and physical were interwoven.

Brück (1999) suggests a way forward is to jettison ‘ritual’ and instead look at rationality. In effect, Brück is arguing that we should try to understand why people are ‘doing things’ without imposing our concepts upon the people. This is a useful suggestion for ABGs, as we need to move away from meta-level explanations of activities. However, ritual as an activity/concept/event does exist and it would also be a very hard task to remove ‘ritual’ from the archaeologist’s mindset. With regards to ABGs, ‘ritual’ is not a problem; it is the use of ‘ritual’ as both a description and an interpretation where the problem lies. To develop our understanding of ABGs we need to start looking at specific explanations regarding their creations

If this is the case, then the use of meta-level interpretations for ABGs is at best unhelpful. Hill (1995) does try to move beyond such interpretations by suggesting that ABGs represent the remains from feasting as well as possible offerings, with the domestic and the ritual intertwined at such events; however the majority of archaeologists have not been as successful, preferring meta-level categories to interpret the data. In effect, we as archaeologists are stuck in a loop of thought regarding ABGs. We recognise that the functional and ritual divide probably did not exist, yet we still need to explain why these deposits are present, and different, to the ‘normal’ faunal assemblage and are therefore constantly drawn back to vague ritual interpretations. This has led to a number of authors offering mixed interpretations (see 11.2). However, such approaches are as unhelpful as meta-level explanations, because they do not try to combine ritual and functional, they are simply offering alternative explanations. Neither meta-level nor mixed interpretations are actually telling us why ABGs were created. 166

Chapter 12. Assigning Meaning; Animal Biographies 12.1

A way forward

At present the interpretation of ABGs is stuck in a false dichotomy between ritual and functional categories. This is made more problematic when we consider that such a dichotomy is of our own making, and unlikely to have existed in many of the time periods studied. This, combined with the difficulty of moving beyond a metalevel ‘ritual’ interpretation has led to a confused mix of explanations for the presence of ABGs. Perhaps a way forward is to start investigating specific explanations for ABGs. To do this we need to change the way we view and study these deposits. At present the majority of ABGs are viewed in a single time frame, i.e. their final resting place prior to archaeological recovery. However, in interpreting the meaning of ABGs, archaeologists discuss activities which occur in a multitude of time frames. For example, Hill (1995) suggests that ABGs may represent animals which have been sacrificed, feasted upon and then possibly deposited as offerings. These are three separate events, which would have resulted in changes to the animal (recovered as an ABG) and all the events could have had different meanings and actions associated with them. In effect the ABG is the end result of an animal’s ‘life history’, as Appadurai (1986, 3) suggests; ‘It is only through the analysis of these trajectories that we can interpret the human transactions and calculation that enliven things. Thus, even though from a theoretical point of view human actors encode things with significance, from a methodological point of view it is the things-in-motion that illuminate their human and social context.’ A possible way forward in investigating ‘why ABGs?’ is to use a biographical approach. This draws on the work of Kopytoff (1986) who felt that ‘things’ could not be looked at just one point in their existence, but that the processes of creation, exchange, consumption etc, need to be looked at as a whole. Normally, archaeology looks at material cultural in what Gell (1998, 11) would describe as supra-biographical manner, looking beyond the ‘life cycle’ at longer chronological trends. This approach has been used in this study to examine the differences in the nature of ABGs between periods and site types. Such an approach is not without merit, as it allows us to develop a broad narrative of long term trends upon which timespecific information can be hung. However, the approach is not helpful in explaining why ABGs are present. As already shown, there are distinct trends in the interpretations offered to ABGs dependant upon the time period they are from; put simply, prehistoric and Roman = ‘ritual’, Medieval = ‘functional’. By using a biographical approach we can examine the activities that took place to create the ABG and their possible associated meanings on an individual basis, which can

then inform our understanding of supra-biographical trends. 12.2

Building ABG biographies

So far, biographical approaches have rarely been utilised for animal remains and when they have it has been on a supra-biographical scale (see Jones and Richards 2003). Most of these studies have been concerned with artefacts, such as pottery and metalwork or more personal objects thought to be heirlooms (for example Gell 1998; Gosden and Marshall 1999; Immonen 2002; Joy 2009; Lillios 1999; Whitley 2002). In general, the biographical approach allows artefacts to become ‘networks of significance’ (Thomas 1996, 159), with artefacts given ‘secondary agency’; they do not have the power to initiate happenings, but are objective embodiments of the power society or individuals have given them (Gell 1998, 2021). Such theories are just as relevant to human-animal relations. For example, consider contemporary western reactions to dogs (man’s best friend) and snakes (association with ‘evil’); both species embody different meanings and their secondary agency will cause very different reactions in humans. The study of the biography of artefacts is also the study of transition, as artefacts acquire different meanings throughout their ‘life’. Animals could be viewed as undergoing a large number of transformations. For example, when alive, a sheep may supply wool which would then be transformed to clothing. The sheep may in later life be slaughtered for meat, at which point part of it becomes food, and the bones or horns may become the raw material for an artefact. When these are removed from the animal, the meanings and agency of its parts are transformed. Therefore when we are examining ABGs we are not viewing the original animal, but the results of a transformation process enacted upon it. In investigating the biographies of animals, we need to look at the transformations that have occurred. In this respect zooarchaeologists are well placed. The study of faunal remains may be seen as atheoretical, a science drawing on aspects of zoology, creating facts which are accepted by archaeologists. Zooarchaeology is defiantly not atheoretical, however, the biological data upon which it is built can offer an advantage when constructing biographies. Humans do not physically create animals, but they can over a long period of time alter their skeletal morphology through domestication and selective breeding, and this can be observed by the zooarchaeologists. In effect the biological nature of animals, compared to other forms of material culture, offers us a baseline, upon which we can view the humanly created transformations. Therefore any alteration to the morphology of animal remains has been caused by either specific human or ‘non-human’ taphonomic action. Such actions can result in markers

Investigating Animal Burials; Ritual, Mundane and Beyond The life of an ABG will begin with the death of its constituting animal. Death may either be due to human hand, or natural causes such as old age, disease or accident. An animal may die naturally within an archaeological feature by simply falling into it. Although such ‘pitfall’ victims are often smaller mammals and amphibians, we have seen a number of larger animals interpreted in this way. An animal that dies naturally may also be subject to a form of human influence. A diseased animal may be buried to stop a disease spreading, as in the modern cases of stock afflicted with BSE and Foot and Mouth disease. Alternatively, non-diseased animals may also be buried whole due to socio-cultural reasons. With such instant burial we would expect a complete ABG to be formed, as no bistratinomic factors such as gnawing would have affected it. However this would not be the only way for a complete ABG to be deposited. An animal may be subject to processing such as skinning and still produce a complete ABG (see below).

upon the osteological material, which can be used to build up a picture of the events behind the ABGs deposition, a biography of the deposit. For a long time zooarchaeologists have been building supra-biographies of animals, investigating aspects such as herd patterns by using species proportions, age and sex patterns. This information can be used to create a background for the animals which became ABG deposits. However, this is assuming that these animals originated from the herds that also supplied the ‘normal’ faunal remains. Although archaeologically difficult to detect, Graeco-Roman literature indicates that sacred animals were kept in the vicinity of temples and used as a source of income for sacrifices (Gilhus 2006, 93). We do not know if such practices took place in prehistory, or indeed during the Romano-British period. However, the very high number of sacrificed goats from the Romano-British shrine at Uley, Gloucestershire (Levitan 1993), may have been drawn from such a herd, as goats only usually form a small percentage of the ovicaprid assemblage. The common factor in these examples is the association with religious buildings. However, with the exception of Hayling Island, all the sites recorded in this study seem to be of a domestic nature. It therefore seems likely that the majority of the domestic animals deposited as ABGs came from the same herds as those represented by the non-ABG deposits. Therefore, information from the ‘normal’ faunal assemblage can be used to provide background information for the ABG deposits.

The main point is that if an animal is buried instantly, we would expect a complete deposit. However, this does not mean that one will be encountered by the archaeologist, as post-burial taphonomic effects such as slumpage and intercutting may result in the separation of the bones. Fortunately, such effects may be visible archaeologically and therefore we can be aware when they are a factor, assuming they are reported. If an animal is subjected to human agency in the form of butchery and/or biostratinomic effects we could expect a range of ABG deposits to be formed on a sliding scale from a complete deposit to none at all (Figure 12.1). We would also anticipate that the more biostratinomic effects an ABG is exposed to and the longer the period of exposure, the less likely its survival.

As previously shown the description and interpretation of ABG deposits are often inseparable. In adopting a biographic approach we must make sure that description and interpretation are separate processes, in effect we must separate the ‘how’ and the ‘why’. We can do this by utilising the biological baseline the study of animal remains offers us. We can consider the processes that would have to occur to transform a living animal to a specific kind of ABG, therefore building a taphonomic history of the deposit. This history forms the foundation of the ABGs biography, upon which we can start to develop ideas concerning the meanings, the ‘why’ of the actions. 12.3

The majority of faunal material on archaeological sites is thought to consist of the remains of animals that have been exploited for their primary products. Such exploitation may also cause the creation of ABGs of varying type. This is not to assume that the models proposed are just concerned with ABGs produced through butchery practices. They are very deliberately taking no stance on the ‘functional’ versus ‘ritual’ dichotomy by removing human purpose and only considering human practice. As Hill (1995, 59) points out;

The taphonomic model

‘..ritual would have used the same technologies and practices as mundane (butchery etc). As such ritual draws from and reproduces the same generative principles as other social practices.’

It is probably safe to assume that the majority of ABGs discussed in this study have at some point been affected by human agency, particularly butchery. Most do not consist of complete skeletons but are partial. To be buried/deposited in this state they must have undergone some form of disarticulation, either naturally or by human hand. Therefore it is the taphonomic process that has created the ABG. To identify which processes are involved requires further investigation of the bones to look for evidence such as butchery marks. What the understanding of the taphonomic process gives us is the ability to make a number of assumptive models concerning the way ABGs can be formed.

Therefore the techniques used for disarticulation of the leg of an animal for deposition within an archaeological feature for ‘ritual’ purposes is the same as the disarticulation of the leg for meat processing. The reasons may be very different, but the actions and practices used for the processes are the same and guided by the technology available and the animals biological morphology. However, language does cause a problem in 168

Chapter 12. Assigning Meaning; Animal Biographies with the rest of the carcass by instant burial, we would expect a possible part-articulated complete ABG to be formed. However, the head and feet may have been taken away still attached to the skin, in which case we might expect a partial ABG to be formed consisting of the trunk and limbs. Also the head and feet may be deposited elsewhere resulting in another ABG (Figure 12.2).

this respect. The term ‘butchery’ is normally associated with the ‘functional’ production of meat for human consumption. Also, the arguments of Wilson (1992; 1999) have further galvanised the term as being associated with a ‘functional’ explanation for the presence of ABGs. It must be made clear at this point that when the term butchery is utilised in the below text, it is not to imply that the practice was taking place for a ‘functional’ or alternatively ‘ritual’ means. It is just being used to describe the physical carcass processing activities taking place. 12.4

If the carcass is not instantly buried, it may be subject to another form of taphonomic decay, which could result in further disarticulation. Therefore, we could expect either no ABG or a partial ABG to be formed. The partial ABG may consist of a number of different parts of the trunk and limbs, possibly damaged and less complete than a partial deposit formed through instant burial. The lower legs, if taken away with the skin, may also be deposited forming a partial ABG, which again may possibly become damaged and not as ‘complete’.

How humans create ABGs

The simplest way for human agency to form an ABG is the instant burial of an animal which has just died or been slaughtered, with no other processing taking place. If this occurred we would expect an ABG to be formed of the complete skeleton.

The skull is an element, which is very susceptible to fragmentation and damage by biostratinomic effects, and may be totally destroyed resulting in no formation of an ABG consisting of elements from the head, although some remains of the skull and jaws may be encountered.

However, other processes may also take place. The first could be primary butchery, normally involving skinning of the carcass (see 2.5). This provides a material for use in the form of skin and allows access to the meat and bone. At this point the carcass may also be subjected to further processes. It may then be deposited in a midden, or buried. If rapidly buried, we would expect the carcass to be protected from other taphonomic effects such as gnawing and therefore produce one of two different types of ABG. Working on the principle that the head and feet would have been removed, but the bones were deposited

The above descriptions have worked on the principle that the animal is first skinned, and a number of ABGs do display butchery marks which have been interpreted as skinning marks (for example, Clark 2002; for example, Coy 1991; Poole 2000b). However, a number of ABGs also consist of articulated complete-limbs, which could indicate that the animal was not skinned, or that skinning

Figure 12.1 General model concerning the creation of ABGs

169

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 12.2 Taphonomic model of the creation of ABGs from primary butchery If primary butchery including skinning does take place, the carcass may be subject to secondary butchery practices involving disarticulation. Therefore we can expect a fully articulated complete ABG not to be formed, but if all the waste bone material is deposited instantly together, it may form a partially articulated complete ABG. (Figure 12.3). A sheep/goat deposit formed in this manner was recently identified from Roman Winchester (Maltby 2010, 152-153). If the waste from the previous primary butchery process is removed and deposited elsewhere on the site, we could expect partial ABGs to be formed only from the trunk and upper limbs. If the elements are deposited and subjected to biostratinomic effects then an ABG may be formed, but it will be one that may be less intact compared to instantly buried ABGs, or so severely damaged and disturbed that it does not survive.

did not result in the disarticulation of the lower feet. The problem is that a skilled butcher may be able to skin an animal without leaving any butchery marks on the bone. Also, some activities do not require skinning. For example, spit roasting will only require evisceration of the carcass. Therefore we may expect elements such as the lower feet and skull to remain in articulation. However, if roasting did take place some of the body parts may become dismembered. Therefore, the animal may be deposited as a part-articulated ABG with burning possibly evident on the bones. Also, the animal may be subject to some secondary and tertiary butchery such as dismemberment, without having been skinned, but the same dismemberment techniques would still have been utilised. We could therefore expect the ABG to have been formed through secondary butchery. The possible difference being the inclusion of the skull or lower feet which may have be removed during skinning.

The dismembered material may then be subject to further tertiary butchery. This involves the reduction of meat

170

Chapter 12. Assigning Meaning; Animal Biographies

Figure 12.3 Taphonomic model of the creation of ABGs from secondary butchery groups, as was noted by Binford (1981, 91-92). Again if all the waste bones are deposited together then a partially articulated ABG may be formed, similar to those formed by secondary butchery, although probably suffering from a greater degree of disarticulation.

packages to smaller portions. As discussed in Chapter 2, tertiary butchery is mainly concerned with producing cuts of meat, sometimes involving stripping meat off the disarticulated joint, as indicated by the presence of filleting marks. If meat is filleted off the bone, with no further disarticulation, then we could expect tertiary butchery to produce similar ABGs to secondary practice. However, the limbs and axial skeleton may be further disarticulated during tertiary butchery. If this was the case we could expect ABGs to be formed from the limbs and trunk, but on a reducedscale compared to ones produced by secondary butchery. Vertebrae and ribs may be dismembered into

Some of the bone may be subjected to further processing such as marrow and grease extraction and bone-working. If this type of processing did take place, we could expect the chances of an ABG being formed to be greatly diminished. If one was formed under these circumstances, it would probably be from the feet, head or trunk. This is because limb bones have the highest marrow content, 171

Investigating Animal Burials; Ritual, Mundane and Beyond It is not possible to re-examine the ABGs from all the sites included in this study. Rather, a number of different deposits have been selected and their interpretations revisited using the biographical approach advocated above.

particularly the upper elements (Munro and Bar-Oz 2004; Outram 2001). Also, the limb bones are often used to produce artefacts, especially if they are required to be long and thin (Lemoine 2001). Although all elements can be utilised for bone tool/ornament construction, the elements which are utilised are dependant upon two factors, the match between the morphology of the bone and the tool that is required, and the cultural choices concerning what tool/ ornament is needed. These factors are likely to vary in each society, but the limb bones do seem to be the elements which are utilised the most in societies around the world. 12.5

12.6.1 Windmill Hill outer ditch section V The largest assemblage of Neolithic ABGs recorded in this study is from Windmill Hill (see 3.4). All come from the ditches of the monument, with three recorded from outer ditch section V, one from fill 227 and two from 210.

Assigned meaning Fill 210 is one of the top-most primary fills of the ditch; it consists mainly of small rounded chalk fragments, and the profile indicates it was formed as part of a silting process as the ditch was left open. A calibrated age range of 3630-3500 and 3420-3380 BC was obtained for fill 229, which 210 overlies. A spread of bone material 4 metres wide was recorded from fill 210. Within this spread, small amounts of flint waste and sherds of plain pottery were recorded (Whittle et al. 1999a).

The above discussion indicates the actions which are required to form ABGs. However, the meaning behind such actions has purposely not been discussed. The above models provide us with a way of constructing the narrative framework of an ABG biography. We therefore know the likely processes that created the ABG, and to this we can add the general background information regarding the herd structure from which the animals came.

Two ABGs were recorded from this bone spread. A partial cow consisting of four foot elements (it is unknown from which leg), and a partial sheep/goat comprised of the femur and tibia (the side is unknown). The fusion of the sheep/goat elements suggests the remains come from a juvenile individual (2-12 months old). The rest of the faunal assemblage is fragmented and disarticulated.

One important point is that the above models do not take into account post-burial taphonomic factors, such as slumpage. These factors may result in post-depositional damage and, in some cases, complete destruction of the ABG. However, zooarchaeologists in a number of cases have been able to reconstruct ABGs in the postexcavation stage. Such effects will also leave traces on the ‘normal’ faunal assemblage, so we are able to take them into account. But when they do occur the ABGs we are examining will bear little resemblance to those originally created and deposited.

Using the models discussed above, the evidence indicates that the two ABGs have probably undergone different levels of processing. The cattle foot bones are likely to have been removed from the carcass during the primary butchery phase. As the rest of the carcass is not present within the deposit, it may have been subject to further processing. The sheep/goat ABG would be from a carcass that had undergone secondary processing, resulting in the disarticulation of the limb from the axial skeleton. Again the other remains from the carcass were not deposited within the same context. Butchery marks are not present on either of the ABGs. The fact that these two ABGs have survived would suggest that they were not present on the ground surface for long. Although fragmented, none of the other faunal material associated with these ABGs displayed evidence of gnawing. Compared to the faunal material from the other contexts in the ditch, the bones have a large mean size of fragment (Grigson 1999). This would suggest that the material is a primary deposition that may have been quickly covered over, explaining the survival of the ABGs.

How then do we go about assigning meaning to ABGs? If we look at the previous interpretations offered for ABGs that have gone beyond meta-level categories, three important characteristic are utilised, composition, context and association. Composition is an important factor in the interpretation of fall and culled victims (see 11.5 and 11.7). Context is important for foundation offerings. Association with other ABGs, and deposits such as human remains, is important in the ‘offering’ interpretation. What the biographical approach gives us is a way of investigating how the composition of the ABG came about and the chronology of its context and other material associations. Therefore, it is important to not only investigate the biological nature of the ABG, but to integrate this with its associated archaeological information. What this approach is also advocating is individual rather than group interpretation of ABGs from a site. 12.6

Fill 227 overlies fill 210 and appears to represent a thin depth of material. Within this context is a larger bone scatter consisting of 179 fragments. The majority of the identified bone is from cattle, including 10 ribs. Present within this scatter was a dog ABG consisting of six foot

Looking again at ABGs

172

Chapter 12. Assigning Meaning; Animal Biographies impose our own sacred/mundane worldview upon the evidence.

bones (not known which leg). It is suggested that the cattle remains are all from the same animal, and have been exposed long enough for dog gnawing to take place and for the elements to become fragmented (Whittle et al. 1999a). The mean fragment size is smaller than in the previous context (Grigson 1999). Also, four of the nonABG bone fragments are burnt. It would appear that the faunal material from this context could have a number of different possible depositional histories. The non-ABG faunal material consists either of primary deposits, which have been exposed on the surface of the ditch fill for a period of time, or secondary deposition. The fragmented and abraded nature of the pottery also indicates a protracted period of above ground exposure.

The problem with interpreting these ABGs from Windmill Hill is linked to the general difficulties we have with interpreting sites from this period. A number of different explanations have been put forward for the functions of causewayed enclosures, varying between the two extremes of purely domestic sites and ceremonial centres (Andersen 1997, 242-267). Commonly, causewayed enclosures are seen as aggregation sites where public events took place as Neolithic society became more diverse (Bradley 2007, 74-75). Such events appear to include the excarnation of human remains within the causewayed ditches (M Smith 2006), for which there is possible evidence at Windmill Hill. Smith (1965, 17) noted that some human bones had been exposed for some time as snail shells have been found inside them. Deposits of both human and animal remains have led authors to suggest that a number of activities took place on causewayed enclosures including feasting, animal sacrifice, offerings of food and celebrations of the dead (Bradley 2007, 74).

If the dog ABG is a primary deposit, then the bone material may have been covered soon after deposition. Alternatively it represents the secondary deposition of ‘midden’ material. Keiller’s archive does suggest middens occurred within the interior of the inner circuit of ditch (Whittle et al. 1999c). The foot bones are some of the first elements to become naturally disarticulated from a carcass. If these elements represent part of a deposited carcass, we might expect other dog remains to be present within the associated fills but none are. This suggests the ABG was a primary deposit and may have been created through primary butchery, possibly associated with skinning. It is not known if the carcass would have been processed further, although dog-meat consumption appears to have taken place infrequently in prehistory.

What we do know is that the ABGs from this ditch are different to the non-ABG faunal material. The reason for this is that ‘normal’ carcass processing activity ceased at a certain point resulting in the formation of the ABGs. Also, it would appear that the ABGs represent primary deposition, while the accompanying non-ABG material may be secondary. The ABGs are also very different in composition. The cow foot and sheep/goat upper hind limb ABGs in fill 210 represent different parts of a carcass. The cow foot-bones would have very little meat value. If this ABG was created during a feasting event, it probably represents the discard of a portion of the animal not required for food. In contrast, the sheep/goat ABG represents a body part with a high meat value. If the meat was left on this ABG, its interpretation as feasting waste should be questioned. Perhaps it does represent a food offering. However, why such an offering was placed with secondary deposits of faunal remains that have clearly been fully processed is unknown. The dog-foot ABG from fill 227 is also unlikely to represent a ‘food offering’ as again very little meat would have been present.

All three ABGs described above were interpreted by Grigson (1999) as deliberate placements, not the remains of meals, since consumption requires dismemberment. However, to create these ABGs, butchery has taken place. However, in two instances only primary butchery is required to create the dog and the cattle ABGs, as both are from areas with very little meat value. Secondary butchery would have taken place on the sheep/goat ABG. However, we do not know if the meat was then filleted from the bone before deposition. The authors (Whittle et al. 1999c) do note that the foot ABGs may have been associated with skinning and suggest that such finds could represent the deposition of hides such as those associated with funerary structures (see 3.5). However, if this was the case, we would expect other foot and possibly head remains to be present with the deposits. The assertion that the remains were ‘placed’ as opposed to dumped within the ditches, is used by the authors to suggest a quasi-spiritual symbolic aspect to the deposits. Both sacrifice and feasting are suggested as interpretations for the bone material (Whittle and Pollard 1999) and indeed the two are not mutually exclusive. The ABGs discussed above may certainly have been from animals killed by humans. However, we have at present no means to identify whether the animals were sacrificed, although we could theorise that meat may not have been regularly consumed and therefore every animal killed for consumption could be interpreted as a sacrifice. Indeed to talk of sacrifice as opposed to ‘normal’ killing is to

Currently we do not know exactly why these ABGs were deposited at Windmill Hill and a biographical approach is not going to provide a fully conclusive explanation for their presence. What it has shown is that there are significant differences between the three ABGs from outer ditch V, as well as the differences between the ABGs and non-ABG faunal material. It would appear therefore to suggest a single interpretation for all the faunal material, as Whittle and Pollard (1999) have done, is to lose sight of the different events which have created the assemblage. As different events and processes created each ABG and the non-ABG faunal material, perhaps they also have different associated meanings as well.

173

Investigating Animal Burials; Ritual, Mundane and Beyond 12.6.2 Suddern farm pit 197/7

were deposited in a concentrated group close to the southwest wall of the pit. Although deposited together, they have all undergone different processing. The partial ABGs in the group would have been created by secondary butchery, and they also all appear to be from adult individuals. In comparison the complete pig is a juvenile, and does not appear to have been subject to any processing.

The middle Iron Age ABGs at Suddern Farm (Cunliffe and Poole 2000b) show a different pattern and process. One of the largest multi-ABG deposits recorded in this study comes from fill 7 of pit 197 at Suddern Farm (see 4.7.1). Fill 7 is the second lowest layer of the pit, situated above basal fill 8, which also contains partial cattle and horse ABGs. In total 17 ABGs were recorded from the fill (Poole 2000d). Seven are of horse, five of cattle, four of sheep/goat and one of pig. The pig and one of the sheep/goat ABGs consist of complete skeletons. Unfortunately, detailed osteological information is not given for the individual ABGs. It would appear that none of them have butchery marks present, although considering the scale of the deposit this is surprising and a re-examination of the faunal material would be of interest. However, we do have element and positional information, which allows us to examine the development of this deposit.

The second group of ABGs were deposited towards the centre of the pit. They are probably later (although by how much is unknown) as deposit A overlies deposit B. These ABGs have a different make up. They are all partial and all comprised of axial elements (Table 12.1). The two sheep/goat both consist of thoracic vertebrae and ribs and would have been created by secondary butchery. The two horses consist of the skull and cervical vertebrae and may have been created during the primary butchery phase. Deposit A, a partial cattle ABG, consists of the ribs, thoracic and lumbar vertebrae, pelvis and right upper hind limb. It would appear to have undergone primary butchery and some aspects of secondary butchery.

Using the excavation records it is possible to split the ABGs into four or possibly five different depositional events. These may represent the sequence of deposition, but they give no indication of the timeframe, and the deposition may have occurred as a single act.

The third group of five ABGs overlay the second group and were therefore deposited afterwards. One of the cattle and both horse ABGs consist of limb bones. Partial cattle ABG T is comprised of thoracic vertebrae and ribs. These four ABGs all have undergone secondary processing. In contrast, the sheep/goat consists of a complete ABG and does not appear to have been butchered.

Some of the first ABGs to be deposited were those labelled in the excavation B, U, C and D (Figure 12.4 & Table 12.1). These consist of (B) a partial cattle ABG made up of a humerus, radius, ulna and metacarpal (the side is known); (U) an articulated leg of a horse (possibly lower hind limb); (C) a partial horse ABG, consisting of lumbar vertebrae still in articulation with the pelvis; and (D) a complete juvenile pig ABG. All of these ABGs

A possible fourth phase of ABG deposition consists of two horse skulls, both of which have the axis and atlas in articulation. Deposit N possibly overlay Deposit M from the previous phase. These ABGs may have been created via primary butchery processes. The phasing of deposit R is unknown. It is a partial sheep/goat ABG comprised of thoracic vertebrae and ribs. This ABG is in close association with four quern stone fragments which have been interpreted by the authors as a ‘special deposit’. Although viewed as separate deposits, some of the ABGs may be from the same animals. Unfortunately, this aspect has not been investigated within the report. Using the body areas represented, it is possible that two horse ABGs in the first group of depositions (Deposits U and C) are from the same individual. If deposition groups 1 and 2 were in fact deposited in a single event, then the partial cattle ABGs, B and A, may also be from the same animal (Table 12.1). It is also a possibility that the cattle ABGs G and T are from the same individual. The horse appendicular ABGs J and M may represent parts from the same animal as one of the partial axial horse ABGs. If this is the case than the ABGs in this layer would represent (a minimum of) five horses, four sheep/goat, two cattle and one pig. The reporting author sees the ABGs within pit 197/7 as one ‘special deposit’, and it indeed may represent one act, for which it is possible to extrapolate the sequence of

Figure 12.4 Plan of layer 7 within pit 197 at Suddern Farm. The letters indicate the separate ABGs (Cunliffe and Poole 2000b, 5:C5) 174

Chapter 12. Assigning Meaning; Animal Biographies Table 12.1 Summary information of ABGs from Suddern Farm pit 197/7 Possible order Microfiche Butchery of deposition No. Species Complete/Partial Body areas stage 1 B Cattle Partial Leg Secondary U Horse Partial Leg Secondary C Horse Partial Axis Secondary D Pig Complete None 2 A Cattle Partial Axis + leg Secondary? E Horse Partial Axis + head Primary? H Horse Partial Axis + head Primary? F S/G Partial Axis Secondary K S/G Partial Axis Secondary 3 G Cattle Partial Leg Secondary T Cattle Partial Axis Secondary J Horse Partial Leg Secondary M Horse Partial Leg Secondary S S/G Complete None 4? L Horse Partial Head Primary? N Horse Partial Head Primary? ? R S/G Partial Axis Secondary We might therefore put forward the possibility that this large deposit of ABGs represents a number of processing activities to produce a large quantity of meat, possibly for feasting. However, this explanation would not account for the complete articulated ABGs which are present. Maltby (1981a) has suggested that large deposits of cattle and horse remains may represent unusually large butchery events with the need to bury the resulting obnoxious waste. This may explain the partial ABGs but again not the complete ones. In addition, the pit is in the middle of the settlement not on the periphery.

depositional events. The justification of the interpretation of the ‘special deposits’ from all the Danebury environs excavations is particularly flimsy. The reader is merely referred to the Danebury report as an explanation of their presence (Poole 2000d). They appear to be interpreted utilising the general meta-level of ritual. The presence of such a large number of ABGs is highly unusual. Interestingly, none of the fills above layer 7 contain any ABG deposits. They do, however, contain over 100 fragments of faunal material, which continued to be deposited within the feature after layer 7. It is notable that a large proportion of the ABGs consist of vertebrae, which is a pattern also seen in the non-ABG assemblage. When the elements from the cattle ABGs and non-ABGs for layer 7 are compared, a similar body area pattern is seen (additional data provided by Julie Hamilton) (Figure 12.5). Both assemblages are dominated by vertebrae, skulls and lower limb elements, indicating that similar body areas are being deposited, some as ABGs, but others as ‘normal’ faunal material. Analyses of the sheep/goat and horse faunal material show a similar correspondence between ABGs and non-ABGs elements.

However, is it necessary for all the ABGs from the same context to have been deposited for the same reason? The previous discussions would suggest it is not. Why then are the complete ABGs present? Several explanations could be put forward. Perhaps they represent animals which for some reason were unfit for consumption; they could be offerings; or they could represent a form of conspicuous consumption. The majority of the partial ABGs would have been produced by secondary butchery processing. Therefore some time and effort has gone into dismembering the animals represented in this deposit. Why was this not carried out on the complete ABGs?

It is possible that the non-ABG faunal material was originally deposited as ABGs and became disassociated through post-depositional activity. If soft tissue was still attached to the carcasses recovered as ABGs, which it likely was for the complete skeletons, the layer contained a large amount of organic material, which would have decomposed over time resulting in some movement. It is worth noting that within the Suddern Farm report, the non-ABG assemblage is seen purely as waste and not examined in association with the ABGs. The common pattern seen in the two types of faunal assemblage is the relative lack of upper limb bones, the areas that have the highest concentrations of meat. We could therefore theorise that these were removed from the carcasses, processed for consumption and then deposited elsewhere.

The most obvious difference between the complete and most of the partial deposits, is the complete ones are from smaller species. Within the report these ABGs are stated to be complete, although it is unknown whether all elements are actually present. It is possible that the complete ABGs represent animals whose meat was consumed, but had been spit roasted. If this did occur, an almost complete ABG may still be deposited. We could expect some of the lower-foot elements to be missing and possibly evidence of charring of the lower extremities. Unfortunately such information is not available and the zooarchaeologist was not able to examine all the ‘special’ deposits from the site. This is because assumptions had 175

Investigating Animal Burials; Ritual, Mundane and Beyond

Figure 12.5 Diagram showing the proportion of cattle ABG and non-ABG elements from Suddern Farm pit 197, calculated to the most common element mail coat had been laid over the inhumation and two partial pig ABGs were also included (Figure 12.6). One consisted of a head and upper and lower forelimbs positioned close to the head of the inhumation. The other of an upper and lower forelimb placed close to the inhumation’s knees, on top of the mail coat. It is probable that the pig remains all came from the same animal as both the right and left forelimbs are present. In total 25 pig elements were present, with tooth wear and epiphyseal fusion indicating they came from a sub-adult animal. Knife marks were present on the upper front limbs, indicating they may have been defleshed before deposition. Legge (1991a) suggests the defleshing of some elements may indicate the offerings are symbolic. Cunliffe (2005, 548) views the ABGs recovered from Arras burials as food offerings, or in the case of the pig heads, offerings for the spirit of the deceased. If we consider the biography of the deposit then a more detailed picture of the events behind its creation can be theorised.

already been made regarding the nature of these deposits. There is currently nothing to disprove that all the ABGs may represent the remains of a feast. This could also explain why the animals were not fully processed, as perhaps they represent an event in which cooking processes different to the norm were used. Such a conclusion is at this point merely speculation; further research would be required to investigate it more fully, including re-examination of the bone material, and possibly experimental archaeology to test the assumptions. What is certain is even with the current amount of data we are able to move beyond our metalevel interpretations and begin to ask more interesting questions, even if we do not yet know the answers. 12.6.3 Kirkburn burial K5 The middle Iron Age site at Kirkburn, Yorkshire produced very different types of ABGs (see 5.4). The site consists of a number of inhumations of the Arras tradition, some with ABGs present. If we take the example from one particular gave (K5) we can see that the ABGs have undergone a number of different transformative process.

During the construction of the grave a young pig was chosen for slaughter. The choice of species must have been significant, as only pig remains appear to be found as ABGs in Arras culture burials (see 5.4). Pig remains are poorly represented in ‘normal’ faunal assemblages from this region during the Iron Age, with cattle and sheep/goat the most commonly recovered species (Hambleton 1999, 47). The pig would have been killed, most probably following the same methods as slaughter for ‘everyday’ consumption. However, we do not know if the act of slaughter would have had a ritualistic meaning

This grave consisted of a chariot burial and the inhumation of a 25-35 year old male. The wheels of the chariot had been dismantled and the body laid at the junction between them. Copper alloy objects including horse equipment were deposited in the grave. An iron 176

Chapter 12. Assigning Meaning; Animal Biographies and Huntington 1991; Oestigaard and Goldhahn 2006). Therefore, if we consider the biography of the deposit it undergoes a number of different transformations from a living domestic animal to food suitable for the living, food for the dead and possible ‘waste’. These transformations allow a greater insight into the activities and possible meanings behind such a deposit. 12.6.4 Rudston Roman villa pit 47 A different example of ABG deposits can be seen at the Rudston Roman Villa site (see 7.4) (Stead 1980). Present on the site are a number of early Romano-British round house structures, the outlines of which are visible via their associated gullies and post-holes (Stead 1980, 2123). Associated with these structures are a number of ABGs, many of which were deposited within the floors of, or adjacent to, structures. Pit 47 is situated at the terminus of one of the gullies. Deposited within the pit is a partial sheep/goat ABG (described in the report as burial 8). The deposit is almost complete consisting of the skull, vertebrae, ribs, pelvis, both fore-limbs and both upper hind-limbs. Only the lower hind-limbs are not present. This indicates that the carcass was subjected to only a limited amount of butchery, if any, before deposition. The lower hind-limbs are the first elements to become naturally disarticulated in sheep/goat. Therefore the ABG may represent secondary deposition of the carcass. However another process has also occurred. Both the lower fore limbs and the left tibia are burnt.

Figure 12.6 Plan of burial K5 from Kirkburn. The pig ABGs are coloured grey (Redrawn from Stead 1991, figure 127)

Burning is not often reported on the ABGs recorded in this study, although this may be due to the quality of the data. Interestingly another sheep/goat ABG from Rudston, also deposited in association with the roundhouse structures, displays similar signs of burning, as do some of the ABGs from Shiptonthorpe in the same region (see 7.3 and 7.4).

attached. The carcass was then skinned and eviscerated. Further processing took place to remove the pig’s head and forelimbs. We do not know what happened to the rest of the carcass. Slaughter of such an animal may have been a relatively rare event and the meat from it may not have been wasted. Therefore most of animal would have been transformed into food for the living and after consumption the axial and hind limbs may have been disposed elsewhere as ‘normal’ rubbish.

The ABG from pit 47 is the main deposit within the feature, and only seven other non-ABG bone fragments were recovered from the pit. It would therefore appear that the pit may have been deliberately created for the sheep/goat. This would tie in with one of the interpretations offered: Stead (1980, 23) views the ABGs as either ‘ritual’ burials or foundation deposits/burials. The association with buildings and the apparent deliberate creation of the pit point towards the latter explanation, the first one being the standard meta-ritual explanation. However, it does not explain the burning on the ABG. The burnt bones are described as charred, which occurs when they have been exposed to flames, rather than deposited in embers (Gilchrist and Mytum 1986). It is therefore very likely that the carcass had been subjected to some form of roasting. The domination of pots in the Iron Age and on rural Romano-British sites, suggests that the staple food was a form of stew (Cool 2006, 165). Roasting appears to have been rare. If whole

However, the head and forelimbs were kept apart. Butchery marks on the upper forelimbs indicate that some of the meat may have been stripped from them. This meat may have been consumed with the rest of the animal, or possibly in a separate event linked to the construction of the grave. Eventually, the remains were carefully placed within the grave, their position on top of the iron coat indicating they were some of the last elements to enter the grave. As suggested, the elements may symbolically represent food for the deceased. However, considering that the rest of the animal was consumed, perhaps these represent the deceased’s share of a feasting event linked to the creation of the grave. Funerary activity is as much about the living as the dead, with events incorporating the creation and modification of social connections (Metcalf

177

Investigating Animal Burials; Ritual, Mundane and Beyond carcasses were roasted, then it is possible only the meat was taken from the carcass and connective tissue left. The ABG may therefore be representative of an event, in which an animal was cooked in a manner not often employed. The association with the structure is also significant. Perhaps the sheep/goat was roasted in association with the structure’s construction, as part of a celebration for the members of the society who aided in its construction. If this is the case, then it is perhaps not the deposition of the carcass that is important, but its treatment prior to burial. However, placing its remains so close to a building in a small pit is not how the majority of faunal material from the site was deposited. The deposition of the remains of the carcass may have been part of the celebrations, a mnemonic activity in association with the structure. Figure 12.7 Model of the changes and process behind the creation of the partial cat ABG from pit fill 6473

Although similar arguments are utilised for some of the ABGs from Suddern Farm, the scale of the event and its possible associations are different. This interpretation of the ABG is not dismissing of Stead’s conclusion that this ABG represents a ‘foundation deposit’. Rather, it utilises the data more deeply to give a picture of the possible events involved in the ABG’s creation.

remained between the skull and vertebrae resulting in the articulated ABG deposit. In comparison all the foot elements were disarticulated, suggesting little connective tissue was present, or they become disarticulated whilst being removed from the pelt. Because of the lack of element data for the ABG from 6947 it is unknown if it underwent the exact same process. Certainly the butchery marks and presence of the skull suggest it was skinned. The complete ABG in pit fill 6347 underwent a slightly different process. Rather than being brought to the site as a pelt the animal represents either a complete carcass brought to the site or an animal that lived in close proximity. The different composition of the deposits shows how slight variations can produce different ABGs.

12.6.5 Coppergate, York The late Anglo-Scandinavian assemblage recovered from Coppergate, York (O'Connor 1985; 1989) (see 9.3) include excellent examples of the transformation processes ABGs undergo. The site is unusual in that it has a number of cat ABG deposits. The cat ABGs consist of a complete skeleton from pit fill 6347, a deposit of skull and three cervical vertebrae from pit fill 6473 and an unknown partial deposit, although we know the skull is present, from pit fill 6947. Butchery was present on the skulls in all three deposits, consisting of parallel knife cuts immediately above and between the orbits. Their composition, together with the butchery marks and the high proportion of foot elements in the non-ABG assemblage led Terry O’Connor to suggest the cat remains, including the ABGs, represent the collection of cat skins at one of the tenements (1989, 186). The lack of cat long bones compared to skull and foot elements within the pit fills on the site would suggest that the cat remains deposited here represent the end of a transformative process.

The one aspect all the ABGs have in common is the skinning process and the removal of the bones, transforming the pelt into a usable cat fur skin. This would have then undergone another series of transformations eventually becoming a garment. It is during these transformations the meanings of the objects are created and changed. Therefore, the cat goes from being a useful commensal species that is possibly tolerated, to a clothing garment and waste. Neither object may have maintained the animal’s original agency and it is the above ground transformations that are of primary importance. 12.6.6 Winnall Down pit 6596 revisited, again!

During this period cats may be viewed not necessarily as pets but as a commensal species. The cats that formed the ABGs had probably been living within the settlement, at which point they either died of natural causes or were killed by humans. The partial ABG from pit fill 6473 was skinned, with the skull and foot bones left attached to the pelt. This process transformed the animal into two products, a waste carcass and a fur pelt. The pelt was transported to the area around Coppergate, where the skull and feet were removed. Some connective tissue

It is perhaps fitting finally to re-examine the ABGs from the same site and feature as Hill did at the end of his study, pit 6596 from Winnall Down, Hampshire. He suggested; ‘Pit 6595 is evidence for a specific practice through which key definitions and knowledge about the world were socially defined through ritual. It is evidence for a single event in which 178

Chapter 12. Assigning Meaning; Animal Biographies that the non-ABG faunal remains Hill is describing come from four separate, identifiable contexts, and we are unaware of the time depth that would have been involved in their creation. Finally, only one bone of the ‘feasted upon’ hare is present in the assemblage. Such a critique shows the dangers of selecting only small proportions of zooarchaeological information, without examining the whole dataset.

relationships between individuals, groups, age sets and genders would have been reproduced’(Hill 1995, 127). He goes on to indicate a communal feast and sacrifice took place which involved the consumption of over twelve cattle and horses, a sheep, a pig and a hare. However, there are a number of problems with Hill’s interpretations of this feature and associated activities. The excavation report indicates two ABGs are present from the pit in layer 6731. One is the complete skeleton of a sow, aged a little over two years old. Butchery marks were present on the right astragalus, suggested to be the result of skinning, or the start of butchery processing that was abandoned. The other ABG within the same layer is that of a complete adult female dog. In Hill’s (1995, 70 & 127) study the dog ABG is discussed during the analysis, but is not mentioned in the revisit. As both ABGs are complete, it would appear that meat was not removed from them for a feast. The twelve cattle Hill refers to are not strictly ABGs. An MNI count of 12 was derived from disarticulated mandibles from the first four layers of the pit, the topmost of which contains the ABGs. The pit does contain the largest number of non-ABG cattle remains from Winnall Down (77). However, examining the elements present shows that the majority consist of skull fragments, mandibles and loose teeth, no other elements make up more than 5% of the assemblage (Maltby 1985a, 100).

By trying to assign a general explanation to all the material within the feature, Hill missed some of the interesting aspects of the ABGs. Both ABGs were deposited in the same layer and in close association. Yet they have very different histories. The female dog had lived into adulthood and may have produced a number of litters. At some point in its life it had fractured its left femur. However, the dog survived and the injury fully healed, although the limb was left distorted and the animal would have had a limp for the reminder of its adult life. The injury may have been so severe it required care, although in aspects of caring for sick and injured animals our interpretations may be clouded by our modern-day viewpoints. Eventually the dog died or was killed and deposited in the pit. In contrast, the pig deposited in association with the dog had only lived a little over two years. Its cause of death is unknown, but the presence of butchery marks may indicate the animal was deliberately killed. Certainly primary butchery of the carcass was started, if not finished, before it was then deposited within the feature. The non-ABG faunal remains present within the fills are of a very different nature. To start with, the bones are of a fragmentary nature, having undergone at least secondary butchery processing. The gnawing of some of the fragments indicates that they were exposed for a period of time before final deposition. Significantly perhaps, a layer of chalk rubble was deposited over the ABGs after deposition.

Hill also discussed the hare remains suggesting; ‘The event involved the hunting and probably consumption of a hare, which as I have suggested, would not have been a common event, but rather one surrounded by ritual and taboo, even one in which certain sorts of people would have been involved’ (1995, 127) No hare ABG is present within the feature, there is, as Hill (1995, 70) did mention in his analysis, one hare bone within the very bottom fill of the feature. We must ask ourselves: Does one bone, a hunt and feast make?

We therefore have a number of different types of faunal material deposited within the same layer. They may all end up in the same place, but they have each undergone very different forms of human agency before getting there.

There are therefore a number of problems with Hill’s interpretation of the feature. He is suggesting that the pig was feasted upon. Yet it consists of a complete skeleton, which appears to have been deposited with flesh still attached. It may have been roasted but we have an accurate report of the bones which indicates none were burnt. It is suggested that at last 12 cattle were eaten, but the majority of the cattle remains are those from the head and loose teeth. If the feasting remains were ‘ritually’ deposited, would we not expect the majority of the skeleton to be represented, especially if this was one event, or find evidence for different cooking methods, as may have been the case in the examples discussed above? Canine gnawing was present on the non-ABG assemblage, suggesting some of the assemblage represents secondary deposition. We must also consider

The deposition of complete dog and pig ABGs in association with each other is of interest. The review of Iron Age ABGs showed that these two species are the most likely to be deposited as complete ABGs. However, pig ABGs are normally neonatal or juvenile (see 4.5). Perhaps it was simply a convenient depository for both animals that died at the same time. Alternatively they could both be sacrificial offerings, but if this was the case, and they are associated with agricultural fertility (Bradley 2003; Cunliffe 1992), why are they not present at the bottom of the pit and why have three separate layers of ‘rubbish’ been deposited before the offering? If they were sacrifices, then they differ from those described in Greek and Roman religion (Gilhus 2006, 115), where a part of the animal was burnt and the rest eaten (or sold). Discussing the Iron Age French site of Gournay, Brunaux 179

Investigating Animal Burials; Ritual, Mundane and Beyond (1988, 120) suggested the entrails were left as offerings and the people ate the rest. On the other hand, instead of attempting to provide a single explanation for all the ABGs, should we not again consider that the ABGs were deposited for different reasons? It has been argued that dogs and humans have a close relationship. Perhaps the dog ABG represents a burial and the pig ABG a suitable offering. 12.7

Through the adoption of a biographical approach to ABGs, we can start to examine the differences in their composition, which in turn leads us to explore the differences in activity and associated agency involved in their formation. The reader may therefore wonder why such a detailed, yet supra-biographical, investigation of the nature of ABGs was conducted in the second part of this study, when a more detailed approach to a small number of sites would be sufficient. However, it is through this detailed study that the inconsistencies in approach and problems associated with the interpretation of ABGs became apparent. This study is also not arguing that trends in the ABG assemblage do not exist, as it has shown that they do. Rather, it has demonstrated that patterns and trends exist for different ABGs, although there is not a set series of ‘types’. There are similarities between some ABGs, which suggest similar practices and meanings are associated with them. Therefore we can discuss the possible reason for dog multiple ABG deposits in Romano-British towns, as long as we are aware that the processes creating them and therefore the associated meanings differ to those of domestic fowl ABGs in a funerary context, or to those of partial cattle ABGs in Neolithic ditches.

What are ABGs?

Such conjecture therefore leads us to the crux of the matter: what are ABGs? In fact, such a question is easy to answer: ABGs are archaeological constructs; they are a category applied to archaeological material. They have been ‘created’ by a scientific approach to archaeology and zooarchaeology in particular, which generates knowledge by engagement with the world through categories. Such an approach is not problematic, as long as we are aware of it. People in past societies did not go out and deposit an ABG. They carried out a number of acts and the associated agency resulted in the deposition of animal remains in variable states of association. Dependant upon the post-depositional taphonomic processes they undergo, such deposits may then survive and be recovered by archaeologists, who categorise them as ‘special deposits’, ‘animal burials’, ‘something weird’ or indeed ‘ABGs’.

Therefore the question we should be asking is not ‘what do ABGs mean’, but rather what does this ABG mean? However, this still leaves us with the problem of applying meaning, albeit on an individual-scale. However, this makes the task less daunting. By using and integrating all the available archaeological data, a ‘life-history’ of a particular deposit can be constructed, which in turn develops explanations for ABG creation that is beyond the meta-level. We must also be aware that the meaning associated with their creation, such as feasting, and the meaning associated with deposition may be entirely separate. We may never get to one true explanation for an individual ABG deposit, and we could argue that no single explanation is likely. The above examples have shown by using a biographical approach it is possible to look beyond meta-level interpretations. The interpretation of such deposits is still not easy and the reader may disagree with those offered above, but will hopefully see that such an approach will lead us to new interpretations and in turn new questions, of which this study has already supplied plenty.

Therefore the question about what ABGs are, and what they mean, is inappropriate. By adopting a wide chronological timeframe to different regions this study has shown that ABGs are infinitely varied. This is not to say that patterns do not exist in the data, but rather many of the previous studies have been looking for one meaning to explain ABGs, when it cannot exist. This is the reason meta-level interpretations such as ‘ritual’ are so commonly used for such deposits. As archaeologists, we have become concerned with applying meaning to a category that does not exist. An example of this is Hamerow’s (2006) use of Grants (1984a) work on Iron Age ABGs to interpret Anglo-Saxon ones. They may be classified as the same archaeological deposit type, but this removes all the associated chronology and context. Iron Age and Anglo-Saxon ABGs may look alike, they may be created by the same basic actions, but the meaning behind such actions and their depositions will likely be very different (Morris and Jervis 2011). Indeed, the meaning behind the deposition of different ABGs within the same period may be very different. In Hill’s (1995, 100) influential study, he argued that ABGs are the result of ‘ritual’ acts, stating he had interpreted them as such by trying to avoid arguing from their composition, more from the nature and shape of activities that formed them. Yet it is their composition that informs us of the activities. Failure to consider variation in ABG composition resulted in Hill applying meta-level interpretations to his data and hence to the entire archaeological category.

The use of categories such as ABG is unavoidable in archaeology, and the language we use is built upon such concepts. However, this study has shown we must not see such concepts as static or concrete entities. ABGs are created via cultural practice, which is a constantly changing process. They are the result of many different processes, some involving culling or natural deaths; others may be the remains of a feast or an offering, but every individual ABG is the result of a separate action, each with its own associations and meanings.

180

Chapter 13. Conclusions The overarching aim of this research was twofold; to investigate the nature and the interpretation of ABGs from the Neolithic to Medieval periods. Upon undertaking the study it became apparent that these two aspects needed separating. Although often noted, the composition of ABGs is rarely discussed in the literature because their nature is not considered to be as important as their meaning. This has resulted in authors reporting such finds as ‘ritual sheep burials’. Such a description is neither informative nor helpful in understanding the behaviour of past societies. If anything, it is hoped that this study has shown the need for further description of these deposits; especially as a biographical methodology relies on detailed information being available. 13.1

Associated Bone Groups; their nature

Although the majority of previous literature regarding ABGs has been concerned with examples from the Iron Age, this study has shown that these deposits are not chronologically limited. In fact, more ABGs were recorded from the Romano-British period than any other. This study has also shown that ABGs are not a ‘Wessex’ phenomenon, as they are also present in the Yorkshire archaeological record, although they display different trends. The data have shown that ABGs are present on around half of Iron Age, Romano-British and early Medieval sites from southern England and Yorkshire. By comparison, deposits from Neolithic and Bronze Age sites in Yorkshire are much rarer than in southern England. This may be due to the monumental and/or funerary nature of many of these earlier sites, as ABGs appear to be more common on ‘domestic’ settlements. In the high and late Medieval periods, the proportion of southern England sites with ABGs reduces, but they are still present on a high proportion of sites from Yorkshire. In fact, there are more ABGs recorded in Yorkshire from the later Medieval period than the Iron Age. Although ABGs are present on a large number of sites (when bone survives), the dataset in most periods is dominated by a small number of large assemblages. For example, 44% (364) of the deposits from the southern England, Romano-British assemblages come from two sites, Owslebury and Oakridge Well (see 6.4 and 6.5). The majority of sites have between one and four ABGs present. This may be due in part to the size of the excavations, or the types of features present. For example, a large number of dog ABGs have been recovered from deep pits and wells within southern English Romano-British towns. Such a pattern is not seen in the data from Yorkshire Romano-British towns, where deep pits and wells have not been excavated in comparable numbers. Domestic mammals dominate ABG assemblages, although there are some variations between periods and regions (see 10.4). The proportions of cattle, sheep/goat

and pig deposits appear to follow the trends seen in the non-ABG assemblages from the Neolithic to the Iron Age. The proportions of dog and horse ABGs show a different pattern. Dog and horse make up a much larger percentage of the ABG assemblage compared to the nonABG dataset in all periods except the Bronze Age. By the Iron Age, dogs are the second most common ABG animal, and by the end of the Romano-British period they dominate the southern England dataset. This pattern changes by the later Medieval period, when domestic fowl are the most common ABG species, although the sample size for this period is small. The Yorkshire dataset shows a very different pattern, with cattle and pig the most common Iron Age ABGs, changing to sheep/goat followed by dog in the Romano-British period. By the later Medieval period domestic fowl and cat are the most common ABGs from Yorkshire sites. The study also showed that wild species, dog, cat and domestic fowl are more commonly recovered as complete skeletons. This is due to these species undergoing different transformation processes compared with the major meat-providing domestic mammals. There is also variability in the completeness of ABGs depending upon the period (see 10.5). The majority of deposits have been recovered from pit features, but this is unsurprising as the most common archaeological features excavated are pits. The southern England Neolithic sample is the exception, but this is due to the large number of ABGs present in the Windmill Hill ditches. However, overall ABGs are also recovered from a number of different feature types and are not exclusively recovered from pits. 13.2

Associated Bone Groups; their meaning

Although the large corpus of ABG data collected and analysed will prove a useful tool for archaeologists, we need to go further. As archaeologists, we strive to investigate the past to find out more about the people who inhabited it. This can take many forms, but always involves applying meaning to material culture. As such the meanings and interpretations applied to ABGs become an increasingly large and important part of this study. This study has shown that interpretations of ABGs have developed alongside broader theoretical paradigms and been influenced by a small number of key texts. The assumptions concerning a site’s functions and human behaviour within a time period are also influential. Prehistoric and, recently, Romano-British ABGs are usually viewed as ‘ritual’ deposits, whereas Medieval ones are nearly always interpreted as ‘functional’ (see 11.2). Current interpretations also adopt a Cartesian dichotomy between ‘functional’ and ‘ritual’. This has

Investigating Animal Burials; Ritual, Mundane and Beyond One of the most problematic areas of this study was the recording of negative results. Of the 493 faunal reports examined, only one from Brickley Lane, Devizes (Charles 2002), clearly stated that ABGs were not present. Absence had to be inferred from the other 279 negative sites. Although it is likely that for reports completed from the 1980’s onwards ABGs would have been noted if present (see 10.2.3), we have no way of being sure that this was always the case. Therefore the assumption had to be made that ABGs would always have been recorded, if present.

resulted in a large number of ABGs only being given meta-level explanations (see 11.8). Therefore the current interpretations are largely generalizations, but this study has shown that the nature of ABGs is varied and diverse. By utilising a biographical approach, it is possible to move away from present-day assumptions concerning these deposits (see 12.1). Such an approach shows that a number of different human actions have created the moments of transformation that result in ABGs. By then contextualizing the data, it is possible to apply meanings to these transformations.

To deal with such a large dataset, a categorical database was required. Such an approach could be criticised, as statistical databases are designed to split data into its component parts for cross-tabulation, and require formal single use classification (Martin 2005). Such an approach may be problematic for complete site databases. However, when examining ABGs we have the advantage of being able to use biological categories, such as species and elements. However, we must bear in mind that such categories use modern western ‘scientific taxonomy’ (see for example Gentry et al. 2004). Past societies would have had their own criteria in separating the taxa encountered within their environment (Ingold 1994, 15; B Morris 2000, 35). Therefore we must be aware that the patterns we see in the data are patterns in our own categories. Both Hill (1995, 104) and Pollard (2006) have suggested that animals may have been classified in the past by their conceptual distance from people.

This method enables archaeologists to move away from the meta-level interpretations and look at the rationale behind the deposits. This study has shown that animal remains can be given many different forms of agency. They can be completely natural without human agency, but when given agency by humans they can also be food, an event, mnemonic agents, a gift, necessary and/or unwanted items. Within and across periods and even within the same feature, there is no one type of ABG and there is no single reason for their deposition; they represent a myriad of human actions and meanings. 13.3

Developing methodologies

From the outset of this research, the decision was made to use publicly available sources of data. Within the timeframe of the project, it was not possible to conduct osteological examinations of individual ABG assemblages without compromising the time depth and geographical breadth. Therefore this study has been largely dependant on the zooarchaeologists’ published reports. This is advantageous, as it has allowed a large dataset to be collected and analysed through the use of modern database technology. It is only through the collection of a large dataset that it has been possible to examine the chronological and regional differences in ABG assemblages.

The advantage of using biological categories is that it has enabled us to investigate both regional and chronological trends in the ABG data. It has also facilitated the comparisons between the ABG and non-ABG assemblages. Therefore we can investigate the treatment of different ‘scientifically categorised’ species, which in turn can help us investigate how different species may have been conceptualised in the past. The investigation of trends in the ABG dataset has produced a supra-biography, showing the changes that occur through time. It has also shown that regional differences are present and we should not be reliant upon ‘Wessex’-based models. However, throughout the creation of this ‘grand narrative’ it became apparent that there is no uniform ABG deposit, nor uniform interpretation. To move beyond such concepts we need to start investigating the series of transformations that created each deposit. To do this, a biographical approach to individual ABGs should be utilised. Such a methodology moves us away from generalising concepts and meta-level interpretations. It is, however, reliant on the availability of a fully integrated dataset. This approach enables us to further develop our understanding of human-animal relationships.

Such a method is not without its difficulties. The major problem encountered has been the variability in the quality and completeness of the reported data. In some reports ABGs get little more than a mention; whereas in others detailed descriptions are offered. The level of detail provided does closely relate to when the report was published. From the 1980’s onwards there is an improvement in the general level of detail given in faunal reports, which is linked to archaeologists gaining a greater understanding of the value of zooarchaeological data and also an increased awareness of ABGs. However, even from the most recent published reports it was often not possible to record all the deposits variables. The three variables most often reported are the type of feature the ABG is recovered from, the species identified, and the general body areas present. To start to build up a clearer picture of these deposits, much more information is required.

13.4

182

Recording recommendations

Chapter 13. Conclusions The problems encountered in this study lead to a number of recommendations. They are grouped into on-site, postexcavation and reporting. These recommendations are for ideal circumstances, (with the time and monetary constraints of commercial archaeology it may not always be possible to enact them all, such as the zooarchaeologist visiting the site). However, it is hoped that archaeologists will recognise the wealth of information available from ABG deposits and their ability to inform us about human actions and therefore act on these recommendations. It is important that future zoo/archaeologists investigating this deposit type have as much information available as possible. This can be achieved by following some of the recommendations below. On site x

x

Reporting x The faunal report should state if ABGs are not present in the assemblage.

If possible a photograph and plan should be made of the deposit.

x

The presence of any associated finds should be noted.

x

The position of the ABG within the feature should be plotted, and its location in relation to other finds should be recorded.

x

Sieving and sampling would greatly increase the recovery of smaller elements and confirm which elements were genuinely missing.

x

The ABG should be bagged separately.

x

If possible, the zooarchaeologist should visit the site to view the deposit/s and offer advice on their retrieval, or the excavation team should include archaeologists with some knowledge of skeletal anatomy.

Post-excavation x The zooarchaeologist should examine the ABG deposits along with the rest of the faunal material. x

The faunal material from the same context as the ABG should be examined to see if any of the elements are from the same individual as the ABG. This can help inform on postdepositional disarticulation.

x

All taphonomic indicators present on the bones of the ABG should be recorded in detail (butchery, gnawing, weathering, burning etc.).

x

x

A detailed description of each ABG should be made available which includes, if possible;  Contextual information about the feature, spatial location, date of the deposit and any associated deposits, including other ABGs.  Whether the ABG was recovered in articulation.  Species.  Elements present (including side information).  Ageing evidence.  Sexing evidence.  Presence and severity of taphonomic indicators and the elements they are present on.  Description of any pathology.  Measurements of elements.

x

Finally the zooarchaeologist should be able to liaise with the site director and other specialists regarding the interpretation of the ABG. By utilising all the available information and adopting a biographical approach to each ABG, a detailed picture of human actions and associated meaning can be built up. This would result in the interpretation of each individual ABG and not a meta-level interpretation of a deposit type.

The presence of all ABGs should be noted.

x

The zooarchaeologist should integrate his/her analysis/interpretation of finds with those of other specialists. This would allow all specialists to be better informed regarding the taphonomic processes present on the site, as well as the human depositional practices.

13.5

Future directions

As this study progressed other avenues for investigation became apparent. Due to the time constraints present in a study of this nature, it was not possible to follow all of them. Those areas that are considered worthy of further sustained development and study are outlined here. 13.5.1 Availability of data One of the main recommendations of this study is that as much data as possible are made available for other researchers. This is a general problem for zooarchaeology. As a data-heavy sub-discipline, it is often not possible for all the faunal information to be made available within a site’s publications. Other

The zooarchaeologist should be made aware of the nature of the context and any associated material recovered within the same context and elsewhere in the feature. 183

Investigating Animal Burials; Ritual, Mundane and Beyond treatment to ABGs. This study has shown there is no pattern of close association between ABGs and other find types. Although as previously stated, the data are very limited and only close associations were examined in this study. No work was carried out looking beyond the contexts ABGs were recovered from. As Hill (1995) also had problems with the availability of data, a useful approach would be to revisit Iron Age pits and indeed expand the study out chronologically, using Hill’s methodology, but utilising recent more reliable archaeological datasets. Such a study would have to tackle interpreting ‘special’ finds from other material types, and the biographical approach advocated for ABGs would be of value here.

specialists would probably make the same point for the materials they study. Specialist reports are often placed towards the end of site publications, and with space at a premium, detailed information can be edited out. This situation is unlikely to change and is driven by the cost of publication. Therefore it is important that zooarchaeologists develop other means to make their data widely available. The database utilised in this study has been placed on the accompanying CD, and after publication will be placed in a suitable internet repository such as the Archaeology Data Service. This will enable other researchers to use the data gathered. The future for zooarchaeological data is undoubtedly digital, and this is an avenue which deserves further development. For example summary zooarchaeological data could continue to be published in site reports, with further information such as detailed accounts of ABGs available from a web-based repository.

13.5.4 Intra-site study Due to limits imposed upon this study it was not possible to investigate intra-site patterns in the deposition of ABGs. Such a study would have required the research to move beyond published data. Although commonly recovered from pit features, this study has shown ABGs are present in a number of different feature types. For example, there were a number of differences in the species deposited in pits and ditches at the RomanoBritish site of Owslebury (see 6.4.2). Intra-site analysis of ABGs may also aid in their interpretation, as we would be able to examine associations between ABGs and specific site areas. This is certainly an avenue which deserves further work. However, it would require a large assemblage from a well recorded site.

13.5.2 Regionality This study has shown that blanket interpretations of ABGs do not work. By contrasting the southern England and Yorkshire regions, it has attempted to move beyond the ‘Wessex’ dataset. This has shown there are a number of regional differences in the types of ABGs deposited. Some of these may be due to differences in the types of archaeological sites excavated. This study could also be expanded to examine other regions within Britain. It appears likely that the nature of ABGs in the rest of Britain will not necessarily follow the ‘Wessex’ pattern. Indeed it could be argued that the Wessex pattern disguises intra-regional variations.

13.5.5 Metrical analysis Due to time constraints metrical aspects were not investigated in this study but are worthy of future research. Such a study would aid us in understanding the choices made by human societies. As this study has shown, some animals were selected for particular events such as food offerings for human burials. At present we do not know whether such animals came from the same groups that produced the non-ABG faunal material, although we do know that in some cases rarer species such as domestic fowl were used. If they did not and were part of a separate group used for such activities, then a metrical comparison between the ABG and non-ABG assemblages may indicate such a difference. Due to the rarity of complete bones in the non-ABG faunal assemblage, such analysis would need to be conducted on a large assemblage. We could theorise that differences in the stature of animals which were deposited as ABGs compared to those in the non-ABG assemblage may indicate that either specific individuals fulfilling a morphological criteria were chosen, or the animals were from a different population group entirely. Such ideas require further development, but have the potential to add to our knowledge.

13.5.3 Associated deposit groups The aspect of this study which suffered most from a lack of available data was the analysis of associated deposit groups (ADGs). In some cases associations between multiple ABGs were reported and were a feature of interpretations such as culling (see 11.5). However, with the exception of formal funerary features, very few associations were recorded between ABGs and other material groups. One of the reasons for this is the separation of specialist reports. Often the zooarchaeologist would be unaware of the other material finds from the same context as the ABGs, such as the example from Greyhound Yard (see 1.2.9). Information regarding ADGs was usually found in the excavations chronological narrative rather than the specialist reports. However, this also relied upon the excavators spotting associations in the field. The interpretation of other material types found in association with ABGs is an area that requires a great deal of work. Hill (1995, 74) suggested that pottery and small finds from Iron Age pits had received similar 184

Chapter 13. Conclusions 13.5.6 Post-depositional taphonomic action

13.9

One of the non-human factors that can influence the composition of the ABG assemblage is post-depositional activity, such as bioturbation and slumpage (see 2.6). On some sites, such as Oakridge Well, such effects are to be expected and can be seen in the stratigraphy of the feature (see 6.5). However, little is known of such effects on smaller features. It is perhaps unsurprising that the majority of ABGs are present in the bottom of features, were such factors may not affect their composition. To remain articulated, ABGs must be deposited with some soft tissue present. One of the current issues that needs resolving is what effect the decomposition of soft tissue has on the composition of an ABG?

In conclusion, this study has proved to be more than the study of animal remains. It is the study of human actions and the meanings behind them. ABGs are unique phenomena in that they are a wholly artificial construct. Humans in the past did not deposit an ABG; they deposited the remains of an animal in a manner they deemed appropriate, although ABGs can also be created without human action, such as the accidental fall of an animal. As a ‘category’ of faunal data they have the ability to inform on the actions that created them. It is these actions which had meaning and agenda and it is these actions that inform us about past societies. This study has shown that a biographical approach to individual ABGs leads to a more informed view, moving away from the sweeping generalizations used so far when dealing with this deposit type. It has also shown the value of not only utilising specialist data but integrating such knowledge with other archaeological material evidence. Zooachaeologists should not be afraid of or be held back from doing the same. This study has shown that as a discipline zooarchaeology can move beyond the economic towards the social.

A possible way of investigating this matter is through the use of experimental archaeology. A series of experiments involving the recreation of archaeological features and the deposition of organic material including ABGs would aid us in understanding the effects post-depositional taphonomic action has upon the ABG assemblage.

Concluding remarks

13.5.7 Animal behaviour There is no standard type of ABG and there is no standard interpretation. In effect, the conclusion of this study could be that there is no general conclusion. This is because each deposit is unique, and to apply a meta-level interpretation to all ABGs, even from the same period, would be inaccurate and inappropriate. ABGs were created by a myriad of human actions and motivations; they are neither ritual nor functional, for such interpretations lead us only to generalizations. There are trends in their creation, but each bone group is created by specific actions and it is the investigation of these individual events that moves us closer to the societies we wish to understand.

This study has made the case for the integration of different specialists when trying to understand ABGs. What is also needed is further co-operation between zooarchaeologists and zoologists, specifically in the fields of natural history and animal behaviour. The review of ABG interpretations have shown that zooarchaeologists make a number of assumptions regarding animal behaviour, especially for wild mammals in regard to pitfall victims. Currently we do not know whether certain species are typically likely to ‘fall down holes’. However, by engaging with zoologists such questions may be answered. 13.5.8 Moments of transition This study has advocated that a biographical approach to ABGs helps us to understand the actions that created them and the meanings behind such actions. Key to this approach is the investigation of moments of transition, in which the nature of the animal as well as its associated meanings are changed. Such a process applies to all faunal remains as well as ABGs. For example, cattle carcass-processing in the Romano-British and Medieval periods involves a number of transformations, from living animal, to food, to raw material, to object. At each transformation the agency of the animal is changed. Although beyond the scope of this study, such ideas deserve further development. They offer a means of looking at animal remains in a different light, moving beyond economics as the focus of zooarchaeological study and starting to look in greater depth at the social dimensions of the zooarchaeological record.

185

Bibliography Abe, Y., Marean, C. W., Nilssen, P. J., Assefa, Z. and Stone, E. C. (2002) The analysis of cutmarks on archaeofauna: a review and critique of quantification procedures and a new image-analysis GIS approach. American Antiquity 67 (4), 643-663. Aitken, G. and Aitken, N. (1982) Excavations on the Library Site, Colliton Park, Dorchester, 1961-3. Proceedings of the Dorset Natural History and Archaeological Society 104, 123-125. Albarella, U. (2005a) Alternate fortunes? The role of domestic ducks and geese from Roman to Medieval times in Britain. In G. Grupe and J. Peters (eds.) Feathers, Grit and Symbolism: Birds and Humans in the Ancient Old and New Worlds, 249-258. Rahden, Verlag Marie Leidorf. Albarella, U. (2005b) Meat production and consumption in town and country. In K. Giles and C. Dyer (eds.) Town and Country in the Middle Ages. Contrasts, Contacts and Interconnections, 1100-1500, 131-148. Leeds, Maney. The Society for Medieval Archaeology Monograph 22. Albarella, U. (2007) The end of the sheep age: people and animals in the Late Iron Age. In C. Haselgrove and T. Moore (eds.) The Later Iron Age in Britain and Beyond, 389-402. Oxford, Oxbow. Albarella, U. and Davis, S. J. M. (1996) Mammals and birds from Launceston Castle, Cornwall: decline in status and the rise of agriculture. Circaea 12 (1), 1-156. Albarella, U. and Serjeantson, D. (2002) A passion for pork: meat consumption at the British late Neolithic site of Durrington Walls. In P. Miracle and N. Milner (eds.) Consuming Passions and Patterns of Consumption, 3349. Cambridge, McDonald Institute for Archaeological Research. Alcock, J. P. (1966) Celtic water cults in Roman Britain. Archaeological Journal 122, 1-12. Alcock, L. (1970) Notes and News: South Cadbury excavations 1969. Antiquity 44, 43-49. Allen, D. and Anderson, S. (1999) Basing House Hampshire. Excavations 1978-1991. Winchester, Hampshire Field Club Monograph 10. Allen, M. J. (1992) Animal bone. In M. J. Heaton. Two Mid-Saxon grain-driers and later Medieval features. At Chantry Fields, Gillingham, Dorset. Proceedings of the Dorset Natural History and Archaeological Society 114, 123-124. Allen, M. J. (1995) Animal bone. In P. A. Harding, L. Mepham and R. J. C. Smith (eds.) The excavation of 1213th century deposits at Howards Lane. Proceedings of

the Dorset Natural History and Archaeological Society 117, 89. Allen, M. J. (2000) Animal bone. In P. S. Bellamy (ed.) Excavations at Penny's Farm, Cranborne. Proceedings of the Dorset Natural History and Archaeological Society 122, 93. Andersen, N. H. (1997) Sarup vol 1. The Sarup Enclosures. Moesgaard, Jutland Archaeological Society Publications XXXIII. Appadurai, A. (1986) Introduction: commodities and the politics of value. In A. Appadurai (ed.) The Social Life of Things: Commodities in Cultural Perspective, 3-63. Cambridge, Cambridge University Press. Armitage, P. L. (1978) Hertfordshire cattle and London meat markets in the 17th and 18th centuries. The London Archaeologist 3, 217-223. Armour-Chelu, M. (1986) The animal bone. In S. M. Davis and D. Grieve. The Poundbury Pipe-line: archaeological observations and excavations. Proceedings of the Dorset Natural History and Archaeological Society 108, 86-87. Armour-Chelu, M. (1991) The animal bones. In N. Sharples (ed.) Maiden Castle; Excavations and Field Survey, 131-151. London, English Heritage Archaeological Reports No.19. Armstrong, P. (1977) Excavations in Sewer Lane, Hull. Kingston Upon Hull, East Riding Archaeologist Vol.3. Hull Old Town Report Series No. 1. Armstrong, P. and Ayers, B. (1987) Excavations in High Street and Blackfriargate. Kingston Upon Hull, East Riding Archaeologist Vol.8. Hull Old Town Report Series No. 5. Ashbee, P. (1966) The Fussell's Lodge Long Barrow Excavations 1957. Archaeologia 100, 1-81. Ashbee, P. (1970) The Earthen Long Barrows in Britain. London, J. M. Dent and Sons Ltd. Ashbee, P. (1989) Archaeological conclusions. In P. Ashbee, M. Bell and E. Proudfoot (eds.) Wilsford Shaft: Excavations 1960-62, 133-138. London, English Heritage Archaeology Report No 11. Ashbee, P., Bell, M. and Proudfoot, E. (eds.) (1989) Wilsford Shaft: Excavations 1960-62. London, English Heritage Archaeology Report No 11. Ashbee, P., Smith, I. F. and Evans, J. G. (1979) Excavation of three long barrows near Avebury,

Bibliography Wiltshire. Proceedings of the Prehistoric Society 45 (207-300)

Behrensmeyer and A. Hill (eds.) Fossils in the Making, 72-92. Chicago, University of Chicago Press.

Bailey, C. J. (1965) Notes on Roman and Medieval Sites in the Bride Valley, Dorset. Proceedings of the Dorset Natural History and Archaeological Society 87, 90-92.

Bell, C. (1992) Ritual Theory, Ritual Practice. Oxford, Oxford University Press. Bell, C. (1997) Ritual Perspectives and Dimensions. Oxford, Oxford University Press.

Bailey, C. J. (1967) An early Iron Age/Romano-British site at Pins Knoll, Litton Cheney. Proceedings of the Dorset Natural History and Archaeological Society 89, 147-159.

Bell, M. (1989) Environmental conclusions. In P. Ashbee, M. Bell and E. Proudfoot (eds.) Wilsford Shaft: Excavations 1960-62, 128-133. London, English Heritage Archaeology Report No 11.

Baker, G. (1983) The animal remains. In J. Collis. Excavations at Silchester, Hants, 1968. Proceedings of the Hampshire Field Club and Archaeological Society 39, 67.

Bendrey, R. (2007) Ossification of the interosseous ligaments between the metapodials in horses: a new recording methodology and preliminary study. International journal of Osteoarchaeology 17 (2), 207213.

Baker, J. R. and Brothwell, D. (1980) Animal diseases in archaeology. London, Academic Press.

Bendrey, R., Leach, S. and Clark, K. (2010) New light on an old rite: reanalysis of an Iron Age burial group from Blewburton Hill, Oxfordshire. In J. Morris and M. Maltby (eds.) Integrating Social and Environmental Archaeologies; Reconsidering Deposition, 33-44. Oxford, British Archaeology Reports British Series.

Barber, M. (2003) Bronze and the Bronze Age. Stroud, Tempus. Barnetson, L. D. P. (1993) The animal bones: inventory. In D. W. Harding, I. M. Blake and P. J. Reynolds (eds.) An Iron Age Settlement in Dorset: Excavation and Reconstruction, 64-72. Edinburgh, University of Edinburgh Department of Archaeology Monograph Series No 1.

Berg, D. (1987) The animal bone. The faunal remains from the Hotham/Celererman and Ousefleet properties. In P. Armstrong and B. Ayers (eds.) Excavations in High Street and Blackfriargate, 245-252. Kingston Upon Hull, East Riding Archaeologist Vol.8. Hull Old Town Report Series No. 5.

Barone, R. (1976) Anatomie Comparée des Mammiféres Domestiques. Paris, Vigot. Batt, C. M. and Dockrill, S. J. (1998) Magnetic moments in prehistory: integrating magnetic measurements with other archaeological data from the Scatness Multiperiod Settlement. Archaeological Prospection 5 (4), 217-227.

Berg, D. (1990a) The mammal bone. In S. Wrathmell and A. Nicholson (eds.) Dalton Parlours Iron Age Settlement and Roman Villa, 174-189. Leeds, West Yorkshire Archaeology Services. Yorkshire Archaeological Report No. 3.

Baxter, I. (2004) Animal bone. In C. Gibson. The Iron Age and Roman site of Viables Two (Jays Close), Basingstoke. Proceedings of the Hampshire Field Club and Archaeological Society 59, 24-26.

Berg, D. (1990b) Mammal bones from well 1. In S. Wrathmell and A. Nicholson (eds.) Dalton Parlours Iron Age Settlement and Roman Villa, 245-259. Leeds, West Yorkshire Archaeology Service. Yorkshire Archaeological Report No. 3.

Baxter, I. (2005) Animal bones. In C. K. Currie and N. S. Rushton. An archaeological excavation at the rear of 36 Milford Street / 34 Gigant Street, Salisbury. Wiltshire Archaeological and Natural History Magazine 98, 232233.

Berg, D. (1995) The animal bone. In I. Roberts. Excavation of a D-shaped enclosure at Upton, West Yorkshire. Yorkshire Archaeological Journal 67, 18.

Behrens, H. (1964) Die neolithisch-frühmetallzeitlichen Tierskelettfunde der Alten Welt. Berlin, Veb Deutscher Verlag der Wissenschaften.

Berg, D. (1999) The mammal bones. In P. Abramson, D. S. Berg and M. R. Fossick (eds.) Roman Castleford Excavations 1974-85. Volume 2. The Structural and Environmental Evidence, 223-279. York, West Yorkshire Archaeology Service, Yorkshire Archaeological Report No. 5.

Behrensmeyer, A. K. (1978) Taphonomic and ecological information from bone weathering. Paleobiology 4, 150162.

Bevan, B. (1999) The landscape context of the Iron Age square-barrow burials, East Yorkshire. In J. Downes and

Behrensmeyer, A. K. and Dechant Boaz, D. E. (1980) The recent bones of Amboseli National Park Kenya, in relation to East African paleoecology. In A. K. 187

Investigating Animal Burials; Ritual, Mundane and Beyond Bourdillon, J. (1984) Animal bones from further contexts on site SOU 169 (Saxon Southampton). English Heritage, Ancient Monuments Laboratory Report 4581.

T. Pollard (eds.) The Loved Body's Corruption, 69-93. Glasgow, Scottish Archaeological Forum/Cruithne Press. Biddick, K. (1984) Field edge, forest edge: early medieval social change and resource allocation. In K. Biddick (ed.) Archaeological Approaches to Medieval Europe. Studies in Medieval Culture 18, 105-118. Michigan, Kalamazoo: Medieval Institue Publications, Western Michigan University.

Bourdillon, J. (1986a) The animal bone from castle contexts in SOU 123 area A. In J. Oxley (ed.) Excavations at Southampton Castle, 37. Southampton, Southampton City Museums. Bourdillon, J. (1986b) The animal bone from the castle ditch. In J. Oxley (ed.) Excavations at Southampton Castle, 45. Southampton, Southampton City Museums.

Bilton, L. (1958) The animal bones. In A. Cotton and P. W. Gathercole (eds.) Excavations at Clausentum, Southampton 1951-1954, 155-157. London, HMSO.

Bourdillon, J. (1986c) The SOU 29 animal bone. In J. Oxley (ed.) Excavations at Southampton Castle, 73-77. Southampton, Southampton City Museums.

Binford, L. R. (1978) Nunamiut Ethnoarchaeology. New York, Academic Press.

Bourdillon, J. (1987) Animal bones scanned from likely street contexts on site SOU 258 (Hamwic Six Dials Excavations Southampton). English Heritage, Ancient Monuments Laboratory Report 104/87.

Binford, L. R. (1981) Bones: Ancient Men and Modern Myths. Orlando, Academic Press. Binford, L. R. (1983) In Pursuit of the Past. Berkeley, University of California Press.

Bourdillon, J. (1990a) The animal bones from La Sagesse (The Presbytery) 1988, Romsey, Hampshire. English Heritage, Ancient Monuments Laboratory Report 106/90.

Birbeck, V. and Moore, C. (2004) Preservation and investigation of Roman and Medieval remains at Hyde street, Winchester. Proceedings of the Hampshire Field Club and Archaeological Society 59, 77-110.

Bourdillon, J. (1990b) Animal bones recovered by coarse water-sieved recovery from middle Saxon Southampton, Hampshire (SOU15, excavated in 1974). English Heritage, Ancient Monuments Laboratory Report 103/90.

Bird, P. F. (1968) Appendix III: animal bone from Tollard Royal. In G. J. Wainwright. The Excavation of a Durotrigian Farmstead near Tollard Royal in Cranbourne Chase, Southern England. Proceedings of the Prehistoric Society 34, 146-147.

Bourdillon, J. (1992) Animal bones from late Saxon sites in Winchester, Hampshire. English Heritage, Ancient Monuments Laboratory Report 42/92.

Blumenschine, R. J. (1988) An experimental model of the timing of hominid and carnivore influence on archaeological bone assemblages. Journal of Archaeological Science 15, 483-502.

Bourdillon, J. (1993a) Animal bones. In A. H. Graham and S. M. Davies (eds.) Excavations in Trowbridge Wiltshire, 1977 and 1986-1988, 127-135. Salisbury, Wessex Archaeology Report No.2.

Bond, J. M. and O'Connor, T. P. (1999) Bones from Medieval Deposits at 16-22 Coppergate and Other Sites in York. The Archaeology of York 15/5. York, Council for British Archaeology.

Bourdillon, J. (1993b) Animal bones. In M. F. Garner (ed.) Middle Saxon Evidence at Cook Street, Southampton (SOU 254). Proceedings of the Hampshire Field Club and Archaeological Society 49, 116-120.

Bourdillon, J. (1978) The animal bone. In J. S. F. Walker. Excavations in Medieval tenements on the Quilter's Vault site in Southampton,. Proceedings of the Hampshire Field Club and Archaeological Society 35, 207-212.

Bourdillon, J. (2006) Animal bones. In P. Williams and R. Newman (eds.) Market Lavington, Wiltshire, An Anglo-Saxon Cemetery and Settlement. Excavations at Grove Farm 1986-90, 150-169. Salisbury, Wessex Archaeology Report No. 19.

Bourdillon, J. (1980a) Animal bone from Saxon contexts. In S. M. Davies. Excavations at Old Down Farm, Andover. Part 1: Saxon. Proceedings of the Hampshire Field Club and Archaeological Society 36, 177-178.

Bourdillon, J. and Andrews, P. (1997) The animal bone. In P. Andrews (ed.) Excavations at Hamwic Volume 2: Excavations at Six Dials, 242-245. London, Council for British Archaeology Research Report 109.

Bourdillon, J. (1980b) Town life and animal husbandry in the Southampton area, as suggested by the excavated bones. Proceedings of the Hampshire Field Club and Archaeological Society 36, 181-191.

Bourdillon, J. and Coy, J. (1980) The animal bones. In P. Holdsworth (ed.) Excavations in Melbourne Street, Southampton, 1971-76, 79-121. London, Council for British Archaeology Research Report 33. 188

Bibliography Yorkshire. Proceedings of the Prehistoric Society 42, 157-158.

Bowen, H. and Wood, P. (1968) Experimental storage of corn underground and its implications for Iron Age settlement. Bulletin of the Institute of Archaeology, London 7 (1-14)

Bratlund, B. (1996) Hunting strategies in the Late Glacial of northern Europe: A survey of the faunal evidence. Journal of World Prehistory 10 (1), 1-48.

Boylan, P. (1997) William Buckland (1784–1856) and the foundations of taphonomy and palaeoecology. Archives of Natural History 24, 361–372.

Brewster, T. C. M. (1980) The Excavation of Garton and Wetwang Slacks. Malton, The East Riding Archaeological Research Committee Prehistoric Excavation Reports No.2.

Bradley, R. (1967) Excavations on Portsdown Hill. Proceedings of the Hampshire Field Club and Archaeological Society 24, 42-58.

Brewster, T. C. M. (1984) The Excavation of Whitegrounds Barrow, Burythorpe. Malton, East Riding Archaeological Research Committee.

Bradley, R. (1975) Maumbury Rings, Dorchester: The Excavations of 1908-1913. Archaeologia 105, 1-98.

Brothwell, D. and Harcourt, R. (1979) Animal remains. In G. Clarke (ed.) Winchester Studies 3. Pre-Roman and Roman Winchester. Part 2, The Roman cemetery at Lankhills, 239-245. Oxford, Clarendon Press.

Bradley, R. (1984) The Social Foundations of British Prehistory. London, Longman. Bradley, R. (1990) The Passage of Arms: an archaeological analysis of prehistoric hoards and votive deposits. Cambridge, Cambridge University Press.

Brown, S. (1991) Vertebrate and marine molluscan remains from the latrine pit (F5300). In G. D. Scobie, J. M. Zant and R. Whinney (eds.) The Brooks, Winchester A Preliminary Report on the Excavations, 1987-88, 68. Winchester, Winchester Museum Services.

Bradley, R. (2003) A life less ordinary: the ritualization of the domestic sphere in later prehistoric Europe. Cambridge Archaeology Journal 13:1 (1), 5-23.

Brück, J. (1999) Ritual and rationality: some problems of interpretation in European archaeology. European Journal of Archaeology 2 (3), 313-344.

Bradley, R. (2005) Ritual and Domestic Life in Prehistoric Europe. London, Routledge. Bradley, R. (2007) The Prehistory of Britian and Ireland. Cambridge, Cambridge University Press.

Brunaux, J. L. (1988) The Celtic Gauls : gods, rites and sanctuaries. London, Seaby Ltd.

Brain, C. K. (1967) Hottentot food remains and their bearing on the interpretation of fossil bone assemblages. Scientific Papers of the Namib Desert Research Station 32, 1-11.

Brunaux, J. L., Meniel, P. and Poplin, F. (1985) Gournay 1 : Les fouilles sur le sanctuaire et l'oppidum. Amiens, Revue Archéologique de Picardie. Buckland-Wright, J. C. (1987) The animal bones. In C. S. Green (ed.) Excavations at Poundbury, Dorchester, Dorset 1966-1982. Vol 1: The Settlements, 129-132, 4:A1-D14. Dorchester, Dorset Natural History and Archaeological Society. Monograph Series No 7.

Brain, C. K. (1981) The Hunters or the Hunted? An Introduction to African Cave Taphonomy. Chicago, University of Chicago Press. Bramwell, D. (1963) Appendix IV: Animal bones. In T. G. Manby. The excavation of the Willerby Wold Long Barrow, East Ridding of Yorkshire. Proceedings of the Prehistoric Society 29, 204.

Buckland-Wright, J. C. (1990) The animal bones. In G. M. Aitken and G. N. Aitken. Excavations at Whitcombe, 1965-1967. Proceedings of the Dorset Natural History and Archaeological Society 112, 86-90.

Bramwell, D. (1972) Animal bones from the 1960 excavation. In A. L. Pacitto. Rudston Barrow LXII: The 1968 Excavation. Yorkshire Archaeological Journal 44, 11-12.

Buckland-Wright, J. C. (1993) The animal bones. In D. E. Farwell and T. L. Molleson (eds.) Excavations at Poundbury 1966-80. Volume II: The Cemeteries, 110111. Dorchester, Dorset Natural History and Archaeology Society Monograph Series 11.

Bramwell, D. (1974) Appendix 1: animal remains from Rudston and Boynton Grooved Ware sites. In T. G. Manby (ed.) Grooved Ware Sites in Yorkshire and the North of England, 103-108. Oxford, British Archaeological Report British Series 9.

Buckland, W. (1823) Reliquiae Diluvianae; or, Observations on the Organic Remains Contained in Caves, Fissures,and Diluvial Gravel, and on other Geological Phenomena, Attesting to the Action of an Universal Deluge. London, Murray.

Bramwell, D. (1976) Animal bones. In T. G. Manby. Excavation of the Kilham Long Barrow, East Riding of 189

Investigating Animal Burials; Ritual, Mundane and Beyond Chapman, A., Deighton, K. and Halstead, P. (2004) Gayhurst. Reconstructing the burial rite of an early Bronze Age lord. Current Archaeology 195, 114-118.

Bullock, A. E. and Allen, M. J. (1997) Animal bones. In R. J. C. Smith, F. Healy, M. J. Allen, E. L. Morris, I. Barnes and P. J. Woodward (eds.) Excavations along the Route of the Dorchester By-pass, Dorset, 1986-8, 190199. Salisbury, Wessex Archaeology Report No 11.

Chapman, J. (2000) Pit-digging and structured deposition in the Neolithic and Copper Age. Proceedings of the Prehistoric Society 66, 61-87.

Bunting, G. H., Verity, D. K. and Cornwall, I. W. (1959) Report on the animal bones. In F. D. M. Morgan (ed.) The excavation of a long Barrow at Nutbane, Hants. Proceedings of the Prehistoric Society 25, 47-49.

Charles, B. (2002) Animal Thomason and A. Brossler. Roman activity at Brickley Wiltshire Archaeological and 95, 228-231.

Bunting, G. H., Verity, D. W. and Cornwall, I. W. (1962) Animal bones from Bury Wood Camp. In D. G. King. Bury Wood Camp, report on excavations, 1960. Wiltshire Archaeological and Natural History Magazine 58, 204208.

bones. In D. Poore, D. Iron Age settlement and Lane, Devizes, Wiltshire. Natural History Magazine

Charles, R. (1997) The exploitation of carnivores and other fur-bearing mammals during the North-western European Late and Upper Palaeolithic and Mesolithic. Oxford Journal of Archaeology 16 (3), 253-277.

Bunting, G. H., Verity, D. W. and Cornwall, I. W. (1968) Report on human and animal bones from pits 15B and 15C and on human skeletons from the Sleepleton Gate. In I. Richmond (ed.) Hod Hill: Volume 2 Excavations carried out between 1951 and 1958, 123-125. London, The Trustees of the British Museum.

Chazan, M. and Horwitz, L. K. (2006) Finding the message in intricacy: The association of lithics and fauna on Lower Palaeolithic Multiple Carcass Sites. Journal of Anthropological Archaeology 25 (4), 436-447.

Butterworth, C. (2005) Excavations at Old George Mall, Salisbury. Wiltshire Archaeological and Natural History Magazine 98, 236-246.

Cherryson, A. K. (2002) The identification of archaeological evidence for hawking in Medieval England. Acta Zoologica Cracoviensia (Proceedings of the 4th Meeting of the ICAZ Bird Working Group) 45, 307-314.

Cardwell, P. (1989) Excavations at Cat Babbleton Farm, Ganton, North Yorkshire. Yorkshire Archaeological Journal 61, 15-27.

Chesney, C. F. (1968) The animal bones from the Iron Age site. In R. Bradley and E. Lewis. Excavations at the George Inn, Portsdown. Proceedings of the Hampshire Field Club and Archaeological Society 25, 39.

Carter, P. L. and Higgs, E. S. (1979) The animal remains. In P. Ashbee, I. F. Smith and J. G. Evans. Excavation of three long barrows near Avebury, Wiltshire. Proceedings of the Prehistoric Society 45, 248-249.

Christie, P. M. (1964) A Bronze Age round barrow on Earl's Farm Down, Amesbury. Wiltshire Archaeological and Natural History Magazine 59, 30-45.

Carver, M. O. H. (1995) Roman to Norman at York Minster. In D. Philips and B. Haywood (eds.) Excavations at York Minster 1: From Roman Fortress to Norman Cathedral, 177-222. London, HMSO.

Christie, P. M. (1967) A barrow cemetery of the second millennium BC in Wiltshire, England. Proceedings of the Prehistoric Society 33, 336-343.

Chaplin, R. E. (1965) Animals in Archaeology. Antiquity 39, 204-211.

Clark, H. (1986) The faunal remains and animal economy. In M. Morris. An Iron Age and RomanoBritish site at Choseley Farm, Odiham: The Excavations of Dorothy Liddell, 1937. Proceedings of the Hampshire Field Club and Archaeological Society 42, 104.

Chaplin, R. E. and Barnetson, L. P. (1976) The animal bones from the Roman well at Rudston, Yorkshire. English Heritage, Ancient Monuments Laboratory Report 2150.

Clark, K. M. (2002) Mammal and Bird bones. In A. H. Graham, D. A. Hinton and D. P. S. Peacock (eds.) The excavation of an Iron Age and Romano-British settlement in Quarry Field, south of Compact Farm, Worth Matravers, Dorset, 59-67. Oxford, Oxbow.

Chaplin, R. E. and Barnetson, L. P. (1980) Animal bones. In I. M. Stead (ed.) Rudston Roman Villa, 149-161. York, Yorkshire Archaeological Society. Chaplin, R. E. and Barnetson, L. P. (1981) The animal bones from the excavation of a Roman site at Rudston, Yorkshire. Part 2. The buildings and other features. English Heritage, Ancient Monuments Laboratory Report 3466.

Clark, K. M. (2006) The dog assemblage. In M. Fulford, A. Clarke and H. Eckardt (eds.) Life and Labour in Late Roman Silchester. Excavations in Insula IX since 1997, 189-195. London, Britannia Monograph Series No. 22.

190

Bibliography Clutton-Brock, J. and Jewell, P. (2005) Remains of wild and domestic animals. In N. Thomas (ed.) Snail Down Wiltshire. The Bronze Age Barrow Cemetery and Related Earthworks, in the Parishes of Collingbourne Ducis and Collingbourne Kingston: Excavations, 1953, 1955 and 1957, 236-243. Salisbury, Wiltshire Archaeological and Natural history Society Monograph No 3.

Clark, R. (1992) The animal bones. In P. Cox. Excavations at the former site of the Wimborne model town, 1991. Proceedings of the Dorset Natural History and Archaeological Society 114, 149. Clark, R. H. (1995) Faunal remains. In M. J. Allen, M. Morris and R. H. Clark. Food for the living: a reassessment of a Bronze Age Barrow at Buckskin, Basingstoke, Hampshire. Proceedings of the Prehistoric Society 61, 170-172.

Clutton-Brock, T. H., Price, O. F., Albon, S. D. and Jewell, P. A. (1992) Early development and population fluctuations in soay sheep. The Journal of Animal Ecology 61 (2), 381-396.

Clarke, A. J. (1966) Appendix II. Identification of bones. In G. King. The Lanhill Long Barrow, Wiltshire, England: An essay in Reconstruction. Proceedings of the Prehistoric Society 32, 84.

Coe, D. and Newman, R. (1992) Excavations of an early Iron Age Building and Romano-British enclosure at Brighton Hill South, Hampshire. Proceedings of the Hampshire Field Club and Archaeological Society 48, 521.

Clarke, S. (1997) Abandonment, rubbish disposal and 'special' deposits at Newstead. In K. Meadows, C. Lemke and J. Heron (eds.) Proceedings of the Sixth Annual Theoretical Roman Archaeology Conference Sheffield 1996, 73-81. Oxford, Oxbow.

Colley, S. M. (1984) Animal bone from SOU 17, Saxon Southampton (Hamwih). English Heritage, Ancient Monuments Laboratory Report 4918.

Clarke, S. (1999) In search of a different Roman period: the finds assemblage at the Newstead military complex. In G. Fincham, G. Harrison, R. Holland and L. Revell (eds.) Proceedings of the Ninth Annual Theoretical Roman Archaeology Conference Durham 1999., 22-29. Oxford, Oxbow.

Collins, A. E. P. (1953) Excavations at Blewburton Hill, 1948 and 1949. Berkshire Archaeological Journal 53, 2166. Collis, J. (1968) Excavations at Owslebury, Hants: An Interim Report. Antiquaries Journal 48, 18-31

Clarke, S. (in press) Identifying the actors and motivations behind pit deposition at the Trimontium Military Complex (Newstead). In J. Morris and C. Randall (eds.) Ritual in Context: Explaining Ritual Complexity in Archaeology, Oxford, Oxbow.

Collis, J. (1978) Winchester Excavations. Excavations in the Suburbs and Western Parts of the Town. Volume II 1949-1960. Winchester, Winchester City Museum. Collis, J. (1979) The animal bones. In J. Collis and P. Fasham. Excavation and field survey at Borough Farm, Micheldever, Hants. Proceedings of the Hampshire Field Club and Archaeological Society 36, 150-151.

Clutton-Brock, J. (1974) Animal remains from Bronze Age Barrows, near Amesbury, Wilts. English Heritage, Ancient Monuments Laboratory Report 1573. Clutton-Brock, J. (1975/1976) Animal remains from Amesbury Barrow 51. In P. Ashbee. Amesbury Barrow 51: Excavation 1960. Wiltshire Archaeological and Natural History Magazine 70/71, 56.

Connah, G. (1965) Excavations at Knap Hill, Alton Priors, 1961. Wiltshire Archaeological and Natural History Magazine 60, 1-23. Cool, H. E. M. (2006) Eating and Drinking in Roman Britain. Cambridge, Cambridge University Press.

Clutton-Brock, J. (1984) Large animal remains. In P. Ashbee. The excavation of Amesbury Barrows 58, 61a, 61, 72. Wiltshire Archaeological and Natural History Magazine 79, 88.

Cool, H. E. M. and Baxter, M. J. (2002) Exploring Romano-British finds assemblages. Oxford Journal of Archaeology 21 (4), 365-380.

Clutton-Brock, J. (1999) A Natural History of Domesticated Mammals. London and Cambridge, British Museum (Natural History) and Cambridge University Press.

Cornwall, I. W. (1958) The animal bone. In A. Cotton and P. W. Gathercole (eds.) Excavations at Clausentum, Southampton 1951-1954, 141-142. London, HMSO.

Clutton-Brock, J. and Jewell, P. (1963) Report on animal remains. In F. M. Vatcher. The excavation of the barrows on Lamb Down, Codford St. Mary. Wiltshire Archaeological and Natural History Magazine 58, 437438.

Cox, P. S. (1989) A seventh Century Inhumation Cemetery at Shepherd's Farm, Ulwell near Swanage, Dorset. Proceedings of the Dorset Natural History and Archaeological Society 110, 37-47.

191

Investigating Animal Burials; Ritual, Mundane and Beyond excavations. English Heritage, Ancient Monuments Laboratory Report 3288.

Coy, J. (1967) Appendix II report on animal bones. In D. G. King. Bury Wood Camp excavations in the area of the South-West opening. Wiltshire Archaeological and Natural History Magazine 62, 13.

Coy, J. (1982a) The animal bones. In C. Gringell. Excavation of an Iron Age enclosure at Groundwell Farm, Blunsdon St. Andrew, 1976-7. Wiltshire Archaeological and Natural History Magazine 76, 68-73.

Coy, J. (1969) Report on the animal bones from the north-west and north-east entrances. In D. G. King. Bury Wood Camp excavations in the North-East and NorthWest areas. Wiltshire Archaeological and Natural History Magazine 64, 47-48.

Coy, J. (1982b) Animal bones. In F. P. J. The excavation of four ring-ditches in central Hampshire (MARC3 sites R17, Feature 1972; R&; R30 and R363). Proceedings of the Hampshire Field Club and Archaeological Society 38, 37.

Coy, J. (1976a) Animal Bones from Groundwell farm, Blunsdon St. Andrew, Wiltshire. Ancient Monuments Laboratory Report 3591. Coy, J. (1976b) Animal bones from Site V, Bishops Waltham, Hants. English Heritage, Ancient Monuments Laboratory Report 2322.

Coy, J. (1982c) Animal bones from excavations W5, W6, W7, W8, W9, W10 by the Wessex Archaeology Committee at Christchurch, 1981. Ancient Monuments Laboratory Report 3877.

Coy, J. (1977a) Animal bones from Ramsbury, Wiltshire. English Heritage, Ancient Monuments Laboratory Report 2429.

Coy, J. (1982d) The animal bones, level IV report for Ower Monograph. English Heritage, Ancient Monuments Laboratory Report 3876.

Coy, J. (1977b) Bird bones and non-domestic species from Melbourne Street sites I, II, V, VI and XX of Saxon Southampton (Hamwih) with statistical appendix. English Heritage, Ancient Monuments Laboratory Report 2323.

Coy, J. (1983a) The animal bone. In K. S. Jarvis (ed.) Excavations in Christchurch 1969-1980, 91-97. Dorchester, Dorset Natural History and Archaeological Society Monograph 5.

Coy, J. (1977c) Bird bones from town cellars, Poole, Dorset. English Heritage, Ancient Monuments Laboratory Report 2327.

Coy, J. (1983b) Animal bone. In S. M. Davis. Excavations at Christchurch, Dorset, 1981 to 1983. Proceedings of the Dorset Natural History and Archaeological Society 105, 43-45.

Coy, J. (1977d) Bird bones from Upper Bulge street, Southampton. English Heritage, Ancient Monuments Laboratory Report 2325.

Coy, J. (1983c) Animal bone identification from Clarendon Park, Vatcher site 10. English Heritage, Ancient Monuments Laboratory Report 4069.

Coy, J. (1977e) Small mammal and bird bones from Thames Street, Poole, Dorset. English Heritage, Ancient Monuments Laboratory Report 2328.

Coy, J. (1983d) The animal bones. In P. Woodward (ed.) Wimborne Minster, Dorset. Excavations in the Town Centre 1975-1980. Proceedings of the Dorset Natural History and Archaeological Society 105, 72-73.

Coy, J. (1977f) Small mammal, bird , amphibian and fish bones from soil samples taken from site V111, Hamwih, Southampton. English Heritage, Ancient Monuments Laboratory Report 2324.

Coy, J. (1983e) Animal bones and marine molluscs from the 1979 excavations at Rope Lake, Hole, Kimmeridge, Dorset. English Heritage, Ancient Monuments Laboratory Report 4070.

Coy, J. (1978a) Animal bones from Christchurch, Dorset. English Heritage, Ancient Monuments Laboratory Report 2643.

Coy, J. (1983f) The animal bones from Lake Farm Roman Fort, Poole Museums site No. 20. English Heritage, Ancient Monuments Laboratory Report 4133.

Coy, J. (1978b) Animal bones from R27, M3 Motorway Excavations. English Heritage, Ancient Monuments Laboratory Report 2648.

Coy, J. (1983g) The Animal Bones from the Pre-Roman and Roman Layers at Winchester, Western Suburbs 1974-9. English Heritage, Ancient Monuments Laboratory Report 4063.

Coy, J. (1980a) The animal bones. In J. Haslam, L. Biek and R. F. Tylecote. A Middle Saxon Iron Smelting Site at Ramsbury, Wiltshire. Medieval Archaeology 24, 41-51.

Coy, J. (1983h) Bird bones from the Maddison Street, 1979, excavations, Southampton site 29. English Heritage, Ancient Monuments Laboratory Report 4064.

Coy, J. (1980b) Animal bones from the Micheldever Wood banjo, Hampshire, R27 M3 motorway rescue

192

Bibliography of the A303 road improvement between Andover, Hampshire and Amesbury, Wiltshire 1984 - 1987. Proceedings of the Hampshire Field Club and Archaeological Society 47, 5-81.

Coy, J. (1984a) Animal bones from Foxcotte, deserted Medieval village, Hants. English Heritage, Ancient Monuments Laboratory Report 4911. Coy, J. (1984b) Animal bones from Saxon, Medieval and Post-medieval phases (10-18) of Winchester Western Suburbs. English Heritage, Ancient Monuments Laboratory Report 4910.

Coy, J. (1992) The animal bone. In I. P. Horsey (ed.) Excavations in Poole 1973-1983, 186-193. Dorchester, Dorset Natural History and Archaeological Society. Monograph Series No 10.

Coy, J. (1984c) The bird bones. In B. Cunliffe (ed.) Danebury: an Iron Age Hillfort in Hampshire. Volume 2. The Excavations 1969-1978: the Finds, 527-531. London, Council for British Archaeology Research Report 52.

Coy, J. (1993) The assemblage from context 320 Bell Street. In H. Rees (ed.) Late Bronze Age and Early Iron Age settlement in the Lower Test Valley. Evidence from excavations and finds, 1981-9. Proceedings of the Hampshire Field Club and Archaeological Society 49, 39-41.

Coy, J. (1985a) Animal bones from excavations at Wickham Glebe, Hampshire, 1976-1980. English Heritage, Ancient Monuments Laboratory Report 4914.

Coy, J. and Maltby, M. (1978) Report on 3 small collections of animal bones from ring ditches at R17, R30 and R36, M3 motorway rescue excavations. English Heritage, Ancient Monuments Laboratory Report 2770.

Coy, J. (1985b) Animal bones from Poole, Dorset. Medieval and Post-Medieval Layers from sites PM2, PM3, PM7, PM9, PM10 and PM11. English Heritage, Ancient Monuments Laboratory Report 1985/30.

Coy, J. and Maltby, M. (1987) Archaeozoology in Wessex. In H. C. M. Keeley (ed.) Environmental Archaeology a Regional Review Vol 2, 204-251. London, Historic Buildings and Monuments Commission for England Occasional Paper No. 1.

Coy, J. (1985c) Faunal remains. In A. D. Russel (ed.) Foxcotte: The archaeology and history of a Hampshire Hamlet. Proceedings of the Hampshire Field Club and Archaeological Society 41, 213-216.

Coy, J. and Winder, J. (1975a) Animal bones from the London Lodge Complex, Sites R3, M3 Archaeology Rescue Committee. English Heritage, Ancient Monuments Laboratory Report 1801.

Coy, J. (1986) Animal bones from Excavations at Osborne House, Romsey RII 1972. English Heritage, Ancient Monuments Laboratory Report 4917.

Coy, J. and Winder, J. (1975b) Animal bones from the Osbourne House Site, Romsey, Hants. English Heritage, Ancient Monuments Laboratory Report 1803.

Coy, J. (1987a) Animal bone. In P. J. Fasham (ed.) A 'Banjo' Enclosure in Micheldever Wood, Hampshire, 4553. Winchester, Hampshire Field Club and Archaeological Society Monograph 5.

Coy, J. and Winder, J. (1976a) Animal bone from a Bronze Age barrow and Iron Age pits at Easton Down, Site R7, M3 Motoway Archaeological Rescue Committee. English Heritage, Ancient Monuments Laboratory Report 2070.

Coy, J. (1987b) The animal bones. In N. Sunter and P. J. Woodward (eds.) Romano-British Industries in Purbeck, 114-118. Dorchester, Dorset Natural History and Archaeological Society Monograph Series No. 6.

Coy, J. and Winder, J. (1976b) Animal bones from an early Bronze Age barrow, site R4, M3 Archaeological Rescue Committee. English Heritage, Ancient Monuments Laboratory Report 2068.

Coy, J. (1987c) The animal bones and marine molluscs. In N. Sunter and P. J. Woodward (eds.) Romano-British Industries in Purbeck, 178-179. Dorchester, Dorset Natural History and Archaeological Society Monograph Series No. 6.

Coy, J. and Winder, J. (1976c) Animal bones from sites R5 and R6, M3 motorway archaeological rescue committee. English Heritage, Ancient Monuments Laboratory Report 2069.

Coy, J. (1987d) Animal bones from Abbots Worthy (Itchen Abbas Road), Hampshire. English Heritage, Ancient Monuments Laboratory Report 156/87.

Coy, J. and Winder, J. (1981) Animal bone. In P. J. Fasham. Fieldwork and excavations at East Stratton along the Roman Road from Winchester to Silchester (MARC3 sites R1 and R3). Proceedings of the Hampshire Field Club and Archaeological Society 37, 182-183.

Coy, J. (1988) Animal bones from medieval contexts (Trenches A, A1, L and M) at Brighton Hill South, Farleigh Wallop, Hampshire 1984-6. English Heritage, Ancient Monuments Laboratory Report 34/1988. Coy, J. (1991) The animal bones. In P. S. Bellamy. The investigation of the prehistoric landscape along the route 193

Investigating Animal Burials; Ritual, Mundane and Beyond Heritage and Oxford Univeristy Archaeology. Monograph No 49.

Crabtree, P. (1989a) Sheep, horses, swine, and kine: A zooarchaeological perspective on the Anglo-Saxon settlement of England. Journal of Field Archaeology 16 (2), 205-213.

Committee

for

Cunliffe, B. and Poole, C. (2000b) The Danebury Environs Programme. The Prehistory of a Wessex landscape. Volume 2, Part 3. Suddern Farm, Middle Wallop, Hants, 1991 and 1996. Oxford, English Heritage and Oxford University Committee for Archaeology Monograph No. 49.

Crabtree, P. (1989b) West Stow, Suffolk: early Anglo Saxon animal husbandry. Bury St Edmunds, East Anglian Archaeology Report No. 47. Crabtree, P. (1995) The symbolic role of animals in Anglo-Saxon England: evidence from burials and cremations. In K. Ryan and P. Crabtree (eds.) The Symbolic Role of Animals in Archaeology, 20-26. Philadelphia, MASCA Research Papers in Science and Archaeology 12.

Currie, T. C. (1993) The animal bone. In C. K. Currie (ed.) Excavations at the Postern Mill site, Malmesbury, 1986-87. Wiltshire Archaeological and Natural History Magazine 86, 72-74. Currie, T. C. (1995) Summary report on bones. In C. K. Currie. Rescue excavation in the town of Wootton Bassett, 1986. Wiltshire Archaeological and Natural History Magazine 88, 80-81.

Crabtree, P. (1996) Production and consumption in an early complex society: Animal use in middle Saxon East Anglia. World Archaeology 28 (1), 58-75.

Daróczi-Szabó, L. (2004) Animal bones as indicators of kosher food refuse from 14th century AD, Buda, Hungary. In S. J. O Day, W. Van Neer and A. Ervynck (eds.) Behaviour Behind Bones. The zooarchaeology of Ritual, Religion, Status and Identity, 252-261. Oxford, Oxbow.

Cunliffe, B. (1974) Iron Age Communities in Britain (1st Edition). London, Routledge Kegan and Paul. Cunliffe, B. (ed.) (1975) Excavations at Porchester Castle. Volume 1: Roman. London, Society of Antiquaries of London.

Darvill, T. (2007) Stonehenge: The Biography of a Landscape. Stroud, Tempus.

Cunliffe, B. (1984) Danebury: an Iron Age Hillfort in Hampshire. Volume 2. The Excavations 1969-1978: the Finds. London, Council for British Archaeology Research Report 52.

Darvill, T. and Russell, B. (2002) Archaeology after PPG16: Archaeological Investigations in England 19901999. London and Bournemouth, English Heritage and Bournemouth University, School of Conservation Sciences Research Report 10.

Cunliffe, B. (1988) Greeks, Romans and Barbarians: Spheres of Interaction London, Batsford. Cunliffe, B. (1991a) The deposition of human remains. In B. Cunliffe and C. Poole (eds.) Danebury: An Iron Age Hillfort in Hampshire. Volume 5, The Excavations 19791988: The Finds, 418-425. London, Council for British Archaeology.

Davies, J. J. (2000) The animal bone. In A. Whittle, J. J. Davies, I. Dennis, A. S. Fairbairn and M. A. Hamilton. Neolithic activity and occupation outside Windmill Hill causewayed enclosure, Wiltshire: survey and excavation 1992-93. Wiltshire Archaeological and Natural History Magazine 93, 154-168.

Cunliffe, B. (1991b) Iron Age Communities in Britain (3rd Edition). London, Routledge.

Davis, S. and Payne, S. (1993) A barrow full of cattle skulls. Antiquity 67, 12-22.

Cunliffe, B. (1992) Pits, preconceptions and propitiation in the British Iron Age. Oxford Journal of Archaeology 11, 69-84.

Davis, S. J. M. (1987) Animal bone from sites A and C at Wilsford. English Heritage, Ancient Monuments Laboratory Report 72/87.

Cunliffe, B. (2005) Iron Age Communities in Britain (4th Edition). London, Routledge. Cunliffe, B. and Poole, C. (1991) Danebury: An Iron Age Hillfort in Hampshire. Volume 5, The Excavations 19791988: The Finds. London, Council for British Archaeology.

Dawkins, W. B. and Jackson, J. W. (1917) The remains of the Mammalia in the Lake Village of Glastonbury. In A. Bulleid and H. G. Gray (eds.) The Glastonbury Lake Village; Volume 2, 641-672. Glastonbury, Glastonbury Antiquarian Society.

Cunliffe, B. and Poole, C. (2000a) The Danebury Environs Programme. The Prehistory of a Wessex landscape. Volume 2, Part 1. Woolbury and Stockbridge Down, Stockbridge, Hants, 1989. Oxford, English

Denston, C. B. (1970) The human and animal remains. In J. X. W. P. Corcoran (ed.) The Giant's Caves, Luckington (Wil 2). Wiltshire Archaeological and Natural History Magazine 65B, 57-62.

194

Bibliography Dietler, M. (2001) Theorising the feast: rituals of consumption, commensal politics and power in African contexts. In M. Dietler and B. Hayden (eds.) Feasts: archaeological and ethnographic perspectives on food, politics and power, 64-114. Washington DC, Smithsonian Institution.

Signifying Animals: Human meaning in the Natural World, 25-36. London, Routledge.

Dobney, K. (2001) A place at the table: the role of vertebrate zooarchaeology within a Roman research agenda for Britan. In S. James and M. Millett (eds.) Britons and Romans: Advancing an Archaeological Agenda, 36-45. London, Council for British Archaeology Research Report 125.

Drew, C. D. (1931) The excavations at Jordan Hill, 1931. Proceedings of the Dorset Natural History and Archaeological Society LIII, 265-273.

Dobney, K. (2005) Animal bone. In A. E. Finney, M. R. Stephens and P. A. Ware. Excavations at West Street, Gargrave, North Yorkshire. Yorkshire Archaeological Journal 77, 135-136.

Eastham, A. (1976) The bird bones. In B. Cunliffe (ed.) Excavations at Portchester Castle Volume II: Saxon, 287296. London, Reports of the Research Committee of the Society of Antiquaries of London No. 33.

Dobney, K., Jaques, D. and Brothwell, D. (1999) Assessment of the bone assemblage from COT93. In J. D. Richards (ed.) Cottam: an Anglo-Scandinavian settlement on the Yorkshire Wolds. The Archaeological Journal 156, 84-85.

Eastham, A. (1977) Birds. In B. Cunliffe (ed.) Excavations at Portchester Castle. Volume III: Medieval, the Outer Bailey and its Defences, 233-238. London, Reports of the Research Committee of the Society of Antiquaries of London. No.34.

Dobney, K., Jaques, D., Carrott, J., Hall, A., Issitt, M. and Large, F. (2000) Biological remains. In P. Ottaway. Excavations on the site of the Roman signal station at Carr Naze, Filey, 1993-94. The Archaeological Journal 157, 148-182.

Eastham, A. (1985) Bird bone. In B. Cunliffe and J. Munby (eds.) Excavations at Portchester Castle. Volume IV: Medieval, the Inner Bailey, 261-269. London, Reports of the Research Committee of the Society of Antiquaries of London. No.4.

Dobney, K., Jaques, D. and Irving, B. (1996) Of Butchers and Breeds. report on vertebrate remains from various sites in the City of Lincoln. Lincoln, Lincoln Archaeological studies 5.

Edwards, A. and Horne, M. (1997) Animal bone from the West Kennet palisade enclosures. In A. Whittle (ed.) Sacred Mound Holy Rings. Silbury Hill and the West Kennet palisade enclosures: a later Neolithic complex in North Wiltshire, 47-49. Oxford, Oxbow Monograph 74.

Draper, J. and Chaplin, C. (1982) Dorchester Excavations Volume 1. Dorchester, Dorset Natural History and Archaeological Society Monograph Series No 2.

Dunbabin, K. (2003) The Roman Banquet: Images of conviviality. Cambridge, Cambridge University Press.

Donaldson, P. (1977) The excavation of a multiple round barrow at Barnack, Cambridgeshire. Antiquaries Journal 57, 197-231.

Efremov, I. A. (1940) Taphonomy: a new branch of paleontology. Pan-American Geologist 74, 81-93.

Done, G. (1986) The animal bones from areas A and B. In M. Millett and D. Graham (eds.) Excavations on the Romano-British Small Town at Neatham Hampshire, 1969-1979, 141-148. Winchester, Hampshire Field Club and Archaeological Society Monograph No 3.

Egerton, J. (1996) Animal bones. In M. Rawling and A. P. Fitzpatrick. Prehistoric sites and a Romano-British Settlement at Butterfield Down, Amesbury. Wiltshire Archaeological and Natural History Magazine 89, 35-36. Ellis, C. J. and Rawlings, M. (2001) Excavations at Balksbury Camp, Andover 1995-97. Proceedings of the Hampshire Field Club and Archaeological Society 56, 76-80.

Dorell, P. and Cornwall, I. W. (1960) Appendix I Down Farm, Pewsey report on the bones. In F. M. Vatcher. The excavation of a group of barrows at Down Farm, Pewsey, Wilts. Wiltshire Archaeological and Natural History Magazine 57, 348.

Ellis, P. (ed.) (2000) Ludgershall Castle. Excavations by Peter Addyman 1964-1972. Salisbury, Wiltshire Archaeological and Natural History Society Monograph series 2.

Douglas, M. (1963) The Lele of the Kasai. Oxford, Oxford University Press. Douglas, M. (1984) Food and Social Order: Studies of food and festivities. New York, Russell Sage Foundation.

Ellison, A. (1980) Natives, Romans and Christians on West Hill, Uley: an interim report on the excavation of a ritual complex of the first millennium AD. In W. Rodwell (ed.) Temples, Churches and Religion in Roman Britain,

Douglas, M. (1990) The pangolin revisited: a new approach to animal symbolism. In R. Willis (ed.)

195

Investigating Animal Burials; Ritual, Mundane and Beyond

Epstein, H. J. (1943) The origin and earliest history of falconry. Isis 34, 497-509.

Ford, S. (1997) Bone. In S. Ford. Excavation of Saxon structures and Bronze Age features at Bentley Green Farm, Bentley, Hampshire. Proceedings of the Hampshire Field Club and Archaeological Society 52, 71.

Esmonde Cleary, A. S. (1989) The Ending of Roman Britain. London, Routledge.

Fowler, P. (2002) Farming in the First Millennium AD. Cambridge, Cambridge University Press.

Evison, V. I. (1994) An Anglo-Saxon Cemetery at Great Chesterford, Essex. London, Council of British Archaeology research report No. 91.

Fowler, P. J., Musty, J. W. G. and Taylor, C. C. (1965) Some earthwork enclosures in Wiltshire. Wiltshire Archaeological and Natural History Magazine 60, 52-74.

Fasham, P. J., Farwell, D. E. and Whinney, R. J. B. (eds.) (1989) The Archaeological Sites at Easton Lane, Winchester. Winchester, Hampshire Field Club and Archaeological Society Monograph No 6.

Fox, C. and Wolseley, G. R. (1928) The early Iron Age sites at Findon Park, Sussex. The Antiquaries Journal 8, 449-460.

305-320. Oxford, British Archaeological Report British series 77.

Fraser, F. C. (1963) Appendix D. Rag Copse animal bones. In. G. M. Knocker (Ed.). Excavation of a Round Barrow in Rag Copse, Near Hurstbourne Tarrant, Hants.

Faull, M. L. (1981) The Roman period. In M. L. Faull and S. A. Moorhouse (eds.) West Yorkshire: An Archaeological Survey To A.D. 1500, 141-169. Leeds, University of Leeds.

Fraser, F. C. and Ryder, M. L. (1968) Animal bones. In L. P. Wenham (ed.) The Romano-British Cemetery at Trentholme Drive, York, 104-109. London, HMSO.

Firth, R. (1963) Offering and sacrifice: problems of organisation. Journal of the Royal Anthropological Institute 93, 12-24.

Fraser, J. C. (1958) Animal bones. In J. S. Wacher. Litton Cheney Excavations. Proceedings of the Dorset Natural History and Archaeological Society 80, 177.

Fisher, C. and Dartnall, D. L. (1987) Roman and later animal remains and Roman molluscs. In H. R. Hurst, D. L. Dartnall and C. Fisher. Excavations at Box Roman Villa, 1967-8. Wiltshire Archaeological and Natural History Magazine 81, 49.

Fulford, M. (2001) Links with the past: pervasive 'ritual' behaviour in Roman Britain. Britannia 32, 119-218. *DEDOyZQD /   3RFKyZNL E\GOĊFH NXOWXU\ DPIRU NXOLVW\FK ]H VWDQRZLVND  Z %U]HĞFLX .XMDZVNLP Z ĞZLHWOH SRGREQ\FK ]QDOH]LVN NXOWXU ĞURGNRZRHXURSHMVNLFK3UDFHL0DWHULDO\/yGĨ 3, 63-108.

Fitzgerald, C. and Egerton, J. (1994) The animal bone. In R. M. J. Cleal and M. J. Allen (eds.) Investigation of treedamaged barrows on King barrow ridge and Luxenborough Plantation, Amesbury. Wiltshire Archaeological and Natural History Magazine 87, 75-76.

Gardner, N. (1997) Vertebrates and small vertebrates. In A. Whittle (ed.) Sacred Mound Holy Rings. Silbury Hill and the West Kennet Palisade Enclosures: a Later Neolithic Complex in North Wiltshire, 47-49. Oxford, Oxbow Monograph 74.

Fitzpatrick, A. P. (1989) The uses of Roman imperialism by the Celtic barbarians in the later Republic. In J. C. Barrett, A. P. Fitzpatrick and L. Macinnes (eds.) Barbarians and Romans in North-West Europe: from late Republic to late Antiquity, 27-54. Oxford, British Archaeological Report International Series 471.

Gaunt, G. D. and Buckland, P. C. (2003) The geological background to Yorkshire's archaeology. In T. G. Manby, S. Moorhouse and P. Ottaway (eds.) The Archaeology of Yorkshire. An Assessment at the Beginning of the 21st Century, 17-24. Leeds, Yorkshire Archaeological Society Occasional Paper No. 3.

Fitzpatrick, A. P. (1991) 'Celtic (Iron Age) religion' Traditional and Timeless? Scottish Archaeology Review 8, 123-128. Fitzpatrick, A. P. (1997) Everyday life in Iron Age Wessex. In A. Gwilt and C. Haselgrove (eds.) Reconstructing Iron Age Societies, 73-86. Oxford, Oxbow.

Gell, A. (1998) Art and Agency. An Anthropological Theory. Oxford, Clarendon Press. Gentry, A., Clutton-Brock, J. and Groves, C. P. (2004) The naming of wild animal species and their domestic derivatives. Journal of Archaeological Science 31 (5), 645-651.

Fitzpatrick, A. P. (2001) Cross channel exchange, Hengistbury Head and the end of hillforts. In J. Collis (ed.) Society and settlement in Iron Age Europe, 82-97. Sheffield, J. R. Collis Publications.

Gidney, L. (1994) The animal bone. In R. L. Fitts, C. C. Haselgrove, P. C. Lowther and P. Turnbull. An Iron Age 196

Bibliography Gordan, B. C. and Buikstra, J. E. (1981) Soil pH, bone preservation and sampling bias at mortuary sites. American Antiquity 46 (566-571)

farmstead at Rock Castle Gilling West, North Yorkshire. Durham Archaeological Journal 10, 31. Gidney, L. (1999a) The animal bone. In. M. C. Bishop (Ed.). An Iron Age and Romano-British 'ladder' settlement at Melton, East Yorkshire.

Gosden, C. and Marshall, Y. (1999) The cultural biography of objects. World Archaeology 31 (2), 169178.

Gidney, L. (1999b) The animal bones. In P. Halkon and M. Millett (eds.) Rural Settlement and Industry: Studies in the Iron Age and Roman Archaeology of Lowland East Yorkshire, 81. York, Yorkshire Archaeological Report No 4.

Graham, A. (2006) The Excavation of Five Beaker Burials, the Iron Age and Romano-British settlements, and the 4th Century Courtyard Villa at Barton Field, Tarrant Hinton. Dorchester, Dorset Natural History and Archaeological Society Monograph Series: 17.

Gidney, L. (1999c) The animal bones. In P. Halkon and M. Millett (eds.) Rural Settlement and Industry: Studies in the Iron Age and Roman Archaeology of Lowland East Yorkshire, 71. York, Yorkshire Archaeological Report No 4.

Graham, A. and Newman, C. (1993) Recent Excavations of Iron Age and Romano-British Enclosures in the Avon Valley, Wiltshire. Wiltshire Archaeological and Natural History Magazine 86, 8-57. Grant, A. (1970) The animal bones. In B. Cunliffe. A Bronze Age settlement at Chalton Hants (Site 78). The Antiquaries Journal 50, 11-12.

Gidney, L. (2004) Animal bone. In P. G. E. Neal and R. A. Fraser. Romano-British Enclosed Farmstead at Billinglet Drive, Thurnscoe, South Yorkshire. Yorkshire Archaeological Journal 77, 58-59.

Grant, A. (1975) The animal bones. In B. Cunliffe (ed.) Excavations at Porchester Castle. Volume 1: Roman, 437-450. London, Report of the Research Committee Society of Antiquaries of London 32.

Gifford, D. P. (1977) Observations of modern human settlements as an aid to archaeological interpretation, University of California, PhD.

Grant, A. (1976a) Animal bone. In E. Lewis and G. Walker. A middle Bronze Age settlement site at Westbury, West Meon, Hampshire. Proceedings of the Hampshire Field Club and Archaeological Society 33, 41.

Gilchrist, R. (1996) The animal bones. In M. Foreman (ed.) Further Excavations at the Dominican Priory, Beverley, 1986-89, 213-228. Sheffield, Sheffield Academic Press. Sheffield Excavation Reports 4. Gilchrist, R. and Mytum, H. C. (1986) Experimental archaeology and burnt animal bone from archaeological sites. Circaea 4 (1), 29-38.

Grant, A. (1976b) The animal bones. In B. Cunliffe (ed.) Excavations at Portchester Castle Volume II: Saxon, 262286. London, Reports of the Research Committee of the Society of Antiquaries of London No. 33.

Gilhus, I. S. (2006) Animals, Gods and Humans. Changing Attitudes to Animals in Greek, Roman and early Christian Ideas. Oxford, Routledge.

Grant, A. (1977) The animal bones. In B. Cunliffe (ed.) Excavations at Portchester Castle. Volume III: Medieval, the Outer Bailey and its Defences, 213-233. London, Reports of the Research Committee of the Society of Antiquaries of London. No.34.

Gingell, C. (ed.) (1992) The Marlborough Downs: A Late Bronze Age Landscape and its Origins. Salisbury, Wiltshire Archaeological and Natural History Society Monograph 1.

Grant, A. (1982) The use of tooth wear as a guide to the age of domestic ungulates. In B. Wilson, C. Grigson and S. Payne (eds.) Ageing and Sexing Animal Bones from Archaeological Sites, 91-108. Oxford, British Archaeological Report British Series 109.

Gooden, D., Hamilton-Dyer, S., Laidlaw, M. and Merpham, L. (2002) Excavation of Saxon Pits at Tidworth, 1999. Wiltshire Archaeological and Natural History Magazine 95, 240-248.

Grant, A. (1984a) Animal husbandry. In B. Cunliffe (ed.) Danebury: an Iron Age Hillfort in Hampshire. Volume 2. The Excavations 1969-1978: the Finds, 496-548. London, Council for British Archaeology Research Report 52.

Goodman, M. (1999) The Roman World 44BC – AD180. London, Routledge. Goody, J. (1982) Cooking, Cuisine and Class: A Study in Comparative Sociology. Cambridge, Cambridge University Press.

Grant, A. (1984b) Animal husbandry in Wessex and the Thames Valley. In B. Cunliffe and D. Miles (eds.)

197

Investigating Animal Burials; Ritual, Mundane and Beyond Aspects of the Iron Age in Central Southern Britain, 102119. Oxford, OUCA Monograph 2.

Natural History and Archaeological Society 104, 170172.

Grant, A. (1985) The large mammals. In B. Cunliffe and J. Munby (eds.) Excavations at Portchester Castle. Volume IV: Medieval, the Inner Bailey, 244-256. London, Reports of the Research Committee of the Society of Antiquaries of London. No.43.

Gregory, I. (1978) Animal bones. In M. Oliver and B. Applin (eds.) Excavations at Ructstalls Hill, Basingstoke. Proceedings of the Hampshire Field Club and Archaeological Society 35, 82-86. Griffith, N. (1976) The animal bone remains from Knights Enham, Andover, Hampshire. English Heritage, Ancient Monuments Laboratory Report 2430.

Grant, A. (1987) The animal bone. In B. Cunliffe (ed.) Hengistbury Head, Dorset. Vol 1, 322. Oxford Oxford University Committee for Archaeology Monograph. No.13.

Griffith, N., Halstead, P. L. J., MacLean, A. C. and Rowley-Conwy, P. A. (1983) Faunal remains and economy. In P. Mayes and L. Butler (eds.) Sandal Castle Excavations 1964-1973, Wakefield, Wakefield Historical Publications.

Grant, A. (1988) Animal resources. In G. Astill and A. Grant (eds.) The Countryside of Medieval England, 149187. Oxford, Blackwell.

Griffiths, N. L. J. (1978) Animal Bones from R27, M3 Motorway Excavations. English Heritage, Ancient Monuments Laboratory Report 2647.

Grant, A. (1989a) Animals and ritual in Early Britain: the visible and the invisible. In P. Meniel (ed.) Animal et Pratiques Religieuses: Les Manifestations Materielles. Anthropozoologica Numero Special, 79-86.

Grigson, C. (1966) Appendix II: the animal remains from Fussell's Lodge Long Barrow. In P. Ashbee. The Fussell's Lodge Long Barrow Excavations 1957. Archaeologia 100, 63-73.

Grant, A. (1989b) Animals in Roman Britain. In M. Todd (ed.) Research on Roman Britain: 1960-1989, 425-500. London, Britannia Monograph 11.

Grigson, C. (1980) The animal bones. In M. E. Robertson-Mackay (ed.) A 'Head and Hooves' Burial Beneath a Round Barrow, with other Neolithic and Bronze Age Sites, on Hemp Knoll, near Avebury, Wiltshire. Proceeding of the Prehistoric Society 46, 161171.

Grant, A. (1991) Animal husbandry. In B. Cunliffe and C. Poole (eds.) Danebury: an Iron Age Hillfort in Hampshire. Volume 5. The Excavations 1979-1988: the Finds, 447-487. London, Council for British Archaeology Research Report 73. Grant, A. (2000) Diet, economy and ritual evidence from the faunal remains. In M. Fulford and J. Timby (eds.) Late Iron Age and Roman Silchester: Excavations on the Site of the Forum-Basilica 1977, 1980-86, 425-482. London, Britannia Monograph 15.

Grigson, C. (1982) Porridge and pannage: pig husbandry in Neolithic England. In M. Bell and S. Limbrey (eds.) Archaeological Aspects of Woodland Ecology, 297-313. Oxford, British Archaeological Report International Series 146.

Grant, A. (2002) Food, status and social hierarchy. In P. Miracle and N. Milner (eds.) Consuming Passions and Patterns of Consumption, 17-24. Cambridge, McDonald Institute Monographs.

Grigson, C. (1983) Mesolithic and Neolithic animal bones. In J. G. Evans and I. F. Smith. Excavations at Cherhill, North Wiltshire. Proceedings of the Prehistoric Society 49, 64-72.

Green, C. and Rollo-Smith, S. (1984) Eighteen Barrows near Shrewton, Wiltshire. Proceedings of the Prehistoric Society 50, 255-318.

Grigson, C. (1986) Animal remains. In P. Ashbee. The excavation of Milton Lilbourne Barrows 1-5. Wiltshire Archaeological and Natural History Magazine 80, 74-80.

Green, C. S. (ed.) (1987) Excavations at Poundbury, Dorchester, Dorset 1966-1982. Vol 1: The Settlements. Dorchester, Dorset Natural History and Archaeological Society. Monograph Series No 7.

Grigson, C. (1989) Large mammals. In P. Ashbee, M. Bell and E. Proudfoot (eds.) Wilsford Shaft: Excavations 1960-62, 106-121. London, English Heritage Archaeology Report No 11.

Green, M. J. (1992) Animals in Celtic Life and Myth. London, Routledge.

Grigson, C. (1999) The mammalian remains. In A. Whittle, J. Pollard and C. Grigson (eds.) The Harmony of Symbols. The Windmill Hill Causewayed Enclosure, 164252. Oxford, Oxbow.

Green, M. T. (1982) Second and Final Interim Report on Excavation of an Unrecorded Pond Barrow at Down Farm, Gussage St Michael. Proceedings of the Dorset

198

Bibliography Hamilton-Dyer, S. (1993) Animal bones. In R. J. C. Smith (ed.) Excavations at County Hall, Dorchester, Dorset, 1988, 77-82. Salisbury, Wessex Archaeology Report No 4.

Grinsell, L. V. (1959) Dorset Barrows. Dorchester, Dorset Natural History and Archaeological Society. Hadley, D. M. (2005) Dining in disharmony in the later Middle Ages. In M. Carroll, D. M. Hadley and H. Willmott (eds.) Consuming Passions. Dining From Antiquity to the Eighteenth Century, 101-120. Stroud, Tempus.

Hamilton-Dyer, S. (1996a) Animal bone. In J. I. Mckinley and M. Heaton. A Romano-British Farmstead and Associted Burials at Maddington Farm, Shrewton. Wiltshire Archaeological and Natural History Magazine 89, 63-65.

Halkon, P. (2003) Researching an ancient landscape: The Foulness Valley, East Yorkshire. In T. G. Manby, S. Moorhouse and P. Ottaway (eds.) The Archaeology of Yorkshire. An Assessment at the Beginning of the 21st Century, 261-272. Leeds, Yorkshire Archaeological Society Occasional Paper No. 3.

Hamilton-Dyer, S. (1996b) Animal remains. In A. D. Crockett. Excavations at Montefiore new halls of residence, Swaythling, Southampton, 1992. Proceedings of the Hampshire Field Club and Archaeological Society 51, 43-46.

Hall, R. (1996) York. London, English Heritage/Basford Ltd.

Hamilton-Dyer, S. (1997a) Animal bone. In C. A. Butterworth and R. Seager Smith. Excavations at the Hermitage, Old town, Swindon. Wiltshire Archaeological and Natural History Magazine 90, 71-72.

Hall, R. A. (2003) Yorkshire AD 700-1066. In T. G. Manby, S. Moorhouse and P. Ottaway (eds.) The Archaeology of Yorkshire. An Assessment at the Beginning of the 21st Century, 171-180. Leeds, Yorkshire Archaeological Society Occasional Paper No. 3.

Hamilton-Dyer, S. (1997b) The animal bones. In M. Fulford, S. Rippon, S. Ford, J. Timby and W. B. Silchester: Excavations at the North Gate, on the North Walls, and in the Northen Suburbs 1988 and 1991-3. Britannia 28, 131-135.

Hambleton, E. (1999) Animal Husbandry Regimes in Iron Age Britain. Oxford, British Archaeological Report British Series 282.

Hamilton-Dyer, S. (1998) Animal bone. In A. P. Fitzpatrick and A. D. Crockett. A Romano-British settlement and inhumation cemetery at Eyewell Farm, Chilmark. Wiltshire Archaeological and Natural History Magazine 91, 29-30.

Hambleton, E. (2001) Animal bone. In G. Walker, B. Langton and N. Oakey (eds.) An Iron Age Site at Groundwell West, Blunsdon St. Andrew, Wiltshire, 31-33. Cirencester, Cotswold Archaeological Trust Ltd.

Hamilton-Dyer, S. (1999a) Animal bone. In J. I. Mckinley. Further Excavations of an Iron Age and Romano-British enclosed settlement at Figheldean, near Netheravon. Wiltshire Archaeological and Natural History Magazine 92, 25-27.

Hambleton, E. (2007) Review of Middle Bronze Age to Late Iron Age faunal Assemblages from southern Britain. Unpublished Report Hambleton, E. and Maltby, M. (2004) The animal bone from Battlesbury Bowl. Unpublished Report

Hamilton-Dyer, S. (1999b) Animal Bones. In C. M. Hearne and V. Birbeck (eds.) A35 Tolpuddle to Puddletown Bypass DBFO, Dorset, 1996-8, 188-202. Salisbury, Wessex Archaeology Report No.15.

Hamerow, H. (2006) 'Special deposits' in Anglo-Saxon settlements. Medieval Archaeology 50, 1-30. Hamilton-Dyer, S. (1986) A note on the sieved material from the Garderobe layers A296-8. In J. Oxley (ed.) Excavations at Southampton Castle, 37-38. Southampton, Southampton City Museums.

Hamilton-Dyer, S. (2000a) Animal bone. In P. Andrews, L. Mepham and R. S. Smith (eds.) Excavations in Wilton 1995-6: St John's Hospital and South Street. Wiltshire Archaeological and Natural History Magazine 93, 198199.

Hamilton-Dyer, S. (1991) Animal bone. In P. W. Cox and C. M. Hearne (eds.) Redeemed from the Heath. The Archaeology of the Wytch Farm Oilfield (1987-90), 209212. Dorchester, Dorset Natural History and Archaeological Society Monograph. No.9.

Hamilton-Dyer, S. (2000b) Animal bones. In M. Rawlings (ed.) Excavations at Ivy Street and Brown Street, Salisbury, 1994. Wiltshire Archaeological and Natural History Magazine 93, 45-51.

Hamilton-Dyer, S. (1992) Animal bone. In R. J. C. Smith, M. Rawlings and I. Barnes. Excavations at Coburg Road and Weymouth Road, Fordington, Dorchester, 1988 and 1989. Proceedings of the Dorset Natural History and Archaeological Society 114, 41-42.

Hamilton-Dyer, S. (2001a) Animal bone. In K. Taylor. The excavation of Medieval and Post-medieval features at 3 Kingsbury Square, Wilton. Wiltshire Archaeological and Natural History Magazine 94, 74.

199

Investigating Animal Burials; Ritual, Mundane and Beyond Hamilton-Dyer, S. (2001b) Animal bone. In J. Pine (ed.) The excavation of a Saxon settlement at Cadley Road, Collingbourne Ducis, Wiltshire. Wiltshire Archaeological and Natural History Magazine 94, 102-109.

Suddern Farm, Middle Wallop, Hants, 1991 and 1996, 175-193. Oxford, English Heritage and Oxford University Committee for Archaeology Monograph No. 49.

Hamilton-Dyer, S. (2001c) Animal bones. In M. Garner (ed.) A middle Saxon cemetery at Cook Street, Southampton (SOU 823). Proceedings of the Hampshire Field Club and Archaeological Society 56, 182-186.

Hamilton, J. (2004) The animal bone. In D. Stansbie and G. Laws. Prehistoric settlement and Medieval to Postmedieval field systems at Latton lands. Wiltshire Archaeological and Natural History Magazine 97, 128133.

Hamilton-Dyer, S. (2002a) Animal bone. In S. D. G. Weaver (ed.) The excavation of Iron Age and early Roman features at Vicking Way, Andover, Hampshire, 1996. Proceedings of the Hampshire Field Club and Archaeological Society 57, 1-19.

Hammon, A. (2006) Vertebrate remains. In T. Wilmott and D. Shipp. The Romano-British roadside settlement of White-Walls, Easton Grey, Wiltshire: recent fieldwork. Wiltshire Archaeological and Natural History Magazine 99, 184.

Hamilton-Dyer, S. (2002b) Animal bones. In M. P. Smith (ed.) A Roman settlement and iron Age working site at St Denys, Southampton (SOU 981). Proceedings of the Hampshire Field Club and Archaeological Society 57, 30-37.

Hammond, P. W. (1993) Food and Feast in Medieval England. London, Stroud. Handelman, D. (2006) Conceptual alternatives to 'ritual'. In J. Kreinath, J. Snoek and M. Stausberg (eds.) Theorizing Rituals: Issues, Topics, Approaches, Concepts, 37-49. Leiden, Brill.

Hamilton-Dyer, S. (2003a) Animal bone. In M. Heaton. Neolithic Pits at the Beehive. Wiltshire Archaeological and Natural History Magazine 96, 58.

Harcourt, R. (1968a) Animal bones from Brough, Yorks. English Heritage, Ancient Monuments Laboratory Report 1549.

Hamilton-Dyer, S. (2003b) The faunal remains. In M. Hall-Torrence and S. Weaver (eds.) The excavation of a Saxon settlement at Riverdene, Basingstoke, Hampshire. 1995. Proceedings of the Hampshire Field Club and Archaeological Society 58, 92-94.

Harcourt, R. (1968b) Animal remains from Cowdown, Longbridge Deverill. English Heritage, Ancient Monuments Laboratory Report 1626.

Hamilton-Dyer, S. (2004a) Animal Bone. In R. M. J. Cleal, M. J. Allen and C. Newman. An archaeological and environmental study of the Neolithic and Later Prehistoric landscape of the Avon Valley and Durrington Walls environs. Wiltshire Archaeological and Natural History Magazine 97, 226-228.

Harcourt, R. (1969a) Animal bones. In J. S. Wacher (ed.) Excavations at Brough-on-Humber 1958-1961, 231-233. Leeds, Reports of the Research Committee of the Society of Antiquaries of London No. 25. Harcourt, R. (1969b) Animal bones from Marden. English Heritage, Ancient Monuments Laboratory Report 1555.

Hamilton-Dyer, S. (2004b) Animal bones. In J. Pine and S. Preston (eds.) Early medieval settlement on land adjoining Froman's, Cow Drove Hill, Kings Somborne, Hampshire. Proceedings of the Hampshire Field Club and Archaeological Society 59, 154-156.

Harcourt, R. (1969c) Animal bones from Weaverthorpe Manor. English Heritage, Ancient Monuments Laboratory Report 1560.

Hamilton-Dyer, S. (2005a) Animal bone. In B. Barber. The development of Trinity Chequer: Excavations at the Anchor Brewery site, Gigant Street, Salisbury. Wiltshire Archaeological and Natural History Magazine 98, 200204.

Harcourt, R. (1969d) Animal remains from Balksbury Camp. English Heritage, Ancient Monuments Laboratory Report 1625. Harcourt, R. (1969e) Animal remains from Balksbury Camp. In G. Wainwright. The excavation at Balksbury Camp, Andover, Hampshire. Proceedings of the Hampshire Field Club and Archaeological Society 26, 21-55.

Hamilton-Dyer, S. (2005b) The animal remains. In V. Birbeck, R. J. C. Smith, P. Andrews and N. Stoodley (eds.) The origins of Mid-Saxon Southampton: Excavations at friends Provident St Mary's Stadium 1998-2000, Salisbury, Wessex Archaeology.

Harcourt, R. (1971a) The animal bones. In. N. P. Thompson (Ed.). Archaeological research in the Pewsey Vale.

Hamilton, J. (2000) The animal bone. In B. Cunliffe and C. Poole (eds.) The Danebury Environs Programme. The Prehistory of a Wessex landscape. Volume 2, Part 3.

200

Bibliography Britain: 1960-89, 1-18. London, Britannia Monograph 11.

Harcourt, R. (1971b) Animal bones from Durrington Walls. In G. J. Wainwright and I. H. Longworth (eds.) Durrington Walls: Excavations 1966-1968, 338-350. London, Reports of the Research Committee of the Society of Antiquaries of London 30.

Hawkes, C. (1954) Archaeological theory and method: some suggestions from the old world. American Anthropologist 56, 155-68.

Harcourt, R. (1971c) Animal bones from Marden. In G. J. Wainwright, J. G. Evans and I. H. Longworth. The excavation of a Late Neolithic enclosure at Marden, Wiltshire. The Antiquaries Journal 51, 234-235.

Hayden, B. (2001) Fabulous feasts: a prolegomenon to the importance of feasting. In M. Dietler and B. Hayden (eds.) Feasts: archaeological and ethnographic perspectives on food, politics and power, 23-64. Washington DC, Smithsonian Institution.

Harcourt, R. (1971d) Animal bones from Mount Pleasant. English Heritage, Ancient Monuments Laboratory Report 1627.

Hayes, R. H. (1983) Levisham Moor Archaeological Investigations, 1957-1978. Scarborough, North York Moors National Park Committee and Scarborough Archaeological and Historical Society.

Harcourt, R. (1971e) Animal bones from Wilsford Down. English Heritage, Ancient Monuments Laboratory Report 1569.

Hayfield, C. (1986) Wharram Grange Roman villa 197582. In P. Rahtz, C. Hayfield and J. Bateman (eds.) Two Roman Villas at Wharram Le Street, 15-28. York, York University Archaeological Publications 2.

Harcourt, R. (1972) Animal bones from Weaverthorpe Manor. In T. C. M. Brewster. An excavation at Weaverthorpe Manor, East riding, 1960. Yorkshire Archaeological Journal 44, 133.

Hayfield, C. and Greig, J. (1990) Excavation and Salvage work on a Moated site at Cowick, South Humberside, 1976. Part Two: The Finds Assemblage. Yorkshire Archaeological Journal 62, 111-124.

Harcourt, R. (1975) The Animal Bones from GussageAll-Saints, Dorset. English Heritage, Ancient Monuments Laboratory Report 1804.

Haynes, G. (1980) Evidence of carnivore gnawing on Pleistocene and recent mammalian bones. Paleobiology 6 (3), 341-351.

Harcourt, R. (1979a) The animal bones. In G. J. Wainwright (ed.) Mount Pleasant, Dorset: Excavations 1970-1971, 214-215. London, Society of Antiquaries of London.

Hedges, J. (1983) Isbister: A Chambered Tomb in Orkney. Oxford, British Archaeological Reports British Series 115.

Harcourt, R. (1979b) The animal bones. In G. J. Wainwright (ed.) Gussage All Saints, 150-160. London, Department of the Environment Archaeology Reports No.10.

Hedges, J. (1984) The Tomb of the Eagles. London, John Murray.

Harcourt, R. (1986) Animal bones. In J. Musty and D. Algar. Excavations at the deserted medieval village of Gomeldon, near Salisbury. Wiltshire Archaeological and Natural History Magazine 80, 166-169.

Henderson, J. (1987) Factors determining the state of preservation of human remains. In A. Boddington, A. N. Garland and R. C. Janaway (eds.) Death, Decay and Reconstruction: approaches to archaeology and forensic science, 43-54. Manchester, Manchester University Press.

Harding, P. (1988) The chalk plaque pit, Amesbury. Proceedings of the Prehistoric Society 54, 320-327. Harding, P. (2007) Two possible Iron Age 'banjo' enclosures and a Romano-British villa and settlement at Beach's Barn, Fittleton, Salisbury Plain. Wiltshire Archaeological and Natural History Magazine 100, 8390.

Hicklin, K. (2006) The Romano-British horse burials. In A. Graham (ed.) The Excavation of Five Beaker Burials, the Iron Age and Romano-British settlements, and the 4th Century Courtyard Villa at Barton Field, Tarrant Hinton, 167-169. Dorchester, Dorset Natural History and Archaeological Society Monograph Series: 17.

Haselgrove, C. (1982) Wealth, prestige and power: the dynamics of late Iron Age political centralisation in south-east England. In A. C. Refrew and S. Shennan (eds.) Ranking, Resource and Exchange, 79-88. Cambridge, Cambridge University Press.

Higbee, L. (2006) Animal bone. In R. Young and A. Hancocks. Early Bronze Age ring ditches and RomanoBritish agriculture at Showell Farm, Chippenham, Excavations in 1999. Wiltshire Archaeological and Natural History Magazine 99, 44-45.

Haselgrove, C. (1989) The later Iron Age in southern Britain and beyond. In M. Todd (ed.) Research on Roman 201

Investigating Animal Burials; Ritual, Mundane and Beyond Higgins, J. (1999) Túnel: A case study of avian zooarchaeology and taphonomy. Journal of Archaeological Science 26, 1449-1457.

Hirst, S. M. (1985) An Anglo-Saxon Inhumation Cemetery at Sewerby, East Yorkshire. York, York University Archaeological Publications 4.

Higgs, E. S. and Jarman, M. (1975) Palaeoeconomy. In E. S. Higgs and M. Jarman (eds.) Palaeoeconomy, 1-7. Cambridge, Cambridge University Press.

Hodder, I. (1992) Theory and Practice in Archaeology. London, Routledge. Hodson, F. (1964) Cultural grouping within the British pre-Roman Iron Age. Proceedings of the Prehistoric Society 30, 99-110.

Highman, C. F. W. and Higgs, E. S. (1979) The fauna. In P. Ashbee, I. F. Smith and J. G. Evans. Excavation of three long barrows near Avebury, Wiltshire. Proceedings of the Prehistoric Society 45, 225-228.

Hope, V. M. (1999) The Iron and Roman Ages: c.600 BC to AD 400. In P. C. Jupp and C. Gitting (eds.) Death in England. An Illustrated History, 40-64. Manchester, Manchester University Press.

Hill, A. (1979a) Butchery and natural disarticulation: an investigatory technique. American Antiquity 44, 739-744.

Hughes, M. (1972) A Romano-British Sacred Well at Norden, Corfe Castle. Proceedings of the Dorset Natural History and Archaeological Society 94, 76-77.

Hill, A. (1979b) Disarticulation and scattering of mammal skeletons. Paleobiology 5, 261-274. Hill, A. and Behrensmeyer, A. K. (1985) Disarticulation patterns of some modern East African mammals. Paleobiology 10, 336-376.

Humphrey, C. and Laidlaw, J. (1994) The Archetypal Actions of Ritual. Oxford, Clarendon Press. Humphrey, J. (2003) The utilization and technology of flint in the British Iron Age. In J. Humphrey (ed.) Researching the Iron Age. Selected papers from the proceedings of the Iron Age Research Student Seminars, 1999 and 2000, 17-23. Leicester, Leicester Archaeology Monographs No.11.

Hill, J. D. (1988) Celtic Castles in the Air, University of Cambridge, Unpublished M.Phil. Hill, J. D. (1995) Ritual and Rubbish in the Iron Age of Wessex. Oxford, British Archaeological Report British Series 242. Hill, J. D. (1996) The identification of ritual deposits of animals. A general perspective from a specific study of 'special animal deposits' from the Southern English Iron Age. In S. Anderson and K. Boyle (eds.) Ritual Treatment of Human and Animal Remains. Proceedings of the First Meeting of the Osteoarchaeological Research Group held in Cambridge on 8th October 1994, 17-32. Oxford, Oxbow.

Hunter, J. R. and Ralston, I. (eds.) (1999) Archaeology of Britain. An introduction from the Upper Palaeolithic to the Industrial revolution. London, Routledge.

Hill, J. D. (2007) The dynamics of social change in later Iron Age eastern and south-eastern England c. 300BCAD 43. In C. Haselgrove and T. Moore (eds.) The Later Iron Age in Britain and Beyond, 16-40. Oxford, Oxbow.

Immonen, V. (2002) Functional ladles or ceremonial cutlery? A cultural biography of prehistoric wooden spoons from Finland. Acta Borealia 19 (1), 27-47.

Iles, M. (1992) The animal bone. In P. Cox. Excavations at the Waitrose supermarket site, Chantry Fields, Gillingham, Dorset 1991-2. Proceedings of the Dorset Natural History and Archaeological Society 114, 134.

Ingold, T. (1994) What is an Animal? London, Routledge.

Hills, C. (1999) Did the people of Spong Hill come from Schleswig-Holstein? Studien zur Sachsenforschung 11, 145-154.

Ingrem, C. (2000) The animal bone. In J. Valentin and S. Robinson (eds.) Archaeological excavations and recording of land between 28 and 30 Bell Street, Shaftesbury, Dorset. Proceedings of the Dorset Natural History and Archaeological Society 122, 108.

Hines, J. (1992) The seriation and chronology of Anglian English women's graves: a critical assessment. In L. Jørgensen (ed.) Chronological Studies of Anglo-Saxon England, Lombard Italy and Vendel-Period Sweden, 8193. Copenhagen, Institute of Prehistoric and Classical Archaeology, University of Copenhagen.

Ingrem, C. (2001) The animal bone. In J. Valentin and S. Robinson (eds.) A medieval site in Gillingham, Dorset: Further excavations at Chantry Fields 1999. Proceedings of the Dorset Natural History and Archaeological Society 123, 43-44.

Hingley, R. (1996) The 'legacy' of Rome: the rise, decline and fall of the theory of Romanization. In J. Webster and N. Cooper (eds.) Roman Imperialism: Post-Colonial Perspectives, 35-48. Leicester, University of Leicester Archaeological Monograph No. 3.

Ingrem, C. (2002) The animal bone. In J. Valentin and S. Robinson. Excavations in 1999 on land adjacent to 202

Bibliography Wayside Farm, Nursteed Road, Devizes. Wiltshire Archaeological and Natural History Magazine 95, 195202.

Jewell, P. A. (1960) Appendix II. Animal remains. In P. Wenham. Seven archaeological discoveries in Yorkshire. Yorkshire Archaeological Journal 40, 326-327.

Ingrem, C. (2006) The animal bone. In M. Fulford, A. Clarke and H. Eckardt (eds.) Life and Labour in Late Roman Silchester. Excavations in Insula IX since 1997, 167-188. London, Britannia Monograph Series No. 22.

Jones, A. and Richards, C. (2003) Animals into ancestors: domestication, food and identity in late Neolithic Orkney. In M. Parker Pearson (ed.) Food, Culture and Identity in the Neolithic and Early Bronze Age, 45-51. Oxford, British Archaeological Reports International Series 1117.

Ingrem, C. (2007) Animal Bone. In C. Malim and A. Martin (eds.) A Romano-British roadside settlement on Chapperton Down, Salisbury Plain Training Area. Wiltshire Archaeological and Natural History Magazine 100, 123-125.

Jones, D. (1993) The animal bones. In M. Russell (ed.) Excavations at New Park Street, Devizes. Wiltshire Archaeological and Natural History Magazine 86, 100. Jones, M. U. (1971) Aldborough, West Riding, 1964: Excavations at the south gate and bastion and at extramural sites. Yorkshire Archaeological Journal 43, 39-78.

Insoll, T. (2004) Archaeology, Ritual, Religion. London, Routledge.

Jones, P. R. (1980) Experimental butchery with modern stone tools and its relevance for palaeolithic archaeology. World Archaeology 12 (2), 153-165.

Jackson, D. A. (1975) An Iron Age site at Twywell, Northants. Northamptonshire Archaeology 10, 31-93. Jackson, J. W. (1925) Report on the animal remains. In R. Clay. An inhabited site of La Tene 1 date on Swallowcliffe Down. Wiltshire Archaeological and Natural History Magazine 43, 90-93.

Jones, R. T. (1974) Brownsea Island animal bones. English Heritage, Ancient Monuments Laboratory Report 1698. Jones, R. T. (1975a) Long Bredy, Dorset. Animal bones. English Heritage, Ancient Monuments Laboratory Report 1925.

Jackson, J. W. (1943) Animal bones. In R. E. M. Wheeler (ed.) Maiden Castle, Dorset, 360-371. London, Report of the Research Committee Society of Antiquaries of London. No. XII.

Jones, R. T. (1975b) Poxwell, Dorset: Animal bones. English Heritage, Ancient Monuments Laboratory Report 1948.

Jackson, J. W. (1948a) Excavations at little Woodbury, Wiltshire (1938-39), Part III: the animal remains. Proceedings of the Prehistoric Society 14, 19-23.

Jones, R. T. (1976a) Avebury animal bones. English Heritage, Ancient Monuments Laboratory Report 2071.

Jackson, J. W. (1948b) Part III: the animal remains. In J. Brailsford and J. W. Jackson. Excavations at Little Woodbury, Wiltshire (1938-39). Proceedings of the Prehistoric Society 14, 19-23.

Jones, R. T. (1976b) Winklebury Animal Bone. English Heritage, Ancient Monuments Laboratory Report 2173.

Jackson, J. W. (1971) Appendix II animal remains. In E. S. Wood. The excavation of a Bronze age Barrow, Green Howe, North Deighton, Yorkshire. Yorkshire Archaeological Journal 43, 30.

Jones, R. T. (1977) Animal bones. In K. Smith. The excavation of Winklebury Camp, Basingstoke, Hampshire. Proceedings of the Prehistoric Society 43, 58-69.

Jackson, W. J. (1948) Report on animal bones. In J. F. S. Stone and W. E. V. Yong. Two pits of grooved ware date near Woodhenge. Wiltshire Archaeological and Natural History Magazine LII, 300-301.

Jones, R. T. (1978) Animal bone. In G. Eogan. The excavation of a round barrow at Long Bredy, Dorset. Proceedings of the Dorset Natural History and Archaeological Society 100, 52.

Jarman, M., Fagg, A. and Higgs, E. S. (1968) Appendix A: Animal remains. In I. M. Stead (ed.) An Iron Age HillFort at Grimthorpe, Yorkshire, England. Proceedings of the Prehistoric Society 34, 182-189.

Jones, R. T. (1979/1980) The animal bone. In P. Ashbee (ed.) Amesbury Barrow 39: Excavations 1960. Wiltshire Archaeological and Natural History Magazine 74/75, 28. Jones, R. T. (1986) The animal bone. In J. D. Hurst and J. S. Wacher (eds.) A multi-period site at Poxwell, Dorset. Proceedings of the Dorset Natural History and Archaeological Society 108, 79.

Jenkins, F. (1957) The role of dog in Romano-Gaulish Religion. Collection Latomus 16, 60-78.

203

Investigating Animal Burials; Ritual, Mundane and Beyond King, J. E. (1962) Report on animal bones. In P. Rahtz. Excavations at Shearplace Hill, Sydling St. Nicholas Dorset, England. Proceedings of the Prehistoric Society 28, 325.

Jope, M. (1965) Faunal remains: frequencies and ages of species. In I. F. Smith (ed.) Windmill Hill and Avebury, 142-45. Oxford, Clarendon Press. Joy, J. (2009) Reinvigorating object biography: reproducing the drama of object lives. World Archaeology 41 (4), 540-556.

King, J. E. (1963a) Report on the animal bones. In T. C. M. Brewster (ed.) The Excavation of Staple Howe, Malton, The East Riding Archaeological Research Committee.

Keith, K. (1997) Animal bone. In M. Adams. Excavation of a pre-conquest cemetery at Addingham, West Yorkshire. Medieval Archaeology 41, 180.

King, J. E. (1963b) Report on the animal bones. In T. C. M. Brewster (ed.) The Excavation of Staple Howe, 136137. Malton, The East Riding Archaeological Research Committee.

King, A. (1991) Food Production and Consumption Meat. In R. Jones (ed.) Britain in the Roman Period: Recent Trends, 15-20. Sheffield, University of Sheffield.

King, J. E. (1965) Bones. Appendix V. In N. H. Field. Romano-British settlement at Studland, Dorset. Proceedings of the Dorset Natural History and Archaeological Society 87, 190-191.

King, A., Jackson, J. W. and Bramwell, D. (1988) Animal bones and shells. In T. B. James and A. M. Robinson (eds.) Clarendon Palace. The History and Archaeology of a Medieval Palace and Hunting Lodge near Salisbury, Wiltshire, 260-265. London, Reports of the Research Committee of the Society of Antiquaries of London No. XLV.

King, J. E. (1970) Winterslow the Animal Bones, Appendix V1. English Heritage, Ancient Monuments Laboratory Report 1535. Kinnes, I. (1979) Round Barrows and Ring-ditches in the British Neolithic. London, British Museum Occasional Papers 7.

King, A. C. (1978) A comparative survey of bone assemblages from Roman sites in Britain. Bulletin of the Institute of Archaeology University of London 15, 207232.

Kinnes, I. (1992) Non-Megalithic Long Barrows and Allied Structures in the British Neolithic. London, British Museum Occasional Paper 52.

King, A. C. (1984) Animal Bones and the Dietary Identity of Military and Civilian Groups in Roman Britain, Germany and Gaul. In T. F. C. Blagg and A. C. King (eds.) Military and Civilian in Roman Britain, 187217. Oxford, British Archaeological Report British series 136.

Knight, S. (2001) Beasts and burial in the interpretation of ritual space: a case study from Danebury. In A. T. Smith and A. Brookes (eds.) Holy Ground: Theoretical Issues Relating to the Landscape and Material Culture of Ritual Space, 49-59. Oxford, British Archaeological Report International Series 956.

King, A. C. (1999a) Animal bone report. In P. Halkon and M. Millett (eds.) Rural Settlement and Industry: Studies in the Iron Age and Roman Archaeology of Lowland East Yorkshire, 183-185. York, Yorkshire Archaeological Report No 4.

Knight, S. (2003) Butchery and Intra-Site Spatial Analysis of Animal Bone: A Case Study from Danebury Hillfort, Hampshire, University of Leicester, PhD.

King, A. C. (1999b) Diet in the Roman world: a regional inter-site comparison of the mammal bones. Journal of Roman Archaeology 12, 220-241.

Kooyman, B. (2006) Boundary theory as a means to understanding social space in archaeological sites. Journal of Anthropological Archaeology 25 (4), 424-435.

King, A. C. (2005) Animal remains from temples in Roman Britain. Britannia 36, 329-369.

Kopytoff, I. (1986) The cultural biography of things: commoditization as process. In A. Appadurai (ed.) The Social Life of Things: Commodities in Cultural Perspective, 64-91. Cambridge, Cambridge University Press.

King, J. E. (1951) Animal bones. In K. M. Richardson. The excavation of Iron Age villages on Boscome Down West. Wiltshire Archaeological and Natural History Magazine 54, 165-166.

Kreinath, J., Snoek, J. and Stausberg, M. (2006) Theorizing Rituals: Issues, Topics, Approaches, Concepts. Leiden, Brill.

King, J. E. (1953) The animal remains. In A. L. Parke. The excavation of a Bell-Barrow, Oakley Down, Wimborne St. Giles. Proceedings of the Dorset Natural History and Archaeological Society 75, 44.

Latour, B. (2004) A prologue in form of a dialogue between a student and his (somewhat) socratic professor. In C. Avgerou, C. Ciborra and F. F. Land (eds.) The 204

Bibliography Locker, A. (1977a) Animal bone; Kings Somborne, Nr Winchester. English Heritage, Ancient Monuments Laboratory Report 2364.

Social Study of Information and Communication Study, 62-76. Oxford, Oxford University Press. Lauwerier, R. C. G. M. (1993) Bird remains in Roman graves. Archaeofauna 2, 75-82.

Locker, A. (1977b) Little Somborne Animal Bone. English Heritage, Ancient Monuments Laboratory Report 2365.

Lawson, A. J. (2000) Potterne 1982-5: Animal Husbandry in Later Prehistoric Wiltshire. Salisbury, Wessex Archaeology Report No. 17. Lee, C. (2007) Feasting the Dead. Food and Drink in Anglo-Saxon Burial Rituals. Woodbridge, The Boydell Press. Anglo-Saxon Studies 9.

Locker, A. (1980a) Animal bone. In D. S. Neal. Bronze Age, Iron Age and Roman settlement sites at Little Somborne and Ashley, Hampshire. Proceedings of the Hampshire Field Club and Archaeological Society 36, 122-124.

Legge, A. J. (1991a) The animal bones. In I. M. Steed (ed.) Iron Age Cemeteries in East Yorkshire, 140-147. London, English Heritage.

Locker, A. (1980b) Animal bones. In. D. S. Neal (Ed.). Bronze Age, Iron Age and Roman settlement sites at Little Somborne and Ashley, Hampshire.

Legge, A. J. (1991b) The animal remains from six sites at Down Farm, Woodcutts. In J. Barrett, R. Bradley and M. Hall (eds.) Papers on the Prehistoric Archaeology of Cranborne Chase, 54-100. Oxford, Oxbow Monograph 11.

Locker, A. (1992a) Animal bone. In M. Papworth (ed.) Excavation and survey of Bronze Age sites in the Badbury area, Kingston Lacy Estate. Proceedings of the Dorset Natural History and Archaeological Society 114, 56-57.

Lemoine, G. (2001) Skeletal technology in context; an optimistic overview. In A. M. Choyke and L. Bartosiewicz (eds.) Crafting Bone: Skeletal technologies through time and space, Oxford, British Archaeological Report International Series 937.

Locker, A. (1992b) Animal bone. In S. Woodiwiss (ed.) Iron Age and Roman Salt Production and the Medieval Town of Droitwich, 122-125. London, CBA Research Report 81. Locker, A. (2000) Animal bone. In A. J. Lawson (ed.) Potterne 1982-5: Animal Husbandry in Later Prehistoric Wiltshire, 101-119. Salisbury, Wessex Archaeology Report No. 17.

Lévi-Strauss, C. (1962) Le Totémism aujourd'hui. Paris, Presses Universitaires de France. Lévi-Strauss, C. (1964) Les Mythologiques: Le Cru et la Cuit. Paris, Plon.

Locker, A. M. (1998) Faunal remains. In M. Papworth (ed.) The Medieval manorial building of Kingston Lacy: Survey and excavation results with an analysis of the Medieval account rolls 1295-1462. Proceedings of the Dorset Natural History and Archaeological Society 120, 59-60.

Levitan, B. (1989) The vertebrate remains from Chichester Cattlemarket. In A. Down (ed.) Chichester Excavations 6, 242–276. Chichester, Phillimore. Levitan, B. (1993) Vertebrate remains. In A. Woodward and P. Leach (eds.) The Uley Shrines: Excavation of a Ritual Complex on West Hill, Uley, Gloucestershire: 1977-9, 257-263. London, British Museum Press.

Lovell, J., Hamilton-Dyer, S., Loader, E. and McKinley, J. I. (1999) Futher investiagations of an Iron Age and Romano-British Farmstead on Cockey Down, near Salisbury. Wiltshire Archaeological and Natural History Magazine 92, 33-38.

Levy, J. (1982) Social and Religious Organisation in Bronze Age Denmark: an Analysis of Ritual Hoard Finds. Oxford, British Archaeological Report International Series 124.

Loveluck, C. (1999) Archaeological expressions of the transition from the late Roman to early Anglo-Saxon period in lowland East Yorkshire. In P. Halkon and M. Millett (eds.) Rural Settlement and Industry:Studies in the Iron Age and Roman Archaeology of Lowland East Yorkshire, 228-36. York, Yorkshire Archaeological Report No. 4.

Lillios, K. T. (1999) Objects of memory: The ethnography and archaeology of heirlooms. Journal of Archaeological Method and Theory 6 (3), 235-262. Loader, E. (2000) Other Finds. Animal bones. In D. Godden, J. Grove and R. J. C. Smith (eds.) Medieval and post-medieval Bridport: Excavations at 43 South Street, 1996. Proceedings of the Dorset Natural History and Archaeological Society 122, 120.

Loveluck, C. (2003) The archaeology of Post-Roman Yorkshire, AD 400 to 700: Overview and future direction for research. In T. G. Manby, S. Moorhouse and P. Ottaway (eds.) The Archaeology of Yorkshire. An Assessment at the Beginning of the 21st Century, 151-

205

Investigating Animal Burials; Ritual, Mundane and Beyond Maltby, M. (1979b) Summary of identification of animal bone from MARC 3 R1, Stratton Park. English Heritage, Ancient Monuments Laboratory Report 2917.

170. Leeds, Yorkshire Archaeological Society Occasional Paper No. 3. Lubbock, J. (1865) Prehistoric Times as Illustrated by Ancient Remains and the Manners and Customs of Modern Savages. London, Williams and Norgate.

Maltby, M. (1981a) The animal bones. In S. Davies. The excavations at Old Down Farm, Andover. Part 2 Prehistoric and Roman. Proceedings of the Hampshire Field Club 37, 147-153.

Lucy, S. (2000) The Anglo-Saxon Way of Death. London, Sutton Publishing.

Maltby, M. (1981b) Iron Age, Romano-British and Anglo-Saxon animal husbandry: a review of the faunal evidence. In M. Jones and G. Dimbleby (eds.) The Environment of Man; the Iron Age to the Anglo-Saxon Period, 155-204. Oxford, British Archaeological Report British Series 87.

Luff, R., M (1993) Animal Bones from Excavations in Colchester, 1971–85. Colchester, Colchester Archaeological Trust Report 12. Lyman, R. L. (1987) Archaeofaunas and butchery studies: a taphonomic perspective. In M. B. Schiffer (ed.) Advances in Archaeological Method and Theory, Volume 10, San Diego, Academic Press.

Maltby, M. (1982a) Animal and bird bones. In R. A. Higham (ed.) Excavations at Oakhampton Castle, Devon. Part 2 The Bailey. Devon Archaeological Society 40, 114-135.

Lyman, R. L. (1994) Vertebrate Taphonomy. Cambridge, Cambridge University Press.

Maltby, M. (1982b) The animal bones. In M. Millett and D. Russell. An Iron Age burial from Viables Farm, Basingstoke. Archaeological Journal 139, 75-81.

Lyman, R. L. and Fox, G. L. (1989) A critical evaluation of bone weathering as an indication of bone assemblage formation. Journal of Archaeological Science 16, 293317.

Maltby, M. (1982c) The animal bones from Cowdery's Down, Basingstoke, Hampshire. English Heritage, Ancient Monuments Laboratory Report 3875.

Lyman, R. L. and O'Brien, M., J (1987) Plow-zone zooarchaeology: Fragmentation and Identifiability. Journal of Field Archaeology 14 (4), 493-500.

Maltby, M. (1982d) The animal bones from the 74, 75 and 78 excavations at Silchester. English Heritage, Ancient Monuments Laboratory Report 3595.

Macdonald, J. D. (1958) The bird bones. In A. Cotton and P. W. Gathercole (eds.) Excavations at Clausentum, Southampton 1951-1954, 143. London, HMSO.

Maltby, M. (1983a) The animal bone. In M. Millett and S. James. Excavations at Cowdery's Down Basingstoke, Hampshire, 1978-81. The Archaeological Journal 140, 187-192, 258-259.

MacDonald, K. C. (1991) Archaeozoology as anthropology. Archaeology Review from Cambridge 10 (1), 60-69.

Maltby, M. (1983b) The animal bone. In M. Millett and S. James. Excavations at Cowdery's Down Basingstoke, Hampshire, 1978-81. The Archaeological Journal 140, 167.

Mainland, I. L. (2006) The mammal and bird bone. In M. Millett (ed.) Shiptonthorpe, East Yorkshire: Archaeological Studies of a Romano-British Roadside Settlement, 258-279. Leeds, Yorkshire Archaeological Society, Roman Antiquities Section and East Riding Archaeological Society. Yorkshire Archaeological Report No. 5.

Maltby, M. (1984a) The animal bones. In M. Fulford (ed.) Silchester Defences 1974-80, 199-212. London, Britannia Monograph Series No. 5. Maltby, M. (1984b) Animal bones and the RomanoBritish economy. In C. Grigson and J. Clutton-Brock (eds.) Animals and Archaeology volume 4: Husbandry in Europe, 125-138. Oxford, British Archaeology Reports 227.

Maltby, M. (1978a) Animal bones from the Iron Age settlement at Chilbolton Down, Hants. English Heritage, Ancient Monuments Laboratory Report 2645. Maltby, M. (1978b) Animal Bones from the Roman Site of Little Somborne, Hants. English Heritage, Ancient Monuments Laboratory Report 2644.

Maltby, M. (1985a) The animal bones. In P. Fasham (ed.) The Prehistoric Settlement at Winnall Down, Winchester, 97-125. Winchester, Hampshire Field Club Monograph 2.

Maltby, M. (1979a) Animal bones from Braishfield. English Heritage, Ancient Monuments Laboratory Report 2916.

Maltby, M. (1985b) Animal bones from four late Bronze Age sites on the Marlborough Downs, Wiltshire. English Heritage, Ancient Monuments Laboratory Report 4927. 206

Bibliography XVI. English Heritage, Ancient Monuments Laboratory Report 125/87.

Maltby, M. (1985c) The Animal Bones from the 1973 Excavations at Balksbury, Hampshire. English Heritage, Ancient Monuments Laboratory Report 4542.

Maltby, M. (1988) The animal bones from a RomanoBritish well from the 1965-66 excavations at Oakridge II, Basingstoke, Hampshire. English Heritage, Ancient Monuments Laboratory Report 64/1988.

Maltby, M. (1985d) Animal bones from the Neolithic and Bronze Age deposits from the excavations at Rowden 1981 (W4) and Cowleaze 1982 (W27). English Heritage, Ancient Monuments Laboratory Report 4928.

Maltby, M. (1989a) Animal bone. In S. Howard. A double ring ditched, Bronze Age barrow at Barford Farm, Pamphill. Proceedings of the Dorset Natural History and Archaeological Society 111, 46-49.

Maltby, M. (1985e) Assessing variations in Iron Age and Roman butchery practices: the need for quantification. In N. R. Fieller, D. D. Gilbertson and N. G. Ralph (eds.) Palaeobiological Investigation: research design, methods and data analysis, 19-31. Oxford, British Archaeological Report International Series 266.

Maltby, M. (1989b) The animal bones. In P. J. Fasham, D. E. Farwell and R. J. B. Whinney (eds.) The Archaeological Sites at Easton Lane, Winchester, 122130. Winchester, Hampshire Field Club and Archaeological Society Monograph No 6.

Maltby, M. (1985f) Patterns in faunal assemblage variability. In G. Barker and C. Gamble (eds.) Beyond domestication in prehistoric Europe, 33-74. London, Academic press.

Maltby, M. (1989c) The animal bones. In P. Addison (ed.) Excavation of Neolithic and Bronze Age Pits and a Section of a Roman Road on a Pipeline near Lodge Farm, Pamphill, Dorset. Proceedings of the Dorset Natural History and Archaeological Society 111, 28??

Maltby, M. (1986a) The animal bones from 1978 excavations of the Late Iron Age and Early RomanoBritish settlement at Abbotstone Down, near New Arlesford, Hampshire. English Heritage, Ancient Monuments Laboratory Report 58/86.

Maltby, M. (1989d) Urban-rural variations in the butchery of cattle in Romano-British Hampshire. In D. Serjeantson and T. Waldron (eds.) Diet and Crafts in Towns, 75-106. Oxford, British Archaeological Report British Series 199.

Maltby, M. (1986b) Animal bones from the Iron Age and Romano-British phases of the Staple Gardens excavations, Winchester, Hampshire. English Heritage, Ancient Monuments Laboratory Report 4908.

Maltby, M. (1990a) Animal bones. In J. Richards (ed.) The Stonehenge Environs Project, 105. London, English Heritage Archaeology Report 16.

Maltby, M. (1986c) Identification of Animal Bones from the Vatcher Excavations of an Earth-Work near Bradbury Rings, Dorset. English Heritage, Ancient Monuments Laboratory Report 4907.

Maltby, M. (1990b) Animal bones. In J. Richards (ed.) The Stonehenge Environs Project, 207-208. London, English Heritage Archaeology Report 16.

Maltby, M. (1986d) Identification of Animal Bones from the Vatcher Excavations of Round Barrows at Winterbourne Stoke and Woodford Road, Wiltshire. English Heritage, Ancient Monuments Laboratory Report 4906.

Maltby, M. (1990c) Animal bones. In J. Richards (ed.) The Stonehenge Environs Project, 57-61. London, English Heritage Archaeology Report 16.

Maltby, M. (1987a) The Animal Bones from Brighton Hill South (Trenches B, C and K), Farleigh Wallop, Hampshire. English Heritage, Ancient Monuments Laboratory Report 155/87.

Maltby, M. (1990d) Animal bones. In J. Richards (ed.) The Stonehenge Environs Project, 121-123. London, English Heritage Archaeology Report 16. Maltby, M. (1990e) Animal bones. In J. Richards (ed.) The Stonehenge Environs Project, 150-154. London, English Heritage Archaeology Report 16.

Maltby, M. (1987b) Animal Bones from the 1981 Excavations at Balksbury, Hampshire. English Heritage, Ancient Monuments Laboratory Report 226/87.

Maltby, M. (1990f) The Animal Bones from the RomanoBritish Deposits at the Greyhound Yard and Methodist Chapel Sites in Dorchester, Dorset. English Heritage, Ancient Monuments Laboratory Report 9/90.

Maltby, M. (1987c) The Animal Bones from the Excavations at Owslebury, Hants. An Iron Age and Early Romano-British settlement. Ancient Monuments Laboratory Report 6/1987.

Maltby, M. (1991a) The animal bone. In P. J. Woodward (ed.) The South Dorset Ridgeway. Survey and Excavations 1977-84, 105-106. Dorchester, Dorset

Maltby, M. (1987d) The Animal Remains from the Later Roman Phases from Winchester Northern Suburbs: 1: The Unsieved Samples from Victoria Road Trenches X -

207

Investigating Animal Burials; Ritual, Mundane and Beyond Natural History and Archaeological Society. Monograph Series No 8.

Dorset, 1984-87, 111-116. Dorchester, Dorset Natural History and Archaeological Society Monograph 15.

Maltby, M. (1991b) The animal bone. In P. S. Bellamy (ed.) The excavation of Fordington Farm Round Barrow. Proceedings of the Dorset Natual History and Archaeological Society 133, 121-122.

Maltby, M. (2002b) Animal bones from prehistoric features. In S. M. Davis, P. S. Bellamy, M. J. Heaton and P. J. Woodward (eds.) Excavations at Arlington Avenue, Fordington, Dorchester, Dorset, 1984-87, 53-55. Dorchester, Dorset Natural History and Archaeological Society Monograph 15.

Maltby, M. (1992) The animal bone. In C. Gingell (ed.) The Marlborough Downs: a Late Bronze Age Landscape and its Origins, 137-142. Salisbury, Wiltshire Archaeological and Natural History Society Monograph 1.

Maltby, M. (2004a) The animal bones. In M. Rawlings, M. J. Allen and F. Healy. Investigation of the Whitesheet Down environs 1989-90: Neolithic causeway enclosure and Iron Age settlement. Wiltshire Archaeological and Natural History Magazine 97, 167-170.

Maltby, M. (1993) Animal bones. In P. J. Woodward, S. M. Davies and A. H. Graham (eds.) Excavations at the Old Methodist Chapel and Greyhound Yard, Dorchester, 1981-1984, 314-340. Dorchester, Dorset Natural History and Archaeological Society Monograph Series 12.

Maltby, M. (2004b) Animal bones Staple Gardens. In K. E. Qualmann, H. Rees, G. D. Scobie and R. Whinney (eds.) Oram's Arbour. The Iron Age enclosure at Winchester. Volume 1:Investigations 1950-1999, 69-73. Winchester, Winchester Museums Service.

Maltby, M. (1994a) The animal bone from a RomanoBritish well at Oakridge II, Basingstoke, Hampshire. Proceedings of the Hampshire Field Club and Archaeological Society 49, 47-76.

Maltby, M. (2007) Chop and change: specialist cattle carcass processing in Roman Britain. In B. Croxford, N. Ray, R. Roth and N. White (eds.) TRAC 2006: Proceedings of the 16th Annual Theoretical Archaeological Conference, Cambridge 2006, 59-76. Oxford, Oxbow.

Maltby, M. (1994b) The meat supply in Roman Dorchester and Winchester. In A. R. Hall and H. K. Kenward (eds.) Urban-rural Connexions: Perspectives from Environmental Archaeology, 85-102. Oxford, Oxbow Monograph 47.

Maltby, M. (2010) Feeding a Roman Town. Environmental evidence from excavations in Winchester 1972-1985. Winchester, Winchester Museums and English Hertiage.

Maltby, M. (1995a) Animal bone. In G. J. Wainwright and S. M. Davies (eds.) Balksbury Camp, Hampshire: Excavations 1973 and 1981, 83-87. London, English Heritage Archaeological Report 4.

Maltby, M. and Foot, R. (1985) The animal bone. In J. Rogers and J. S. F. Walker. A detached Romano-British Bath-House at Braishfield, near Romsey, Hampshire. Proceedings of the Hampshire Field Club and Archaeological Society 41, 76-77.

Maltby, M. (1995b) Animal bone. In P. J. Fasham, G. Keevill and D. Coe (eds.) Brighton Hill South (Hatch Warren): an Iron Age Farmstead and Deserted Medieval Village in Hampshire, 49-56. Salisbury, Trust for Wessex Archaeology.

Maltby, M. and Richards, J. (1990) Animal bones and worked bone. In J. Richards (ed.) The Stonehenge environs project, 83-88. London, English Heritage Archaeology report No. 16.

Maltby, M. (1996) The exploitation of animals in the Iron Age: the archaeozoological evidence. In T. C. Champion and J. Collis (eds.) The Iron Age in Britain and Ireland: Recent trends, 17-28. Sheffied, J. R. Collis Publications.

Manby, T. G. (1975) Neolithic Occupation sites on the Yorkshire Wolds. Yorkshire Archaeological Journal 47, 23-60.

Maltby, M. (1997) Domestic Fowl on Romano-British Sites: Inter-site Comparisons of Abundance. International Journal of Osteoarchaeology 7, 402-414.

Manby, T. G. (1980) Excavation of barrows at Grindale and Boynton, East Yorkshire, 1972. Yorkshire Archaeological Journal 52, 19-48.

Maltby, M. (2001) Animal bones. In C. J. Ellis and M. Rawlings (eds.) Excavations at Balksbury Camp, Andover 1995-97. Proceedings of the Hampshire Field Club and Archaeological Society 56, 76-80.

Manby, T. G., King, A. and Vyner, B. (2003a) The Neolithic and Bronze Ages: a time of early agriculture. In T. G. Manby, S. Moorhouse and P. Ottaway (eds.) The Archaeology of Yorkshire. An Assessment at the Beginning of the 21st Century, 35-113. Leeds, Yorkshire Archaeological Society Occasional Paper No. 3.

Maltby, M. (2002a) Animal bones. In S. M. Davis, P. S. Bellamy, M. J. Heaton and P. J. Woodward (eds.) Excavations at Alington Avenue, Fordington, Dorchester,

208

Bibliography Millett, M. (ed.) (2006) Shiptonthorpe, East Yorkshire: Archaeological Studies of a Romano-British Roadside Settlement. Leeds, Yorkshire Archaeological Society, Roman Antiquities Section and East Riding Archaeological Society. Yorkshire Archaeological Report No. 5.

Manby, T. G., Moorhouse, S. and Ottaway, P. (eds.) (2003b) The Archaeology of Yorkshire. An Assessment at the Beginning of the 21st Century. Leeds, Yorkshire Archaeological Society Occasional Paper No. 3. Marciniak, A. (2005) Placing Animals in the Neolithic. London, UCL Press.

Millett, M. and McGrail, S. (1987) The archaeology of the Hasholme logboat. The Archaeological Journal 144, 69-155.

Marean, C. W. and Spencer, L. M. (1991) Impact of carnivore ravaging on zooarchaeological measures of element abundance. American Antiquity 56 (4), 645-658.

Milner, N. and Fuller, D. (1999) Contending with animal bones. Archaeology Review from Cambridge 16 (1), 1-12.

Martin, A. (2005) Agents in inter-action: Bruno Latour and agency. Journal of Archaeological Method and Theory 12 (4), 283-311.

Moody, M. S. and Moody, R. (1997) Animal bone. In P. R. Saunders. The excavation of an Iron Age settlement site at Stockton. Wiltshire Archaeological and Natural History Magazine 90, 22-23.

Mays, S. A. (1992) Human Bones. In M. Oliver. Excavation of an Iron Age and Romano-British settlement site at Oakridge, Basingstoke, Hampshire. Proceedings of the Hampshire Field Club and Archaeological Society 48, 87-88.

Moorhouse, S. (2003) Medieval Yorkshire: a Rural Landscape for the Future. In T. G. Manby, S. Moorhouse and P. Ottaway (eds.) The Archaeology of Yorkshire. An Assessment at the Beginning of the 21st Century, 181214. Leeds, Yorkshire Archaeological Society Occasional Paper No. 3.

Meddens, B. (1990a) Animal Bones from Bainesse Farm, a Roman Roadside Settlement near Catterick (Yorkshire), Excavated in 1980 and 1981. English Heritage, Ancient Monuments Laboratory Report 98/90.

Morris, B. (2000) Animals and Ancestors: Ethnography. London, Berg Publishers.

Meddens, B. (1990b) Animal Bones from Catterick Bridge (CEU 240) a Roman Town (North Yorkshire) Excavated in 1983. English Heritage, Ancient Monuments Laboratory Report 31/90.

An

Morris, J. (2000) A re-examination of the domestic mammal remains from the 1892-1907 Glastonbury lake Village excavation, Bournemouth University, Undergraduate Dissertation.

Meddens, B. (2002a) Animal bones from Bainesse (site 46). In P. R. Wilson (ed.) Cataractonium: Roman Catterick and its Hinterland. Excavations and Research, 1958-1997, Part II, 419-425. London, Council for British Archaeology Research Report 129.

Morris, J. (2008) Associated bone groups; one archaeologist's rubbish is another's ritual deposition. In O. Davis, N. Sharples and K. Waddington (eds.) Changing Perspectives on the First Millenium BC, 83-98. Oxford, Oxbow.

Meddens, B. (2002b) Animal bones from Catterick Bridge (site 240). In P. R. Wilson (ed.) Cataractonium: Roman Catterick and its Hinterland. Excavations and Research, 1958-1997, Part II, 425-431. London, Council for British Archaeology Research Report 129.

Morris, J. (2010a) Associated bone groups: beyond the Iron Age. In J. Morris and M. Maltby (eds.) Integrating Social and Environmental Archaeologies; Reconsidering Deposition, 12-23. Oxford, British Archaeological Reports, British Series.

Méniel, P. (1989) Preambule. In P. Meniel (ed.) Animal et Pratiques Religieuses: Les Manifestations Materielles. Anthropozoologica Numero Special, 7.

Morris, J. (2010b) The composition and interpretation of associated bone groups from Wessex. In D. Campana, A. Choyke, P. Crabtree, S. Defrance and J. Lev-Tov (eds.) Anthropological Approaches to Zooarchaeology: Colonialism, Complexity and Animal Transformations, 261-271. Oxford, Oxbow.

Méniel, P. (1992) Les Sacrifices d'Animaux chez les Gaulois. Paris, Editions Errance. Metcalf, P. and Huntington, R. (1991) Celebrations of Death. Cambridge, Cambridge University Press. Micozzi, M. S. (1991) Post-mortem Change in Human and Animal Remains: a systematic approach. Springfield, Charles C. Thomas.

Morris, J. (in press-a) Animal 'ritual' killing; from remains to meanings. In A. Pluskowski (ed.) Animal Ritual Killing and Burial: European Perspectives Oxford, Oxbow.

Millett, M. (1990) The Romanization of Britain. Cambridge, Cambridge University Press.

Morris, J. (in press-b) The problem with ritual. In J. Morris and C. Randall (eds.) Ritual in Context:

209

Investigating Animal Burials; Ritual, Mundane and Beyond Explaining Ritual Complexity in Archaeology, Oxford, Oxbow.

Winterbourne Monkton, North Wiltshire. Wiltshire Archaeological and Natural History Magazine 87, 24-36.

Morris, J. and Jervis, B. (2011) Whats's so special? A reinterpretaion of Anglo-Saxon 'special deposits'. Medieval Archaeology 55

O'Connor, T. P. (1983a) Report on bone spot-findings from 16-22 Coppegate, York. English Heritage, Ancient Monuments Laboratory Report 4100.

Mountenay, G. (1981) Faunal attrition and subsistence reconstruction at Thwing. In G. Barker (ed.) Prehistoric Communities in Northern England : essays in social and economic reconstruction, 73-86. Sheffield, University of Sheffield Department of Prehistory and Archaeology.

O'Connor, T. P. (1983b) Snails and animal bones. In D. Williams (ed.) Excavations at Gargrave (1977-1981), 50. Skipton, Craven District Council. O'Connor, T. P. (1984a) Bones from Aldwark, York. English Heritage, Ancient Monuments Laboratory Report 4391.

Mounteney, G. A. and Cumberpatch, C. G. (1996) The animal bones. In J. A. Dunkley, C. G. Cumberpatch, I. D. Latham and R. Thorpe (eds.) Excavations at 16-20 Church Street Bawtry, South Yorkshire, 180-183. Oxford, British Archaeological Report British Series 248.

O'Connor, T. P. (1984b) Selected Groups of Bones from Skeldergate and Walmgate. The Archaeology of York 15/1. York, Council for British Archaeology.

Mulkeen, S. and O'Connor, T. (1997) Raptors in towns: Towards an ecological model. International Journal of Osteoarchaeology 7, 440-449.

O'Connor, T. P. (1985) Bones from pre-Viking levels at 16-22 Coppergate, York. Ancient Monuments Laboratory Report 4560.

Munro, N. D. and Bar-Oz, G. (2004) Gazelle bone fat processing in the Levantine Epipalaeolithic. Journal of Archaeological Science 32 (2), 223-239.

O'Connor, T. P. (1987) Bones from the Roman and Medieval deposits at the city garage, 9 Blake Street, York. English Heritage, Ancient Monuments Laboratory Report 196/97.

Musty, J. and Rahtz, P. A. (1964) The suburbs of Old Sarum. Wiltshire Archaeological and Natural History Magazine 59, 130-154.

O'Connor, T. P. (1988) Bones from the General Accident Site, Tanner Row. The Archaeology of York 15/5. York, Council for British Archaeology.

Nadel, D., Weiss, E., Simchoni, O., Tsatskin, A., Danin, A., Kislev, M. and Flannery, K. V. (2004) Stone Age Hut in Israel Yields World's Oldest Evidence of Bedding. Proceedings of the National Academy of Sciences of the United States of America 101 (17), 6821-6826.

O'Connor, T. P. (1989) Bones from Anglo-Scandinavian Levels at 16-22 Coppergate. The Archaeology of York 15/3. York, Council for British Archaeology. O'Connor, T. P. (1998) Environmental Archaeology: a matter of definition. Environmental Archaeology 2, 1-6.

Newton, P. N. (1979/1980) Animal bones. In C. J. Gingell and J. H. Gringell. Excavation of a Medieval 'Highworth Circle' at Stratton St. Margaret. Wiltshire Archaeological and Natural History Magazine 74/75, 66.

O'Connor, T. P. (2000) The archaeology of animal bones. Stroud, Sutton Publishing.

Noddle, B. (1975) The animal bones. In C. Platt and R. Coleman-Smith (eds.) Excavations in Medieval Southampton 1953-1969, 332-341. Leicester, Leicester University Press.

O'Connor, T. P. (2003) The Analysis of Urban Animal Bone Assemblages. A Handbook for Archaeologists. York, Council of British Archaeology. O'Connor, T. P. (in press) Wild or domestic? Biometric variation in the cat Felis silvestris schreber. International journal of Osteoarchaeology

Noddle, B. (1979) Animal bones from Garton Slack. English Heritage, Ancient Monuments Laboratory Report 2754.

Oestigaard, T. and Goldhahn, J. (2006) From the dead to the living: Death as transactions and re-negotiations. Norwegian Archaeological Review 39 (1), 27-48.

Noddle, B. (1993) Animal bone. In A. Whittle, A. J. Rouse and J. G. Evans (eds.) A Neolithic Downland Monument in its Environment: Excavations at the Easton Down Long Barrow, Bishops Cannings, North Wiltshire. Proceedings of the Prehistoric Society 59, 222.

Okkens, A. C., Hekerman, T. W., de Vogel, J. W. and van Haaften, B. (1993) Influence of litter size and breed on variation length of gestation in the dog. The Veterinary Quarterly 15 (4), 160-1.

Noddle, B. (1994) Animal bone. In A. Whittle. Excavations at Millbarrow Neolithic Chambered Tomb,

210

Bibliography 1974-1976. London, Council for British Archaeology Research Report 28.

Oliver, M. (1992) Excavation of an Iron Age and Romano-British settlement site at Oakridge, Basingstoke, Hampshire. Proceedings of the Hampshire Field Club and Archaeological Society 48, 55-94.

Payne, S. (1972) Partial recovery and sample bias: the results of some sieving experiments. In E. S. Higgs (ed.) Papers in Economic Prehistory, 49-64. Cambridge, Cambridge University Press.

Oswald, A., Dyer, C. and Barber, M. (2001) The Creation of Monuments: Neolithic Causewayed Enclosures in the British Isles. Swindon, English Heritage.

Payne, S. (1975) Partial recovery and sample bias. In A. T. Clason (ed.) Archaeozoological Studies, 7-17. Amsterdam, North-Holland Publishing Company.

Ottaway, P. (2003) The archaeology of the Roman period in the Yorkshire region: a rapid resource assessment. In T. G. Manby, S. Moorhouse and P. Ottaway (eds.) The Archaeology of Yorkshire. An Assessment at the Beginning of the 21st Century, 125-149. Leeds, Yorkshire Archaeological Society Occasional Paper No. 3.

Payne, S. (1990) Animal Bones from Excavations in 1972 at Catterick Site 434, North Yorkshire. English Heritage, Ancient Monuments Laboratory Report 5/90.

Outram, A. K. (2001) A new approach to identifying bone marrow and grease exploitation: Why the 'indeterminate' fragments should not be ignored. Journal of Archaeological Science 28 (4), 401-410.

Payne, S. (1997) Animal remains. In E. Hostetter and T. N. Howe (eds.) The Romano-British villa at Castle Copse, Great Bedwyn, 322-344. Indiana, Indiana University Press.

Papworth, M. (1992) Excavation and Survey of Bronze Age Sites in the Badbury area, Kingston Lacy Estate. Proceedings of the Dorset Natural History and Archaeological Society 114, 47-76.

Payne, S. (2002) Animal bones from the 1972 excavations (site 434). In P. R. Wilson (ed.) Cataractonium: Roman Catterick and its hinterland. Excavations and research, 1958-1997, Part II, 415-419. London, Council for British Archaeology Research Report 129.

Parker-Pearson, M. (1993) Bronze Age Britain. London, B.T. Batsford Ltd/English Heritage.

Peck, R. W. (1993) The animal bones. In R. N. Lucas (ed.) The Romano-British villa at Halstock, Dorset. Excavations 1967-1985, 111-114. Dorchester, Dorset Natural History and Archaeological Society Monograph Series 13.

Parker-Pearson, M. (2003) Food, culture and identity: an introduction and overview. In M. Parker Pearson (ed.) Food, Culture and Identity in the Neolithic and Early Bronze Age, 1-31. Oxford, British Archaeological Reports International Series 1117.

Peck, R. W. (2001) A Discussion on the Animal Economy and Butchery Practices at Tarrant Hilton Roman Villa. Unpublished Report

Parker-Pearson, M., Pollard, J., Richards, C., Thomas, J., Tilley, C., Welham, K. and Albarella, U. (2006) Materializing Stonehenge. The Stonehenge riverside project and new discoveries. Journal of Material Culture 11, 227-261.

Peck, R. W. and Maltby, M. (2006) A discussion of the animal economy and butchery practices. In A. Graham (ed.) The Excavation of Five Beaker Burials, the Iron Age and Romano-British settlements, and the 4th Century Courtyard Villa at Barton Field, Tarrant Hinton, 164167. Dorchester, Dorset Natural History and Archaeological Society Monograph Series: 17.

Parker-Pearson, M., Pollard, J., Richards, C., Thomas, J., Tilley, C., Welham, K., Allen, M., Bennett, W., Field, D., French, C., Linford, N., Payne, A., Robinson, D. and Ruggles, C. (2007) The Stonehenge Riverside Project Summary Interim Report on the 2006 Season. Sheffield, Sheffield University.

Petres, É., F (1972) On Celtic and Human Sacrifices. Acta Archaeologica Academiae Scientiarum Hungaricae 24, 365-383.

Parker-Pearson, M., Pollard, J., Tilley, C., Thomas, J., Richards, C. and Welham, K. (2005) The Stonehenge Riverside Project Interim Report 2005. Sheffield, Sheffield University.

Philips, J. T. (1960) An Iron Age site at Driffield, East Riding, Yorks. Yorkshire Archaeological Journal 40 (158), 183-191.

Parker, A. (1988) The Birds of Roman Britain. Oxford Journal of Archaeology 7, 197-203.

Phillips, G. (1979) The mammal and bird bones. In B. Ayers (ed.) Excavations at Chapel Lane Staith 1978, 5864. Kingston Upon Hull, East Riding Archaeologist Vol.5. Hull Old Town Report Series No.3.

Parrington, M. (1978) The Excavation of an Iron Age Settlement, Bronze Age Ring Ditches and Roman Features at Ashville Trading Estate, Abingdon, Oxon,

211

Investigating Animal Burials; Ritual, Mundane and Beyond Pollard, J. (2006) A community of beings: animals and people in the Neolithic of southern Britain. In D. Serjeantson and D. Field (eds.) Animals in the Neolithic of Britian and Europe. Neolithic Studies Group Seminar Papers 7, 135-148. Oxford, Oxbow Books.

Phillips, G. (1980) The fauna. In P. Armstrong (ed.) Excavations in Scale Lane/Lowgate 1974, 77-85. Kingston Upon Hull, East Riding Archaeologist Vol.6. Hull Old Town Report Series No. 4. Phillips, G. (1993) The Vicar Lane bone. In D. H. Evans (ed.) Excavations in Hull 1975-76, 192. Kingston Upon Hull, East Riding Archaeologist Vol.4. Hull Old Town Report Series No. 2.

Pollex, A. (1999) Comments on the interpretation of socalled cattle burials of Neolithic Central Europe. Antiquity 73, 542-550. Poole, C. (2000a) Special deposits. In B. Cunliffe and C. Poole (eds.) The Danebury Environs Programme. The Prehistory of a Wessex landscape. Volume 2, Part 4. New Buildings, Longstock, Hants, 1992 and Fiveways, Longstock, Hants, 1996, 7:A7-B4. Oxford, English Heritage and Oxford University Committee for Archaeology Monograph No 49.

Phillipson, D. W. (1968) Part III animal bones. In B. Cunliffe. Excavations at Eldons Seat, Encombe, Dorset. Proceedings of the Prehistoric Society 34, 226-229. Philpott, R. (1991) Burial practies in Roman Britain: a survey of grave treatment and furnishing AD 43-410. Oxford, British Archaeological Reports. British Series 219.

Poole, C. (2000b) Special deposits. In B. Cunliffe and C. Poole (eds.) The Danebury Environs Programme. The Prehistory of a Wessex Landscape. Vol 2. Part 5. Nettlebank Copse, Wherwell, Hants, 1993, Microfiche 9:E10-G14. Oxford, English Heritage and Oxford University Committee for Archaeology Monograph 49.

Piggott, S. (1962a) Heads and Hoofs. Antiquity 36, 110118. Piggott, S. (1962b) The West Kennet Long Barrow. Excavations 1955-56. London, HMSO.

Poole, C. (2000c) Special deposits. In B. Cunliffe and C. Poole (eds.) The Danebury Environs Programme. The Prehistory of a Wessex landscape. Volume 2, Part 6. Houghton Down, Stockbridge, Hants, 1994, 11:G112:C6. Oxford, English Heritage and Oxford University Committee for Archaeology Monograph No 49.

Pilgram, T. and Marshall, F. (1995) Bone counts and statisticians: A reply to Ringrose. Journal of Archaeological Science 22 (1), 93-97. Pinter-Bellows, S. (1992) The vertebrate remains from sites 94 and 95. In G. Milne and J. D. Richards (eds.) Two Anglo-Saxon Buildings and Associated Finds, 69-79. York, York University Archaeological Publications 9.

Poole, C. (2000d) Special deposits in pits. In B. Cunliffe and C. Poole (eds.) The Danebury Environs Programme. The Prehistory of a Wessex Landscape. Volume 2, Part 3. Suddern Farm, Middle Wallop, Hants, 1991 and 1996, Microfiche 4:G5-5:C10. Oxford, English Heritage and Oxford University Committee for Archaeology Monograph 49.

Pitt-Rivers, A. (1888) Excavations in Chranborne Chase II. Privately Published. Platt, M. I. (1951) Bird bones. In K. M. Richardson. The excavation of Iron Age villages on Boscome Down West. Wiltshire Archaeological and Natural History Magazine 54, 166.

Poole, C. (2000e) Special deposits in pits. In B. Cunliffe and C. Poole (eds.) The Danebury Environs Programme. The Prehistory of a Wessex landscape. Volume 2, Part 2. Bury Hill, Upper Clatford, Hants, 1990, 2:B7-C8. Oxford, English Heritage and Oxford University Committee for Archaeology Monograph No 49.

Pluskowski, A. (2002) Hares with crossbows and rabbit bones: intergrating physical and conceptual studies of Medieval fauna. Archaeological Review from Cambridge 18, 153-182.

Potter, S. L. (2005) The physics of cutmarks. Journal of Taphonomy 3 (2), 91-106.

Pollard, J. (1995) Inscibing space: formal deposition at the later Neolithic monument of Woodhenge, Wiltshire. Proceedings of the Prehistoric Society 61 (137-156)

Powell, A. B., Allen, M. J., Chapman, J., Every, R., Gale, R., Harding, P., Knight, S., McKinley, K. I. and Stevens, C. (2005) Excavations along the Old Sarum Water Pipeline, North of Salisbury. Wiltshire Archaeological and Natural History Magazine 98, 250-280.

Pollard, J. (1999) The Keiller excavations. In A. Whittle, J. Pollard and C. Grigson (eds.) The Harmony of Symbols. The Windmill Hill Causewayed Enclosure, 2472. Oxford, Oxbow Books.

Powell, A. B., Smith, P., Clark, K. M., Serjeantson, D. and Knight, S. (2006) Animal Bone. In M. Fulford, A. B. Powell, R. Entwistle and F. Raymond (eds.) Iron Age and Romano-British Settlements and Landscapes of Salisbury

Pollard, J. (2001) The Aesthetics of Depositional Practice. World Archaeology 33 (2), 315-333.

212

Bibliography Plain, Salisbury, Wessex Archaeology and University of Reading, Wessex Archaeology Report 20.

Renfrew, C. (1985) The Archaeology of Cult: the sanctuary at Phylakopi. London, Thames and Hudson.

Prummel, W. (1997) Evidence of hawking (Falconry) from bird and mammal bones. International journal of Osteoarchaeology 7, 333-338.

Renfrew, C. (1994) Towards a cognitive archaeology. In C. Renfrew and E. Zubrow (eds.) The Ancient Mind, 112. Cambridge, Cambridge University Press.

Pryor, F. (1988) Etton, near Maxey, Cambridgeshire: a causewayed enclosure on the Fen-edge. In C. Burgess, P. Topping, C. Mordant and M. Maddison (eds.) Enclosures and Defences in the Neolithic of Western Europe, 107126. Oxford, British Archaeological Report International Series 403.

Reynolds, P. (1974) Experimental Iron Age storage pits: an interim report. Proceedings of the Prehistoric Society 40, 118-131. Richards, C. and Thomas, J. (1984) Ritual activity and structured deposition in Later Neolithic Wessex. In R. Bradley and J. Gardiner (eds.) Neolithic Studies: a Review of some Current Research, 189-218. Oxford, British Archaeological Report British Series 133.

Rackham, D. J. (1977a) The fauna from a 9th century farmstead at Ribblehead, West Yorkshire. English Heritage, Ancient Monuments Laboratory Report 2970.

Richards, J. (1990) The Stonehenge Environs Project. London, English Hertiage.

Rackham, D. J. (1977b) The faunal remains. In P. V. Addyman and J. Priestley. Baile Hill, York: A report on the Institute's excavations. The Archaeological Journal 134, 146-152.

Richardson, J. (2001) The animal bone. In I. Roberts, A. Burgess and D. Berg (eds.) A New Link to the Past The Archaeological Landscape of the M1-A1 Link Road, 214219. Leeds, West Yorkshire Archaeology Service. Yorkshire Archaeological Report No. 7.

Rackham, D. J. (1985) The animal bone. In D. H. Evans and D. H. Heslop. Two Medieval Sites in Yarm. Yorkshire Archaeological Journal 57, 76-77.

Richardson, J. (2002) The mammal bone. In I. Roberts (ed.) Pontefract Castle Archaeological Excavations 1982-86, 363-384. Leeds, West Yorkshire Archaeology Service. Yorkshire Archaeological Report No. 8.

Rackham, D. J. (forthcoming) The faunal remains. In C. Haselgrove (ed.) Stanwick, North Yorkshire IV: Excavations in the Tofts 1984-9,

Richardson, J. (2003a) Animal remains. In C. Moloney, R. Holbrey, P. Wheelhouse and I. Roberts (eds.) Catterick Racecourse, North Yorkshire, the reuse and adaptation of a monument from prehistoric to Anglian times, 43. Leeds, West Yorkshire Archaeology Service publication No. 4.

Radley, J. (1972) Excavations in the defences of the city of York: An early medieval stone tower and the successive earth ramparts. Yorkshire Archaeological Journal 44, 38-64. Rahtz, P. A. (1963) A Roman Villa at Downton. Wiltshire Archaeological and Natural History Magazine 58, 303-341.

Richardson, J. (2003b) The environmental evidence. In I. Robert (ed.) Excavations at Topham Farm, Sykehouse, South Yorkshire. A Late Iron Age and Romano-British Settlement in the Humberhead Levels, 26-27. Leeds, West Yorkshire Archaeological Services Publication 5.

Rahtz, P. A. and Bateman, J. (1986) Wharram Le Street Roman villa 1978-1980. In P. Rahtz, C. Hayfield and J. Bateman (eds.) Two Roman Villas at Wharram Le Street, 1-14. York, York University Archaeological Publications 2.

Richardson, J. (2005) The animal remains. In I. Roberts (ed.) Ferrybridge Henge: The Ritual Landscape, 180183. Leeds, West Yorkshire Archaeology Service. Yorkshire Archaeological Report No. 10.

Ralph, S. (2007) Feasting and Social Complexity in Later Iron Age East Anglia. Oxford, British Archaeological Reports British Series 451.

Richardson, J. E. (1997) Economy and ritual; the use of animal bone in the interpretation of the Iron Age to Roman cultural transition. In K. Meadows, C. Lemke and J. Heron (eds.) Proceedings of the Sixth Annual Theoretical Roman Archaeology Conference Sheffield 1996, 82-90. Oxford Oxbow.

Ray, K. and Thomas, J. (2003) In the kinship of cows: the social centrality of cattle in the earlier Neolithic of Southern Britain. In M. Parker-Pearson (ed.) Food, Culture and Identity in the Neolithic and Early Bronze Age, 37-44. Oxford, British Archaeological Reports International Series 1117.

Richardson, K. M. (1951) The excavation of Iron Age villages on Boscome Down West. Wiltshire Archaeological and Natural History Magazine 54, 123168.

Reitz, E. J. and Wing, E. S. (1999) Zooarchaeology. Cambridge, Cambridge University Press.

213

Investigating Animal Burials; Ritual, Mundane and Beyond Hill, Avebury. Wiltshire Archaeological and Natural History Magazine 94, 15.

Riggott, J. M. and Williams, I. C. (1984) Report on the animal bones. In T. C. M. Brewster (ed.) The Excavation of Whitegrounds Barrow, Burythorpe, 18-19. Malton, East Riding Archaeological Research Committee.

Roux, V. (2007) Ethnoarchaeology: A non historical science of reference necessary for interpreting the past. Journal of Archaeological Method and Theory 14 (2), 153-178.

Rixson, D. (1982) The animal bones. In N. H. Field (ed.) The Iron Age and Romano-British Settlement on Bradford Down, Pamphill, Dorset. Proceedings of the Dorset Natual History and Archaeological Society 104, 91-92.

Rütimeyer, L. (1862) Die fauna der Pfahlbauten der Schweiz. Zurich, Druck von Zürcher und Furrer.

Rixson, D. (1988) Butchery evidence on animal bones. Circaea 6 (1), 49-62.

Rushe, C. M., Smith, H. and Halstead, P. (1994) The animal bone. In T. C. M. Brewster and C. Hayfield. Excavations at Sherburn, East Yorkshire. Yorkshire Archaeological Journal 66, 143-146.

Rixson, D. (2000) The History of Meat Trading. Nottingham, Nottingham University Press.

Russell, A. (2010) Structured deposition or casual deposition of human remains? A case study of four Iron Age sites from southern England. In J. Morris and M. Maltby (eds.) Integrating Social and Environmental Archaeologies; Reconsidering Deposition, 45-65. Oxford, British Archaeological Reports.

Roberts, C. and Cox, M. (2003) Health and Disease in Britain: from Early Prehistory to the Present day. Stroud, Sutton Publishing. Roberts, C. and Manchester, K. (1997) The Archaeology of Disease. Stroud, Sutton Publishing Ltd.

Russell, M. (2002) Monuments of the British Neolithic. Stroud, Tempus.

Roberts, I., Stead, I. M., Rush, P., Sitch, B. J., McHugh, M. and Milles, A. (2005) Appendix 1: excavation of the Henge Bank and Ditches, 1991; A Summary report. In I. Roberts (ed.) Ferrybridge Henge: The Ritual Landscape, 223-235. Leeds, West Yorkshire Archaeology Service. Yorkshire Archaeological Report No. 10.

Rutter, J. G. and Duke, G. (1958) Excavations at Crossgates near Scarborough 1947-56. Scarborough, Scarborough and District Archaeological Society Research report No. 1. Ryder, M. L. (1961) Report on the animal bones. In C. M. Mitchell (ed.) Kirkstall Abbey Excavations 19551959, 41-54, 67-78, 130-132. Leeds, Publications of the Thoresby Society Vol. 48.

Robertson-Mackay, M. E. (1980) A 'Head and Hooves' Burial Beneath a Round Barrow, with other Neolithic and Bronze Age Sites, on Hemp Knoll, near Avebury, Wiltshire. Proceedings of the Prehistoric Society 46, 123176.

Ryder, M. L. (1965) Appendix H. Animal remains. In C. V. Bellamy (ed.) Pontefract Priory Excavations 19571961, 132-136. Leeds, The Publications of the Thoresby Society Vol 49.

Roncaglia, N. and Grant, A. (2000) Animal Husbandry. In B. Cunliffe and C. Poole (eds.) The Danebury environs programme: the prehistory of a Wessex landscape. Vol.2. Part 1: Woolbury and Stockbridge Down, Stockbridge, Hampshire. 1989, 65-74. Oxford, English Heritage and Oxford University Committee for Archaeology Monograph No. 49.

Ryder, M. L. (1970) The animal remains from Petergate, York, 1957-58. Yorkshire Archaeological Journal 42, 418-428.

Ross, A. (1967) Pagan Celtic Britain. Studies in Iconography and Tradition. London, Routledge & Kegan Paul.

Ryder, M. L. (1974) Animal remains from Wharram Percy. Yorkshire Archaeological Journal 46, 42-52. Sadler, P. (1990) Faunal remains. In J. R. Fairbrother (ed.) Faccombe Netherton. Excavations of a Saxon and Medieval Manorial Complex. Vol 2, 462-508. London, British Museum Occasional Paper 74.

Ross, A. (1968) Shafts, pits, wells - sanctuaries of the Belgic Britons? In J. M. Coles and D. D. A. Simpson (eds.) Studies in Ancient Europe, 255-285. Leicester, Leicester University Press.

Saunder, A. and Phillips, G. (1993) The mammal and bird bones from Mytongate. In D. H. Evans (ed.) Excavations in Hull 1975-76, 191-192. Kingston Upon Hull, East Riding Archaeologist Vol.4. Hull Old Town Report Series No. 2.

Ross, A. and Feachem, R. (1976) Ritual rubbish? The Newstead pits. In J. V. S. Megaw (ed.) To Illustrate the Monuments, 229-237. London, Thames and Hudson. Rouse, A. (2001) Animal remains. In M. Pitts. Excavating the sanctuary: New investigations on Overton 214

Bibliography Serjeantson, D. (1993) The animal bone and shell. In M. S. Ross. Melbury Abbas: Medieval pottery in perspective. Proceedings of the Dorset Natural History and Archaeological Society 115, 119.

Schäefer, W. (1972) Ecology and Palaeoecology of Marine Environments. Chicago, University of Chicago Press. Scott, S. (1987) The faunal remains from 15 pits in the Wytelard property. In P. Armstrong and B. Ayers (eds.) Excavations in High Street and Blackfriargate, 252-253. Kingston Upon Hull, East Riding Archaeologist Vol.8. Hull Old Town Report Series No. 5.

Serjeantson, D. (2002) Goose husbandry in medieval England, and the problem of ageing goose bones. Acta Zoologica Cracoviensia (Proceedings of the 4th Meeting of the ICAZ Bird Working Group) 45, 39-54. Serjeantson, D. (2006) Food or feast at Neolithic Runnymede? In D. Serjeantson and D. Field (eds.) Animals in the Neolithic of Britian and Europe. Neolithic Studies Group Seminar Papers 7, 113-134. Oxford, Oxbow Books.

Scott, S. (1991) The animal bones. In P. Armstrong, D. Tomlinson and D. H. Evans (eds.) Excavations at Lurk Lane Beverley 1979-82, 216-233. Sheffield, Sheffield Excavation Reports 1. Scott, S. (1992) The animal bone. In D. H. Evans and D. G. Tomlinson (eds.) Excavations at 33-35 Eastgate, Beverley 1983-86, 236-251. Sheffield, Sheffield Excavation Reports 3.

Serjeantson, D. and Morris, J. (2011) Ravens and crows in Iron Age and Roman Britain. Oxford Journal of Archaeology 30 (1), 85-107. Shennan, S. (1997) Quantifying Archaeology. 2nd Edition. Edinburgh, Edinburgh University Press.

Scott, S. (1993) The animal bone from the Augustinian Friary garden. In D. H. Evans (ed.) Excavations in Hull 1975-76, 194-195. Kingston Upon Hull, East Riding Archaeologist Vol.4. Hull Old Town Report Series No. 2.

Shipman, P. (1986) Studies of hominid-faunal interactions at Olduvai Gorge. Journal of Human Evolution 15, 691-706.

Screeton, A. and Spratt, D. A. (2001) Skelton Village, History and Excavation. Yorkshire Archaeological Journal 73, 3-18.

Silver, I. A. (1963) The ageing of domestic animals. In D. Brothwell, E. S. Higgs and G. Clark (eds.) Science in Archaeology, 283-302. London, Thames and Hudson.

Scull, C. (1993) Archaeology, early Anglo-Saxon society and the origins of Anglo-Saxon kingdoms. Anglo-Saxon Studies in Archaeology and History 6, 65-82.

Simms, C. (1979) The animal bone. In J. S. Dent. Bronze Age Burials from Wetwang Slack. Yorkshire Archaeological Journal 51, 39.

Seetah, K. (2005) Butchery as a tool for understanding the changing views of animals: Cattle in Roman Britain. In A. Pluskowski (ed.) Just Skin and Bones? New Perspectives on Human-Animal Relations in the Historical Past, 1-8. Oxford, British Archaeological Report International Series 1410.

Smith, I. F. (1965) Excavation of a Bell Barrow, Avebury G55. Wiltshire Archaeological and Natural History Magazine 60, 24-46. Smith, I. F. (1965) Windmill Hill and Avebury. Oxford, Clarendon Press.

Seetah, K. (2006) Multidisciplinary approach to RomanoBritish cattle butchery. In M. Maltby (ed.) Integrating Zooarchaeology, 109-116. Oxford, Oxbow.

Smith, I. F. and Simpson, D. D. A. (1966) Excavation of a Round Barrow on Overton Hill, North Wiltshire. Proceedings of the Prehistoric Society 32, 122-155.

Serjeantson, D. (1985) Evidence of 16th-17th century diet from Shaftesbury, Dorset: The animal remains from No.8 Gold Hill. In. M. Cox (Ed.). Excavations within No. 8 Gold Hill, Shaftesbury, Dorset.

Smith, K. (1977) The Excavation of Winklebury Camp, Basingstoke, Hampshire. Proceedings of the Prehistoric Society 43, 58-69.

Serjeantson, D. (1991a) The bird bones. In B. Cunliffe and C. Poole (eds.) Danebury: an Iron Age Hillfort in Hampshire. Volume 5. The Excavations 1979-1988: the Finds, 479-481. London, Council for British Archaeology Research Report 73.

Smith, K. (2006) Guides, Guards and Gifts to the Gods: Domesticated Dogs in Art and Archaeology of Iron Age and Roman Britain. Oxford, British Archaeology Reports British Series 422.

Serjeantson, D. (1991b) 'Rid grasse of bone': A taphonomic study of the bones from midden deposits at the Neolithic and Bronze Age site of Runnymede, Surrey, England. International Journal of Osteoarchaeology 1, 73-89.

Smith, M. (2006) Bones chewed by canids as evidence for human excarnation: a British case study. Antiquity 80 (309), 671-685.

215

Investigating Animal Burials; Ritual, Mundane and Beyond Stallibrass, S. (1997) Animal bones from CAS sites 452 and 482, Thornbrough Farm, Catterick, North Yorkshire. English Heritage, Ancient Monuments Laboratory Report 104/97.

Smith, P. (1996) Animal bone. In A. P. Fitzpatrick (ed.) A 1st-century AD 'Durotrigian' Inhumation Burial with a Decorated Iron Age Mirror from Portesham, Dorset. Proceedings of the Dorset Natural History and Archaeological Society 118, 54-55.

Stallibrass, S. (1999) The animal bone. In P. Halkon and M. Millett (eds.) Rural Settlement and Industry: Studies in the Iron Age and Roman Archaeology of Lowland East Yorkshire, 129. York, Yorkshire Archaeological Report No 4.

Smith, P. (2004) Animal bone. In J. Wright (ed.) Excavation of early Saxon settlement and Mesolithic activity at Goch Way, near Charlton, Andover. Proceedings of the Hampshire Field Club and Archaeological Society 59, 126. Smith, P. (2006) Animal bone. In P. Andres. A Middle to Late Bronze Age settlement at Dunch Hill, Tidworth. Wiltshire Archaeological and Natural History Magazine 99, 70.

Stallibrass, S. (2000) Cattle, culture, status and soldiers in northern England. In G. Fincham, G. Harrison, R. Holland and L. Revell (eds.) Proceedings of the Ninth Annual Theoretical Roman Archaeology Conference Durham 1999, 64-73. Oxford, Oxbow.

Smith, P. and Allen, M. J. (2000) Animal bone. In M. J. Allen and J. Gardiner (eds.) Our Changing coast a survey of the intertidal archaeology of Langstone Harbour, Hampshire, 166-167. London, CBA Research Report 124.

Stallibrass, S. (2002a) Animal bones. In P. R. Wilson (ed.) Cataractonium: Roman Catterick and its Hinterland. Excavations and Research, 1958-1997, Part II, 392-438. London, Council for British Archaeology Research Report 129.

Smith, R. J. C. (2000) Excavation of Neolithic and Bronze Age Features at the Thomas Hardye (formerly Castlefield) School, Coburg Road, Dorchester, 1994. Proceedings of the Dorset Natual History and Archaeological Society 122, 73-82.

Stallibrass, S. (2002b) Animal bones from Thornbrough Farm (sites 452 and 482). In P. R. Wilson (ed.) Cataractonium: Roman Catterick and its Hinterland. Excavations and Research, 1958-1997, Part II, 431-438. London, Council for British Archaeology Research Report 129.

Smith, R. W. (1997) Excavations at Emwell Street, Warminster: the Early Economy and Environment of a Wiltshire Market Town. Salisbury, Wessex Archaeology.

Startin, J. (1976) The animal bones from two PRIA sites at Chalton, Hampshire. In B. W. Cunliffe (ed.) Iron Age sites in Central Southern England, 47-48. London, CBA Research report 16.

Smith, S. R., Brisbane, M. A. and Cooper, N. (1984) An Iron Age site at Maddison Street, Southampton. Proceedings of the Hampshire Field Club and Archaeological Society 40, 35-47.

Startin, J. (1981) Animal bones. In D. W. A. Startin (ed.) Excavations at South Grove Cottage, Dorchester. Proceedings of the Dorset Natural History and Archaeological Society 103, 42.

Snoek, J. (2006) Defining 'Rituals'. In J. Kreinath, J. Snoek and M. Stausberg (eds.) Theorizing Rituals. Issues, Topics, Approaches, Concepts, 3-14. Leiden, Brill.

Stead, I. M. (ed.) (1980) Rudston Roman Villa. York, Yorkshire Archaeological Society.

Stallibrass, S. (1987) The animal bones. In M. Millett and S. Mcgrail (eds.) The archaeology of the Hasholme logboat. The Archaeological Journal 144, 139-144.

Stead, I. M. (1991) Iron Age Cemeteries in East Yorkshire. London, English Heritage. Stein, J. K. (1983) Earthworm activity: a source of potential disturbance of archaeological sediments. American Antiquity 48 (2), 277-289.

Stallibrass, S. (1995a) Animal bone. In P. Cardwell (ed.) Excavation of the hospital of St Giles by Brompton Bridge, North Yorkshire. The Archaeological Journal 152, 220-224.

Stevens, P. M. (1987) The vertebrate remains from Wharram Percy North Manor Boundary East, Site 39, North Yorkshire. English Heritage, Ancient Monuments Laboratory Report 99/87.

Stallibrass, S. (1995b) Review of the vertebrate remains. In J. P. Huntley and S. Stallibrass (eds.) Plant and Vertebrate Remains from Archaeological Sites in Northern England: Data Reviews and Future Directions, 84-198. Durham, Architectural and Archaeological Society of Durham and Northumberland Research Report 4.

Stevens, P. M. (1992) The vertebrate remains from site 39. In G. Milne and J. D. Richards (eds.) Two AngloSaxon Buildings and Associated Finds, 67-69. York, York University Archaeological Publications 9.

216

Bibliography Thomas, N. (1964) The Neolithic Causewayed Camp at Robin Hood's Ball, Shrewton. Wiltshire Archaeological and Natural History Magazine 59, 1-27.

Stickler, T. (2000) Animal bone. In C. Bateman (ed.) Excavations along the Littleton Drew to Chippenham Gas pipeline. Wiltshire Archaeological and Natural History Magazine 93, 96.

Thomas, R. (2005) Animals, Economy and Status: Integrating Zooarchaeological and Historical Data in the Study of Dudley Castle, West Midlands (c.1100-1750). Oxford, British Archaeological Report British Series 392.

Stoddart, S. (2002) Late Iron Age sacred space in western Europe. Antiquity 76, 1143-1144. Stoertz, C. (1997) Ancient Landscapes of the Yorkshire Wolds. Swindon, RCHME.

Thomas, R. (2007) Chasing the ideal? Ritualism, pragmatism and the Late Medieval hunt in England. In A. Pluskowski (ed.) Breaking and Shaping Beastly Bodies, 125-148. Oxford, Oxbow.

Stone, J. F. S., Piggott, S. and Booth, A. S. J. (1954) Durrington walls, Wiltshire: Recent excavations at a ceremonial site of the early second millennium BC. The Antiquaries Journal 36, 155-177.

Thomas, R. and Mainland, I. (2005) Introduction: animal diet and health – current perspectives and future directions. In J. Davies, M. Fabiã, I. Mainland, M. Richards and R. Thomas (eds.) Diet And Health in Past Animal Populations: Current Research and Future Directions, 1-7. Oxford, Oxbow.

Stopes, M. C., Oakley, K. P. and Wells, L. H. (1952) A discovery of human skulls, with stone artifacts and animal bones, in a fissure at Portland. Proceedings of the Dorset Natural History and Archaeological Society 74, 39-47.

Thomas, R. and Mikliková, Z. (2008) Introduction: current research in animal palaeopathology. In Z. Mikliková and R. Thomas (eds.) Current Research in Animal Palaeopathology. Proceedings of the Second ICAZ Animal Palaeopathology Working Group Conference, 1-3. Oxford, British Archaeological Reports International Series 1844.

Sykes, N. (2003) Animal remains. In J. Valentin. Manor Farm, Portesham, Dorset: Excavations on a multi-period religious and settlement site. . Proceedings of the Dorset Natural History and Archaeological Society 125, 23-69. Sykes, N. (2005) Hunting for the Anglo-Normans: Zooarchaeological evidence for Medieval identity. In A. Pluskowski (ed.) Just Skin and Bones? New Perspectives on Human-Animal Relations in the Historical past, 7380. Oxford, British Archaeological Report International Series 1410.

Thurnam, J. (1868) On ancient British barrows. Part 1. Long barrows. Achaeologia 42, 161-244. Todd, L. C. (1983) The Horner site: taphonomy of an early Holocene bison bonebed, New Mexico, Unpublished PhD

Sykes, N. (2006a) The animal bones. In T. Longman. Iron Age and later defences at Malmesbury: Excavations 1998-2000. Wiltshire Archaeological and Natural History Magazine 99, 138-145.

Todd, L. C. (1987) Taphonomy of the Horner II bone bed. In G. C. Frison and L. C. Todd (eds.) The Horner Site: the type site of the Cody Culture complex, 107-198. Orlando, Academic Press.

Sykes, N. (2006b) The impact of Normans on hunting practices in England. In C. M. Woolgar, D. Serjeantson and T. Waldron (eds.) Food in Medieval England: Diet and Nutrition, 162-175. Oxford, Oxford University Press.

Toynbee, J. M. C. (1971) Death and Burial in the Roman World. London, Johns Hopkins University Press.

Sykes, N. (2007) The Norman Conquest: A Zooarchaeological Perspective. Oxford, British Archaeological Reports International Series 1656.

Toynbee, J. M. C. (1973) Animals in Roman Life and Art. New York, Cornell University Press. Turnbull, P. (1983) Excavations at Rillington, 1980. Yorkshire Archaeological Journal 55, 1-9.

Taylor, I. and Chrismas, M. (1998) Faunal remains. In. D. E. Jonston (Ed.). A Roman and Anglo-Saxon site at Northbrook, Micheldever, Hampshire.

Turner, R. C. (1986) The animal bone from the Roman civil settlement, Doncaster. In P. C. Buckland and J. R. Magilton (eds.) The Archaeology of Doncaster 1. The Roman Civil Settlement, 202-207. Oxford, British Archaeological Report British Series 148.

Thomas, J. (1996) Time, Culture, and Identity: An Interpretive Archaeology. London, Routledge. Thomas, J. (1999) Understanding the Neolithic. London, Routledge.

Turner, V. (1967) The Forest of Symbols. New York, Cornell University Press.

217

Investigating Animal Burials; Ritual, Mundane and Beyond Walker, L. (1984) The deposition of the human remains. In B. Cunliffe (ed.) Danebury: An Iron Age Hillfort in Hampshire. Volume 2, The Excavations 1969-1978: The Finds, 442-461. London, Council for British Archaeology.

Vann, S. and Thomas, R. (2006) Humans, Other Animals and Disease: A comparative approach towards the development of a standardised recording protocol for animal palaeopathology. Internet Archaeology. http://intarch.ac.uk/journal/issue20/vannthomas_toc.html . 20

Walker, W. H. (2002) Stratigraphy and Practical Reason. American Anthropologist 104 (1), 159-177.

Vatcher, F. D. M. (1963) The excavation of the Roman Earthwork at Winterslow, Wiltshire. The Antiquaries Journal 43, 197-214.

Watkins, J. G. and Williams, R. A. H. (1981) An excavation in Highgate, Beverley, 1977. East Riding Archaeologist 7, 71-84.

Vatcher, F. D. M. and Vatcher, H. L. (1965) An excavation of an earthwork near Badbury rings, Dorset. Proceedings of the Dorset Natual History and Archaeological Society 87, 101-102.

Watts, L. and Rahtz, P. (1984) Cowlam Wold Barrows. York, York Univeristy Archaeological Publications No. 3.

Waddington, K. (2010) The politics of the everyday: exploring ‘midden’ space in Late Bronze Age Wiltshire. In J. Morris and M. Maltby (eds.) Integrating Social and Environmental Archaeologies; Reconsidering Deposition, 103-118. Oxford, British Archaeological Reports.

Webster, G. (1964) Further investigations on the site of the Roman fort at Waddon Hill, Stoke Abbott, 1960-62. Proceedings of the Dorset Natural History and Archaeological Society 86, 135-149.

Wahl, J. and Kokabi, M. (1987) Osteologische Untersuchungen an den im Rosgartenmuseum aufbewahrten spätrömischen Skelettresten aus Konstanz. Fundberiche aus Baden-Württemberg 12, 456-461.

Weinstock, J. (2005) Animal bone. In C. Hayfield and T. Pacitto. Excavations of the Great Hall or 'Kyngeshalle' at Scarborough Castle, North Yorkhire. Yorkshire Archaeological Journal 77, 75-82.

Wainwright, G. (1969) The excavation at Balksbury Camp, Andover, Hampshire. Proceedings of the Hampshire Field Club and Archaeological Society 26, 21-55.

Wells, P. S. (2001) The Barbarians Speak: How the Conquered Peoples shaped Roman Europe. Princeton, Princeton University Press. Wenham, L. P. (1968) The Romano-British Cemetery at Trentholme Drive, York. London, HMSO.

Wainwright, G., J (1975) Religion and Settlement in Wessex 300-100 BC. In P. Fowler (ed.) Recent work in Rural Archaeology, 57-71. Bradford-on-Avon, Moonraker Press.

Westley, B. (1970a) The animal bones from Budbury. In G. J. Wainwright (ed.) An Iron Age promontory fort at Budbury, Bradford-on-Avon, Wiltshire. Wiltshire Archaeological and Natural History Magazine 65B, 152154.

Wainwright, G., J and Longworth, I. (1971) Durrington Walls: Excavation 1966-1968. London, Society of Antiquaries Reports of the Research Committee No. 29.

Westley, B. (1970b) Bones from Budbury, 1979. English Heritage, Ancient Monuments Laboratory Report 1552.

Wainwright, G. J. (1979) Gussage All Saints. London, Department of the Environment Archaeology Reports No. 10.

Westley, B. (1971) The animal bones from Durrington Walls, 1970. In. G. J. Wainwright (Ed.). The excavation of Prehistoric and Romano-British settlements near Durrington Walls.

Wainwright, G. J. (1979) Gussage All Saints : an Iron Age settlement in Dorset. London, H.M.S.O. Wainwright, G. J. (1989) The Henge Monuments: ceremony and society in prehistoric Britain. London, Thames and Hudson.

Wheeler, M. (1954) The Stanwick Fortifications North Riding of Yorkshire. Oxford, Reports of the Research Committee of the Society of Antiquaries of London No. XVII.

Wainwright, G. J. and Davies, S. M. (1995) Balksbury Camp, Hampshire: Excavations 1973 and 198. London, English Heritage Archaeological Report 4.

Wheeler, R. E. M. (1943) Maiden Castle, Dorset. London, Report of the Research Committee of the Society of Antiquaries 12.

Wait, G. (1985) Ritual and Religion in Iron Age Britain. Oxford, British Archaeological Report British Series 149.

Whimster, R. (1981) Burial Practices in Iron Age Britain. Oxford, British Archaeological Report British Series 90. 218

Bibliography White, D. A. (1970) Excavation of an Iron Age Round Barrow, Near Handley, Dorset, 1969. The Antiquaries Journal 50, 26-36.

Wilson, B. (1992) Considerations for the identification of ritual deposits of animal bones in Iron Age pits. International Journal of Osteoarchaeology 2, 341-349.

Whitley, J. (2002) Objects with attitude: Biographical facts and fallacies in the study of Late Bronze Age and Early Iron Age warrior graves. Cambridge Archaeological Journal 12 (02), 217-232.

Wilson, B. (1996) Spatial Patterning among Animal Bones in Settlement Archaeology. Oxford, British Archaeological Report British Series 251. Wilson, B. (1999) Displayed or Concealed? Crosscultural evidence for symbolic and ritual activity depositing Iron Age animal bones. Oxford Journal of Archaeology 18 (3), 297-305.

Whittle, A., Grigson, C. and Pollard, J. (1999a) The 1988 Excavations. In A. Whittle, J. Pollard and C. Grigson (eds.) The Harmony of Symbols. The Windmill Hill Causewayed enclosure, 73-115. Oxford, Oxbow.

Wilson, B. and Edwards, P. (1993) Butchery of horse and dog at Witney Palace, Oxfordshire, and the knackering and feeding of meat to hounds during the post-medieval period. Postmedieval Archaeology 27, 43-56.

Whittle, A. and Pollard, J. (1999) The harmony of symbols: Wider meanings. In A. Whittle, J. Pollard and C. Grigson (eds.) The Harmony of Symbols. The Windmill Hill Causewayed enclosure, 381-390. Oxford, Oxbow.

Winder, N. (1991) The animal bone. In. C. M. Hearne and R. J. C. Smith (Eds.). A Late Iron Age settlement and Black-Burnished Ware (BB1) production site at Worgret, near Wareham, Dorset (1986-7).

Whittle, A., Pollard, J. and Grigson, C. (1999b) The Harmony of Symbols. The Windmill Hill Causewayed enclosure. Oxford, Oxbow.

Winder, N. P. (1991) How many bones make five? The art and science of guesstimation in archaeozoology. International journal of Osteoarchaeology 1, 111-126.

Whittle, A., Pollard, J. and Grigson, C. (1999c) Interpretations. In A. Whittle, J. Pollard and C. Grigson (eds.) The Harmony of Symbols. The Windmill Hill Causewayed enclosure, 347-380. Oxford, Oxbow.

Wood, G. W. and Barrett, R. H. (1979) Status of wild pigs in the United States. Wildlife Society Bulletin 7, 237246.

Wille, F. and Kampp, K. (1983) Food of the white-tailed eagle Haliaeetus albicilla in Greenland. Holarctic Ecology 6, 81-88.

Wood, J. W., George, R. M., Henry, C. H. and Kenneth, M. W. (1992) The osteological paradox: Problems of inferring health from skeletal samples. Current Anthropology 33, 343-370.

Williams, H. (2005) Animals, ashes and ancestors. In A. Pluskowski (ed.) Just Skin and Bones? New Perspectives on Human-Animal Relations in the Historical Past, 1940. Oxford, British Archaeological Report International Series 1410.

Woodward, A. (2000) British Barrows. A Matter of Life and Death. Stroud, Tempus.

Willis, S. H. (1994) Roman imports into late Iron Age British societies: towards a critique of existing models. In S. Cottam, D. Dungworth, S. Scott and J. Taylor (eds.) TRAC 94: Proceedings of the Fourth Theoritical Roman Archaeology Conference, Durham 1994, 141-150. Oxford, Oxbow.

Woodward, P. and Woodward, A. (2004) Dedicating the town: urban foundation deposits in Roman Britain. World Archaeology 36 (1), 68-86. Woodward, P. J. (ed.) (1991) The South Dorset Ridgeway. Survey and Excavations 1977-84. Dorchester, Dorset Natural History and Archaeological Society. Monograph Series No 8.

Wilson, B. (1978) Methods and results of bone analysis. In M. Parrington (ed.) The Excavation of an Iron Age Settlement, Bronze Age Ring Ditches and Roman Feature at Ashville Trading Estate, Abingdon (Oxfordshire) 197476, 110-126. Oxford, CBA Research Report 28.

Woodward, P. J. (1993) The discussion. In P. J. Woodward, S. M. Davies and A. J. Graham (eds.) Excavations at Greyhound Yard, Dorchester 1981–4, 351–80. Dorchester, Dorset Natural History and Archaeological Society Monograph Series 12.

Wilson, B. (1985) Degraded bones, feature type and spatial patterning on an Iron Age occupation site in Oxfordshire, England. In N. R. Fieller, D. D. Gilbertson and N. G. Ralph (eds.) Palaeobiological Investigations, 81-100. Oxford, British Archaeological Report International Series 266.

Woodward, P. J., Davies, S. M. and Graham, A. J. (eds.) (1993) Excavations at Greyhound Yard, Dorchester 1981–4. Dorchester, Dorset Natural History and Archaeological Society Monograph Series 12.

219

Investigating Animal Burials; Ritual, Mundane and Beyond Wrathmell, S. and Nicholson, A. (eds.) (1990) Dalton Parlours Iron Age Settlement and Roman Villa. Leeds, West Yorkshire Archaeology Servces. Yorkshire Archaeology No. 3. Wright, S. J. and Ashton, B. L. (1978) The water consumption of sheep and beef cattle on Northern Eyre Peninsula. Agricultural Record 5, 9-12. Yalden, D. and Albarella, A. (2009) The History of British Birds. Oxford, Oxford University Press. Yellen, J. E. (1977a) Archaeological Approaches to the Present: models for reconstructing the past. New York, Academic press. Yellen, J. E. (1977b) Cultural patterning in faunal remains: Evidence from the !Kung Bushmen. In D. Ingersoll and J. E. Yellen (eds.) Experimental Archaeology, 271-331. New York, Columbia University Press. Yonge, E. M. (1983) The animal bone (Sites: X4, X5, X7, X11.1 and X11.2). In K. S. Jarvis (ed.) Excavations in Christchurch 1969-1980, 97. Dorchester, Dorset Natural History and Archaeological Society Monograph 5. Youngson, C. (1978) Animal bones. In G. Coppack. An excavation at Chapel Garth, Bolton, Fangfoss, Humberside. Yorkshire Archaeological Journal 50, 147148.

220

Appendix 1: Sites with ABGs from Southern England Site Name

County

Site type

Knap Hill

Wiltshire

Coneybury 'anomaly'

Wiltshire

Causeway enclosure Pit complex

Fussells Lodge

Wiltshire

Period Earliest

Period Latest

Number of ABGs

Reference

Early Neolithic

Middle Neolithic Middle Neolithic

Long barrow

Early Neolithic

Late Neolithic

3

(Grigson 1966)

Dorset

Causeway enclosure

Early Neolithic

Early Iron Age

1

(Armour-Chelu 1991)

Wiltshire

Rural Settlement

Early Neolithic

Late RomanoBritish

1

(King 1970)

Dorset

Long barrow

Early Neolithic

Late Iron Age

2

(Jackson 1943)

Robin Hoods ball

Wiltshire

Causeway enclosure

Early Neolithic

Middle Neolithic

2

(Thomas 1964)

Windmill Hill

Wiltshire

Causeway enclosure

Early Neolithic

Late Neolithic

26

(Grigson 1999; Jope 1965)

Silbury Hill

Wiltshire

Other

Early Neolithic

Late Neolithic

1

(Gardner 1997)

Whitesheet Hill

Wiltshire

Causeway enclosure

Early Neolithic

Early Neolithic

4

(Maltby 2004a)

Rowden Pasture Neolithic (W4)

Dorset

Pit complex

Early Neolithic

Early Neolithic

1

(Maltby 1985d; 1991a)

Alington Avenue long barrow

Dorset

Long barrow

Middle Neolithic

Middle Neolithic

1

(Maltby 2002b)

Old Sarum Water Pipe: Old Sarum Spur

Wiltshire

Pit complex

Middle Neolithic

Middle Bronze Age

7

(Powell et al. 2005)

Dorset

Round barrow

Late Neolithic

Early Bronze Age

4

(Legge 1991b)

Wiltshire

Round barrow

Late Neolithic

1

(Green and Rollo-Smith 1984)

Thomas Hardye School

Dorset

Enclosure

Late Neolithic

Early Bronze Age Late Bronze Age

1

(Smith 2000)

Alington Avenue land enclosures

Dorset

Ditch complex

Late Neolithic

Early Bronze Age

1

(Maltby 2002b)

Snail down site 1

Wiltshire

Round barrow

Late Neolithic

Middle Bronze Age

1

(Clutton-Brock and Jewell 2005)

Flagstones barrow

Dorset

Round barrow

Late Neolithic

Early Bronze Age

1

(Bullock and Allen 1997)

Coneybury Henge

Wiltshire

Henge

Late Neolithic

Early Bronze Age

2

(Maltby 1990e)

Hampshire

Rural Settlement

Late Neolithic

High Medieval

29

(Maltby 1989b)

Wiltshire

Enclosure

Late Neolithic

Late Neolithic

1

(Harcourt 1969b; 1971c)

Site A. A354 pipe line

Wiltshire

Enclosed Settlement

Early Bronze Age

Late RomanoBritish

2

(Graham and Newman 1993; Hamilton-Dyer 1999a)

Vatcher excavations of an earth-work

Dorset

Enclosure

Early Bronze Age

Late Bronze Age

1

(Maltby 1986c)

Dorset

Enclosed Settlement

Early Bronze Age

Early RomanoBritish

6

(Locker 1992a)

Wiltshire

Enclosure

Early Bronze Age

Late Bronze Age

1

(Legge 1991b)

Dorset

Rural Settlement

Early Bronze Age

Middle Bronze Age

2

(Bullock and Allen 1997)

Maiden Castle Causeway enclosure Winterslow NW Maiden Castle bank barrow

Down Farm Pond Barrow Winterbourne Stoke barrow 44

Easton Lane Marden Enclosure

Crab Farm Enclosure South Lodge Camp Middle Farm

Early Neolithic

3

(Connah 1965)

5

(Maltby 1990c)

Investigating Animal Burials; Ritual, Mundane and Beyond Site Name

County

Site type

Period Earliest

Period Latest

Number of ABGs

Wiltshire

Rural Settlement

Early Bronze Age

Late RomanoBritish

4

(Egerton 1996)

Poundbury settlement

Dorset

Hillfort

Early Bronze Age

Early AngloSaxon

54

(Buckland-Wright 1987)

Bishop Cannings 81, Hemp knoll

Wiltshire

Round barrow

Early Bronze Age

Middle Bronze Age

2

(Grigson 1980)

Dorset

Rural Settlement

Middle Bronze Age

Late Bronze Age

2

(King 1962)

Wiltshire

Shaft

Middle Bronze Age

Early Iron Age

17

(Grigson 1989)

Dorset

Round barrow

Middle Bronze Age

Late Bronze Age

1

(Grinsell 1959)

Hampshire

Enclosed Settlement

(Poole 2000a)

Wiltshire

Midden

Middle Iron Age Late Bronze Age

15

Potterne

Middle Bronze Age Late Bronze Age

2

(Locker 2000)

Dean Bottom

Wiltshire

Rural Settlement

Late Bronze Age

Late Bronze Age

2

(Maltby 1985b; 1992)

Hampshire

Midden

Late Bronze Age

Early Iron Age

4

(Bourdillon 1990a)

Poundbury

Dorset

Cemetery

Late Bronze Age

Late RomanoBritish

21

(Buckland-Wright 1993)

Bell street

Hampshire

Rural Settlement

Late Bronze Age

Early Iron Age

1

(Coy 1993)

Wiltshire

Rural Settlement

Late Bronze Age

Late Iron Age

2

(J. W Jackson 1948b)

Hampshire

Hillfort

Late Bronze Age

Early RomanoBritish

1

(Cunliffe and Poole 2000a)

Wiltshire

Rural Settlement

Late Bronze Age

Late Bronze Age

3

(Maltby 1985b; 1992)

Hampshire

Enclosed Settlement

Late Bronze Age

Early RomanoBritish

58

(Maltby 1985a)

Hampshire

Hillfort

Late Bronze Age

Early RomanoBritish

169

(Harcourt 1969d;1969e; Maltby 1985c; 1987b; 1995a; 2001)

Dorset

Enclosed Settlement

Early Iron Age

Early RomanoBritish

7

(Clark 2002)

Hyde street

Hampshire

Town

Early Iron Age

Late Medieval

1

(Birbeck and Moore 2004)

Nettlebank copse

Hampshire

Banjo

Early Iron Age

Early RomanoBritish

30

(Poole 2000b)

Old Down Farm

Hampshire

Enclosed Settlement

Early Iron Age

Early RomanoBritish

48

(Maltby 1981a)

Owslebury

Hampshire

Banjo

Early Iron Age

Late RomanoBritish

237

(Maltby 1987c)

Dorset

Hillfort

Early Iron Age

Early RomanoBritish

13

(Jackson 1943)

Little Somborne

Hampshire

Enclosed Settlement

Early Iron Age

Late Iron Age

3

(Locker 1977b; 1980a)

Lains Farm, A303 Road inprovment

Hampshire

Enclosure

Early Iron Age

Late Iron Age

11

(Coy 1991)

Dorset

Enclosed Settlement

Early Iron Age

Early Iron Age

1

(Barnetson 1993)

Butterfield down

Shearplace Hill Wilsford Shaft Barrow 23, East of North Down Barn New buildings

La Sagesse (the presbytery)

Little Woodbury Woolbury Bishop Cannings Down Winnall Down

Balksbury camp Compact farm

Maiden Castle wheeler war graves

Pimperne Down

222

Reference

Appendix 1 Site Name

County

Site type

Period Earliest

Period Latest

Number of ABGs

Hampshire

Rural Settlement

Early Iron Age

Late RomanoBritish

2

(Griffith 1976)

Pins Knoll

Dorset

Rural Settlement

Early Iron Age

Late RomanoBritish

3

(Bailey 1967)

Manor farm

Dorset

Rural Settlement

Early Iron Age

High Medieval

14

(Sykes 2003) (Coy 1984c; Grant 1984a; Serjeantson 1991a)

Knights Enham

Danebury

Reference

Hampshire

Hillfort

Early Iron Age

Late Iron Age

279

Dorset

Rural Settlement

Early Iron Age

Early RomanoBritish

4

(Graham 2006)

Wiltshire

Rural Settlement

Early Iron Age

Late Iron Age

3

(King 1951; Platt 1951)

Hampshire

Hillfort

Early Iron Age

Late Iron Age

32

(Jones 1976b; 1977)

Dorset

Rural Settlement

Early Iron Age

Late RomanoBritish

2

(Rixson 1982)

Wiltshire

Hillfort

Early Iron Age

Early Iron Age

2

(Westley 1970a)

Hampshire

Hillfort

Early Iron Age

Late Iron Age

7

(Poole 2000e)

Wiltshire

Hillfort

Early Iron Age

Late Iron Age

2

(Bunting et al. 1962; Coy 1969)

Hod Hill

Dorset

Hillfort

Early Iron Age

Late Iron Age

2

(Bunting et al. 1968)

Cowdown

Wiltshire

Rural Settlement

Early Iron Age

Late Iron Age

3

(Harcourt 1968b)

Hampshire

Enclosed Settlement

Early Iron Age

Late RomanoBritish

32

(Poole 2000c)

Wiltshire

Enclosed Settlement

Early Iron Age

Late Iron Age

4

(Hambleton 2001)

Hampshire

Enclosed Settlement

Early Iron Age

Late RomanoBritish

61

(Poole 2000d)

Maiden Castle internal occupation

Dorset

Hillfort

Early Iron Age

Early RomanoBritish

8

(Armour-Chelu 1991)

Gussage All Saints

Dorset

Enclosed Settlement

Early Iron Age

Late Iron Age

7

(Harcourt 1975; 1979b)

Hampshire

Rural Settlement

Middle Iron Age

Early RomanoBritish

1

(Birbeck and Moore 2004)

Wiltshire

Pit complex

Middle Iron Age

Early RomanoBritish

7

(Powell et al. 2005)

Oakridge well

Hampshire

Enclosed Settlement

Middle Iron Age

Late RomanoBritish

177

(Maltby 1988; 1993)

Tolpuddle Ball

Dorset

Enclosure

Middle Iron Age

Late Iron Age

12

(Hamilton-Dyer 1999b)

2

(Coy 1983g)

Barton Field Boscombe Down West RAF Station Winklebury camp Bradford Down Budbury fort Bury Hill Bury Wood Camp

Houghton Down Groundwell West Suddern Farm

Vicking way Old Sarum Water Pipe: Castle hill

New road Western Suburbs

Hampshire

Town

Middle Iron Age

Late RomanoBritish

Micheldever Wood

Hampshire

Banjo

Middle Iron Age

Late Iron Age

4

(Coy 1987a; Griffiths 1978)

Brighton Hill south

Hampshire

Enclosed Settlement

Middle Iron Age

Early RomanoBritish

8

(Maltby 1987a; 1995b)

Dorset

Rural Settlement

Middle Iron Age

Late RomanoBritish

27

(Buckland-Wright 1990)

Wiltshire

Enclosed Settlement

Middle Iron Age

Early RomanoBritish

1

(Coy 1976a; 1982a)

Hampshire

Rural Settlement

Late Iron Age

Early RomanoBritish

1

(Collis 1979)

Whitcombe Groundwell farm Borough Farm

223

Investigating Animal Burials; Ritual, Mundane and Beyond Site Name

County

Site type

Period Earliest

Period Latest

Number of ABGs

Cowdery's Down

Hampshire

Enclosed Settlement

Late Iron Age

Middle AngloSaxon

5

(Maltby 1983a)

Alington Avenue settlement

Dorset

Cemetery

Late Iron Age

Late RomanoBritish

17

(Maltby 2002a)

Hampshire

Enclosed Settlement

Late Iron Age

Late Iron Age

9

(Maltby 1982b)

Wiltshire

Enclosed Settlement

Late Iron Age

Late Iron Age

1

(Harcourt 1971b)

Wiltshire

Enclosed Settlement

Late Iron Age

Early RomanoBritish

2

(Bird 1968)

Dorset

Cemetery

Late Iron Age

Late Iron Age

2

(Smith 1996)

Hampshire

Town

Late Iron Age

Late RomanoBritish

2

(Maltby 1986b)

St Georges Road

Dorset

Enclosed Settlement

Late Iron Age

Late Medieval

1

(Bullock and Allen 1997)

Silchester ForumBasilica

Hampshire

Town

Late Iron Age

Early AngloSaxon

4

(Grant 2000)

Poundbury pipe-line

Dorset

Rural Settlement

Late Iron Age

Late RomanoBritish

6

(Armour-Chelu 1986)

Flagstones enclosure

Dorset

Enclosed Settlement

Late Iron Age

Early RomanoBritish

22

(Bullock and Allen 1997)

Hampshire

Villa

Early RomanoBritish

Late RomanoBritish

1

(Maltby 1979a; Maltby and Foot 1985)

Dorset

Pit complex

Early RomanoBritish

Late RomanoBritish

1

(Bailey 1965)

Hampshire

Rural Settlement

Early RomanoBritish

Late RomanoBritish

9

(Maltby 1978b)

Hampshire

Cemetery

Early RomanoBritish

Late RomanoBritish

8

(Brothwell and Harcourt 1979)

Wiltshire

Villa

Early RomanoBritish

Late RomanoBritish

3

(Payne 1997)

Hampshire

Town

Early RomanoBritish

Late RomanoBritish

3

(Hamilton-Dyer 1997b)

South Grove Cottage

Dorset

Town

Early RomanoBritish

Late RomanoBritish

2

(Startin 1981)

Greyhound Yard

Dorset

Town

Early RomanoBritish

Early AngloSaxon

172

(Maltby 1990f; 1993)

Dorchester Prison

Dorset

Town

Early RomanoBritish

Late RomanoBritish

1

(Draper and Chaplin 1982)

Colliton park

Dorset

Town

Early RomanoBritish

Late RomanoBritish

3

(Aitken and Aitken 1982)

Portchester Castle Roman

Hampshire

Military

Early RomanoBritish

Late RomanoBritish

55

(Grant 1975)

Neatham

Hampshire

Town

Early RomanoBritish

Late RomanoBritish

10

(Done 1986)

Late RomanoBritish

6

(Hicklin 2006; Peck 2001; Peck and Maltby 2006)

Viables Farm Durrington, Parkway Enclosure Berwick Down Durotrigian Inhumation Portesham Staple Gardens

Braishfield Bath House White way hill Little Somborne Lankhills Castle Copse Roman Villa Silchester South of the North gate (Area 4)

Reference

Barton Field

Dorset

Villa

Early RomanoBritish

Norden

Dorset

Shrine

Early RomanoBritish

Late RomanoBritish

1

(Hughes 1972)

Downton Villa

Wiltshire

Villa

Middle Romano-British

Late RomanoBritish

2

(Rahtz 1963)

Maddington Farm

Wiltshire

Rural Settlement

Middle Romano-British

Late RomanoBritish

11

(Hamilton-Dyer 1996a)

224

Appendix 1 Site Name

County

Site type

Period Earliest

Period Latest

Number of ABGs

Dorset

Enclosed Settlement

Middle Romano-British

Late RomanoBritish

12

(Bullock and Allen 1997)

Hampshire

Town

Late RomanoBritish

Late RomanoBritish

50

(Maltby 1987d)

Chapperton Down

Wiltshire

Rural Settlement

Late RomanoBritish

Late RomanoBritish

1

(Ingrem 2007)

Silchester Insula IX

Hampshire

Town

Late RomanoBritish

Late RomanoBritish

22

(Clark 2006; Ingrem 2006)

Wiltshire

Pit complex

Early AngloSaxon

Middle AngloSaxon

1

(Gooden et al. 2002)

Hampshire

Enclosed Settlement

Early AngloSaxon

Late AngloSaxon

1

(Eastham 1976; Grant 1976b)

Wiltshire

Rural Settlement

Early AngloSaxon

Late Medieval

3

(Bourdillon 2006)

Hampshire

Manorial

Early AngloSaxon

Late Medieval

49

(Sadler 1990)

Wiltshire

Town

Early AngloSaxon

Late Medieval

1

(Smith 1997)

Middle AngloSaxon

7

(Bourdillon 1984; 1987; Bourdillon and Andrews 1997)

Maiden Castle Road Nothern suburbs, victoria road

Matthew estate Porchester Castle Saxon occupation Grove Farm Faccombe Emwell Street

Reference

Six Dials

Hampshire

Town

Middle AngloSaxon

Clifford Street (SOU 15)

Hampshire

Town

Middle AngloSaxon

Middle AngloSaxon

10

(Bourdillon 1990b)

Cook Street

Hampshire

Town

Middle AngloSaxon

Middle AngloSaxon

2

(Bourdillon 1993b)

High Street

Wiltshire

Rural Settlement

Middle AngloSaxon

Late AngloSaxon

1

(Coy 1977a; 1980a)

Cadley road

Wiltshire

Rural Settlement

Middle AngloSaxon

Middle AngloSaxon

4

(Hamilton-Dyer 2001b)

City defences, eastern suburbs and northern suburbs

Hampshire

Town

Late AngloSaxon

Late AngloSaxon

3

(Bourdillon 1992)

Crowder terrace

Hampshire

Town

Late AngloSaxon

Late Medieval

1

(Coy 1984b)

Town

Late AngloSaxon

High Medieval

9

(Coy 1984b)

Late Medieval

1

(Coy 1985a)

New road

Hampshire

Wickham Glebe

Hampshire

Manorial

Late AngloSaxon

Sussex Street

Hampshire

Town

Late AngloSaxon

High Medieval

33

(Coy 1984b)

Osborne house

Hampshire

Manor

High Medieval

Late Medieval

4

(Coy 1986)

Wiltshire

Manorial

High Medieval

High Medieval

1

(Fisher and Dartnall 1987)

Hampshire

Castle

High Medieval

High Medieval

6

(Eastham 1977; 1985; Grant 1977; 1985)

Thames Street

Dorset

Town

High Medieval

Late Medieval

1

(Coy 1977e; 1985b; 1992)

Christchurch Staggs site (X8, X9, X12)

Dorset

Town

High Medieval

Late Medieval

2

(Coy 1978a; 1983a)

West Mead

Dorset

Ditch complex

High Medieval

Late Medieval

3

(Hamilton-Dyer 1999b)

Christchurch easten defences

Dorset

Town

Late Medieval

Late Medieval

1

(Coy 1982c; 1983b)

Selwyn Hall Portchester Castle Medieval

225

Appendix 2: Sites with ABGs from Yorkshire Site Name

County

Site type

Period earliest

Period Latest

Number of ABGs

Reference

Whitegrounds barrow 1

North Yorkshire

Long barrow

Early Neolithic

Late Neolithic

1

(Riggott and Williams 1984)

Burton Fleming, bell slack

East Yorkshire

Cemetery

Middle Iron Age

Middle Iron Age

3

(Legge 1991a)

Grindale Barrow II

North Yorkshire

Square barrow

Middle Iron Age

Middle Iron Age

1

(Manby 1980)

Ferrybridge

West Yorkshire

Enclosed Settlement

Middle Iron Age

Early AngloSaxon

3

(Richardson 2005)

Garton Station

East Yorkshire

Cemetery

Middle Iron Age

Late Iron Age

1

(Legge 1991a)

Hasholme Logboat

East Yorkshire

Other

Middle Iron Age

Middle Iron Age

4

(Stallibrass 1987)

Hawling Road

East Yorkshire

Rural Settlement

Middle Iron Age

Late RomanoBritish

1

(King 1999a)

Kirkburn

East Yorkshire

Cemetery

Middle Iron Age

Middle Iron Age

4

(Legge 1991a)

Rudston Makeshift

East Yorkshire

Cemetery

Middle Iron Age

Late Iron Age

9

(Legge 1991a)

Dalton Parlours

West Yorkshire

Rural Settlement

Late Iron Age

Late Iron Age

4

(Berg 1990a)

Rudston Roman Villa

East Yorkshire

Villa

Late Iron Age

16

(Chaplin and Barnetson 1976; 1980; 1981)

Garton and Wetwang Slack

East Yorkshire

Rural Settlement

Late Iron Age

Late RomanoBritish Early RomanoBritish

13

(Noddle 1979)

Parlington Hollins

West Yorkshire

Enclosed Settlement

Late Iron Age

Early AngloSaxon

5

(Richardson 2001)

9 Blake Street, City Garage, York

North Yorkshire

Town

Early RomanoBritish

Late Medieval

2

(Bond and O'Connor 1999; O'Connor 1987)

Catterick Bainesse farm (site 46)

North Yorkshire

Rural Settlement

Early RomanoBritish

Late RomanoBritish

14

(Meddens 1990a; 2002a; Stallibrass 2002a)

Castleford

West Yorkshire

Fort

Early RomanoBritish

Middle RomanoBritish

1

(Berg 1999)

Aldborough, Chapel hill

North Yorkshire

Town

Early RomanoBritish

Late RomanoBritish

1

(Jones 1971)

16-22 Coppergate, York

North Yorkshire

Town

Early RomanoBritish

Late Medieval

12

(O'Connor 1983a; 1985; 1989)

Catterick Dere Street (site 434)

North Yorkshire

Town

Late RomanoBritish

2

(Payne 1990; 2002)

Catterick Thornbrough Farm

North Yorkshire

Town

Late RomanoBritish

2

(Stallibrass 1997; 2002b)

Catterick Bridge (site 240)

North Yorkshire

Town

Late RomanoBritish

2

(Meddens 1990b; 2002b)

Dalton Parlours (well)

West Yorkshire

Villa

Late RomanoBritish

5

(Berg 1990b)

58-59, Skeldergate, York

North Yorkshire

Town

Late Medieval

16

(O'Connor 1984b)

Shiptonthorpe Roman Road side settlement

East Yorkshire

Rural Settlement

Late RomanoBritish

19

(Mainland 2006)

Early RomanoBritish Middle RomanoBritish Middle RomanoBritish Middle RomanoBritish Middle RomanoBritish Middle RomanoBritish

Appendix 2 Site Name

County

Site type

Trentholme drive, york

North Yorkshire

Cemetery

Wharram Le Street Roman Villa

North Yorkshire

Villa

Period earliest Middle RomanoBritish Middle RomanoBritish

General Accident site, 24-30 Tanner Row, York

North Yorkshire

Town

Middle RomanoBritish

High Medieval

1

(O'Connor 1988)

Sewerby

East Yorkshire

Cemetery

Middle AngloScandinavian

Middle AngloScandinavian

1

(Hirst 1985)

Wharram Site 94-95

North Yorkshire

Rural Settlement

Middle AngloScandinavian

Middle AngloScandinavian

1

(Pinter-Bellows 1992)

Addingham

West Yorkshire

Cemetery

Middle AngloScandinavian

High Medieval

2

(Keith 1997)

Pontefract Castle

West Yorkshire

Castle

Late AngloScandinavian

Late Medieval

1

(Richardson 2002)

46-54 Fishergate, York

North Yorkshire

Monastic

Late AngloScandinavian

Late Medieval

3

(Bond and O'Connor 1999)

Fox Inn, Low Petergate, York

North Yorkshire

Town

High Medieval

Late Medieval

3

(Ryder 1970)

Higher Land

North Yorkshire

Manor House

High Medieval

Late Medieval

4

(O'Connor 1983b)

Kirkstall Abbey

North Yorkshire

Monastic

High Medieval

Late Medieval

1

(Ryder 1961)

16-20 Church Street

South Yorkshire

Town

High Medieval

Late Medieval

7

(Mounteney and Cumberpatch 1996)

West Street

North Yorkshire

Town

High Medieval

Late Medieval

1

(Dobney 2005)

1-5 Aldwark, York

North Yorkshire

Town

Late Medieval

Late Medieval

2

(Bond and O'Connor 1999; O'Connor 1984a)

118-126 Walmgate, York

North Yorkshire

Town

Late Medieval

Late Medieval

1

(O'Connor 1984b)

Ousefleet Property, High Street, Hull

East Yorkshire

Town

Late Medieval

Late Medieval

3

(Berg 1987)

The Bedern Foundry, York

North Yorkshire

Town

Late Medieval

Late Medieval

6

(Bond and O'Connor 1999)

Hotham Property, Blackfriargate, Hull

East Yorkshire

Town

Late Medieval

Late Medieval

1

(Berg 1987)

Scale Lane/Lowergate, Hull

East Yorkshire

Town

Late Medieval

Late Medieval

6

(Phillips 1980)

227

Period Latest

Number of ABGs

Reference

Late RomanoBritish

14

(Fraser and Ryder 1968)

Late RomanoBritish

1

(Rahtz and Bateman 1986)

Appendix 3: Sites from southern England with no ABGs reported Site Name Beehive

County

Site Type

Earliest Period

Latest Period

Reference

Wiltshire

Pit complex

Early Neolithic

Late Neolithic

(Hamilton-Dyer 2003a)

Wiltshire

Pit complex

Early Neolithic

Late Neolithic

(Powell et al. 2005)

Wiltshire

Pit complex

Early Neolithic

Late Neolithic

(Davies 2000)

Easton Down Long Barrow

Wiltshire

Long barrow

Early Neolithic

Late Neolithic

(Noddle 1993)

West Kennet palisade enclosures

Wiltshire

Enclosure

Early Neolithic

Early Bronze Age

(Edwards and Horne 1997)

Stonehenge Lesser Cursus

Wiltshire

Cursus

Early Neolithic

Middle Neolithic

(Maltby and Richards 1990)

long Barrow at Nutbane

Hampshire

Long barrow

Early Neolithic

Late Neolithic

(Bunting et al. 1959)

Lanhill Long Barrow

Wiltshire

Long barrow

Early Neolithic

Late Neolithic

(Clarke 1966)

Amesbury Long Barrow W58

Wiltshire

Long barrow

Early Neolithic

Middle Neolithic

(Maltby 1990a)

Poxwell (Neolithic)

Dorset

Ditch complex

Early Neolithic

Early Neolithic

(Jones 1986) (Ashbee et al. 1979)

Old Sarum Water Pipeline: The Portway Windmill Hill Occupation outside

South Street long Barrow

Wiltshire

Long barrow

Early Neolithic

Middle Neolithic

Beckhampton road long barrow

Wiltshire

Long barrow

Early Neolithic

Middle Neolithic

(Carter and Higgs 1979)

Horslip long barrow

Wiltshire

Long barrow

Early Neolithic

Middle Neolithic

(Highman and Higgs 1979)

King Barrow ridge

Wiltshire

Long barrow

Early Neolithic

Late Neolithic

(Maltby 1990d)

Possible Neolithic settlement Cherhill

Wiltshire

Pit complex

Early Neolithic

Early Neolithic

(Grigson 1983)

Pipeline near Lodge Farm

Dorset

Pit complex

Early Neolithic

Late Neolithic

(Maltby 1989c)

West kennet Long Barrow

Wiltshire

Long barrow

Early Neolithic

Late Neolithic

(Piggott 1962b)

Giants Caves

Wiltshire

chambered cairn

Early Neolithic

Early Neolithic

(Denston 1970)

Enclosure

Early Neolithic

Early Bronze Age

(Maltby 1993)

Middle Neolithic

(W. J Jackson 1948)

Greyhound yard enclosure

Dorset

Woodlands pits (near woodhenge)

Wiltshire

Pit complex

Middle Neolithic

Chalk Plaque pit

Wiltshire

Pit complex

Late Neolithic

Late Neolithic

(Harding 1988)

Durrington Walls

Wiltshire

Henge

Late Neolithic

Early Bronze Age

(Stone et al. 1954)

Bincombe 60a barrow

Dorset

Round barrow

Late Neolithic

Early Bronze Age

(Grinsell 1959)

Millbarrow

Wiltshire

Long barrow

Late Neolithic

Early Bronze Age

(Noddle 1994)

Larkhill near Durrington walls (Neolithic)

Wiltshire

Pit complex

Late Neolithic

Late Neolithic

(Westley 1971)

Fargo Wood II, W34

Wiltshire

Cursus

Late Neolithic

Early Bronze Age

(Maltby 1990b)

Maumbury Rings

Dorset

Henge

Late Neolithic

Early Bronze Age

(Bradley 1975)

Late Neolithic

Middle Bronze Age

(I. F Smith 1965)

Barrow G55, Avebury

Wiltshire

Round barrow

Appendix 3 Site Name

County

Site Type

Earliest Period

Latest Period

Reference

Durrington Walls environs

Wiltshire

Ditch complex

Late Neolithic

Late Neolithic

(Hamilton-Dyer 2004a)

Overton Hill

Wiltshire

Enclosure

Late Neolithic

Early Bronze Age

(Rouse 2001)

Mount Pleasant

Dorset

Henge

Late Neolithic

Early Bronze Age

(Harcourt 1971d; 1979a)

Bumper's Lane Second Quarry

Dorset

Other

Late Neolithic

Middle Bronze Age

(Stopes et al. 1952) (Smith and Simpson 1966)

Overton Hill

Wiltshire

Round barrow

Late Neolithic

Middle Bronze Age

Showell Farm

Wiltshire

Ditch complex

Early Bronze Age

Early Bronze Age

(Higbee 2006)

Wilsford Barrow A

Wiltshire

Round barrow

Early Bronze Age

Late Bronze Age

(Davis 1987)

Barford Farm

Dorset

Round barrow

Early Bronze Age

Early Bronze Age

(Maltby 1989a)

Barrows 5d, 5e Net Down

Wiltshire

Round barrow

Early Bronze Age

Early Bronze Age

(Green and Rollo-Smith 1984)

Milton Lilbourne Barrow 4

Wiltshire

Round barrow

Early Bronze Age

Early Bronze Age

(Grigson 1986)

Barrow, Avebury G70

Wiltshire

Round barrow

Early Bronze Age

Early Bronze Age

(Christie 1964)

Round barrow

Early Bronze Age

Late Bronze Age

(Harcourt 1971e)

Early Bronze Age

(Grigson 1986)

Wilsford Down Group 71 Barrows

Wiltshire

Milton Lilbourne Barrow 2

Wiltshire

Round barrow

Early Bronze Age

Milton Lilbourne Barrow 1

Wiltshire

Round barrow

Early Bronze Age

Early Bronze Age

(Grigson 1986)

Overton Hill, Barrow

Wiltshire

Round barrow

Early Bronze Age

Middle Bronze Age

(Jones 1976a)

Amesbury barrow 39

Wiltshire

Round barrow

Early Bronze Age

Early Bronze Age

(Jones 1979/1980)

Amesbury barrow 51

Wiltshire

Round barrow

Early Bronze Age

Middle Bronze Age

(Clutton-Brock 1975/1976)

Barrow 72, Amesbury

Wiltshire

Round barrow

Early Bronze Age

Early Bronze Age

(Clutton-Brock 1984)

Barrow 61, Amesbury

Wiltshire

Round barrow

Early Bronze Age

Early Bronze Age

(Clutton-Brock 1984)

Barrow 58, Amesbury

Wiltshire

Round barrow

Early Bronze Age

Middle Bronze Age

(Clutton-Brock 1984)

Litton Cheney Barrow

Dorset

Round barrow

Early Bronze Age

Late Bronze Age

(Fraser 1958)

Round barrow

Early Bronze Age

Early Bronze Age

(Coy 1983c)

Early Bronze Age

(Jones 1975a; 1978)

Clarendon Park (Vatcher site 10)

Wiltshire

Long Bredy Round Barrow

Dorset

Round barrow

Early Bronze Age

Fordington Farm

Dorset

Round barrow

Early Bronze Age

Early Bronze Age

(Maltby 1991b)

Rag Copse

Hampshire

Round barrow

Early Bronze Age

Late Bronze Age

(Fraser 1963)

Cowdery's Down

Hampshire

ring ditches

Early Bronze Age

Late Bronze Age

(Maltby 1983b)

Milton Lilbourne Barrow 5

Wiltshire

Round barrow

Early Bronze Age

Early Bronze Age

(Grigson 1986)

229

Investigating Animal Burials; Ritual, Mundane and Beyond Site Name

County

Site Type

Earliest Period

Latest Period

Reference

Bell-Barrow Oakley Down

Dorset

Round barrow

Early Bronze Age

Late Bronze Age

(King 1953)

Bronze Age Barrows, near Amesbury

Wiltshire

Round barrow

Early Bronze Age

Late Bronze Age

(Clutton-Brock 1974)

Down Farm Goddards barrow 4

Wiltshire

Round barrow

Early Bronze Age

Middle Bronze Age

(Dorell and Cornwall 1960)

Micheldever Wood site (R4)

Hampshire

Round barrow

Early Bronze Age

Late Bronze Age

(Coy and Winder 1976b)

Early Bronze Age

(Maltby 1986d)

Winterbourne Stoke Barrow G39

Wiltshire

Round barrow

Early Bronze Age

Latton land

Wiltshire

Ring Ditch

Early Bronze Age

Late Bronze Age

(Hamilton 2004)

Wilsford Barrow C

Wiltshire

Round barrow

Early Bronze Age

Late Bronze Age

(Davis 1987)

Poxwell Barrow

Dorset

Round Barrow

Early Bronze Age

Early Bronze Age

(Jones 1986)

R17, R30, R36 Graces Farm

Hampshire

Ring Ditch

Early Bronze Age

Late Bronze Age

(Coy and Maltby 1978)

King barrow ridge

Wiltshire

Round barrow

Early Bronze Age

Late Bronze Age

(Fitzgerald and Egerton 1994)

Milton Lilbourne Barrow 3

Wiltshire

Round barrow

Early Bronze Age

Early Bronze Age

(Grigson 1986)

Barrows 5e Net Down

Wiltshire

Round barrow

Early Bronze Age

Early Bronze Age

(Green and Rollo-Smith 1984)

Lamb Down site F

Wiltshire

Round barrow

Middle Bronze Age

Middle Bronze Age

(Clutton-Brock and Jewell 1963)

Lamb Down Grinsell barrow 4

Wiltshire

Round barrow

Middle Bronze Age

Middle Bronze Age

(Clutton-Brock and Jewell 1963)

Lamb Down Grinsell barrow 2

Wiltshire

Round barrow

Middle Bronze Age

Middle Bronze Age

(Clutton-Brock and Jewell 1963)

Old Sarum Water Pipeline: South of Ford Road

Wiltshire

Pit complex

Middle Bronze Age

Late Bronze Age

(Powell et al. 2005)

Cowleaze Pasture (W27)

Dorset

Enclosure

Middle Bronze Age

Late Bronze Age

(Maltby 1985d; 1991a)

Winterbourne Stoke Barrow G47

Wiltshire

Round barrow

Middle Bronze Age

Late Bronze Age

(Maltby 1986d)

Dunch Hill

Hampshire

Rural Settlement

Middle Bronze Age

Late Bronze Age

(P Smith 2006)

Westbury

Hampshire

Rural Settlement

Middle Bronze Age

Middle Bronze Age

(Grant 1976a)

Bronze Age Barrow at Buckskin

Wiltshire

Round barrow

Late Bronze Age

Late Bronze Age

(Clark 1995)

Winterbourne Stoke Barrow G46

Wiltshire

Round barrow

Late Bronze Age

Late Bronze Age

(Maltby 1986d)

Coburg road

Dorset

Rural Settlement

Late Bronze Age

Late Bronze Age

(Hamilton-Dyer 1992)

Easton Down

Hampshire

Ring Ditch

Late Bronze Age

Early Iron Age

(Coy 1982b; Coy and Winder 1976a)

Chalton bronze age settlement

Hampshire

Rural Settlement

Late Bronze Age

Late Bronze Age

(Grant 1970)

Burderop Down

Wiltshire

Rural Settlement

Late Bronze Age

Late Bronze Age

(Maltby 1985b; 1992)

Woodford G12 Barrow

Wiltshire

Round barrow

Late Bronze Age

Late Bronze Age

(Maltby 1986d)

230

Appendix 3 Site Name

County

Site Type

Earliest Period

Latest Period

Reference

Rockley Down enclosure

Wiltshire

Enclosure

Late Bronze Age

Late Bronze Age

(Maltby 1985b; 1992)

Rowden Pasture Bronze age (W4)

Dorset

Rural Settlement

Late Bronze Age

Late Bronze Age

(Maltby 1985d; 1991a)

Worgret

Dorset

Rural Settlement

Early Iron Age

Late Iron Age

(N Winder 1991)

Poxwell

Dorset

Pit complex

Early Iron Age

Late Iron Age

(Jones 1975b)

Stockton earthworks

Wiltshire

Hillfort

Early Iron Age

Late Iron Age

(Moody and Moody 1997)

Pewsey Hill Enclosure

Wiltshire

Enclosure

Early Iron Age

Early Iron Age

(Harcourt 1971a)

Eldons Seat

Dorset

Rural Settlement

Early Iron Age

Late Iron Age

(Phillipson 1968)

Gussage Hill

Dorset

Round barrow

Early Iron Age

Late Iron Age

(White 1970)

Hengistbury Head

Dorset

Other

Early Iron Age

Late Iron Age

(Grant 1987)

Chalton (Site 15)

Hampshire

Rural Settlement

Early Iron Age

Late Iron Age

(Startin 1976)

Rope lake

Dorset

Rural Settlement

Early Iron Age

Late Iron Age

(Coy 1983e; 1987c)

George Inn, Portsdown

Hampshire

Rural Settlement

Early Iron Age

Late Iron Age

(Chesney 1968)

Poxwell (Iron Age)

Dorset

Rural settlement

Early Iron Age

Late Iron Age

(Jones 1986)

Littleton Drew to Chippenham Gas pipeline

Wiltshire

Rural Settlement

Early Iron Age

Early RomanoBritish

(Stickler 2000)

Mancombe down settlement enclosure

Wiltshire

Enclosure

Early Iron Age

Early Iron Age

(Fowler et al. 1965)

Bury Wood Camp

Wiltshire

Hillfort

Early Iron Age

Late Iron Age

(Coy 1967)

Down Barn West

Wiltshire

Enclosure

Early Iron Age

Early Iron Age

(Fowler et al. 1965)

Larkhill near Durrington walls (Iron Age)

Wiltshire

Pit complex

Early Iron Age

Late Iron Age

(Westley 1971)

Staple gardens, Orams Arbour

Hampshire

Enclosure

Early Iron Age

Late Iron Age

(Maltby 2004b)

Budbury hillfort

Wiltshire

Hillfort

Early Iron Age

Late Iron Age

(Westley 1970b)

Brickley Lane

Wiltshire

Rural Settlement

Early Iron Age

Early RomanoBritish

(Charles 2002)

Chilbolton Down

Hampshire

Enclosure

Early Iron Age

Late Iron Age

(Maltby 1978a)

Chalton (Site 50)

Hampshire

Rural Settlement

Early Iron Age

Late Iron Age

(Startin 1976)

Woolbury

Hampshire

Hillfort

Early Iron Age

Late Iron Age

(Roncaglia and Grant 2000)

Abbotstone Down

Hampshire

Rural Settlement

Early Iron Age

Late Iron Age

(Maltby 1986a)

Nuns' Walk

Wiltshire

Hillfort

Early Iron Age

Late Iron Age

(Sykes 2006a)

Portsdown Hill

Hampshire

Enclosure

Early Iron Age

Late Iron Age

(Bradley 1967)

West Creech

Dorset

Rural Settlement

Middle Iron Age

Late Iron Age

(Hamilton-Dyer 1991)

Furzy Island

Dorset

Rural Settlement

Middle Iron Age

Late Iron Age

(Hamilton-Dyer 1991)

Hampshire

Pit complex

Middle Iron Age

(Smith et al. 1984)

Hampshire

Banjo

Late Iron Age

(Coy 1978b; 1980b)

Maddison Street, Iron Age, (SOU 29) R27 Micheldever Wood banjo

Middle Iron Age Middle Iron Age

231

Investigating Animal Burials; Ritual, Mundane and Beyond Site Name

County

Site Type

Earliest Period

Latest Period

Reference

Beach's Barn

Wiltshire

Rural Settlement

Late Iron Age

Late RomanoBritish

(Harding 2007)

Waside Farm

Wiltshire

Rural Settlement

Late Iron Age

Late RomanoBritish

(Ingrem 2002)

Ructstalls Hill

Hampshire

Rural Settlement

Late Iron Age

Middle Romano-British

(Gregory 1978)

Viables two (Jays Close)

Hampshire

Rural Settlement

Late Iron Age

Early RomanoBritish

(Baxter 2004) (Coy 1982d; 1987b)

Ower Farm Iron Age and Romano-British settlement Christchurch Druitt Gardens (X5)

Dorset

Rural Settlement

Late Iron Age

Early RomanoBritish

Dorset

Pit complex

Late Iron Age

Late Iron Age

(Yonge 1983)

Romano-British settlement Corfe Castle Estate

Dorset

Rural Settlement

Late Iron Age

Late RomanoBritish

(King 1965) (Clark 1986)

Choseley Farm

Hampshire

Rural Settlement

Late Iron Age

Late RomanoBritish

Ower Peninsula industral area

Dorset

Rural Settlement

Late Iron Age

Early RomanoBritish

(Hamilton-Dyer 1991)

Brighton Hill South RomanoBritish enclosure

Hampshire

Enclosure

Late Iron Age

Late RomanoBritish

(Coe and Newman 1992)

North of London Lodge MARC3, site R3

Hampshire

Road

Late Iron Age

Late RomanoBritish

(Coy and Winder 1975a; 1981)

Kings Somborne (BGC SITE BS/M28);

Hampshire

Rural Settlement

Early RomanoBritish

Late RomanoBritish

(Locker 1977a)

Halstock villa

Dorset

Villa

Early RomanoBritish

Late RomanoBritish

(Peck 1993)

Silchester Manor Farm

Hampshire

Town

Early RomanoBritish

Early RomanoBritish

(Maltby 1982d; 1984a)

St Denys (SOU 981)

Hampshire

Rural Settlement

Early RomanoBritish

Late RomanoBritish

(Hamilton-Dyer 2002b)

Late RomanoBritish

(Cornwall 1958; Macdonald 1958)

Steuart Road fosse Clausentum

Hampshire

Town

Early RomanoBritish

Bridget's Farm and Burntwood Farm

Hampshire

Ditch complex

Early RomanoBritish

Late RomanoBritish

(Coy and Winder 1976c)

Penny's Farm

Dorset

Pit complex

Early RomanoBritish

Late RomanoBritish

(Allen 2000)

Steuart/Bitterne Road Clausentum

Hampshire

Town

Early RomanoBritish

Late RomanoBritish

(Bilton 1958)

County Hall

Dorset

Town

Early RomanoBritish

Late RomanoBritish

(Hamilton-Dyer 1993)

Langstone Harbour

Hampshire

Pit complex

Early RomanoBritish

Late RomanoBritish

(Smith and Allen 2000)

Lake Farm

Dorset

Military

Early RomanoBritish

Late RomanoBritish

(Coy 1983f)

Silchester Defences

Hampshire

Town

Early RomanoBritish

Late RomanoBritish

(Maltby 1982d; 1984a)

Late RomanoBritish

(Vatcher 1963)

Possible roman settlement Winterslow

Wiltshire

Enclosure

Early RomanoBritish

Hermitage

Wiltshire

Rural Settlement

Early RomanoBritish

Late RomanoBritish

(Hamilton-Dyer 1997a)

Poxwell (Romano-British)

Dorset

Ditch complex

Early RomanoBritish

Late RomanoBritish

(Jones 1986)

Cockey Down farmstead

Wiltshire

Rural Settlement

Early RomanoBritish

Late RomanoBritish

(Lovell et al. 1999)

Stratton park R1

Hampshire

Ditch complex

Early RomanoBritish

Late RomanoBritish

(Maltby 1979b)

232

Appendix 3 Site Name

County

Site Type

Earliest Period

Latest Period

Reference

White-Walls

Wiltshire

Rural Settlement

Early RomanoBritish

Late RomanoBritish

(Hammon 2006)

Old Sarum Water Pipeline: Camp Hill

Wiltshire

Rural Settlement

Early RomanoBritish

Early RomanoBritish

(Powell et al. 2005)

Eyewell Farm

Wiltshire

Rural Settlement

Early RomanoBritish

Late RomanoBritish

(Hamilton-Dyer 1998)

Waddon Hill

Dorset

Military

Early RomanoBritish

Late RomanoBritish

(Webster 1964)

Middle Romano-British

(Locker 1980b)

Ashley enclosures

Hampshire

Enclosure

Middle Romano-British

Silchester defences 1968

Hampshire

Town

Middle Romano-British

Late RomanoBritish

(Baker 1983)

Box Roman Villa

Wiltshire

Villa

Middle Romano-British

Late RomanoBritish

(Fisher and Dartnall 1987)

Showell Farm

Wiltshire

Ditch complex

Late RomanoBritish

Late RomanoBritish

(Higbee 2006)

Larkhill near Durrington walls (Romano-British)

Wiltshire

Rural Settlement

Late RomanoBritish

Late RomanoBritish

(Westley 1971)

Northbrook

Hampshire

Rural Settlement

Early AngloSaxon

Late AngloSaxon

(Taylor and Chrismas 1998)

Shepherd's Farm

Dorset

Cemetery

Early AngloSaxon

Late AngloSaxon

(Cox 1989)

Rural Settlement

Early AngloSaxon

Late AngloSaxon

(Ford 1997)

Late AngloSaxon

(Hamilton-Dyer 2003b)

Bentley Green Farm

Hampshire

Riverdene

Hampshire

Rural Settlement

Early AngloSaxon

Abbots Worthy

Hampshire

Rural Settlement

Early AngloSaxon

Late AngloSaxon

(Coy 1987d)

Nuns' Walk

Wiltshire

Rural settlement

Early AngloSaxon

Late AngloSaxon

(Sykes 2006a)

Montefiore (SOU 503)

Hampshire

Ditch complex

Early AngloSaxon

Late AngloSaxon

(Hamilton-Dyer 1996b)

St John's Hospital and South Street

Wiltshire

Town

Early AngloSaxon

High Medieval

(Hamilton-Dyer 2000a)

St Mary's Stadium Hamwic

Hampshire

Rural Settlement

Early AngloSaxon

Middle AngloSaxon

(Hamilton-Dyer 2005b)

Goch way

Hampshire

Rural Settlement

Early AngloSaxon

Late AngloSaxon

(Smith 2004)

Old Down Farm (AngloSaxon)

Hampshire

Rural Settlement

Middle AngloSaxon

Middle AngloSaxon

(Bourdillon 1980a)

Middle AngloSaxon

(Colley 1984)

Deanery School Chapel Road (SOU 9, 17)

Hampshire

Town

Middle AngloSaxon

Saxon cemetery at Cook Street (SOU 823)

Hampshire

Cemetery

Middle AngloSaxon

Middle AngloSaxon

(Hamilton-Dyer 2001c)

Melbourne Street (SOU 1, 4, 5, 6, 20)

Hampshire

Rural Settlement

Middle AngloSaxon

Middle AngloSaxon

(Bourdillon and Coy 1980; Coy 1977b)

Chantry Fields

Dorset

Rural Settlement

Middle AngloSaxon

Middle AngloSaxon

(Allen 1992)

Christchurch Burh defences (w9)

Dorset

Town

Late AngloSaxon

Late AngloSaxon

(Coy 1982c; 1983b)

West Gate Street/41 Bugle Street (SOU 111)

Hampshire

Town

High Mediaeval

Late Mediaeval

(Coy 1977f)

Quilters Vault (SOU 129)

Hampshire

Town

High Mediaeval

High Mediaeval

(Bourdillon 1978)

Melbury Abbas pond

Dorset

Rural Settlement

High Mediaeval

High Mediaeval

(Serjeantson 1993)

233

Investigating Animal Burials; Ritual, Mundane and Beyond Site Name

County

Site Type

Earliest Period

Latest Period

Reference

Bugle Hall (SOU 164) Christchurch Northen Defences (w5)

Hampshire

Town

High Mediaeval

High Mediaeval

(Noddle 1975)

Dorset

Town

High Mediaeval

High Mediaeval

(Coy 1982c; 1983b)

Gomeldon DMV

Wiltshire

Rural Settlement

High Mediaeval

High Mediaeval

(Harcourt 1986)

Statton Park

Wiltshire

Enclosure

High Mediaeval

High Mediaeval

(Newton 1979/1980)

Clarendon Palace

Wiltshire

Manorial

High Mediaeval

Late Mediaeval

(King et al. 1988)

Cuckoo Lane (SOU 163)

Hampshire

Town

High Mediaeval

High Mediaeval

(Noddle 1975)

Ower Farm Medieval settlement

Dorset

Rural Settlement

High Mediaeval

High Mediaeval

(Hamilton-Dyer 1991)

High street and Broad Lane (SOU161)

Hampshire

Town

High Mediaeval

Late Mediaeval

(Noddle 1975)

Wiltshire

Rural Settlement

High Mediaeval

High Mediaeval

(Jones 1976a)

Dorset

Town

High Mediaeval

Late Mediaeval

(Coy 1982c; 1983b)

Brownsea island

Dorset

Rural Settlement

High Mediaeval

High Mediaeval

(Jones 1974)

Old Sarum Farm

Wiltshire

Town

High Mediaeval

High Mediaeval

(Musty and Rahtz 1964)

New Park street

Wiltshire

Town

High Medieval

Late Medieval

(Jones 1993)

Postern Mill site

Wiltshire

Industrial

High Medieval

Late Medieval

(Currie 1993)

Brighton Hill South Medieval settlement

Hampshire

Rural Settlement

High Medieval

Late Medieval

(Coy 1988)

Maddison Street, castle (SOU 29)

Hampshire

Castle

High Medieval

High Mediaeval

(Bourdillon 1986c; Coy 1983h)

Foxcotte

Hampshire

Rural Settlement

High Medieval

Late Medieval

(Coy 1984a; 1985c)

3 Kingsbury square

Wiltshire

Town

High Medieval

Late Medieval

(Hamilton-Dyer 2001a)

Froman's Cow drove hill

Hampshire

Rural Settlement

High Medieval

Late Medieval

(Hamilton-Dyer 2004b)

West Kennet Village oil pipe line Christchurch Sainsbury car park (w6)

Upper Bugle Street (SOU 123)

Hampshire

Castle

High Medieval

Late Medieval

(Bourdillon 1986a; Coy 1977d; Hamilton-Dyer 1986)

Osbourne House

Hampshire

Other

High Medieval

High Medieval

(Coy and Winder 1975b)

Basing House

Hampshire

Other

High Medieval

Late Medieval

(Allen and Anderson 1999)

8 Gold Hill

Dorset

Town

High Medieval

Late Medieval

(Serjeantson 1985)

Howards Lane

Dorset

Town

High Medieval

Late Medieval

(Allen 1995)

Medieval Manorial Building of Kingston Lacy

Dorset

Rural Settlement

High Medieval

Late Medieval

(Locker 1998)

43 South Street

Dorset

Town

High Medieval

Late Medieval

(Loader 2000)

Bell Street

Dorset

Town

High Medieval

Late Medieval

(Ingrem 2000)

Chantry Fields

Dorset

Rural Settlement

High Medieval

Late Medieval

(Ingrem 2001)

Church Street (PM3)

Dorset

Town

High Medieval

Late Medieval

(Coy 1985b; 1992)

Ivy Street and Brown Street

Wiltshire

Town

High Medieval

Late Medieval

(Hamilton-Dyer 2000b)

Wootton Bassett

Wiltshire

Town

High Medieval

High Medieval

(Currie 1995)

Town Centre

Dorset

Town

High Medieval

Late Medieval

(Coy 1983d)

234

Appendix 3 Site Name

County

Site Type

Earliest Period

Latest Period

Reference

Anchor Brewery site 36 Milford Street / 34 Gigant Street

Wiltshire

Industrial

High Medieval

Late Medieval

(Hamilton-Dyer 2005a)

Wiltshire

Town

High Medieval

Late Medieval

(Baxter 2005)

Trowbridge center

Wiltshire

Town

High Medieval

Late Medieval

(Bourdillon 1993a)

Ludgershall Castle

Wiltshire

Castle

High Medieval

Late Medieval

(Ellis 2000)

New Street (PM7)

Dorset

Town

Late Mediaeval

Late Mediaeval

(Coy 1985b; 1992)

Chantry Field

Dorset

Town

Late Mediaeval

Late Mediaeval

(Iles 1992)

Orchard Car park (PM2)

Dorset

Town

Late Mediaeval

Late Mediaeval

(Coy 1985b; 1992)

Town cellars (PM11)

Dorset

Town

Late Mediaeval

Late Medieval

(Coy 1977c)

Christchurch Dolpin Development (X11.1)

Dorset

Town

Late Mediaeval

Late Mediaeval

(Yonge 1983)

Hampshire

Town

Late Mediaeval

Late Mediaeval

(Noddle 1975)

Dorset

Town

Late Mediaeval

Late Mediaeval

(Yonge 1983)

Dorset

Town

Late Mediaeval

Late Mediaeval

(Coy 1983a)

Dorset

Town

Late Mediaeval

Late Mediaeval

(Coy 1982c; 1983b)

Dorset

Town

Late Mediaeval

Late Mediaeval

(Yonge 1983)

Winkle Street (SOU 162) Christchurch Trokes Garden (X7) Christchurch Keith Motors (X13) Christchurch Old Town Hall (w8) Christchurch Millhams Street (X4) The Brooks, Winchester Christchurch Priory Garderode (X3)

Hampshire

Town

Late Mediaeval

Late Mediaeval

(Brown 1991)

Dorset

Monastic

Late Mediaeval

Late Mediaeval

(Coy 1983a)

Wimborne Model Town

Dorset

Town

Late Mediaeval

Late Mediaeval

(Clark 1992)

Upper Bugle Street castle Ditch (SOU 124)

Hampshire

Castle

Late Mediaeval

Late Mediaeval

(Bourdillon 1986b)

Site V, Bishops Waltham

Hampshire

Town

Late Medieval

Late Medieval

(Coy 1976b)

Old George Mall

Wiltshire

Town

Late Medieval

Late Medieval

(Butterworth 2005)

235

Appendix 4: Sites from Yorkshire with no ABGs reported Site Name

County

Site Type

Earliest Period

Latest Period

Reference

Rudston Wold, Corner Field, Site 8

North Yorkshire

Castle

Early Neolithic

Late Neolithic

(Manby 1975)

Kilham Long Barrow

East Yorkshire

Rural Settlement

Early Neolithic

Late Neolithic

(Bramwell 1976)

Willerby Wold Long Barrow

East Yorkshire

Rural Settlement

Early Neolithic

Late Neolithic

(Bramwell 1963)

Ferrybridge Henge

West Yorkshire

Mannor House

Middle Neolithic

Early Bronze Age

(Roberts et al. 2005)

Rudston East site 3

East Yorkshire

Rural Settlement

Late Neolithic

Late Neolithic

(Bramwell 1974)

North Carnaby Temple

East Yorkshire

Town

Late Neolithic

Late Neolithic

(Bramwell 1974)

Carnaby Top Site 20

East Yorkshire

Rural Settlement

Late Neolithic

Late Neolithic

(Bramwell 1974)

Catterick Racecourse Cairn

North Yorkshire

Round barrow

Late Neolithic

Early Bronze Age

(Richardson 2003a)

Low Caythorpe

East Yorkshire

Pit complex

Late Neolithic

Late Neolithic

(Bramwell 1974)

Late Bronze Age

(Watts and Rahtz 1984)

Cowlam Wold Barrow 3

East Yorkshire

Enclosure

Early Bronze Age

Wetwang Slack Barrow C

North Yorkshire

Long barrow

Early Bronze Age

Late Bronze Age

(Simms 1979)

Green Howe

North Yorkshire

Hillfort

Early Bronze Age

Late Bronze Age

(Jackson 1971)

Rudston Barrow LXII

East Yorkshire

Long barrow

Early Bronze Age

Middle Bronze Age

(Bramwell 1972)

Wetwang Slack Barrow A

North Yorkshire

Rural Settlement

Early Bronze Age

Late Bronze Age

(Simms 1979)

Cowlam Wold Barrow 1

East Yorkshire

Industrial

Late Bronze Age

(Watts and Rahtz 1984)

Staple Howe

North Yorkshire

Industrial

Early Bronze Age Late Bronze Age

Late Iron Age

(King 1963b)

Driffield RAF Station

East Yorkshire

Town

Early Iron Age

Late Iron Age

(Philips 1960)

Bursea Grange

East Yorkshire

Enclosure

Early Iron Age

Late Iron Age

(Gidney 1999c)

Grimthorpe Hillfort

East Yorkshire

Town

Early Iron Age

Late Iron Age

(Jarman et al. 1968)

Catterick Racecourse Settlement

North Yorkshire

Rural Settlement

Early Iron Age

Late Iron Age

(Richardson 2003a)

Rillington

North Yorkshire

Round barrow

Middle Iron Age

Late Iron Age

(Turnbull 1983)

Staple Howe

North Yorkshire

Round barrow

Middle Iron Age

Late Iron Age

(King 1963a)

Rock Castle

North Yorkshire

Town

Middle Iron Age

Late Iron Age

(Gidney 1994)

Late Iron Age

Early RomanoBritish

(Richardson 2003b) (Gidney 1999a)

Topham Farm

South Yorkshire

Town

South Lawn 'Ladder' settlement

East Yorkshire

Other

Late Iron Age

Late RomanoBritish

Bursea House

East Yorkshire

Henge

Late Iron Age

Late RomanoBritish

(Stallibrass 1999)

Stanwick camp

North Yorkshire

Round barrow

Late Iron Age

Early RomanoBritish

(Wheeler 1954)

Levisham moor enclosure A

North Yorkshire

Fort

Early RomanoBritish

Early RomanoBritish

(Hayes 1983)

Brough-on-Humber

East Yorkshire

Manor

Early RomanoBritish

Late RomanoBritish

(Harcourt 1969a)

Cat Babbleton Farm

North Yorkshire

Round barrow

Early RomanoBritish

Late RomanoBritish

(Cardwell 1989)

Railway Offices, Holgate Road, York

North Yorkshire

Castle

Early RomanoBritish

Late RomanoBritish

(Jewell 1960)

Appendix 4 Site Name Kings manor, York

North Yorkshire

Monastic

Early RomanoBritish

Latest Period Middle RomanoBritish

Doncaster civil settlement

South Yorkshire

Military

Early RomanoBritish

Late RomanoBritish

(Turner 1986)

Billingley Drive

South Yorkshire

Town

Early RomanoBritish

Late RomanoBritish

(Gidney 2004)

Welham Bridge

East Yorkshire

Rural Settlement

Early RomanoBritish

Late RomanoBritish

(Gidney 1999b)

Town

Early RomanoBritish

Late AngloSaxon

(Rutter and Duke 1958)

Late RomanoBritish

(Hayfield 1986)

Crossgates

County

North Yorkshire

Site Type

Earliest Period

Reference (Radley 1972)

Wharram Grange Roman Villa

North Yorkshire

Monastic

Middle Romano-British

Roman Signal Station Carr Naze

East Yorkshire

Town

Late RomanoBritish

Early AngloSaxon

(Dobney et al. 2000)

D-Shaped enclosure Upton

West Yorkshire

Town

Late RomanoBritish

Late RomanoBritish

(Berg 1995)

Cottam COT93

East Yorkshire

Monastic

Middle AngloSaxon

Late AngloSaxon

(Dobney et al. 1999)

Wharram Site 39

North Yorkshire

Town

Middle AngloSaxon

Middle AngloSaxon

(Stevens 1987; 1992)

Ribblehead

West Yorkshire

Enclosure

Late AngloSaxon

Late AngloSaxon

(Rackham 1977a)

Lurk Lane, Beverley

East Yorkshire

Manorial

Late AngloSaxon

Late Medieval

(Scott 1991)

Sandal Castle

West Yorkshire

Pit complex

High Medieval

Late Medieval

(Griffith et al. 1983)

Baile Hill, York

North Yorkshire

Enclosure

High Medieval

High Medieval

(Rackham 1977b)

Augustinian Friary garden, Hull

East Yorkshire

Town

High Medieval

Late Medieval

(Scott 1993)

33-35 Eastgate, Beverley

East Yorkshire

Pit complex

High Medieval

Late Medieval

(Scott 1992)

6-14 Highgate

East Yorkshire

Pit complex

High Medieval

Late Medieval

(Watkins and Williams 1981)

Sherburn Manor

East Yorkshire

Pit complex

High Medieval

High Medieval

(Rushe et al. 1994)

Dominican Priory Beverley

East Yorkshire

Rural Settlement

High Medieval

High Medieval

(Gilchrist 1996)

Weaverthorpe Manor

East Yorkshire

Monastic

High Medieval

Late Medieval

(Harcourt 1969c; 1972)

Cowick moat

North Yorkshire

Rural Settlement

High Medieval

Late Medieval

(Hayfield and Greig 1990)

Scarborough Castle

North Yorkshire

Rural Settlement

High Medieval

Late Medieval

(Weinstock 2005)

Mytongate, Hull

East Yorkshire

Monastic

Late Medieval

Late Medieval

(Saunder and Phillips 1993)

Vicar Lane, Hull

East Yorkshire

Rural Settlement

Late Medieval

Late Medieval

(Phillips 1993)

1-2 Tower Street, York

North Yorkshire

Town

Late Medieval

Late Medieval

(Bond and O'Connor 1999)

The Bedern, York

North Yorkshire

Town

Late Medieval

Late Medieval

(Bond and O'Connor 1999)

Chapel Lane Staith, Hull

East Yorkshire

Rural Settlement

Late Medieval

Late Medieval

(Phillips 1979)

Sewer Lane, Hull

East Yorkshire

Town

Late Medieval

Late Medieval

(Armstrong 1977)

Hospital of St Giles

North Yorkshire

Villa

Late Medieval

Late Medieval

(Stallibrass 1995a)

21-33 Aldwark

North Yorkshire

Castle

Late Medieval

Late Medieval

(Bond and O'Connor 1999)

Wytelard property, Blackfriargate, Hull

East Yorkshire

Enclosure

Late Medieval

Late Medieval

(Scott 1987)

Wharram Percey

North Yorkshire

Pit complex

Late Medieval

Late Medieval

(Ryder 1974)

237

Investigating Animal Burials; Ritual, Mundane and Beyond Site Name

County

Site Type

Earliest Period

Latest Period

Reference

St John's Priory

West Yorkshire

Hillfort

Late Medieval

Late Medieval

(Ryder 1965)

Brough-on-Humber

East Yorkshire

Enclosure

Late Medieval

Late Medieval

(Harcourt 1968a)

High Street

North Yorkshire

Rural Settlement

Late Medieval

Late Medieval

(Rackham 1985)

Chapel Grath

East Yorkshire

Round barrow

Late Medieval

Late Medieval

(Youngson 1978)

Skelton Orchard Field

North Yorkshire

Enclosure

Late Medieval

Late Medieval

(Screeton and Spratt 2001)

238